Merge pull request #9583 from containerd/dependabot/go_modules/github.com/containerd/cgroups/v3-3.0.3

build(deps): bump github.com/containerd/cgroups/v3 from 3.0.2 to 3.0.3

go.mod

@@ -10,7 +10,7 @@ require (
 	github.com/Microsoft/hcsshim v0.12.0-rc.2
 	github.com/container-orchestrated-devices/container-device-interface v0.6.1
 	github.com/containerd/btrfs/v2 v2.0.0
-	github.com/containerd/cgroups/v3 v3.0.2
+	github.com/containerd/cgroups/v3 v3.0.3
 	github.com/containerd/console v1.0.3
 	github.com/containerd/continuity v0.4.3
 	github.com/containerd/fifo v1.1.0
@@ -84,7 +84,7 @@ require (
 	github.com/beorn7/perks v1.0.1 // indirect
 	github.com/cenkalti/backoff/v4 v4.2.1 // indirect
 	github.com/cespare/xxhash/v2 v2.2.0 // indirect
-	github.com/cilium/ebpf v0.9.1 // indirect
+	github.com/cilium/ebpf v0.11.0 // indirect
 	github.com/containerd/containerd v1.7.8 // indirect
 	github.com/cpuguy83/go-md2man/v2 v2.0.2 // indirect
 	github.com/emicklei/go-restful/v3 v3.10.2 // indirect
@@ -114,12 +114,13 @@ require (
 	go.opencensus.io v0.24.0 // indirect
 	go.opentelemetry.io/otel/metric v1.21.0 // indirect
 	go.opentelemetry.io/proto/otlp v1.0.0 // indirect
+	golang.org/x/exp v0.0.0-20231214170342-aacd6d4b4611 // indirect
 	golang.org/x/net v0.19.0 // indirect
 	golang.org/x/oauth2 v0.13.0 // indirect
 	golang.org/x/term v0.15.0 // indirect
 	golang.org/x/text v0.14.0 // indirect
 	golang.org/x/time v0.3.0 // indirect
-	golang.org/x/tools v0.14.0 // indirect
+	golang.org/x/tools v0.16.0 // indirect
 	google.golang.org/appengine v1.6.8 // indirect
 	google.golang.org/genproto/googleapis/api v0.0.0-20231002182017-d307bd883b97 // indirect
 	gopkg.in/inf.v0 v0.9.1 // indirect

go.sum

@@ -33,8 +33,8 @@ github.com/cespare/xxhash/v2 v2.2.0/go.mod h1:VGX0DQ3Q6kWi7AoAeZDth3/j3BFtOZR5XL
 github.com/chzyer/logex v1.1.10/go.mod h1:+Ywpsq7O8HXn0nuIou7OrIPyXbp3wmkHB+jjWRnGsAI=
 github.com/chzyer/readline v0.0.0-20180603132655-2972be24d48e/go.mod h1:nSuG5e5PlCu98SY8svDHJxuZscDgtXS6KTTbou5AhLI=
 github.com/chzyer/test v0.0.0-20180213035817-a1ea475d72b1/go.mod h1:Q3SI9o4m/ZMnBNeIyt5eFwwo7qiLfzFZmjNmxjkiQlU=
-github.com/cilium/ebpf v0.9.1 h1:64sn2K3UKw8NbP/blsixRpF3nXuyhz/VjRlRzvlBRu4=
+github.com/cilium/ebpf v0.11.0 h1:V8gS/bTCCjX9uUnkUFUpPsksM8n1lXBAvHcpiFk1X2Y=
-github.com/cilium/ebpf v0.9.1/go.mod h1:+OhNOIXx/Fnu1IE8bJz2dzOA+VSfyTfdNUVdlQnxUFY=
+github.com/cilium/ebpf v0.11.0/go.mod h1:WE7CZAnqOL2RouJ4f1uyNhqr2P4CCvXFIqdRDUgWsVs=
 github.com/client9/misspell v0.3.4/go.mod h1:qj6jICC3Q7zFZvVWo7KLAzC3yx5G7kyvSDkc90ppPyw=
 github.com/cncf/udpa/go v0.0.0-20191209042840-269d4d468f6f/go.mod h1:M8M6+tZqaGXZJjfX53e64911xZQV5JYwmTeXPW+k8Sc=
 github.com/cncf/xds/go v0.0.0-20230607035331-e9ce68804cb4 h1:/inchEIKaYC1Akx+H+gqO04wryn5h75LSazbRlnya1k=
@@ -42,8 +42,8 @@ github.com/container-orchestrated-devices/container-device-interface v0.6.1 h1:m
 github.com/container-orchestrated-devices/container-device-interface v0.6.1/go.mod h1:40T6oW59rFrL/ksiSs7q45GzjGlbvxnA4xaK6cyq+kA=
 github.com/containerd/btrfs/v2 v2.0.0 h1:FN4wsx7KQrYoLXN7uLP0vBV4oVWHOIKDRQ1G2Z0oL5M=
 github.com/containerd/btrfs/v2 v2.0.0/go.mod h1:swkD/7j9HApWpzl8OHfrHNxppPd9l44DFZdF94BUj9k=
-github.com/containerd/cgroups/v3 v3.0.2 h1:f5WFqIVSgo5IZmtTT3qVBo6TzI1ON6sycSBKkymb9L0=
+github.com/containerd/cgroups/v3 v3.0.3 h1:S5ByHZ/h9PMe5IOQoN7E+nMc2UcLEM/V48DGDJ9kip0=
-github.com/containerd/cgroups/v3 v3.0.2/go.mod h1:JUgITrzdFqp42uI2ryGA+ge0ap/nxzYgkGmIcetmErE=
+github.com/containerd/cgroups/v3 v3.0.3/go.mod h1:8HBe7V3aWGLFPd/k03swSIsGjZhHI2WzJmticMgVuz0=
 github.com/containerd/console v1.0.3 h1:lIr7SlA5PxZyMV30bDW0MGbiOPXwc63yRuCP0ARubLw=
 github.com/containerd/console v1.0.3/go.mod h1:7LqA/THxQ86k76b8c/EMSiaJ3h1eZkMkXar0TQ1gf3U=
 github.com/containerd/containerd v1.7.8 h1:RkwgOW3AVUT3H/dyT0W03Dc8AzlpMG65lX48KftOFSM=
@@ -94,7 +94,7 @@ github.com/envoyproxy/protoc-gen-validate v0.1.0/go.mod h1:iSmxcyjqTsJpI2R4NaDN7
 github.com/envoyproxy/protoc-gen-validate v1.0.2 h1:QkIBuU5k+x7/QXPvPPnWXWlCdaBFApVqftFV6k087DA=
 github.com/felixge/httpsnoop v1.0.3 h1:s/nj+GCswXYzN5v2DpNMuMQYe+0DDwt5WVCU6CWBdXk=
 github.com/felixge/httpsnoop v1.0.3/go.mod h1:m8KPJKqk1gH5J9DgRY2ASl2lWCfGKXixSwevea8zH2U=
-github.com/frankban/quicktest v1.14.0 h1:+cqqvzZV87b4adx/5ayVOaYZ2CrvM4ejQvUdBzPPUss=
+github.com/frankban/quicktest v1.14.5 h1:dfYrrRyLtiqT9GyKXgdh+k4inNeTvmGbuSgZ3lx3GhA=
 github.com/fsnotify/fsnotify v1.4.7/go.mod h1:jwhsz4b93w/PPRr/qN1Yymfu8t87LnFCMoQvtojpjFo=
 github.com/fsnotify/fsnotify v1.4.9/go.mod h1:znqG4EE+3YCdAaPaxE2ZRY/06pZUdp0tY4IgpuI1SZQ=
 github.com/fsnotify/fsnotify v1.7.0 h1:8JEhPFa5W2WU7YfeZzPNqzMP6Lwt7L2715Ggo0nosvA=
@@ -362,6 +362,8 @@ golang.org/x/crypto v0.0.0-20191011191535-87dc89f01550/go.mod h1:yigFU9vqHzYiE8U
 golang.org/x/crypto v0.0.0-20200622213623-75b288015ac9/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto=
 golang.org/x/crypto v0.0.0-20210921155107-089bfa567519/go.mod h1:GvvjBRRGRdwPK5ydBHafDWAxML/pGHZbMvKqRZ5+Abc=
 golang.org/x/exp v0.0.0-20190121172915-509febef88a4/go.mod h1:CJ0aWSM057203Lf6IL+f9T1iT9GByDxfZKAQTCR3kQA=
+golang.org/x/exp v0.0.0-20231214170342-aacd6d4b4611 h1:qCEDpW1G+vcj3Y7Fy52pEM1AWm3abj8WimGYejI3SC4=
+golang.org/x/exp v0.0.0-20231214170342-aacd6d4b4611/go.mod h1:iRJReGqOEeBhDZGkGbynYwcHlctCvnjTYIamk7uXpHI=
 golang.org/x/lint v0.0.0-20181026193005-c67002cb31c3/go.mod h1:UVdnD1Gm6xHRNCYTkRU2/jEulfH38KcIWyp/GAMgvoE=
 golang.org/x/lint v0.0.0-20190227174305-5b3e6a55c961/go.mod h1:wehouNa3lNwaWXcvxsM5YxQ5yQlVC4a0KAMCusXpPoU=
 golang.org/x/lint v0.0.0-20190313153728-d0100b6bd8b3/go.mod h1:6SW0HCj/g11FgYtHlgUYUwCkIfeOF89ocIRzGO/8vkc=
@@ -457,8 +459,8 @@ golang.org/x/tools v0.0.0-20200619180055-7c47624df98f/go.mod h1:EkVYQZoAsY45+roY
 golang.org/x/tools v0.0.0-20201224043029-2b0845dc783e/go.mod h1:emZCQorbCU4vsT4fOWvOPXz4eW1wZW4PmDk9uLelYpA=
 golang.org/x/tools v0.0.0-20210106214847-113979e3529a/go.mod h1:emZCQorbCU4vsT4fOWvOPXz4eW1wZW4PmDk9uLelYpA=
 golang.org/x/tools v0.1.12/go.mod h1:hNGJHUnrk76NpqgfD5Aqm5Crs+Hm0VOH/i9J2+nxYbc=
-golang.org/x/tools v0.14.0 h1:jvNa2pY0M4r62jkRQ6RwEZZyPcymeL9XZMLBbV7U2nc=
+golang.org/x/tools v0.16.0 h1:GO788SKMRunPIBCXiQyo2AaexLstOrVhuAL5YwsckQM=
-golang.org/x/tools v0.14.0/go.mod h1:uYBEerGOWcJyEORxN+Ek8+TT266gXkNlHdJBwexUsBg=
+golang.org/x/tools v0.16.0/go.mod h1:kYVVN6I1mBNoB1OX+noeBjbRk4IUEPa7JJ+TJMEooJ0=
 golang.org/x/xerrors v0.0.0-20190717185122-a985d3407aa7/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
 golang.org/x/xerrors v0.0.0-20191011141410-1b5146add898/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
 golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=

vendor/github.com/cilium/ebpf/.clang-format

@@ -14,4 +14,6 @@ KeepEmptyLinesAtTheStartOfBlocks: false
 TabWidth:        4
 UseTab:          ForContinuationAndIndentation
 ColumnLimit:     1000
+# Go compiler comments need to stay unindented.
+CommentPragmas: '^go:.*'
 ...

vendor/github.com/cilium/ebpf/.golangci.yaml

@@ -9,7 +9,6 @@ issues:
 linters:
   disable-all: true
   enable:
-    - deadcode
     - errcheck
     - goimports
     - gosimple
@@ -17,10 +16,9 @@ linters:
     - ineffassign
     - misspell
     - staticcheck
-    - structcheck
     - typecheck
     - unused
-    - varcheck
+    - gofmt
 
     # Could be enabled later:
     # - gocyclo

vendor/github.com/cilium/ebpf/ARCHITECTURE.md

@@ -1,7 +1,21 @@
 Architecture of the library
 ===
 
-    ELF -> Specifications -> Objects -> Links
+```mermaid
+graph RL
+    Program --> ProgramSpec --> ELF
+    btf.Spec --> ELF
+    Map --> MapSpec --> ELF
+    Links --> Map & Program
+    ProgramSpec -.-> btf.Spec
+    MapSpec -.-> btf.Spec
+    subgraph Collection
+        Program & Map
+    end
+    subgraph CollectionSpec
+        ProgramSpec & MapSpec & btf.Spec
+    end
+```
 
 ELF
 ---
@@ -11,7 +25,7 @@ an ELF file which contains program byte code (aka BPF), but also metadata for
 maps used by the program. The metadata follows the conventions set by libbpf
 shipped with the kernel. Certain ELF sections have special meaning
 and contain structures defined by libbpf. Newer versions of clang emit
-additional metadata in BPF Type Format (aka BTF).
+additional metadata in [BPF Type Format](#BTF).
 
 The library aims to be compatible with libbpf so that moving from a C toolchain
 to a Go one creates little friction. To that end, the [ELF reader](elf_reader.go)
@@ -20,41 +34,33 @@ if possible.
 
 The output of the ELF reader is a `CollectionSpec` which encodes
 all of the information contained in the ELF in a form that is easy to work with
-in Go.
+in Go. The returned `CollectionSpec` should be deterministic: reading the same ELF
-
+file on different systems must produce the same output.
-### BTF
+As a corollary, any changes that depend on the runtime environment like the
-
+current kernel version must happen when creating [Objects](#Objects).
-The BPF Type Format describes more than just the types used by a BPF program. It
-includes debug aids like which source line corresponds to which instructions and
-what global variables are used.
-
-[BTF parsing](internal/btf/) lives in a separate internal package since exposing
-it would mean an additional maintenance burden, and because the API still
-has sharp corners. The most important concept is the `btf.Type` interface, which
-also describes things that aren't really types like `.rodata` or `.bss` sections.
-`btf.Type`s can form cyclical graphs, which can easily lead to infinite loops if
-one is not careful. Hopefully a safe pattern to work with `btf.Type` emerges as
-we write more code that deals with it.
 
 Specifications
 ---
 
-`CollectionSpec`, `ProgramSpec` and `MapSpec` are blueprints for in-kernel
+`CollectionSpec` is a very simple container for `ProgramSpec`, `MapSpec` and
-objects and contain everything necessary to execute the relevant `bpf(2)`
+`btf.Spec`. Avoid adding functionality to it if possible.
-syscalls. Since the ELF reader outputs a `CollectionSpec` it's possible to
-modify clang-compiled BPF code, for example to rewrite constants. At the same
-time the [asm](asm/) package provides an assembler that can be used to generate
-`ProgramSpec` on the fly.
 
-Creating a spec should never require any privileges or be restricted in any way,
+`ProgramSpec` and `MapSpec` are blueprints for in-kernel
-for example by only allowing programs in native endianness. This ensures that
+objects and contain everything necessary to execute the relevant `bpf(2)`
-the library stays flexible.
+syscalls. They refer to `btf.Spec` for type information such as `Map` key and
+value types.
+
+The [asm](asm/) package provides an assembler that can be used to generate
+`ProgramSpec` on the fly.
 
 Objects
 ---
 
-`Program` and `Map` are the result of loading specs into the kernel. Sometimes
+`Program` and `Map` are the result of loading specifications into the kernel.
-loading a spec will fail because the kernel is too old, or a feature is not
+Features that depend on knowledge of the current system (e.g kernel version)
+are implemented at this point.
+
+Sometimes loading a spec will fail because the kernel is too old, or a feature is not
 enabled. There are multiple ways the library deals with that:
 
 * Fallback: older kernels don't allow naming programs and maps. The library
@@ -73,7 +79,7 @@ useful when our higher-level API doesn't support a particular use case.
 Links
 ---
 
-BPF can be attached to many different points in the kernel and newer BPF hooks
+Programs can be attached to many different points in the kernel and newer BPF hooks
 tend to use bpf_link to do so. Older hooks unfortunately use a combination of
 syscalls, netlink messages, etc. Adding support for a new link type should not
 pull in large dependencies like netlink, so XDP programs or tracepoints are

vendor/github.com/cilium/ebpf/CONTRIBUTING.md

@@ -5,15 +5,23 @@ the form of pull requests and issues reporting bugs or suggesting new features
 are welcome. Please take a look at [the architecture](ARCHITECTURE.md) to get
 a better understanding for the high-level goals.
 
-New features must be accompanied by tests. Before starting work on any large
+## Adding a new feature
-feature, please [join](https://ebpf.io/slack) the
-[#ebpf-go](https://cilium.slack.com/messages/ebpf-go) channel on Slack to
-discuss the design first.
 
-When submitting pull requests, consider writing details about what problem you
+1. [Join](https://ebpf.io/slack) the
-are solving and why the proposed approach solves that problem in commit messages
+[#ebpf-go](https://cilium.slack.com/messages/ebpf-go) channel to discuss your requirements and how the feature can be implemented. The most important part is figuring out how much new exported API is necessary. **The less new API is required the easier it will be to land the feature.**
-and/or pull request description to help future library users and maintainers to
+2. (*optional*) Create a draft PR if you want to discuss the implementation or have hit a problem. It's fine if this doesn't compile or contains debug statements.
-reason about the proposed changes.
+3. Create a PR that is ready to merge. This must pass CI and have tests.
+
+### API stability
+
+The library doesn't guarantee the stability of its API at the moment.
+
+1. If possible avoid breakage by introducing new API and deprecating the old one
+   at the same time. If an API was deprecated in v0.x it can be removed in v0.x+1.
+2. Breaking API in a way that causes compilation failures is acceptable but must
+   have good reasons.
+3. Changing the semantics of the API without causing compilation failures is
+   heavily discouraged.
 
 ## Running the tests
 
@@ -35,6 +43,6 @@ Examples:
 ./run-tests.sh 5.4
 
 # Run a subset of tests:
-./run-tests.sh 5.4 go test ./link
+./run-tests.sh 5.4 ./link
 ```
 

vendor/github.com/cilium/ebpf/MAINTAINERS.md

@@ -1,8 +1,3 @@
 # Maintainers
 
- * [Lorenz Bauer] 
+Maintainers can be found in the [Cilium Maintainers file](https://github.com/cilium/community/blob/main/roles/Maintainers.md)
- * [Timo Beckers] (Isovalent)
-
-
-[Lorenz Bauer]: https://github.com/lmb
-[Timo Beckers]: https://github.com/ti-mo

vendor/github.com/cilium/ebpf/Makefile

@@ -28,6 +28,7 @@ TARGETS := \
 	testdata/loader-clang-7 \
 	testdata/loader-clang-9 \
 	testdata/loader-$(CLANG) \
+	testdata/manyprogs \
 	testdata/btf_map_init \
 	testdata/invalid_map \
 	testdata/raw_tracepoint \
@@ -39,9 +40,15 @@ TARGETS := \
 	testdata/map_spin_lock \
 	testdata/subprog_reloc \
 	testdata/fwd_decl \
+	testdata/kconfig \
+	testdata/kconfig_config \
+	testdata/kfunc \
+	testdata/invalid-kfunc \
+	testdata/kfunc-kmod \
 	btf/testdata/relocs \
 	btf/testdata/relocs_read \
-	btf/testdata/relocs_read_tgt
+	btf/testdata/relocs_read_tgt \
+	cmd/bpf2go/testdata/minimal
 
 .PHONY: all clean container-all container-shell generate
 
@@ -49,12 +56,12 @@ TARGETS := \
 
 # Build all ELF binaries using a containerized LLVM toolchain.
 container-all:
-	${CONTAINER_ENGINE} run --rm ${CONTAINER_RUN_ARGS} \
+	+${CONTAINER_ENGINE} run --rm -ti ${CONTAINER_RUN_ARGS} \
 		-v "${REPODIR}":/ebpf -w /ebpf --env MAKEFLAGS \
 		--env CFLAGS="-fdebug-prefix-map=/ebpf=." \
 		--env HOME="/tmp" \
 		"${IMAGE}:${VERSION}" \
-		$(MAKE) all
+		make all
 
 # (debug) Drop the user into a shell inside the container as root.
 container-shell:
@@ -77,9 +84,7 @@ all: format $(addsuffix -el.elf,$(TARGETS)) $(addsuffix -eb.elf,$(TARGETS)) gene
 generate: export BPF_CLANG := $(CLANG)
 generate: export BPF_CFLAGS := $(CFLAGS)
 generate:
-	go generate ./cmd/bpf2go/test
+	go generate ./...
-	go generate ./internal/sys
-	cd examples/ && go generate ./...
 
 testdata/loader-%-el.elf: testdata/loader.c
 	$* $(CFLAGS) -target bpfel -c $< -o $@
@@ -98,11 +103,11 @@ testdata/loader-%-eb.elf: testdata/loader.c
 	$(STRIP) -g $@
 
 .PHONY: generate-btf
-generate-btf: KERNEL_VERSION?=5.18
+generate-btf: KERNEL_VERSION?=5.19
 generate-btf:
 	$(eval TMP := $(shell mktemp -d))
 	curl -fL "$(CI_KERNEL_URL)/linux-$(KERNEL_VERSION).bz" -o "$(TMP)/bzImage"
-	./testdata/extract-vmlinux "$(TMP)/bzImage" > "$(TMP)/vmlinux"
+	/lib/modules/$(uname -r)/build/scripts/extract-vmlinux "$(TMP)/bzImage" > "$(TMP)/vmlinux"
 	$(OBJCOPY) --dump-section .BTF=/dev/stdout "$(TMP)/vmlinux" /dev/null | gzip > "btf/testdata/vmlinux.btf.gz"
 	curl -fL "$(CI_KERNEL_URL)/linux-$(KERNEL_VERSION)-selftests-bpf.tgz" -o "$(TMP)/selftests.tgz"
 	tar -xf "$(TMP)/selftests.tgz" --to-stdout tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.ko | \

vendor/github.com/cilium/ebpf/README.md

@@ -4,29 +4,33 @@
 
 ![HoneyGopher](.github/images/cilium-ebpf.png)
 
-eBPF is a pure Go library that provides utilities for loading, compiling, and
+ebpf-go is a pure Go library that provides utilities for loading, compiling, and
 debugging eBPF programs. It has minimal external dependencies and is intended to
 be used in long running processes.
 
-The library is maintained by [Cloudflare](https://www.cloudflare.com) and
+See [ebpf.io](https://ebpf.io) for complementary projects from the wider eBPF
-[Cilium](https://www.cilium.io).
+ecosystem.
-
-See [ebpf.io](https://ebpf.io) for other projects from the eBPF ecosystem.
 
 ## Getting Started
 
 A small collection of Go and eBPF programs that serve as examples for building
 your own tools can be found under [examples/](examples/).
 
-Contributions are highly encouraged, as they highlight certain use cases of
+[Contributions](CONTRIBUTING.md) are highly encouraged, as they highlight certain use cases of
 eBPF and the library, and help shape the future of the project.
 
 ## Getting Help
 
-Please
+The community actively monitors our [GitHub Discussions](https://github.com/cilium/ebpf/discussions) page.
-[join](https://ebpf.io/slack) the
+Please search for existing threads before starting a new one. Refrain from
+opening issues on the bug tracker if you're just starting out or if you're not
+sure if something is a bug in the library code.
+
+Alternatively, [join](https://ebpf.io/slack) the
 [#ebpf-go](https://cilium.slack.com/messages/ebpf-go) channel on Slack if you
-have questions regarding the library.
+have other questions regarding the project. Note that this channel is ephemeral
+and has its history erased past a certain point, which is less helpful for
+others running into the same problem later.
 
 ## Packages
 
@@ -49,6 +53,7 @@ This library includes the following packages:
   of `bpftool feature probe` for discovering BPF-related kernel features using native Go.
 * [rlimit](https://pkg.go.dev/github.com/cilium/ebpf/rlimit) provides a convenient API to lift
   the `RLIMIT_MEMLOCK` constraint on kernels before 5.11.
+* [btf](https://pkg.go.dev/github.com/cilium/ebpf/btf) allows reading the BPF Type Format.
 
 ## Requirements
 

vendor/github.com/cilium/ebpf/asm/func.go

@@ -14,7 +14,7 @@ func (_ BuiltinFunc) Max() BuiltinFunc {
 // You can regenerate this list using the following gawk script:
 //
 //	/FN\(.+\),/ {
-//      match($1, /\((.+)\)/, r)
+//	  match($1, /\(([a-z_0-9]+),/, r)
 //	  split(r[1], p, "_")
 //	  printf "Fn"
 //	  for (i in p) {
@@ -229,6 +229,14 @@ const (
 	FnDynptrRead
 	FnDynptrWrite
 	FnDynptrData
+	FnTcpRawGenSyncookieIpv4
+	FnTcpRawGenSyncookieIpv6
+	FnTcpRawCheckSyncookieIpv4
+	FnTcpRawCheckSyncookieIpv6
+	FnKtimeGetTaiNs
+	FnUserRingbufDrain
+	FnCgrpStorageGet
+	FnCgrpStorageDelete
 
 	maxBuiltinFunc
 )

vendor/github.com/cilium/ebpf/asm/func_string.go

@@ -212,12 +212,20 @@ func _() {
 	_ = x[FnDynptrRead-201]
 	_ = x[FnDynptrWrite-202]
 	_ = x[FnDynptrData-203]
-	_ = x[maxBuiltinFunc-204]
+	_ = x[FnTcpRawGenSyncookieIpv4-204]
+	_ = x[FnTcpRawGenSyncookieIpv6-205]
+	_ = x[FnTcpRawCheckSyncookieIpv4-206]
+	_ = x[FnTcpRawCheckSyncookieIpv6-207]
+	_ = x[FnKtimeGetTaiNs-208]
+	_ = x[FnUserRingbufDrain-209]
+	_ = x[FnCgrpStorageGet-210]
+	_ = x[FnCgrpStorageDelete-211]
+	_ = x[maxBuiltinFunc-212]
 }
 
-const _BuiltinFunc_name = "FnUnspecFnMapLookupElemFnMapUpdateElemFnMapDeleteElemFnProbeReadFnKtimeGetNsFnTracePrintkFnGetPrandomU32FnGetSmpProcessorIdFnSkbStoreBytesFnL3CsumReplaceFnL4CsumReplaceFnTailCallFnCloneRedirectFnGetCurrentPidTgidFnGetCurrentUidGidFnGetCurrentCommFnGetCgroupClassidFnSkbVlanPushFnSkbVlanPopFnSkbGetTunnelKeyFnSkbSetTunnelKeyFnPerfEventReadFnRedirectFnGetRouteRealmFnPerfEventOutputFnSkbLoadBytesFnGetStackidFnCsumDiffFnSkbGetTunnelOptFnSkbSetTunnelOptFnSkbChangeProtoFnSkbChangeTypeFnSkbUnderCgroupFnGetHashRecalcFnGetCurrentTaskFnProbeWriteUserFnCurrentTaskUnderCgroupFnSkbChangeTailFnSkbPullDataFnCsumUpdateFnSetHashInvalidFnGetNumaNodeIdFnSkbChangeHeadFnXdpAdjustHeadFnProbeReadStrFnGetSocketCookieFnGetSocketUidFnSetHashFnSetsockoptFnSkbAdjustRoomFnRedirectMapFnSkRedirectMapFnSockMapUpdateFnXdpAdjustMetaFnPerfEventReadValueFnPerfProgReadValueFnGetsockoptFnOverrideReturnFnSockOpsCbFlagsSetFnMsgRedirectMapFnMsgApplyBytesFnMsgCorkBytesFnMsgPullDataFnBindFnXdpAdjustTailFnSkbGetXfrmStateFnGetStackFnSkbLoadBytesRelativeFnFibLookupFnSockHashUpdateFnMsgRedirectHashFnSkRedirectHashFnLwtPushEncapFnLwtSeg6StoreBytesFnLwtSeg6AdjustSrhFnLwtSeg6ActionFnRcRepeatFnRcKeydownFnSkbCgroupIdFnGetCurrentCgroupIdFnGetLocalStorageFnSkSelectReuseportFnSkbAncestorCgroupIdFnSkLookupTcpFnSkLookupUdpFnSkReleaseFnMapPushElemFnMapPopElemFnMapPeekElemFnMsgPushDataFnMsgPopDataFnRcPointerRelFnSpinLockFnSpinUnlockFnSkFullsockFnTcpSockFnSkbEcnSetCeFnGetListenerSockFnSkcLookupTcpFnTcpCheckSyncookieFnSysctlGetNameFnSysctlGetCurrentValueFnSysctlGetNewValueFnSysctlSetNewValueFnStrtolFnStrtoulFnSkStorageGetFnSkStorageDeleteFnSendSignalFnTcpGenSyncookieFnSkbOutputFnProbeReadUserFnProbeReadKernelFnProbeReadUserStrFnProbeReadKernelStrFnTcpSendAckFnSendSignalThreadFnJiffies64FnReadBranchRecordsFnGetNsCurrentPidTgidFnXdpOutputFnGetNetnsCookieFnGetCurrentAncestorCgroupIdFnSkAssignFnKtimeGetBootNsFnSeqPrintfFnSeqWriteFnSkCgroupIdFnSkAncestorCgroupIdFnRingbufOutputFnRingbufReserveFnRingbufSubmitFnRingbufDiscardFnRingbufQueryFnCsumLevelFnSkcToTcp6SockFnSkcToTcpSockFnSkcToTcpTimewaitSockFnSkcToTcpRequestSockFnSkcToUdp6SockFnGetTaskStackFnLoadHdrOptFnStoreHdrOptFnReserveHdrOptFnInodeStorageGetFnInodeStorageDeleteFnDPathFnCopyFromUserFnSnprintfBtfFnSeqPrintfBtfFnSkbCgroupClassidFnRedirectNeighFnPerCpuPtrFnThisCpuPtrFnRedirectPeerFnTaskStorageGetFnTaskStorageDeleteFnGetCurrentTaskBtfFnBprmOptsSetFnKtimeGetCoarseNsFnImaInodeHashFnSockFromFileFnCheckMtuFnForEachMapElemFnSnprintfFnSysBpfFnBtfFindByNameKindFnSysCloseFnTimerInitFnTimerSetCallbackFnTimerStartFnTimerCancelFnGetFuncIpFnGetAttachCookieFnTaskPtRegsFnGetBranchSnapshotFnTraceVprintkFnSkcToUnixSockFnKallsymsLookupNameFnFindVmaFnLoopFnStrncmpFnGetFuncArgFnGetFuncRetFnGetFuncArgCntFnGetRetvalFnSetRetvalFnXdpGetBuffLenFnXdpLoadBytesFnXdpStoreBytesFnCopyFromUserTaskFnSkbSetTstampFnImaFileHashFnKptrXchgFnMapLookupPercpuElemFnSkcToMptcpSockFnDynptrFromMemFnRingbufReserveDynptrFnRingbufSubmitDynptrFnRingbufDiscardDynptrFnDynptrReadFnDynptrWriteFnDynptrDatamaxBuiltinFunc"
+const _BuiltinFunc_name = "FnUnspecFnMapLookupElemFnMapUpdateElemFnMapDeleteElemFnProbeReadFnKtimeGetNsFnTracePrintkFnGetPrandomU32FnGetSmpProcessorIdFnSkbStoreBytesFnL3CsumReplaceFnL4CsumReplaceFnTailCallFnCloneRedirectFnGetCurrentPidTgidFnGetCurrentUidGidFnGetCurrentCommFnGetCgroupClassidFnSkbVlanPushFnSkbVlanPopFnSkbGetTunnelKeyFnSkbSetTunnelKeyFnPerfEventReadFnRedirectFnGetRouteRealmFnPerfEventOutputFnSkbLoadBytesFnGetStackidFnCsumDiffFnSkbGetTunnelOptFnSkbSetTunnelOptFnSkbChangeProtoFnSkbChangeTypeFnSkbUnderCgroupFnGetHashRecalcFnGetCurrentTaskFnProbeWriteUserFnCurrentTaskUnderCgroupFnSkbChangeTailFnSkbPullDataFnCsumUpdateFnSetHashInvalidFnGetNumaNodeIdFnSkbChangeHeadFnXdpAdjustHeadFnProbeReadStrFnGetSocketCookieFnGetSocketUidFnSetHashFnSetsockoptFnSkbAdjustRoomFnRedirectMapFnSkRedirectMapFnSockMapUpdateFnXdpAdjustMetaFnPerfEventReadValueFnPerfProgReadValueFnGetsockoptFnOverrideReturnFnSockOpsCbFlagsSetFnMsgRedirectMapFnMsgApplyBytesFnMsgCorkBytesFnMsgPullDataFnBindFnXdpAdjustTailFnSkbGetXfrmStateFnGetStackFnSkbLoadBytesRelativeFnFibLookupFnSockHashUpdateFnMsgRedirectHashFnSkRedirectHashFnLwtPushEncapFnLwtSeg6StoreBytesFnLwtSeg6AdjustSrhFnLwtSeg6ActionFnRcRepeatFnRcKeydownFnSkbCgroupIdFnGetCurrentCgroupIdFnGetLocalStorageFnSkSelectReuseportFnSkbAncestorCgroupIdFnSkLookupTcpFnSkLookupUdpFnSkReleaseFnMapPushElemFnMapPopElemFnMapPeekElemFnMsgPushDataFnMsgPopDataFnRcPointerRelFnSpinLockFnSpinUnlockFnSkFullsockFnTcpSockFnSkbEcnSetCeFnGetListenerSockFnSkcLookupTcpFnTcpCheckSyncookieFnSysctlGetNameFnSysctlGetCurrentValueFnSysctlGetNewValueFnSysctlSetNewValueFnStrtolFnStrtoulFnSkStorageGetFnSkStorageDeleteFnSendSignalFnTcpGenSyncookieFnSkbOutputFnProbeReadUserFnProbeReadKernelFnProbeReadUserStrFnProbeReadKernelStrFnTcpSendAckFnSendSignalThreadFnJiffies64FnReadBranchRecordsFnGetNsCurrentPidTgidFnXdpOutputFnGetNetnsCookieFnGetCurrentAncestorCgroupIdFnSkAssignFnKtimeGetBootNsFnSeqPrintfFnSeqWriteFnSkCgroupIdFnSkAncestorCgroupIdFnRingbufOutputFnRingbufReserveFnRingbufSubmitFnRingbufDiscardFnRingbufQueryFnCsumLevelFnSkcToTcp6SockFnSkcToTcpSockFnSkcToTcpTimewaitSockFnSkcToTcpRequestSockFnSkcToUdp6SockFnGetTaskStackFnLoadHdrOptFnStoreHdrOptFnReserveHdrOptFnInodeStorageGetFnInodeStorageDeleteFnDPathFnCopyFromUserFnSnprintfBtfFnSeqPrintfBtfFnSkbCgroupClassidFnRedirectNeighFnPerCpuPtrFnThisCpuPtrFnRedirectPeerFnTaskStorageGetFnTaskStorageDeleteFnGetCurrentTaskBtfFnBprmOptsSetFnKtimeGetCoarseNsFnImaInodeHashFnSockFromFileFnCheckMtuFnForEachMapElemFnSnprintfFnSysBpfFnBtfFindByNameKindFnSysCloseFnTimerInitFnTimerSetCallbackFnTimerStartFnTimerCancelFnGetFuncIpFnGetAttachCookieFnTaskPtRegsFnGetBranchSnapshotFnTraceVprintkFnSkcToUnixSockFnKallsymsLookupNameFnFindVmaFnLoopFnStrncmpFnGetFuncArgFnGetFuncRetFnGetFuncArgCntFnGetRetvalFnSetRetvalFnXdpGetBuffLenFnXdpLoadBytesFnXdpStoreBytesFnCopyFromUserTaskFnSkbSetTstampFnImaFileHashFnKptrXchgFnMapLookupPercpuElemFnSkcToMptcpSockFnDynptrFromMemFnRingbufReserveDynptrFnRingbufSubmitDynptrFnRingbufDiscardDynptrFnDynptrReadFnDynptrWriteFnDynptrDataFnTcpRawGenSyncookieIpv4FnTcpRawGenSyncookieIpv6FnTcpRawCheckSyncookieIpv4FnTcpRawCheckSyncookieIpv6FnKtimeGetTaiNsFnUserRingbufDrainFnCgrpStorageGetFnCgrpStorageDeletemaxBuiltinFunc"
 
-var _BuiltinFunc_index = [...]uint16{0, 8, 23, 38, 53, 64, 76, 89, 104, 123, 138, 153, 168, 178, 193, 212, 230, 246, 264, 277, 289, 306, 323, 338, 348, 363, 380, 394, 406, 416, 433, 450, 466, 481, 497, 512, 528, 544, 568, 583, 596, 608, 624, 639, 654, 669, 683, 700, 714, 723, 735, 750, 763, 778, 793, 808, 828, 847, 859, 875, 894, 910, 925, 939, 952, 958, 973, 990, 1000, 1022, 1033, 1049, 1066, 1082, 1096, 1115, 1133, 1148, 1158, 1169, 1182, 1202, 1219, 1238, 1259, 1272, 1285, 1296, 1309, 1321, 1334, 1347, 1359, 1373, 1383, 1395, 1407, 1416, 1429, 1446, 1460, 1479, 1494, 1517, 1536, 1555, 1563, 1572, 1586, 1603, 1615, 1632, 1643, 1658, 1675, 1693, 1713, 1725, 1743, 1754, 1773, 1794, 1805, 1821, 1849, 1859, 1875, 1886, 1896, 1908, 1928, 1943, 1959, 1974, 1990, 2004, 2015, 2030, 2044, 2066, 2087, 2102, 2116, 2128, 2141, 2156, 2173, 2193, 2200, 2214, 2227, 2241, 2259, 2274, 2285, 2297, 2311, 2327, 2346, 2365, 2378, 2396, 2410, 2424, 2434, 2450, 2460, 2468, 2487, 2497, 2508, 2526, 2538, 2551, 2562, 2579, 2591, 2610, 2624, 2639, 2659, 2668, 2674, 2683, 2695, 2707, 2722, 2733, 2744, 2759, 2773, 2788, 2806, 2820, 2833, 2843, 2864, 2880, 2895, 2917, 2938, 2960, 2972, 2985, 2997, 3011}
+var _BuiltinFunc_index = [...]uint16{0, 8, 23, 38, 53, 64, 76, 89, 104, 123, 138, 153, 168, 178, 193, 212, 230, 246, 264, 277, 289, 306, 323, 338, 348, 363, 380, 394, 406, 416, 433, 450, 466, 481, 497, 512, 528, 544, 568, 583, 596, 608, 624, 639, 654, 669, 683, 700, 714, 723, 735, 750, 763, 778, 793, 808, 828, 847, 859, 875, 894, 910, 925, 939, 952, 958, 973, 990, 1000, 1022, 1033, 1049, 1066, 1082, 1096, 1115, 1133, 1148, 1158, 1169, 1182, 1202, 1219, 1238, 1259, 1272, 1285, 1296, 1309, 1321, 1334, 1347, 1359, 1373, 1383, 1395, 1407, 1416, 1429, 1446, 1460, 1479, 1494, 1517, 1536, 1555, 1563, 1572, 1586, 1603, 1615, 1632, 1643, 1658, 1675, 1693, 1713, 1725, 1743, 1754, 1773, 1794, 1805, 1821, 1849, 1859, 1875, 1886, 1896, 1908, 1928, 1943, 1959, 1974, 1990, 2004, 2015, 2030, 2044, 2066, 2087, 2102, 2116, 2128, 2141, 2156, 2173, 2193, 2200, 2214, 2227, 2241, 2259, 2274, 2285, 2297, 2311, 2327, 2346, 2365, 2378, 2396, 2410, 2424, 2434, 2450, 2460, 2468, 2487, 2497, 2508, 2526, 2538, 2551, 2562, 2579, 2591, 2610, 2624, 2639, 2659, 2668, 2674, 2683, 2695, 2707, 2722, 2733, 2744, 2759, 2773, 2788, 2806, 2820, 2833, 2843, 2864, 2880, 2895, 2917, 2938, 2960, 2972, 2985, 2997, 3021, 3045, 3071, 3097, 3112, 3130, 3146, 3165, 3179}
 
 func (i BuiltinFunc) String() string {
 	if i < 0 || i >= BuiltinFunc(len(_BuiltinFunc_index)-1) {

vendor/github.com/cilium/ebpf/asm/instruction.go

@@ -226,6 +226,13 @@ func (ins *Instruction) IsFunctionCall() bool {
 	return ins.OpCode.JumpOp() == Call && ins.Src == PseudoCall
 }
 
+// IsKfuncCall returns true if the instruction calls a kfunc.
+//
+// This is not the same thing as a BPF helper call.
+func (ins *Instruction) IsKfuncCall() bool {
+	return ins.OpCode.JumpOp() == Call && ins.Src == PseudoKfuncCall
+}
+
 // IsLoadOfFunctionPointer returns true if the instruction loads a function pointer.
 func (ins *Instruction) IsLoadOfFunctionPointer() bool {
 	return ins.OpCode.IsDWordLoad() && ins.Src == PseudoFunc
@@ -318,10 +325,14 @@ func (ins Instruction) Format(f fmt.State, c rune) {
 	case cls.IsJump():
 		switch jop := op.JumpOp(); jop {
 		case Call:
-			if ins.Src == PseudoCall {
+			switch ins.Src {
+			case PseudoCall:
 				// bpf-to-bpf call
 				fmt.Fprint(f, ins.Constant)
-			} else {
+			case PseudoKfuncCall:
+				// kfunc call
+				fmt.Fprintf(f, "Kfunc(%d)", ins.Constant)
+			default:
 				fmt.Fprint(f, BuiltinFunc(ins.Constant))
 			}
 
@@ -354,6 +365,13 @@ func (ins Instruction) Size() uint64 {
 	return uint64(InstructionSize * ins.OpCode.rawInstructions())
 }
 
+// WithMetadata sets the given Metadata on the Instruction. e.g. to copy
+// Metadata from another Instruction when replacing it.
+func (ins Instruction) WithMetadata(meta Metadata) Instruction {
+	ins.Metadata = meta
+	return ins
+}
+
 type symbolMeta struct{}
 
 // WithSymbol marks the Instruction as a Symbol, which other Instructions

vendor/github.com/cilium/ebpf/asm/register.go

@@ -39,6 +39,7 @@ const (
 	PseudoMapValue  = R2 // BPF_PSEUDO_MAP_VALUE
 	PseudoCall      = R1 // BPF_PSEUDO_CALL
 	PseudoFunc      = R4 // BPF_PSEUDO_FUNC
+	PseudoKfuncCall = R2 // BPF_PSEUDO_KFUNC_CALL
 )
 
 func (r Register) String() string {

vendor/github.com/cilium/ebpf/attachtype_string.go

@@ -51,11 +51,12 @@ func _() {
 	_ = x[AttachSkReuseportSelect-39]
 	_ = x[AttachSkReuseportSelectOrMigrate-40]
 	_ = x[AttachPerfEvent-41]
+	_ = x[AttachTraceKprobeMulti-42]
 }
 
-const _AttachType_name = "NoneCGroupInetEgressCGroupInetSockCreateCGroupSockOpsSkSKBStreamParserSkSKBStreamVerdictCGroupDeviceSkMsgVerdictCGroupInet4BindCGroupInet6BindCGroupInet4ConnectCGroupInet6ConnectCGroupInet4PostBindCGroupInet6PostBindCGroupUDP4SendmsgCGroupUDP6SendmsgLircMode2FlowDissectorCGroupSysctlCGroupUDP4RecvmsgCGroupUDP6RecvmsgCGroupGetsockoptCGroupSetsockoptTraceRawTpTraceFEntryTraceFExitModifyReturnLSMMacTraceIterCgroupInet4GetPeernameCgroupInet6GetPeernameCgroupInet4GetSocknameCgroupInet6GetSocknameXDPDevMapCgroupInetSockReleaseXDPCPUMapSkLookupXDPSkSKBVerdictSkReuseportSelectSkReuseportSelectOrMigratePerfEvent"
+const _AttachType_name = "NoneCGroupInetEgressCGroupInetSockCreateCGroupSockOpsSkSKBStreamParserSkSKBStreamVerdictCGroupDeviceSkMsgVerdictCGroupInet4BindCGroupInet6BindCGroupInet4ConnectCGroupInet6ConnectCGroupInet4PostBindCGroupInet6PostBindCGroupUDP4SendmsgCGroupUDP6SendmsgLircMode2FlowDissectorCGroupSysctlCGroupUDP4RecvmsgCGroupUDP6RecvmsgCGroupGetsockoptCGroupSetsockoptTraceRawTpTraceFEntryTraceFExitModifyReturnLSMMacTraceIterCgroupInet4GetPeernameCgroupInet6GetPeernameCgroupInet4GetSocknameCgroupInet6GetSocknameXDPDevMapCgroupInetSockReleaseXDPCPUMapSkLookupXDPSkSKBVerdictSkReuseportSelectSkReuseportSelectOrMigratePerfEventTraceKprobeMulti"
 
-var _AttachType_index = [...]uint16{0, 4, 20, 40, 53, 70, 88, 100, 112, 127, 142, 160, 178, 197, 216, 233, 250, 259, 272, 284, 301, 318, 334, 350, 360, 371, 381, 393, 399, 408, 430, 452, 474, 496, 505, 526, 535, 543, 546, 558, 575, 601, 610}
+var _AttachType_index = [...]uint16{0, 4, 20, 40, 53, 70, 88, 100, 112, 127, 142, 160, 178, 197, 216, 233, 250, 259, 272, 284, 301, 318, 334, 350, 360, 371, 381, 393, 399, 408, 430, 452, 474, 496, 505, 526, 535, 543, 546, 558, 575, 601, 610, 626}
 
 func (i AttachType) String() string {
 	if i >= AttachType(len(_AttachType_index)-1) {

vendor/github.com/cilium/ebpf/btf/btf.go

@@ -2,7 +2,6 @@ package btf
 
 import (
 	"bufio"
-	"bytes"
 	"debug/elf"
 	"encoding/binary"
 	"errors"
@@ -11,6 +10,7 @@ import (
 	"math"
 	"os"
 	"reflect"
+	"sync"
 
 	"github.com/cilium/ebpf/internal"
 	"github.com/cilium/ebpf/internal/sys"
@@ -24,31 +24,38 @@ var (
 	ErrNotSupported    = internal.ErrNotSupported
 	ErrNotFound        = errors.New("not found")
 	ErrNoExtendedInfo  = errors.New("no extended info")
+	ErrMultipleMatches = errors.New("multiple matching types")
 )
 
 // ID represents the unique ID of a BTF object.
 type ID = sys.BTFID
 
-// Spec represents decoded BTF.
+// Spec allows querying a set of Types and loading the set into the
+// kernel.
 type Spec struct {
-	// Data from .BTF.
+	// All types contained by the spec, not including types from the base in
-	rawTypes []rawType
+	// case the spec was parsed from split BTF.
-	strings  *stringTable
+	types []Type
-
-	// All types contained by the spec. For the base type, the position of
-	// a type in the slice is its ID.
-	types types
 
 	// Type IDs indexed by type.
 	typeIDs map[Type]TypeID
 
+	// The ID of the first type in types.
+	firstTypeID TypeID
+
 	// Types indexed by essential name.
 	// Includes all struct flavors and types with the same name.
 	namedTypes map[essentialName][]Type
 
+	// String table from ELF, may be nil.
+	strings *stringTable
+
+	// Byte order of the ELF we decoded the spec from, may be nil.
 	byteOrder binary.ByteOrder
 }
 
+var btfHeaderLen = binary.Size(&btfHeader{})
+
 type btfHeader struct {
 	Magic   uint16
 	Version uint8
@@ -73,6 +80,18 @@ func (h *btfHeader) stringStart() int64 {
 	return int64(h.HdrLen + h.StringOff)
 }
 
+// newSpec creates a Spec containing only Void.
+func newSpec() *Spec {
+	return &Spec{
+		[]Type{(*Void)(nil)},
+		map[Type]TypeID{(*Void)(nil): 0},
+		0,
+		make(map[essentialName][]Type),
+		nil,
+		nil,
+	}
+}
+
 // LoadSpec opens file and calls LoadSpecFromReader on it.
 func LoadSpec(file string) (*Spec, error) {
 	fh, err := os.Open(file)
@@ -92,10 +111,7 @@ func LoadSpecFromReader(rd io.ReaderAt) (*Spec, error) {
 	file, err := internal.NewSafeELFFile(rd)
 	if err != nil {
 		if bo := guessRawBTFByteOrder(rd); bo != nil {
-			// Try to parse a naked BTF blob. This will return an error if
+			return loadRawSpec(io.NewSectionReader(rd, 0, math.MaxInt64), bo, nil)
-			// we encounter a Datasec, since we can't fix it up.
-			spec, err := loadRawSpec(io.NewSectionReader(rd, 0, math.MaxInt64), bo, nil, nil)
-			return spec, err
 		}
 
 		return nil, err
@@ -106,7 +122,7 @@ func LoadSpecFromReader(rd io.ReaderAt) (*Spec, error) {
 
 // LoadSpecAndExtInfosFromReader reads from an ELF.
 //
-// ExtInfos may be nil if the ELF doesn't contain section metadta.
+// ExtInfos may be nil if the ELF doesn't contain section metadata.
 // Returns ErrNotFound if the ELF contains no BTF.
 func LoadSpecAndExtInfosFromReader(rd io.ReaderAt) (*Spec, *ExtInfos, error) {
 	file, err := internal.NewSafeELFFile(rd)
@@ -119,7 +135,7 @@ func LoadSpecAndExtInfosFromReader(rd io.ReaderAt) (*Spec, *ExtInfos, error) {
 		return nil, nil, err
 	}
 
-	extInfos, err := loadExtInfosFromELF(file, spec.types, spec.strings)
+	extInfos, err := loadExtInfosFromELF(file, spec)
 	if err != nil && !errors.Is(err, ErrNotFound) {
 		return nil, nil, err
 	}
@@ -127,40 +143,40 @@ func LoadSpecAndExtInfosFromReader(rd io.ReaderAt) (*Spec, *ExtInfos, error) {
 	return spec, extInfos, nil
 }
 
-// variableOffsets extracts all symbols offsets from an ELF and indexes them by
+// symbolOffsets extracts all symbols offsets from an ELF and indexes them by
 // section and variable name.
 //
 // References to variables in BTF data sections carry unsigned 32-bit offsets.
 // Some ELF symbols (e.g. in vmlinux) may point to virtual memory that is well
 // beyond this range. Since these symbols cannot be described by BTF info,
 // ignore them here.
-func variableOffsets(file *internal.SafeELFFile) (map[variable]uint32, error) {
+func symbolOffsets(file *internal.SafeELFFile) (map[symbol]uint32, error) {
 	symbols, err := file.Symbols()
 	if err != nil {
 		return nil, fmt.Errorf("can't read symbols: %v", err)
 	}
 
-	variableOffsets := make(map[variable]uint32)
+	offsets := make(map[symbol]uint32)
-	for _, symbol := range symbols {
+	for _, sym := range symbols {
-		if idx := symbol.Section; idx >= elf.SHN_LORESERVE && idx <= elf.SHN_HIRESERVE {
+		if idx := sym.Section; idx >= elf.SHN_LORESERVE && idx <= elf.SHN_HIRESERVE {
 			// Ignore things like SHN_ABS
 			continue
 		}
 
-		if symbol.Value > math.MaxUint32 {
+		if sym.Value > math.MaxUint32 {
 			// VarSecinfo offset is u32, cannot reference symbols in higher regions.
 			continue
 		}
 
-		if int(symbol.Section) >= len(file.Sections) {
+		if int(sym.Section) >= len(file.Sections) {
-			return nil, fmt.Errorf("symbol %s: invalid section %d", symbol.Name, symbol.Section)
+			return nil, fmt.Errorf("symbol %s: invalid section %d", sym.Name, sym.Section)
 		}
 
-		secName := file.Sections[symbol.Section].Name
+		secName := file.Sections[sym.Section].Name
-		variableOffsets[variable{secName, symbol.Name}] = uint32(symbol.Value)
+		offsets[symbol{secName, sym.Name}] = uint32(sym.Value)
 	}
 
-	return variableOffsets, nil
+	return offsets, nil
 }
 
 func loadSpecFromELF(file *internal.SafeELFFile) (*Spec, error) {
@@ -190,7 +206,7 @@ func loadSpecFromELF(file *internal.SafeELFFile) (*Spec, error) {
 		return nil, fmt.Errorf("btf: %w", ErrNotFound)
 	}
 
-	vars, err := variableOffsets(file)
+	offsets, err := symbolOffsets(file)
 	if err != nil {
 		return nil, err
 	}
@@ -199,51 +215,66 @@ func loadSpecFromELF(file *internal.SafeELFFile) (*Spec, error) {
 		return nil, fmt.Errorf("compressed BTF is not supported")
 	}
 
-	rawTypes, rawStrings, err := parseBTF(btfSection.ReaderAt, file.ByteOrder, nil)
+	spec, err := loadRawSpec(btfSection.ReaderAt, file.ByteOrder, nil)
 	if err != nil {
 		return nil, err
 	}
 
-	err = fixupDatasec(rawTypes, rawStrings, sectionSizes, vars)
+	err = fixupDatasec(spec.types, sectionSizes, offsets)
 	if err != nil {
 		return nil, err
 	}
 
-	return inflateSpec(rawTypes, rawStrings, file.ByteOrder, nil)
+	return spec, nil
 }
 
-func loadRawSpec(btf io.ReaderAt, bo binary.ByteOrder,
+func loadRawSpec(btf io.ReaderAt, bo binary.ByteOrder, base *Spec) (*Spec, error) {
-	baseTypes types, baseStrings *stringTable) (*Spec, error) {
+	var (
+		baseStrings *stringTable
+		firstTypeID TypeID
+		err         error
+	)
+
+	if base != nil {
+		if base.firstTypeID != 0 {
+			return nil, fmt.Errorf("can't use split BTF as base")
+		}
+
+		if base.strings == nil {
+			return nil, fmt.Errorf("parse split BTF: base must be loaded from an ELF")
+		}
+
+		baseStrings = base.strings
+
+		firstTypeID, err = base.nextTypeID()
+		if err != nil {
+			return nil, err
+		}
+	}
 
 	rawTypes, rawStrings, err := parseBTF(btf, bo, baseStrings)
 	if err != nil {
 		return nil, err
 	}
 
-	return inflateSpec(rawTypes, rawStrings, bo, baseTypes)
+	types, err := inflateRawTypes(rawTypes, rawStrings, base)
-}
-
-func inflateSpec(rawTypes []rawType, rawStrings *stringTable, bo binary.ByteOrder,
-	baseTypes types) (*Spec, error) {
-
-	types, err := inflateRawTypes(rawTypes, baseTypes, rawStrings)
 	if err != nil {
 		return nil, err
 	}
 
-	typeIDs, typesByName := indexTypes(types, TypeID(len(baseTypes)))
+	typeIDs, typesByName := indexTypes(types, firstTypeID)
 
 	return &Spec{
-		rawTypes:   rawTypes,
 		namedTypes:  typesByName,
 		typeIDs:     typeIDs,
 		types:       types,
+		firstTypeID: firstTypeID,
 		strings:     rawStrings,
 		byteOrder:   bo,
 	}, nil
 }
 
-func indexTypes(types []Type, typeIDOffset TypeID) (map[Type]TypeID, map[essentialName][]Type) {
+func indexTypes(types []Type, firstTypeID TypeID) (map[Type]TypeID, map[essentialName][]Type) {
 	namedTypes := 0
 	for _, typ := range types {
 		if typ.TypeName() != "" {
@@ -261,7 +292,7 @@ func indexTypes(types []Type, typeIDOffset TypeID) (map[Type]TypeID, map[essenti
 		if name := newEssentialName(typ.TypeName()); name != "" {
 			typesByName[name] = append(typesByName[name], typ)
 		}
-		typeIDs[typ] = TypeID(i) + typeIDOffset
+		typeIDs[typ] = firstTypeID + TypeID(i)
 	}
 
 	return typeIDs, typesByName
@@ -272,20 +303,70 @@ func indexTypes(types []Type, typeIDOffset TypeID) (map[Type]TypeID, map[essenti
 // Defaults to /sys/kernel/btf/vmlinux and falls back to scanning the file system
 // for vmlinux ELFs. Returns an error wrapping ErrNotSupported if BTF is not enabled.
 func LoadKernelSpec() (*Spec, error) {
+	spec, _, err := kernelSpec()
+	if err != nil {
+		return nil, err
+	}
+	return spec.Copy(), nil
+}
+
+var kernelBTF struct {
+	sync.RWMutex
+	spec *Spec
+	// True if the spec was read from an ELF instead of raw BTF in /sys.
+	fallback bool
+}
+
+// FlushKernelSpec removes any cached kernel type information.
+func FlushKernelSpec() {
+	kernelBTF.Lock()
+	defer kernelBTF.Unlock()
+
+	kernelBTF.spec, kernelBTF.fallback = nil, false
+}
+
+func kernelSpec() (*Spec, bool, error) {
+	kernelBTF.RLock()
+	spec, fallback := kernelBTF.spec, kernelBTF.fallback
+	kernelBTF.RUnlock()
+
+	if spec == nil {
+		kernelBTF.Lock()
+		defer kernelBTF.Unlock()
+
+		spec, fallback = kernelBTF.spec, kernelBTF.fallback
+	}
+
+	if spec != nil {
+		return spec, fallback, nil
+	}
+
+	spec, fallback, err := loadKernelSpec()
+	if err != nil {
+		return nil, false, err
+	}
+
+	kernelBTF.spec, kernelBTF.fallback = spec, fallback
+	return spec, fallback, nil
+}
+
+func loadKernelSpec() (_ *Spec, fallback bool, _ error) {
 	fh, err := os.Open("/sys/kernel/btf/vmlinux")
 	if err == nil {
 		defer fh.Close()
 
-		return loadRawSpec(fh, internal.NativeEndian, nil, nil)
+		spec, err := loadRawSpec(fh, internal.NativeEndian, nil)
+		return spec, false, err
 	}
 
 	file, err := findVMLinux()
 	if err != nil {
-		return nil, err
+		return nil, false, err
 	}
 	defer file.Close()
 
-	return loadSpecFromELF(file)
+	spec, err := loadSpecFromELF(file)
+	return spec, true, err
 }
 
 // findVMLinux scans multiple well-known paths for vmlinux kernel images.
@@ -388,57 +469,102 @@ func parseBTF(btf io.ReaderAt, bo binary.ByteOrder, baseStrings *stringTable) ([
 	return rawTypes, rawStrings, nil
 }
 
-type variable struct {
+type symbol struct {
 	section string
 	name    string
 }
 
-func fixupDatasec(rawTypes []rawType, rawStrings *stringTable, sectionSizes map[string]uint32, variableOffsets map[variable]uint32) error {
+// fixupDatasec attempts to patch up missing info in Datasecs and its members by
-	for i, rawType := range rawTypes {
+// supplementing them with information from the ELF headers and symbol table.
-		if rawType.Kind() != kindDatasec {
+func fixupDatasec(types []Type, sectionSizes map[string]uint32, offsets map[symbol]uint32) error {
+	for _, typ := range types {
+		ds, ok := typ.(*Datasec)
+		if !ok {
 			continue
 		}
 
-		name, err := rawStrings.Lookup(rawType.NameOff)
+		name := ds.Name
-		if err != nil {
+
+		// Some Datasecs are virtual and don't have corresponding ELF sections.
+		switch name {
+		case ".ksyms":
+			// .ksyms describes forward declarations of kfunc signatures.
+			// Nothing to fix up, all sizes and offsets are 0.
+			for _, vsi := range ds.Vars {
+				_, ok := vsi.Type.(*Func)
+				if !ok {
+					// Only Funcs are supported in the .ksyms Datasec.
+					return fmt.Errorf("data section %s: expected *btf.Func, not %T: %w", name, vsi.Type, ErrNotSupported)
+				}
+			}
+
+			continue
+		case ".kconfig":
+			// .kconfig has a size of 0 and has all members' offsets set to 0.
+			// Fix up all offsets and set the Datasec's size.
+			if err := fixupDatasecLayout(ds); err != nil {
 				return err
 			}
 
-		if name == ".kconfig" || name == ".ksyms" {
+			// Fix up extern to global linkage to avoid a BTF verifier error.
-			return fmt.Errorf("reference to %s: %w", name, ErrNotSupported)
+			for _, vsi := range ds.Vars {
+				vsi.Type.(*Var).Linkage = GlobalVar
 			}
 
-		if rawTypes[i].SizeType != 0 {
 			continue
 		}
 
-		size, ok := sectionSizes[name]
+		if ds.Size != 0 {
+			continue
+		}
+
+		ds.Size, ok = sectionSizes[name]
 		if !ok {
 			return fmt.Errorf("data section %s: missing size", name)
 		}
 
-		rawTypes[i].SizeType = size
+		for i := range ds.Vars {
-
+			symName := ds.Vars[i].Type.TypeName()
-		secinfos := rawType.data.([]btfVarSecinfo)
+			ds.Vars[i].Offset, ok = offsets[symbol{name, symName}]
-		for j, secInfo := range secinfos {
-			id := int(secInfo.Type - 1)
-			if id >= len(rawTypes) {
-				return fmt.Errorf("data section %s: invalid type id %d for variable %d", name, id, j)
-			}
-
-			varName, err := rawStrings.Lookup(rawTypes[id].NameOff)
-			if err != nil {
-				return fmt.Errorf("data section %s: can't get name for type %d: %w", name, id, err)
-			}
-
-			offset, ok := variableOffsets[variable{name, varName}]
 			if !ok {
-				return fmt.Errorf("data section %s: missing offset for variable %s", name, varName)
+				return fmt.Errorf("data section %s: missing offset for symbol %s", name, symName)
+			}
+		}
 	}
 
-			secinfos[j].Offset = offset
+	return nil
+}
+
+// fixupDatasecLayout populates ds.Vars[].Offset according to var sizes and
+// alignment. Calculate and set ds.Size.
+func fixupDatasecLayout(ds *Datasec) error {
+	var off uint32
+
+	for i, vsi := range ds.Vars {
+		v, ok := vsi.Type.(*Var)
+		if !ok {
+			return fmt.Errorf("member %d: unsupported type %T", i, vsi.Type)
 		}
+
+		size, err := Sizeof(v.Type)
+		if err != nil {
+			return fmt.Errorf("variable %s: getting size: %w", v.Name, err)
 		}
+		align, err := alignof(v.Type)
+		if err != nil {
+			return fmt.Errorf("variable %s: getting alignment: %w", v.Name, err)
+		}
+
+		// Align the current member based on the offset of the end of the previous
+		// member and the alignment of the current member.
+		off = internal.Align(off, uint32(align))
+
+		ds.Vars[i].Offset = off
+
+		off += uint32(size)
+	}
+
+	ds.Size = off
 
 	return nil
 }
@@ -446,82 +572,19 @@ func fixupDatasec(rawTypes []rawType, rawStrings *stringTable, sectionSizes map[
 // Copy creates a copy of Spec.
 func (s *Spec) Copy() *Spec {
 	types := copyTypes(s.types, nil)
-
+	typeIDs, typesByName := indexTypes(types, s.firstTypeID)
-	typeIDOffset := TypeID(0)
-	if len(s.types) != 0 {
-		typeIDOffset = s.typeIDs[s.types[0]]
-	}
-	typeIDs, typesByName := indexTypes(types, typeIDOffset)
 
 	// NB: Other parts of spec are not copied since they are immutable.
 	return &Spec{
-		s.rawTypes,
-		s.strings,
 		types,
 		typeIDs,
+		s.firstTypeID,
 		typesByName,
+		s.strings,
 		s.byteOrder,
 	}
 }
 
-type marshalOpts struct {
-	ByteOrder        binary.ByteOrder
-	StripFuncLinkage bool
-}
-
-func (s *Spec) marshal(opts marshalOpts) ([]byte, error) {
-	var (
-		buf       bytes.Buffer
-		header    = new(btfHeader)
-		headerLen = binary.Size(header)
-	)
-
-	// Reserve space for the header. We have to write it last since
-	// we don't know the size of the type section yet.
-	_, _ = buf.Write(make([]byte, headerLen))
-
-	// Write type section, just after the header.
-	for _, raw := range s.rawTypes {
-		switch {
-		case opts.StripFuncLinkage && raw.Kind() == kindFunc:
-			raw.SetLinkage(StaticFunc)
-		}
-
-		if err := raw.Marshal(&buf, opts.ByteOrder); err != nil {
-			return nil, fmt.Errorf("can't marshal BTF: %w", err)
-		}
-	}
-
-	typeLen := uint32(buf.Len() - headerLen)
-
-	// Write string section after type section.
-	stringsLen := s.strings.Length()
-	buf.Grow(stringsLen)
-	if err := s.strings.Marshal(&buf); err != nil {
-		return nil, err
-	}
-
-	// Fill out the header, and write it out.
-	header = &btfHeader{
-		Magic:     btfMagic,
-		Version:   1,
-		Flags:     0,
-		HdrLen:    uint32(headerLen),
-		TypeOff:   0,
-		TypeLen:   typeLen,
-		StringOff: typeLen,
-		StringLen: uint32(stringsLen),
-	}
-
-	raw := buf.Bytes()
-	err := binary.Write(sliceWriter(raw[:headerLen]), opts.ByteOrder, header)
-	if err != nil {
-		return nil, fmt.Errorf("can't write header: %v", err)
-	}
-
-	return raw, nil
-}
-
 type sliceWriter []byte
 
 func (sw sliceWriter) Write(p []byte) (int, error) {
@@ -532,12 +595,31 @@ func (sw sliceWriter) Write(p []byte) (int, error) {
 	return copy(sw, p), nil
 }
 
+// nextTypeID returns the next unallocated type ID or an error if there are no
+// more type IDs.
+func (s *Spec) nextTypeID() (TypeID, error) {
+	id := s.firstTypeID + TypeID(len(s.types))
+	if id < s.firstTypeID {
+		return 0, fmt.Errorf("no more type IDs")
+	}
+	return id, nil
+}
+
 // TypeByID returns the BTF Type with the given type ID.
 //
 // Returns an error wrapping ErrNotFound if a Type with the given ID
 // does not exist in the Spec.
 func (s *Spec) TypeByID(id TypeID) (Type, error) {
-	return s.types.ByID(id)
+	if id < s.firstTypeID {
+		return nil, fmt.Errorf("look up type with ID %d (first ID is %d): %w", id, s.firstTypeID, ErrNotFound)
+	}
+
+	index := int(id - s.firstTypeID)
+	if index >= len(s.types) {
+		return nil, fmt.Errorf("look up type with ID %d: %w", id, ErrNotFound)
+	}
+
+	return s.types[index], nil
 }
 
 // TypeID returns the ID for a given Type.
@@ -598,17 +680,19 @@ func (s *Spec) AnyTypeByName(name string) (Type, error) {
 	return types[0], nil
 }
 
-// TypeByName searches for a Type with a specific name. Since multiple
+// TypeByName searches for a Type with a specific name. Since multiple Types
-// Types with the same name can exist, the parameter typ is taken to
+// with the same name can exist, the parameter typ is taken to narrow down the
-// narrow down the search in case of a clash.
+// search in case of a clash.
 //
-// typ must be a non-nil pointer to an implementation of a Type.
+// typ must be a non-nil pointer to an implementation of a Type. On success, the
-// On success, the address of the found Type will be copied to typ.
+// address of the found Type will be copied to typ.
 //
-// Returns an error wrapping ErrNotFound if no matching
+// Returns an error wrapping ErrNotFound if no matching Type exists in the Spec.
-// Type exists in the Spec. If multiple candidates are found,
+// Returns an error wrapping ErrMultipleTypes if multiple candidates are found.
-// an error is returned.
 func (s *Spec) TypeByName(name string, typ interface{}) error {
+	typeInterface := reflect.TypeOf((*Type)(nil)).Elem()
+
+	// typ may be **T or *Type
 	typValue := reflect.ValueOf(typ)
 	if typValue.Kind() != reflect.Ptr {
 		return fmt.Errorf("%T is not a pointer", typ)
@@ -620,7 +704,12 @@ func (s *Spec) TypeByName(name string, typ interface{}) error {
 	}
 
 	wanted := typPtr.Type()
-	if !wanted.AssignableTo(reflect.TypeOf((*Type)(nil)).Elem()) {
+	if wanted == typeInterface {
+		// This is *Type. Unwrap the value's type.
+		wanted = typPtr.Elem().Type()
+	}
+
+	if !wanted.AssignableTo(typeInterface) {
 		return fmt.Errorf("%T does not satisfy Type interface", typ)
 	}
 
@@ -636,14 +725,14 @@ func (s *Spec) TypeByName(name string, typ interface{}) error {
 		}
 
 		if candidate != nil {
-			return fmt.Errorf("type %s: multiple candidates for %T", name, typ)
+			return fmt.Errorf("type %s(%T): %w", name, typ, ErrMultipleMatches)
 		}
 
 		candidate = typ
 	}
 
 	if candidate == nil {
-		return fmt.Errorf("type %s: %w", name, ErrNotFound)
+		return fmt.Errorf("%s %s: %w", wanted, name, ErrNotFound)
 	}
 
 	typPtr.Set(reflect.ValueOf(candidate))
@@ -656,12 +745,12 @@ func (s *Spec) TypeByName(name string, typ interface{}) error {
 // Types from base are used to resolve references in the split BTF.
 // The returned Spec only contains types from the split BTF, not from the base.
 func LoadSplitSpecFromReader(r io.ReaderAt, base *Spec) (*Spec, error) {
-	return loadRawSpec(r, internal.NativeEndian, base.types, base.strings)
+	return loadRawSpec(r, internal.NativeEndian, base)
 }
 
 // TypesIterator iterates over types of a given spec.
 type TypesIterator struct {
-	spec  *Spec
+	types []Type
 	index int
 	// The last visited type in the spec.
 	Type Type
@@ -669,229 +758,112 @@ type TypesIterator struct {
 
 // Iterate returns the types iterator.
 func (s *Spec) Iterate() *TypesIterator {
-	return &TypesIterator{spec: s, index: 0}
+	// We share the backing array of types with the Spec. This is safe since
+	// we don't allow deletion or shuffling of types.
+	return &TypesIterator{types: s.types, index: 0}
 }
 
 // Next returns true as long as there are any remaining types.
 func (iter *TypesIterator) Next() bool {
-	if len(iter.spec.types) <= iter.index {
+	if len(iter.types) <= iter.index {
 		return false
 	}
 
-	iter.Type = iter.spec.types[iter.index]
+	iter.Type = iter.types[iter.index]
 	iter.index++
 	return true
 }
 
-// Handle is a reference to BTF loaded into the kernel.
+// haveBTF attempts to load a BTF blob containing an Int. It should pass on any
-type Handle struct {
+// kernel that supports BPF_BTF_LOAD.
-	fd *sys.FD
+var haveBTF = internal.NewFeatureTest("BTF", "4.18", func() error {
-
+	// 0-length anonymous integer
-	// Size of the raw BTF in bytes.
+	err := probeBTF(&Int{})
-	size uint32
-}
-
-// NewHandle loads BTF into the kernel.
-//
-// Returns ErrNotSupported if BTF is not supported.
-func NewHandle(spec *Spec) (*Handle, error) {
-	if err := haveBTF(); err != nil {
-		return nil, err
-	}
-
-	if spec.byteOrder != internal.NativeEndian {
-		return nil, fmt.Errorf("can't load %s BTF on %s", spec.byteOrder, internal.NativeEndian)
-	}
-
-	btf, err := spec.marshal(marshalOpts{
-		ByteOrder:        internal.NativeEndian,
-		StripFuncLinkage: haveFuncLinkage() != nil,
-	})
-	if err != nil {
-		return nil, fmt.Errorf("can't marshal BTF: %w", err)
-	}
-
-	if uint64(len(btf)) > math.MaxUint32 {
-		return nil, errors.New("BTF exceeds the maximum size")
-	}
-
-	attr := &sys.BtfLoadAttr{
-		Btf:     sys.NewSlicePointer(btf),
-		BtfSize: uint32(len(btf)),
-	}
-
-	fd, err := sys.BtfLoad(attr)
-	if err != nil {
-		logBuf := make([]byte, 64*1024)
-		attr.BtfLogBuf = sys.NewSlicePointer(logBuf)
-		attr.BtfLogSize = uint32(len(logBuf))
-		attr.BtfLogLevel = 1
-		// NB: The syscall will never return ENOSPC as of 5.18-rc4.
-		_, _ = sys.BtfLoad(attr)
-		return nil, internal.ErrorWithLog(err, logBuf)
-	}
-
-	return &Handle{fd, attr.BtfSize}, nil
-}
-
-// NewHandleFromID returns the BTF handle for a given id.
-//
-// Prefer calling [ebpf.Program.Handle] or [ebpf.Map.Handle] if possible.
-//
-// Returns ErrNotExist, if there is no BTF with the given id.
-//
-// Requires CAP_SYS_ADMIN.
-func NewHandleFromID(id ID) (*Handle, error) {
-	fd, err := sys.BtfGetFdById(&sys.BtfGetFdByIdAttr{
-		Id: uint32(id),
-	})
-	if err != nil {
-		return nil, fmt.Errorf("get FD for ID %d: %w", id, err)
-	}
-
-	info, err := newHandleInfoFromFD(fd)
-	if err != nil {
-		_ = fd.Close()
-		return nil, err
-	}
-
-	return &Handle{fd, info.size}, nil
-}
-
-// Spec parses the kernel BTF into Go types.
-//
-// base is used to decode split BTF and may be nil.
-func (h *Handle) Spec(base *Spec) (*Spec, error) {
-	var btfInfo sys.BtfInfo
-	btfBuffer := make([]byte, h.size)
-	btfInfo.Btf, btfInfo.BtfSize = sys.NewSlicePointerLen(btfBuffer)
-
-	if err := sys.ObjInfo(h.fd, &btfInfo); err != nil {
-		return nil, err
-	}
-
-	var baseTypes types
-	var baseStrings *stringTable
-	if base != nil {
-		baseTypes = base.types
-		baseStrings = base.strings
-	}
-
-	return loadRawSpec(bytes.NewReader(btfBuffer), internal.NativeEndian, baseTypes, baseStrings)
-}
-
-// Close destroys the handle.
-//
-// Subsequent calls to FD will return an invalid value.
-func (h *Handle) Close() error {
-	if h == nil {
-		return nil
-	}
-
-	return h.fd.Close()
-}
-
-// FD returns the file descriptor for the handle.
-func (h *Handle) FD() int {
-	return h.fd.Int()
-}
-
-// Info returns metadata about the handle.
-func (h *Handle) Info() (*HandleInfo, error) {
-	return newHandleInfoFromFD(h.fd)
-}
-
-func marshalBTF(types interface{}, strings []byte, bo binary.ByteOrder) []byte {
-	const minHeaderLength = 24
-
-	typesLen := uint32(binary.Size(types))
-	header := btfHeader{
-		Magic:     btfMagic,
-		Version:   1,
-		HdrLen:    minHeaderLength,
-		TypeOff:   0,
-		TypeLen:   typesLen,
-		StringOff: typesLen,
-		StringLen: uint32(len(strings)),
-	}
-
-	buf := new(bytes.Buffer)
-	_ = binary.Write(buf, bo, &header)
-	_ = binary.Write(buf, bo, types)
-	buf.Write(strings)
-
-	return buf.Bytes()
-}
-
-var haveBTF = internal.FeatureTest("BTF", "5.1", func() error {
-	var (
-		types struct {
-			Integer btfType
-			Var     btfType
-			btfVar  struct{ Linkage uint32 }
-		}
-		strings = []byte{0, 'a', 0}
-	)
-
-	// We use a BTF_KIND_VAR here, to make sure that
-	// the kernel understands BTF at least as well as we
-	// do. BTF_KIND_VAR was introduced ~5.1.
-	types.Integer.SetKind(kindPointer)
-	types.Var.NameOff = 1
-	types.Var.SetKind(kindVar)
-	types.Var.SizeType = 1
-
-	btf := marshalBTF(&types, strings, internal.NativeEndian)
-
-	fd, err := sys.BtfLoad(&sys.BtfLoadAttr{
-		Btf:     sys.NewSlicePointer(btf),
-		BtfSize: uint32(len(btf)),
-	})
 	if errors.Is(err, unix.EINVAL) || errors.Is(err, unix.EPERM) {
-		// Treat both EINVAL and EPERM as not supported: loading the program
-		// might still succeed without BTF.
 		return internal.ErrNotSupported
 	}
-	if err != nil {
 	return err
-	}
-
-	fd.Close()
-	return nil
 })
 
-var haveFuncLinkage = internal.FeatureTest("BTF func linkage", "5.6", func() error {
+// haveMapBTF attempts to load a minimal BTF blob containing a Var. It is
+// used as a proxy for .bss, .data and .rodata map support, which generally
+// come with a Var and Datasec. These were introduced in Linux 5.2.
+var haveMapBTF = internal.NewFeatureTest("Map BTF (Var/Datasec)", "5.2", func() error {
 	if err := haveBTF(); err != nil {
 		return err
 	}
 
-	var (
+	v := &Var{
-		types struct {
+		Name: "a",
-			FuncProto btfType
+		Type: &Pointer{(*Void)(nil)},
-			Func      btfType
 	}
-		strings = []byte{0, 'a', 0}
-	)
 
-	types.FuncProto.SetKind(kindFuncProto)
+	err := probeBTF(v)
-	types.Func.SetKind(kindFunc)
+	if errors.Is(err, unix.EINVAL) || errors.Is(err, unix.EPERM) {
-	types.Func.SizeType = 1 // aka FuncProto
+		// Treat both EINVAL and EPERM as not supported: creating the map may still
-	types.Func.NameOff = 1
+		// succeed without Btf* attrs.
-	types.Func.SetLinkage(GlobalFunc)
+		return internal.ErrNotSupported
+	}
+	return err
+})
 
-	btf := marshalBTF(&types, strings, internal.NativeEndian)
+// haveProgBTF attempts to load a BTF blob containing a Func and FuncProto. It
+// is used as a proxy for ext_info (func_info) support, which depends on
+// Func(Proto) by definition.
+var haveProgBTF = internal.NewFeatureTest("Program BTF (func/line_info)", "5.0", func() error {
+	if err := haveBTF(); err != nil {
+		return err
+	}
 
-	fd, err := sys.BtfLoad(&sys.BtfLoadAttr{
+	fn := &Func{
-		Btf:     sys.NewSlicePointer(btf),
+		Name: "a",
-		BtfSize: uint32(len(btf)),
+		Type: &FuncProto{Return: (*Void)(nil)},
-	})
+	}
+
+	err := probeBTF(fn)
+	if errors.Is(err, unix.EINVAL) || errors.Is(err, unix.EPERM) {
+		return internal.ErrNotSupported
+	}
+	return err
+})
+
+var haveFuncLinkage = internal.NewFeatureTest("BTF func linkage", "5.6", func() error {
+	if err := haveProgBTF(); err != nil {
+		return err
+	}
+
+	fn := &Func{
+		Name:    "a",
+		Type:    &FuncProto{Return: (*Void)(nil)},
+		Linkage: GlobalFunc,
+	}
+
+	err := probeBTF(fn)
 	if errors.Is(err, unix.EINVAL) {
 		return internal.ErrNotSupported
 	}
+	return err
+})
+
+func probeBTF(typ Type) error {
+	b, err := NewBuilder([]Type{typ})
 	if err != nil {
 		return err
 	}
 
+	buf, err := b.Marshal(nil, nil)
+	if err != nil {
+		return err
+	}
+
+	fd, err := sys.BtfLoad(&sys.BtfLoadAttr{
+		Btf:     sys.NewSlicePointer(buf),
+		BtfSize: uint32(len(buf)),
+	})
+
+	if err == nil {
 		fd.Close()
-	return nil
+	}
-})
+
+	return err
+}

vendor/github.com/cilium/ebpf/btf/btf_types.go

@@ -4,35 +4,41 @@ import (
 	"encoding/binary"
 	"fmt"
 	"io"
+	"unsafe"
 )
 
-//go:generate stringer -linecomment -output=btf_types_string.go -type=FuncLinkage,VarLinkage
+//go:generate stringer -linecomment -output=btf_types_string.go -type=FuncLinkage,VarLinkage,btfKind
 
 // btfKind describes a Type.
 type btfKind uint8
 
 // Equivalents of the BTF_KIND_* constants.
 const (
-	kindUnknown btfKind = iota
+	kindUnknown  btfKind = iota // Unknown
-	kindInt
+	kindInt                     // Int
-	kindPointer
+	kindPointer                 // Pointer
-	kindArray
+	kindArray                   // Array
-	kindStruct
+	kindStruct                  // Struct
-	kindUnion
+	kindUnion                   // Union
-	kindEnum
+	kindEnum                    // Enum
-	kindForward
+	kindForward                 // Forward
-	kindTypedef
+	kindTypedef                 // Typedef
-	kindVolatile
+	kindVolatile                // Volatile
-	kindConst
+	kindConst                   // Const
-	kindRestrict
+	kindRestrict                // Restrict
 	// Added ~4.20
-	kindFunc
+	kindFunc      // Func
-	kindFuncProto
+	kindFuncProto // FuncProto
 	// Added ~5.1
-	kindVar
+	kindVar     // Var
-	kindDatasec
+	kindDatasec // Datasec
 	// Added ~5.13
-	kindFloat
+	kindFloat // Float
+	// Added 5.16
+	kindDeclTag // DeclTag
+	kindTypeTag // TypeTag
+	// Added 6.0
+	kindEnum64 // Enum64
 )
 
 // FuncLinkage describes BTF function linkage metadata.
@@ -63,6 +69,8 @@ const (
 	btfTypeKindFlagMask  = 1
 )
 
+var btfTypeLen = binary.Size(btfType{})
+
 // btfType is equivalent to struct btf_type in Documentation/bpf/btf.rst.
 type btfType struct {
 	NameOff uint32
@@ -85,47 +93,6 @@ type btfType struct {
 	SizeType uint32
 }
 
-func (k btfKind) String() string {
-	switch k {
-	case kindUnknown:
-		return "Unknown"
-	case kindInt:
-		return "Integer"
-	case kindPointer:
-		return "Pointer"
-	case kindArray:
-		return "Array"
-	case kindStruct:
-		return "Struct"
-	case kindUnion:
-		return "Union"
-	case kindEnum:
-		return "Enumeration"
-	case kindForward:
-		return "Forward"
-	case kindTypedef:
-		return "Typedef"
-	case kindVolatile:
-		return "Volatile"
-	case kindConst:
-		return "Const"
-	case kindRestrict:
-		return "Restrict"
-	case kindFunc:
-		return "Function"
-	case kindFuncProto:
-		return "Function Proto"
-	case kindVar:
-		return "Variable"
-	case kindDatasec:
-		return "Section"
-	case kindFloat:
-		return "Float"
-	default:
-		return fmt.Sprintf("Unknown (%d)", k)
-	}
-}
-
 func mask(len uint32) uint32 {
 	return (1 << len) - 1
 }
@@ -164,10 +131,43 @@ func (bt *btfType) SetVlen(vlen int) {
 	bt.setInfo(uint32(vlen), btfTypeVlenMask, btfTypeVlenShift)
 }
 
-func (bt *btfType) KindFlag() bool {
+func (bt *btfType) kindFlagBool() bool {
 	return bt.info(btfTypeKindFlagMask, btfTypeKindFlagShift) == 1
 }
 
+func (bt *btfType) setKindFlagBool(set bool) {
+	var value uint32
+	if set {
+		value = 1
+	}
+	bt.setInfo(value, btfTypeKindFlagMask, btfTypeKindFlagShift)
+}
+
+// Bitfield returns true if the struct or union contain a bitfield.
+func (bt *btfType) Bitfield() bool {
+	return bt.kindFlagBool()
+}
+
+func (bt *btfType) SetBitfield(isBitfield bool) {
+	bt.setKindFlagBool(isBitfield)
+}
+
+func (bt *btfType) FwdKind() FwdKind {
+	return FwdKind(bt.info(btfTypeKindFlagMask, btfTypeKindFlagShift))
+}
+
+func (bt *btfType) SetFwdKind(kind FwdKind) {
+	bt.setInfo(uint32(kind), btfTypeKindFlagMask, btfTypeKindFlagShift)
+}
+
+func (bt *btfType) Signed() bool {
+	return bt.kindFlagBool()
+}
+
+func (bt *btfType) SetSigned(signed bool) {
+	bt.setKindFlagBool(signed)
+}
+
 func (bt *btfType) Linkage() FuncLinkage {
 	return FuncLinkage(bt.info(btfTypeVlenMask, btfTypeVlenShift))
 }
@@ -181,6 +181,10 @@ func (bt *btfType) Type() TypeID {
 	return TypeID(bt.SizeType)
 }
 
+func (bt *btfType) SetType(id TypeID) {
+	bt.SizeType = uint32(id)
+}
+
 func (bt *btfType) Size() uint32 {
 	// TODO: Panic here if wrong kind?
 	return bt.SizeType
@@ -190,13 +194,22 @@ func (bt *btfType) SetSize(size uint32) {
 	bt.SizeType = size
 }
 
+func (bt *btfType) Marshal(w io.Writer, bo binary.ByteOrder) error {
+	buf := make([]byte, unsafe.Sizeof(*bt))
+	bo.PutUint32(buf[0:], bt.NameOff)
+	bo.PutUint32(buf[4:], bt.Info)
+	bo.PutUint32(buf[8:], bt.SizeType)
+	_, err := w.Write(buf)
+	return err
+}
+
 type rawType struct {
 	btfType
 	data interface{}
 }
 
 func (rt *rawType) Marshal(w io.Writer, bo binary.ByteOrder) error {
-	if err := binary.Write(w, bo, &rt.btfType); err != nil {
+	if err := rt.btfType.Marshal(w, bo); err != nil {
 		return err
 	}
 
@@ -275,7 +288,13 @@ type btfVariable struct {
 
 type btfEnum struct {
 	NameOff uint32
-	Val     int32
+	Val     uint32
+}
+
+type btfEnum64 struct {
+	NameOff uint32
+	ValLo32 uint32
+	ValHi32 uint32
 }
 
 type btfParam struct {
@@ -283,12 +302,16 @@ type btfParam struct {
 	Type    TypeID
 }
 
+type btfDeclTag struct {
+	ComponentIdx uint32
+}
+
 func readTypes(r io.Reader, bo binary.ByteOrder, typeLen uint32) ([]rawType, error) {
 	var header btfType
 	// because of the interleaving between types and struct members it is difficult to
 	// precompute the numbers of raw types this will parse
 	// this "guess" is a good first estimation
-	sizeOfbtfType := uintptr(binary.Size(btfType{}))
+	sizeOfbtfType := uintptr(btfTypeLen)
 	tyMaxCount := uintptr(typeLen) / sizeOfbtfType / 2
 	types := make([]rawType, 0, tyMaxCount)
 
@@ -325,6 +348,11 @@ func readTypes(r io.Reader, bo binary.ByteOrder, typeLen uint32) ([]rawType, err
 		case kindDatasec:
 			data = make([]btfVarSecinfo, header.Vlen())
 		case kindFloat:
+		case kindDeclTag:
+			data = new(btfDeclTag)
+		case kindTypeTag:
+		case kindEnum64:
+			data = make([]btfEnum64, header.Vlen())
 		default:
 			return nil, fmt.Errorf("type id %v: unknown kind: %v", id, header.Kind())
 		}

vendor/github.com/cilium/ebpf/btf/btf_types_string.go

@@ -1,4 +1,4 @@
-// Code generated by "stringer -linecomment -output=btf_types_string.go -type=FuncLinkage,VarLinkage"; DO NOT EDIT.
+// Code generated by "stringer -linecomment -output=btf_types_string.go -type=FuncLinkage,VarLinkage,btfKind"; DO NOT EDIT.
 
 package btf
 
@@ -42,3 +42,39 @@ func (i VarLinkage) String() string {
 	}
 	return _VarLinkage_name[_VarLinkage_index[i]:_VarLinkage_index[i+1]]
 }
+func _() {
+	// An "invalid array index" compiler error signifies that the constant values have changed.
+	// Re-run the stringer command to generate them again.
+	var x [1]struct{}
+	_ = x[kindUnknown-0]
+	_ = x[kindInt-1]
+	_ = x[kindPointer-2]
+	_ = x[kindArray-3]
+	_ = x[kindStruct-4]
+	_ = x[kindUnion-5]
+	_ = x[kindEnum-6]
+	_ = x[kindForward-7]
+	_ = x[kindTypedef-8]
+	_ = x[kindVolatile-9]
+	_ = x[kindConst-10]
+	_ = x[kindRestrict-11]
+	_ = x[kindFunc-12]
+	_ = x[kindFuncProto-13]
+	_ = x[kindVar-14]
+	_ = x[kindDatasec-15]
+	_ = x[kindFloat-16]
+	_ = x[kindDeclTag-17]
+	_ = x[kindTypeTag-18]
+	_ = x[kindEnum64-19]
+}
+
+const _btfKind_name = "UnknownIntPointerArrayStructUnionEnumForwardTypedefVolatileConstRestrictFuncFuncProtoVarDatasecFloatDeclTagTypeTagEnum64"
+
+var _btfKind_index = [...]uint8{0, 7, 10, 17, 22, 28, 33, 37, 44, 51, 59, 64, 72, 76, 85, 88, 95, 100, 107, 114, 120}
+
+func (i btfKind) String() string {
+	if i >= btfKind(len(_btfKind_index)-1) {
+		return "btfKind(" + strconv.FormatInt(int64(i), 10) + ")"
+	}
+	return _btfKind_name[_btfKind_index[i]:_btfKind_index[i+1]]
+}

vendor/github.com/cilium/ebpf/btf/core.go

@@ -156,16 +156,25 @@ func (k coreKind) String() string {
 	}
 }
 
-// CORERelocate calculates the difference in types between local and target.
+// CORERelocate calculates changes needed to adjust eBPF instructions for differences
+// in types.
 //
 // Returns a list of fixups which can be applied to instructions to make them
 // match the target type(s).
 //
 // Fixups are returned in the order of relos, e.g. fixup[i] is the solution
 // for relos[i].
-func CORERelocate(local, target *Spec, relos []*CORERelocation) ([]COREFixup, error) {
+func CORERelocate(relos []*CORERelocation, target *Spec, bo binary.ByteOrder) ([]COREFixup, error) {
-	if local.byteOrder != target.byteOrder {
+	if target == nil {
-		return nil, fmt.Errorf("can't relocate %s against %s", local.byteOrder, target.byteOrder)
+		var err error
+		target, _, err = kernelSpec()
+		if err != nil {
+			return nil, fmt.Errorf("load kernel spec: %w", err)
+		}
+	}
+
+	if bo != target.byteOrder {
+		return nil, fmt.Errorf("can't relocate %s against %s", bo, target.byteOrder)
 	}
 
 	type reloGroup struct {
@@ -185,15 +194,14 @@ func CORERelocate(local, target *Spec, relos []*CORERelocation) ([]COREFixup, er
 				return nil, fmt.Errorf("%s: unexpected accessor %v", relo.kind, relo.accessor)
 			}
 
-			id, err := local.TypeID(relo.typ)
-			if err != nil {
-				return nil, fmt.Errorf("%s: %w", relo.kind, err)
-			}
-
 			result[i] = COREFixup{
 				kind:  relo.kind,
-				local:  uint32(id),
+				local: uint32(relo.id),
-				target: uint32(id),
+				// NB: Using relo.id as the target here is incorrect, since
+				// it doesn't match the BTF we generate on the fly. This isn't
+				// too bad for now since there are no uses of the local type ID
+				// in the kernel, yet.
+				target: uint32(relo.id),
 			}
 			continue
 		}
@@ -214,7 +222,7 @@ func CORERelocate(local, target *Spec, relos []*CORERelocation) ([]COREFixup, er
 		}
 
 		targets := target.namedTypes[newEssentialName(localTypeName)]
-		fixups, err := coreCalculateFixups(local, target, localType, targets, group.relos)
+		fixups, err := coreCalculateFixups(group.relos, target, targets, bo)
 		if err != nil {
 			return nil, fmt.Errorf("relocate %s: %w", localType, err)
 		}
@@ -229,34 +237,29 @@ func CORERelocate(local, target *Spec, relos []*CORERelocation) ([]COREFixup, er
 
 var errAmbiguousRelocation = errors.New("ambiguous relocation")
 var errImpossibleRelocation = errors.New("impossible relocation")
+var errIncompatibleTypes = errors.New("incompatible types")
 
-// coreCalculateFixups calculates the fixups for the given relocations using
+// coreCalculateFixups finds the target type that best matches all relocations.
-// the "best" target.
+//
+// All relos must target the same type.
 //
 // The best target is determined by scoring: the less poisoning we have to do
 // the better the target is.
-func coreCalculateFixups(localSpec, targetSpec *Spec, local Type, targets []Type, relos []*CORERelocation) ([]COREFixup, error) {
+func coreCalculateFixups(relos []*CORERelocation, targetSpec *Spec, targets []Type, bo binary.ByteOrder) ([]COREFixup, error) {
-	localID, err := localSpec.TypeID(local)
-	if err != nil {
-		return nil, fmt.Errorf("local type ID: %w", err)
-	}
-	local = Copy(local, UnderlyingType)
-
 	bestScore := len(relos)
 	var bestFixups []COREFixup
-	for i := range targets {
+	for _, target := range targets {
-		targetID, err := targetSpec.TypeID(targets[i])
+		targetID, err := targetSpec.TypeID(target)
 		if err != nil {
 			return nil, fmt.Errorf("target type ID: %w", err)
 		}
-		target := Copy(targets[i], UnderlyingType)
 
 		score := 0 // lower is better
 		fixups := make([]COREFixup, 0, len(relos))
 		for _, relo := range relos {
-			fixup, err := coreCalculateFixup(localSpec.byteOrder, local, localID, target, targetID, relo)
+			fixup, err := coreCalculateFixup(relo, target, targetID, bo)
 			if err != nil {
-				return nil, fmt.Errorf("target %s: %w", target, err)
+				return nil, fmt.Errorf("target %s: %s: %w", target, relo.kind, err)
 			}
 			if fixup.poison || fixup.isNonExistant() {
 				score++
@@ -303,9 +306,11 @@ func coreCalculateFixups(localSpec, targetSpec *Spec, local Type, targets []Type
 	return bestFixups, nil
 }
 
+var errNoSignedness = errors.New("no signedness")
+
 // coreCalculateFixup calculates the fixup for a single local type, target type
 // and relocation.
-func coreCalculateFixup(byteOrder binary.ByteOrder, local Type, localID TypeID, target Type, targetID TypeID, relo *CORERelocation) (COREFixup, error) {
+func coreCalculateFixup(relo *CORERelocation, target Type, targetID TypeID, bo binary.ByteOrder) (COREFixup, error) {
 	fixup := func(local, target uint32) (COREFixup, error) {
 		return COREFixup{kind: relo.kind, local: local, target: target}, nil
 	}
@@ -320,18 +325,20 @@ func coreCalculateFixup(byteOrder binary.ByteOrder, local Type, localID TypeID,
 	}
 	zero := COREFixup{}
 
+	local := relo.typ
+
 	switch relo.kind {
 	case reloTypeIDTarget, reloTypeSize, reloTypeExists:
 		if len(relo.accessor) > 1 || relo.accessor[0] != 0 {
-			return zero, fmt.Errorf("%s: unexpected accessor %v", relo.kind, relo.accessor)
+			return zero, fmt.Errorf("unexpected accessor %v", relo.accessor)
 		}
 
 		err := coreAreTypesCompatible(local, target)
-		if errors.Is(err, errImpossibleRelocation) {
+		if errors.Is(err, errIncompatibleTypes) {
 			return poison()
 		}
 		if err != nil {
-			return zero, fmt.Errorf("relocation %s: %w", relo.kind, err)
+			return zero, err
 		}
 
 		switch relo.kind {
@@ -339,7 +346,7 @@ func coreCalculateFixup(byteOrder binary.ByteOrder, local Type, localID TypeID,
 			return fixup(1, 1)
 
 		case reloTypeIDTarget:
-			return fixup(uint32(localID), uint32(targetID))
+			return fixup(uint32(relo.id), uint32(targetID))
 
 		case reloTypeSize:
 			localSize, err := Sizeof(local)
@@ -361,7 +368,7 @@ func coreCalculateFixup(byteOrder binary.ByteOrder, local Type, localID TypeID,
 			return poison()
 		}
 		if err != nil {
-			return zero, fmt.Errorf("relocation %s: %w", relo.kind, err)
+			return zero, err
 		}
 
 		switch relo.kind {
@@ -372,21 +379,8 @@ func coreCalculateFixup(byteOrder binary.ByteOrder, local Type, localID TypeID,
 			return fixup(uint32(localValue.Value), uint32(targetValue.Value))
 		}
 
-	case reloFieldSigned:
+	case reloFieldByteOffset, reloFieldByteSize, reloFieldExists, reloFieldLShiftU64, reloFieldRShiftU64, reloFieldSigned:
-		switch local.(type) {
+		if _, ok := as[*Fwd](target); ok {
-		case *Enum:
-			return fixup(1, 1)
-		case *Int:
-			return fixup(
-				uint32(local.(*Int).Encoding&Signed),
-				uint32(target.(*Int).Encoding&Signed),
-			)
-		default:
-			return fixupWithoutValidation(0, 0)
-		}
-
-	case reloFieldByteOffset, reloFieldByteSize, reloFieldExists, reloFieldLShiftU64, reloFieldRShiftU64:
-		if _, ok := target.(*Fwd); ok {
 			// We can't relocate fields using a forward declaration, so
 			// skip it. If a non-forward declaration is present in the BTF
 			// we'll find it in one of the other iterations.
@@ -398,7 +392,7 @@ func coreCalculateFixup(byteOrder binary.ByteOrder, local Type, localID TypeID,
 			return poison()
 		}
 		if err != nil {
-			return zero, fmt.Errorf("target %s: %w", target, err)
+			return zero, err
 		}
 
 		maybeSkipValidation := func(f COREFixup, err error) (COREFixup, error) {
@@ -427,7 +421,7 @@ func coreCalculateFixup(byteOrder binary.ByteOrder, local Type, localID TypeID,
 
 		case reloFieldLShiftU64:
 			var target uint32
-			if byteOrder == binary.LittleEndian {
+			if bo == binary.LittleEndian {
 				targetSize, err := targetField.sizeBits()
 				if err != nil {
 					return zero, err
@@ -451,10 +445,40 @@ func coreCalculateFixup(byteOrder binary.ByteOrder, local Type, localID TypeID,
 			}
 
 			return fixupWithoutValidation(0, uint32(64-targetSize))
+
+		case reloFieldSigned:
+			switch local := UnderlyingType(localField.Type).(type) {
+			case *Enum:
+				target, ok := as[*Enum](targetField.Type)
+				if !ok {
+					return zero, fmt.Errorf("target isn't *Enum but %T", targetField.Type)
+				}
+
+				return fixup(boolToUint32(local.Signed), boolToUint32(target.Signed))
+			case *Int:
+				target, ok := as[*Int](targetField.Type)
+				if !ok {
+					return zero, fmt.Errorf("target isn't *Int but %T", targetField.Type)
+				}
+
+				return fixup(
+					uint32(local.Encoding&Signed),
+					uint32(target.Encoding&Signed),
+				)
+			default:
+				return zero, fmt.Errorf("type %T: %w", local, errNoSignedness)
+			}
 		}
 	}
 
-	return zero, fmt.Errorf("relocation %s: %w", relo.kind, ErrNotSupported)
+	return zero, ErrNotSupported
+}
+
+func boolToUint32(val bool) uint32 {
+	if val {
+		return 1
+	}
+	return 0
 }
 
 /* coreAccessor contains a path through a struct. It contains at least one index.
@@ -516,7 +540,7 @@ func (ca coreAccessor) String() string {
 }
 
 func (ca coreAccessor) enumValue(t Type) (*EnumValue, error) {
-	e, ok := t.(*Enum)
+	e, ok := as[*Enum](t)
 	if !ok {
 		return nil, fmt.Errorf("not an enum: %s", t)
 	}
@@ -555,6 +579,10 @@ type coreField struct {
 }
 
 func (cf *coreField) adjustOffsetToNthElement(n int) error {
+	if n == 0 {
+		return nil
+	}
+
 	size, err := Sizeof(cf.Type)
 	if err != nil {
 		return err
@@ -597,7 +625,7 @@ func (cf *coreField) sizeBits() (Bits, error) {
 	// between kernel versions. Synthesise the size to make the shifts work.
 	size, err := Sizeof(cf.Type)
 	if err != nil {
-		return 0, nil
+		return 0, err
 	}
 	return Bits(size * 8), nil
 }
@@ -611,6 +639,10 @@ func coreFindField(localT Type, localAcc coreAccessor, targetT Type) (coreField,
 	local := coreField{Type: localT}
 	target := coreField{Type: targetT}
 
+	if err := coreAreMembersCompatible(local.Type, target.Type); err != nil {
+		return coreField{}, coreField{}, fmt.Errorf("fields: %w", err)
+	}
+
 	// The first index is used to offset a pointer of the base type like
 	// when accessing an array.
 	if err := local.adjustOffsetToNthElement(localAcc[0]); err != nil {
@@ -621,13 +653,9 @@ func coreFindField(localT Type, localAcc coreAccessor, targetT Type) (coreField,
 		return coreField{}, coreField{}, err
 	}
 
-	if err := coreAreMembersCompatible(local.Type, target.Type); err != nil {
-		return coreField{}, coreField{}, fmt.Errorf("fields: %w", err)
-	}
-
 	var localMaybeFlex, targetMaybeFlex bool
 	for i, acc := range localAcc[1:] {
-		switch localType := local.Type.(type) {
+		switch localType := UnderlyingType(local.Type).(type) {
 		case composite:
 			// For composite types acc is used to find the field in the local type,
 			// and then we try to find a field in target with the same name.
@@ -638,21 +666,21 @@ func coreFindField(localT Type, localAcc coreAccessor, targetT Type) (coreField,
 
 			localMember := localMembers[acc]
 			if localMember.Name == "" {
-				_, ok := localMember.Type.(composite)
+				localMemberType, ok := as[composite](localMember.Type)
 				if !ok {
 					return coreField{}, coreField{}, fmt.Errorf("unnamed field with type %s: %s", localMember.Type, ErrNotSupported)
 				}
 
 				// This is an anonymous struct or union, ignore it.
 				local = coreField{
-					Type:   localMember.Type,
+					Type:   localMemberType,
 					offset: local.offset + localMember.Offset.Bytes(),
 				}
 				localMaybeFlex = false
 				continue
 			}
 
-			targetType, ok := target.Type.(composite)
+			targetType, ok := as[composite](target.Type)
 			if !ok {
 				return coreField{}, coreField{}, fmt.Errorf("target not composite: %w", errImpossibleRelocation)
 			}
@@ -698,7 +726,7 @@ func coreFindField(localT Type, localAcc coreAccessor, targetT Type) (coreField,
 
 		case *Array:
 			// For arrays, acc is the index in the target.
-			targetType, ok := target.Type.(*Array)
+			targetType, ok := as[*Array](target.Type)
 			if !ok {
 				return coreField{}, coreField{}, fmt.Errorf("target not array: %w", errImpossibleRelocation)
 			}
@@ -792,7 +820,7 @@ func coreFindMember(typ composite, name string) (Member, bool, error) {
 				continue
 			}
 
-			comp, ok := member.Type.(composite)
+			comp, ok := as[composite](member.Type)
 			if !ok {
 				return Member{}, false, fmt.Errorf("anonymous non-composite type %T not allowed", member.Type)
 			}
@@ -811,7 +839,7 @@ func coreFindEnumValue(local Type, localAcc coreAccessor, target Type) (localVal
 		return nil, nil, err
 	}
 
-	targetEnum, ok := target.(*Enum)
+	targetEnum, ok := as[*Enum](target)
 	if !ok {
 		return nil, nil, errImpossibleRelocation
 	}
@@ -828,6 +856,13 @@ func coreFindEnumValue(local Type, localAcc coreAccessor, target Type) (localVal
 	return nil, nil, errImpossibleRelocation
 }
 
+// CheckTypeCompatibility checks local and target types for Compatibility according to CO-RE rules.
+//
+// Only layout compatibility is checked, ignoring names of the root type.
+func CheckTypeCompatibility(localType Type, targetType Type) error {
+	return coreAreTypesCompatible(localType, targetType)
+}
+
 /* The comment below is from bpf_core_types_are_compat in libbpf.c:
  *
  * Check local and target types for compatibility. This check is used for
@@ -849,25 +884,26 @@ func coreFindEnumValue(local Type, localAcc coreAccessor, target Type) (localVal
  * These rules are not set in stone and probably will be adjusted as we get
  * more experience with using BPF CO-RE relocations.
  *
- * Returns errImpossibleRelocation if types are not compatible.
+ * Returns errIncompatibleTypes if types are not compatible.
  */
 func coreAreTypesCompatible(localType Type, targetType Type) error {
+
 	var (
 		localTs, targetTs typeDeque
 		l, t              = &localType, &targetType
 		depth             = 0
 	)
 
-	for ; l != nil && t != nil; l, t = localTs.shift(), targetTs.shift() {
+	for ; l != nil && t != nil; l, t = localTs.Shift(), targetTs.Shift() {
 		if depth >= maxTypeDepth {
 			return errors.New("types are nested too deep")
 		}
 
-		localType = *l
+		localType = UnderlyingType(*l)
-		targetType = *t
+		targetType = UnderlyingType(*t)
 
 		if reflect.TypeOf(localType) != reflect.TypeOf(targetType) {
-			return fmt.Errorf("type mismatch: %w", errImpossibleRelocation)
+			return fmt.Errorf("type mismatch: %w", errIncompatibleTypes)
 		}
 
 		switch lv := (localType).(type) {
@@ -876,18 +912,18 @@ func coreAreTypesCompatible(localType Type, targetType Type) error {
 
 		case *Pointer, *Array:
 			depth++
-			localType.walk(&localTs)
+			walkType(localType, localTs.Push)
-			targetType.walk(&targetTs)
+			walkType(targetType, targetTs.Push)
 
 		case *FuncProto:
 			tv := targetType.(*FuncProto)
 			if len(lv.Params) != len(tv.Params) {
-				return fmt.Errorf("function param mismatch: %w", errImpossibleRelocation)
+				return fmt.Errorf("function param mismatch: %w", errIncompatibleTypes)
 			}
 
 			depth++
-			localType.walk(&localTs)
+			walkType(localType, localTs.Push)
-			targetType.walk(&targetTs)
+			walkType(targetType, targetTs.Push)
 
 		default:
 			return fmt.Errorf("unsupported type %T", localType)
@@ -931,6 +967,9 @@ func coreAreTypesCompatible(localType Type, targetType Type) error {
  * Returns errImpossibleRelocation if the members are not compatible.
  */
 func coreAreMembersCompatible(localType Type, targetType Type) error {
+	localType = UnderlyingType(localType)
+	targetType = UnderlyingType(targetType)
+
 	doNamesMatch := func(a, b string) error {
 		if a == "" || b == "" {
 			// allow anonymous and named type to match

vendor/github.com/cilium/ebpf/btf/ext_info.go

@@ -24,7 +24,7 @@ type ExtInfos struct {
 // loadExtInfosFromELF parses ext infos from the .BTF.ext section in an ELF.
 //
 // Returns an error wrapping ErrNotFound if no ext infos are present.
-func loadExtInfosFromELF(file *internal.SafeELFFile, ts types, strings *stringTable) (*ExtInfos, error) {
+func loadExtInfosFromELF(file *internal.SafeELFFile, spec *Spec) (*ExtInfos, error) {
 	section := file.Section(".BTF.ext")
 	if section == nil {
 		return nil, fmt.Errorf("btf ext infos: %w", ErrNotFound)
@@ -34,11 +34,11 @@ func loadExtInfosFromELF(file *internal.SafeELFFile, ts types, strings *stringTa
 		return nil, fmt.Errorf("compressed ext_info is not supported")
 	}
 
-	return loadExtInfos(section.ReaderAt, file.ByteOrder, ts, strings)
+	return loadExtInfos(section.ReaderAt, file.ByteOrder, spec, spec.strings)
 }
 
 // loadExtInfos parses bare ext infos.
-func loadExtInfos(r io.ReaderAt, bo binary.ByteOrder, ts types, strings *stringTable) (*ExtInfos, error) {
+func loadExtInfos(r io.ReaderAt, bo binary.ByteOrder, spec *Spec, strings *stringTable) (*ExtInfos, error) {
 	// Open unbuffered section reader. binary.Read() calls io.ReadFull on
 	// the header structs, resulting in one syscall per header.
 	headerRd := io.NewSectionReader(r, 0, math.MaxInt64)
@@ -60,7 +60,7 @@ func loadExtInfos(r io.ReaderAt, bo binary.ByteOrder, ts types, strings *stringT
 
 	funcInfos := make(map[string][]funcInfo, len(btfFuncInfos))
 	for section, bfis := range btfFuncInfos {
-		funcInfos[section], err = newFuncInfos(bfis, ts)
+		funcInfos[section], err = newFuncInfos(bfis, spec)
 		if err != nil {
 			return nil, fmt.Errorf("section %s: func infos: %w", section, err)
 		}
@@ -93,7 +93,7 @@ func loadExtInfos(r io.ReaderAt, bo binary.ByteOrder, ts types, strings *stringT
 
 	coreRelos := make(map[string][]coreRelocationInfo, len(btfCORERelos))
 	for section, brs := range btfCORERelos {
-		coreRelos[section], err = newRelocationInfos(brs, ts, strings)
+		coreRelos[section], err = newRelocationInfos(brs, spec, strings)
 		if err != nil {
 			return nil, fmt.Errorf("section %s: CO-RE relocations: %w", section, err)
 		}
@@ -114,7 +114,7 @@ func (ei *ExtInfos) Assign(insns asm.Instructions, section string) {
 	iter := insns.Iterate()
 	for iter.Next() {
 		if len(funcInfos) > 0 && funcInfos[0].offset == iter.Offset {
-			iter.Ins.Metadata.Set(funcInfoMeta{}, funcInfos[0].fn)
+			*iter.Ins = WithFuncMetadata(*iter.Ins, funcInfos[0].fn)
 			funcInfos = funcInfos[1:]
 		}
 
@@ -132,17 +132,37 @@ func (ei *ExtInfos) Assign(insns asm.Instructions, section string) {
 
 // MarshalExtInfos encodes function and line info embedded in insns into kernel
 // wire format.
-func MarshalExtInfos(insns asm.Instructions, typeID func(Type) (TypeID, error)) (funcInfos, lineInfos []byte, _ error) {
+//
+// Returns ErrNotSupported if the kernel doesn't support BTF-associated programs.
+func MarshalExtInfos(insns asm.Instructions) (_ *Handle, funcInfos, lineInfos []byte, _ error) {
+	// Bail out early if the kernel doesn't support Func(Proto). If this is the
+	// case, func_info will also be unsupported.
+	if err := haveProgBTF(); err != nil {
+		return nil, nil, nil, err
+	}
+
 	iter := insns.Iterate()
-	var fiBuf, liBuf bytes.Buffer
 	for iter.Next() {
+		_, ok := iter.Ins.Source().(*Line)
+		fn := FuncMetadata(iter.Ins)
+		if ok || fn != nil {
+			goto marshal
+		}
+	}
+
+	return nil, nil, nil, nil
+
+marshal:
+	var b Builder
+	var fiBuf, liBuf bytes.Buffer
+	for {
 		if fn := FuncMetadata(iter.Ins); fn != nil {
 			fi := &funcInfo{
 				fn:     fn,
 				offset: iter.Offset,
 			}
-			if err := fi.marshal(&fiBuf, typeID); err != nil {
+			if err := fi.marshal(&fiBuf, &b); err != nil {
-				return nil, nil, fmt.Errorf("write func info: %w", err)
+				return nil, nil, nil, fmt.Errorf("write func info: %w", err)
 			}
 		}
 
@@ -151,12 +171,18 @@ func MarshalExtInfos(insns asm.Instructions, typeID func(Type) (TypeID, error))
 				line:   line,
 				offset: iter.Offset,
 			}
-			if err := li.marshal(&liBuf); err != nil {
+			if err := li.marshal(&liBuf, &b); err != nil {
-				return nil, nil, fmt.Errorf("write line info: %w", err)
+				return nil, nil, nil, fmt.Errorf("write line info: %w", err)
 			}
 		}
+
+		if !iter.Next() {
+			break
 		}
-	return fiBuf.Bytes(), liBuf.Bytes(), nil
+	}
+
+	handle, err := NewHandle(&b)
+	return handle, fiBuf.Bytes(), liBuf.Bytes(), err
 }
 
 // btfExtHeader is found at the start of the .BTF.ext section.
@@ -311,8 +337,8 @@ type bpfFuncInfo struct {
 	TypeID  TypeID
 }
 
-func newFuncInfo(fi bpfFuncInfo, ts types) (*funcInfo, error) {
+func newFuncInfo(fi bpfFuncInfo, spec *Spec) (*funcInfo, error) {
-	typ, err := ts.ByID(fi.TypeID)
+	typ, err := spec.TypeByID(fi.TypeID)
 	if err != nil {
 		return nil, err
 	}
@@ -333,10 +359,10 @@ func newFuncInfo(fi bpfFuncInfo, ts types) (*funcInfo, error) {
 	}, nil
 }
 
-func newFuncInfos(bfis []bpfFuncInfo, ts types) ([]funcInfo, error) {
+func newFuncInfos(bfis []bpfFuncInfo, spec *Spec) ([]funcInfo, error) {
 	fis := make([]funcInfo, 0, len(bfis))
 	for _, bfi := range bfis {
-		fi, err := newFuncInfo(bfi, ts)
+		fi, err := newFuncInfo(bfi, spec)
 		if err != nil {
 			return nil, fmt.Errorf("offset %d: %w", bfi.InsnOff, err)
 		}
@@ -349,8 +375,8 @@ func newFuncInfos(bfis []bpfFuncInfo, ts types) ([]funcInfo, error) {
 }
 
 // marshal into the BTF wire format.
-func (fi *funcInfo) marshal(w io.Writer, typeID func(Type) (TypeID, error)) error {
+func (fi *funcInfo) marshal(w *bytes.Buffer, b *Builder) error {
-	id, err := typeID(fi.fn)
+	id, err := b.Add(fi.fn)
 	if err != nil {
 		return err
 	}
@@ -358,10 +384,14 @@ func (fi *funcInfo) marshal(w io.Writer, typeID func(Type) (TypeID, error)) erro
 		InsnOff: uint32(fi.offset),
 		TypeID:  id,
 	}
-	return binary.Write(w, internal.NativeEndian, &bfi)
+	buf := make([]byte, FuncInfoSize)
+	internal.NativeEndian.PutUint32(buf, bfi.InsnOff)
+	internal.NativeEndian.PutUint32(buf[4:], uint32(bfi.TypeID))
+	_, err = w.Write(buf)
+	return err
 }
 
-// parseLineInfos parses a func_info sub-section within .BTF.ext ito a map of
+// parseFuncInfos parses a func_info sub-section within .BTF.ext ito a map of
 // func infos indexed by section name.
 func parseFuncInfos(r io.Reader, bo binary.ByteOrder, strings *stringTable) (map[string][]bpfFuncInfo, error) {
 	recordSize, err := parseExtInfoRecordSize(r, bo)
@@ -428,12 +458,6 @@ type Line struct {
 	line       string
 	lineNumber uint32
 	lineColumn uint32
-
-	// TODO: We should get rid of the fields below, but for that we need to be
-	// able to write BTF.
-
-	fileNameOff uint32
-	lineOff     uint32
 }
 
 func (li *Line) FileName() string {
@@ -496,8 +520,6 @@ func newLineInfo(li bpfLineInfo, strings *stringTable) (*lineInfo, error) {
 			line,
 			lineNumber,
 			lineColumn,
-			li.FileNameOff,
-			li.LineOff,
 		},
 		asm.RawInstructionOffset(li.InsnOff),
 	}, nil
@@ -519,7 +541,7 @@ func newLineInfos(blis []bpfLineInfo, strings *stringTable) ([]lineInfo, error)
 }
 
 // marshal writes the binary representation of the LineInfo to w.
-func (li *lineInfo) marshal(w io.Writer) error {
+func (li *lineInfo) marshal(w *bytes.Buffer, b *Builder) error {
 	line := li.line
 	if line.lineNumber > bpfLineMax {
 		return fmt.Errorf("line %d exceeds %d", line.lineNumber, bpfLineMax)
@@ -529,13 +551,30 @@ func (li *lineInfo) marshal(w io.Writer) error {
 		return fmt.Errorf("column %d exceeds %d", line.lineColumn, bpfColumnMax)
 	}
 
+	fileNameOff, err := b.addString(line.fileName)
+	if err != nil {
+		return fmt.Errorf("file name %q: %w", line.fileName, err)
+	}
+
+	lineOff, err := b.addString(line.line)
+	if err != nil {
+		return fmt.Errorf("line %q: %w", line.line, err)
+	}
+
 	bli := bpfLineInfo{
 		uint32(li.offset),
-		line.fileNameOff,
+		fileNameOff,
-		line.lineOff,
+		lineOff,
 		(line.lineNumber << bpfLineShift) | line.lineColumn,
 	}
-	return binary.Write(w, internal.NativeEndian, &bli)
+
+	buf := make([]byte, LineInfoSize)
+	internal.NativeEndian.PutUint32(buf, bli.InsnOff)
+	internal.NativeEndian.PutUint32(buf[4:], bli.FileNameOff)
+	internal.NativeEndian.PutUint32(buf[8:], bli.LineOff)
+	internal.NativeEndian.PutUint32(buf[12:], bli.LineCol)
+	_, err = w.Write(buf)
+	return err
 }
 
 // parseLineInfos parses a line_info sub-section within .BTF.ext ito a map of
@@ -605,9 +644,16 @@ type bpfCORERelo struct {
 }
 
 type CORERelocation struct {
+	// The local type of the relocation, stripped of typedefs and qualifiers.
 	typ      Type
 	accessor coreAccessor
 	kind     coreKind
+	// The ID of the local type in the source BTF.
+	id TypeID
+}
+
+func (cr *CORERelocation) String() string {
+	return fmt.Sprintf("CORERelocation(%s, %s[%s], local_id=%d)", cr.kind, cr.typ, cr.accessor, cr.id)
 }
 
 func CORERelocationMetadata(ins *asm.Instruction) *CORERelocation {
@@ -620,8 +666,8 @@ type coreRelocationInfo struct {
 	offset asm.RawInstructionOffset
 }
 
-func newRelocationInfo(relo bpfCORERelo, ts types, strings *stringTable) (*coreRelocationInfo, error) {
+func newRelocationInfo(relo bpfCORERelo, spec *Spec, strings *stringTable) (*coreRelocationInfo, error) {
-	typ, err := ts.ByID(relo.TypeID)
+	typ, err := spec.TypeByID(relo.TypeID)
 	if err != nil {
 		return nil, err
 	}
@@ -641,15 +687,16 @@ func newRelocationInfo(relo bpfCORERelo, ts types, strings *stringTable) (*coreR
 			typ,
 			accessor,
 			relo.Kind,
+			relo.TypeID,
 		},
 		asm.RawInstructionOffset(relo.InsnOff),
 	}, nil
 }
 
-func newRelocationInfos(brs []bpfCORERelo, ts types, strings *stringTable) ([]coreRelocationInfo, error) {
+func newRelocationInfos(brs []bpfCORERelo, spec *Spec, strings *stringTable) ([]coreRelocationInfo, error) {
 	rs := make([]coreRelocationInfo, 0, len(brs))
 	for _, br := range brs {
-		relo, err := newRelocationInfo(br, ts, strings)
+		relo, err := newRelocationInfo(br, spec, strings)
 		if err != nil {
 			return nil, fmt.Errorf("offset %d: %w", br.InsnOff, err)
 		}

vendor/github.com/cilium/ebpf/btf/format.go

@@ -63,39 +63,25 @@ func (gf *GoFormatter) writeTypeDecl(name string, typ Type) error {
 		return fmt.Errorf("need a name for type %s", typ)
 	}
 
-	switch v := skipQualifiers(typ).(type) {
+	typ = skipQualifiers(typ)
-	case *Enum:
 	fmt.Fprintf(&gf.w, "type %s ", name)
-		switch v.Size {
+	if err := gf.writeTypeLit(typ, 0); err != nil {
-		case 1:
+		return err
-			gf.w.WriteString("int8")
-		case 2:
-			gf.w.WriteString("int16")
-		case 4:
-			gf.w.WriteString("int32")
-		case 8:
-			gf.w.WriteString("int64")
-		default:
-			return fmt.Errorf("%s: invalid enum size %d", typ, v.Size)
 	}
 
-		if len(v.Values) == 0 {
+	e, ok := typ.(*Enum)
+	if !ok || len(e.Values) == 0 {
 		return nil
 	}
 
 	gf.w.WriteString("; const ( ")
-		for _, ev := range v.Values {
+	for _, ev := range e.Values {
 		id := gf.enumIdentifier(name, ev.Name)
 		fmt.Fprintf(&gf.w, "%s %s = %d; ", id, name, ev.Value)
 	}
 	gf.w.WriteString(")")
 
 	return nil
-
-	default:
-		fmt.Fprintf(&gf.w, "type %s ", name)
-		return gf.writeTypeLit(v, 0)
-	}
 }
 
 // writeType outputs the name of a named type or a literal describing the type.
@@ -133,10 +119,24 @@ func (gf *GoFormatter) writeTypeLit(typ Type, depth int) error {
 	var err error
 	switch v := skipQualifiers(typ).(type) {
 	case *Int:
-		gf.writeIntLit(v)
+		err = gf.writeIntLit(v)
 
 	case *Enum:
+		if !v.Signed {
+			gf.w.WriteRune('u')
+		}
+		switch v.Size {
+		case 1:
+			gf.w.WriteString("int8")
+		case 2:
+			gf.w.WriteString("int16")
+		case 4:
 			gf.w.WriteString("int32")
+		case 8:
+			gf.w.WriteString("int64")
+		default:
+			err = fmt.Errorf("invalid enum size %d", v.Size)
+		}
 
 	case *Typedef:
 		err = gf.writeType(v.Type, depth)
@@ -166,19 +166,36 @@ func (gf *GoFormatter) writeTypeLit(typ Type, depth int) error {
 	return nil
 }
 
-func (gf *GoFormatter) writeIntLit(i *Int) {
+func (gf *GoFormatter) writeIntLit(i *Int) error {
-	// NB: Encoding.IsChar is ignored.
-	if i.Encoding.IsBool() && i.Size == 1 {
-		gf.w.WriteString("bool")
-		return
-	}
-
 	bits := i.Size * 8
-	if i.Encoding.IsSigned() {
+	switch i.Encoding {
-		fmt.Fprintf(&gf.w, "int%d", bits)
+	case Bool:
-	} else {
+		if i.Size != 1 {
-		fmt.Fprintf(&gf.w, "uint%d", bits)
+			return fmt.Errorf("bool with size %d", i.Size)
 		}
+		gf.w.WriteString("bool")
+	case Char:
+		if i.Size != 1 {
+			return fmt.Errorf("char with size %d", i.Size)
+		}
+		// BTF doesn't have a way to specify the signedness of a char. Assume
+		// we are dealing with unsigned, since this works nicely with []byte
+		// in Go code.
+		fallthrough
+	case Unsigned, Signed:
+		stem := "uint"
+		if i.Encoding == Signed {
+			stem = "int"
+		}
+		if i.Size > 8 {
+			fmt.Fprintf(&gf.w, "[%d]byte /* %s%d */", i.Size, stem, i.Size*8)
+		} else {
+			fmt.Fprintf(&gf.w, "%s%d", stem, bits)
+		}
+	default:
+		return fmt.Errorf("can't encode %s", i.Encoding)
+	}
+	return nil
 }
 
 func (gf *GoFormatter) writeStructLit(size uint32, members []Member, depth int) error {
@@ -199,11 +216,15 @@ func (gf *GoFormatter) writeStructLit(size uint32, members []Member, depth int)
 			gf.writePadding(n)
 		}
 
-		size, err := Sizeof(m.Type)
+		fieldSize, err := Sizeof(m.Type)
 		if err != nil {
 			return fmt.Errorf("field %d: %w", i, err)
 		}
-		prevOffset = offset + uint32(size)
+
+		prevOffset = offset + uint32(fieldSize)
+		if prevOffset > size {
+			return fmt.Errorf("field %d of size %d exceeds type size %d", i, fieldSize, size)
+		}
 
 		if err := gf.writeStructField(m, depth); err != nil {
 			return fmt.Errorf("field %d: %w", i, err)
@@ -272,7 +293,11 @@ func (gf *GoFormatter) writeDatasecLit(ds *Datasec, depth int) error {
 
 	prevOffset := uint32(0)
 	for i, vsi := range ds.Vars {
-		v := vsi.Type.(*Var)
+		v, ok := vsi.Type.(*Var)
+		if !ok {
+			return fmt.Errorf("can't format %s as part of data section", vsi.Type)
+		}
+
 		if v.Linkage != GlobalVar {
 			// Ignore static, extern, etc. for now.
 			continue

vendor/github.com/cilium/ebpf/btf/handle.go

@@ -1,14 +1,142 @@
 package btf
 
 import (
+	"bytes"
 	"errors"
 	"fmt"
+	"math"
 	"os"
 
+	"github.com/cilium/ebpf/internal"
 	"github.com/cilium/ebpf/internal/sys"
 	"github.com/cilium/ebpf/internal/unix"
 )
 
+// Handle is a reference to BTF loaded into the kernel.
+type Handle struct {
+	fd *sys.FD
+
+	// Size of the raw BTF in bytes.
+	size uint32
+
+	needsKernelBase bool
+}
+
+// NewHandle loads the contents of a [Builder] into the kernel.
+//
+// Returns an error wrapping ErrNotSupported if the kernel doesn't support BTF.
+func NewHandle(b *Builder) (*Handle, error) {
+	small := getByteSlice()
+	defer putByteSlice(small)
+
+	buf, err := b.Marshal(*small, KernelMarshalOptions())
+	if err != nil {
+		return nil, fmt.Errorf("marshal BTF: %w", err)
+	}
+
+	return NewHandleFromRawBTF(buf)
+}
+
+// NewHandleFromRawBTF loads raw BTF into the kernel.
+//
+// Returns an error wrapping ErrNotSupported if the kernel doesn't support BTF.
+func NewHandleFromRawBTF(btf []byte) (*Handle, error) {
+	if uint64(len(btf)) > math.MaxUint32 {
+		return nil, errors.New("BTF exceeds the maximum size")
+	}
+
+	attr := &sys.BtfLoadAttr{
+		Btf:     sys.NewSlicePointer(btf),
+		BtfSize: uint32(len(btf)),
+	}
+
+	fd, err := sys.BtfLoad(attr)
+	if err == nil {
+		return &Handle{fd, attr.BtfSize, false}, nil
+	}
+
+	if err := haveBTF(); err != nil {
+		return nil, err
+	}
+
+	logBuf := make([]byte, 64*1024)
+	attr.BtfLogBuf = sys.NewSlicePointer(logBuf)
+	attr.BtfLogSize = uint32(len(logBuf))
+	attr.BtfLogLevel = 1
+
+	// Up until at least kernel 6.0, the BTF verifier does not return ENOSPC
+	// if there are other verification errors. ENOSPC is only returned when
+	// the BTF blob is correct, a log was requested, and the provided buffer
+	// is too small.
+	_, ve := sys.BtfLoad(attr)
+	return nil, internal.ErrorWithLog("load btf", err, logBuf, errors.Is(ve, unix.ENOSPC))
+}
+
+// NewHandleFromID returns the BTF handle for a given id.
+//
+// Prefer calling [ebpf.Program.Handle] or [ebpf.Map.Handle] if possible.
+//
+// Returns ErrNotExist, if there is no BTF with the given id.
+//
+// Requires CAP_SYS_ADMIN.
+func NewHandleFromID(id ID) (*Handle, error) {
+	fd, err := sys.BtfGetFdById(&sys.BtfGetFdByIdAttr{
+		Id: uint32(id),
+	})
+	if err != nil {
+		return nil, fmt.Errorf("get FD for ID %d: %w", id, err)
+	}
+
+	info, err := newHandleInfoFromFD(fd)
+	if err != nil {
+		_ = fd.Close()
+		return nil, err
+	}
+
+	return &Handle{fd, info.size, info.IsModule()}, nil
+}
+
+// Spec parses the kernel BTF into Go types.
+//
+// base must contain type information for vmlinux if the handle is for
+// a kernel module. It may be nil otherwise.
+func (h *Handle) Spec(base *Spec) (*Spec, error) {
+	var btfInfo sys.BtfInfo
+	btfBuffer := make([]byte, h.size)
+	btfInfo.Btf, btfInfo.BtfSize = sys.NewSlicePointerLen(btfBuffer)
+
+	if err := sys.ObjInfo(h.fd, &btfInfo); err != nil {
+		return nil, err
+	}
+
+	if h.needsKernelBase && base == nil {
+		return nil, fmt.Errorf("missing base types")
+	}
+
+	return loadRawSpec(bytes.NewReader(btfBuffer), internal.NativeEndian, base)
+}
+
+// Close destroys the handle.
+//
+// Subsequent calls to FD will return an invalid value.
+func (h *Handle) Close() error {
+	if h == nil {
+		return nil
+	}
+
+	return h.fd.Close()
+}
+
+// FD returns the file descriptor for the handle.
+func (h *Handle) FD() int {
+	return h.fd.Int()
+}
+
+// Info returns metadata about the handle.
+func (h *Handle) Info() (*HandleInfo, error) {
+	return newHandleInfoFromFD(h.fd)
+}
+
 // HandleInfo describes a Handle.
 type HandleInfo struct {
 	// ID of this handle in the kernel. The ID is only valid as long as the
@@ -59,7 +187,7 @@ func newHandleInfoFromFD(fd *sys.FD) (*HandleInfo, error) {
 	}, nil
 }
 
-// IsModule returns true if the BTF is for the kernel itself.
+// IsVmlinux returns true if the BTF is for the kernel itself.
 func (i *HandleInfo) IsVmlinux() bool {
 	return i.IsKernel && i.Name == "vmlinux"
 }
@@ -71,51 +199,89 @@ func (i *HandleInfo) IsModule() bool {
 
 // HandleIterator allows enumerating BTF blobs loaded into the kernel.
 type HandleIterator struct {
-	// The ID of the last retrieved handle. Only valid after a call to Next.
+	// The ID of the current handle. Only valid after a call to Next.
 	ID ID
+	// The current Handle. Only valid until a call to Next.
+	// See Take if you want to retain the handle.
+	Handle *Handle
 	err    error
 }
 
-// Next retrieves a handle for the next BTF blob.
+// Next retrieves a handle for the next BTF object.
 //
-// [Handle.Close] is called if *handle is non-nil to avoid leaking fds.
+// Returns true if another BTF object was found. Call [HandleIterator.Err] after
-//
-// Returns true if another BTF blob was found. Call [HandleIterator.Err] after
 // the function returns false.
-func (it *HandleIterator) Next(handle **Handle) bool {
+func (it *HandleIterator) Next() bool {
-	if *handle != nil {
-		(*handle).Close()
-		*handle = nil
-	}
-
 	id := it.ID
 	for {
 		attr := &sys.BtfGetNextIdAttr{Id: id}
 		err := sys.BtfGetNextId(attr)
 		if errors.Is(err, os.ErrNotExist) {
 			// There are no more BTF objects.
-			return false
+			break
 		} else if err != nil {
 			it.err = fmt.Errorf("get next BTF ID: %w", err)
-			return false
+			break
 		}
 
 		id = attr.NextId
-		*handle, err = NewHandleFromID(id)
+		handle, err := NewHandleFromID(id)
 		if errors.Is(err, os.ErrNotExist) {
 			// Try again with the next ID.
 			continue
 		} else if err != nil {
 			it.err = fmt.Errorf("retrieve handle for ID %d: %w", id, err)
-			return false
+			break
 		}
 
-		it.ID = id
+		it.Handle.Close()
+		it.ID, it.Handle = id, handle
 		return true
 	}
+
+	// No more handles or we encountered an error.
+	it.Handle.Close()
+	it.Handle = nil
+	return false
+}
+
+// Take the ownership of the current handle.
+//
+// It's the callers responsibility to close the handle.
+func (it *HandleIterator) Take() *Handle {
+	handle := it.Handle
+	it.Handle = nil
+	return handle
 }
 
 // Err returns an error if iteration failed for some reason.
 func (it *HandleIterator) Err() error {
 	return it.err
 }
+
+// FindHandle returns the first handle for which predicate returns true.
+//
+// Requires CAP_SYS_ADMIN.
+//
+// Returns an error wrapping ErrNotFound if predicate never returns true or if
+// there is no BTF loaded into the kernel.
+func FindHandle(predicate func(info *HandleInfo) bool) (*Handle, error) {
+	it := new(HandleIterator)
+	defer it.Handle.Close()
+
+	for it.Next() {
+		info, err := it.Handle.Info()
+		if err != nil {
+			return nil, fmt.Errorf("info for ID %d: %w", it.ID, err)
+		}
+
+		if predicate(info) {
+			return it.Take(), nil
+		}
+	}
+	if err := it.Err(); err != nil {
+		return nil, fmt.Errorf("iterate handles: %w", err)
+	}
+
+	return nil, fmt.Errorf("find handle: %w", ErrNotFound)
+}

vendor/github.com/cilium/ebpf/btf/marshal.go

@@ -0,0 +1,543 @@
+package btf
+
+import (
+	"bytes"
+	"encoding/binary"
+	"errors"
+	"fmt"
+	"math"
+	"sync"
+
+	"github.com/cilium/ebpf/internal"
+
+	"golang.org/x/exp/slices"
+)
+
+type MarshalOptions struct {
+	// Target byte order. Defaults to the system's native endianness.
+	Order binary.ByteOrder
+	// Remove function linkage information for compatibility with <5.6 kernels.
+	StripFuncLinkage bool
+}
+
+// KernelMarshalOptions will generate BTF suitable for the current kernel.
+func KernelMarshalOptions() *MarshalOptions {
+	return &MarshalOptions{
+		Order:            internal.NativeEndian,
+		StripFuncLinkage: haveFuncLinkage() != nil,
+	}
+}
+
+// encoder turns Types into raw BTF.
+type encoder struct {
+	MarshalOptions
+
+	pending internal.Deque[Type]
+	buf     *bytes.Buffer
+	strings *stringTableBuilder
+	ids     map[Type]TypeID
+	lastID  TypeID
+}
+
+var bufferPool = sync.Pool{
+	New: func() any {
+		buf := make([]byte, btfHeaderLen+128)
+		return &buf
+	},
+}
+
+func getByteSlice() *[]byte {
+	return bufferPool.Get().(*[]byte)
+}
+
+func putByteSlice(buf *[]byte) {
+	*buf = (*buf)[:0]
+	bufferPool.Put(buf)
+}
+
+// Builder turns Types into raw BTF.
+//
+// The default value may be used and represents an empty BTF blob. Void is
+// added implicitly if necessary.
+type Builder struct {
+	// Explicitly added types.
+	types []Type
+	// IDs for all added types which the user knows about.
+	stableIDs map[Type]TypeID
+	// Explicitly added strings.
+	strings *stringTableBuilder
+}
+
+// NewBuilder creates a Builder from a list of types.
+//
+// It is more efficient than calling [Add] individually.
+//
+// Returns an error if adding any of the types fails.
+func NewBuilder(types []Type) (*Builder, error) {
+	b := &Builder{
+		make([]Type, 0, len(types)),
+		make(map[Type]TypeID, len(types)),
+		nil,
+	}
+
+	for _, typ := range types {
+		_, err := b.Add(typ)
+		if err != nil {
+			return nil, fmt.Errorf("add %s: %w", typ, err)
+		}
+	}
+
+	return b, nil
+}
+
+// Add a Type and allocate a stable ID for it.
+//
+// Adding the identical Type multiple times is valid and will return the same ID.
+//
+// See [Type] for details on identity.
+func (b *Builder) Add(typ Type) (TypeID, error) {
+	if b.stableIDs == nil {
+		b.stableIDs = make(map[Type]TypeID)
+	}
+
+	if _, ok := typ.(*Void); ok {
+		// Equality is weird for void, since it is a zero sized type.
+		return 0, nil
+	}
+
+	if ds, ok := typ.(*Datasec); ok {
+		if err := datasecResolveWorkaround(b, ds); err != nil {
+			return 0, err
+		}
+	}
+
+	id, ok := b.stableIDs[typ]
+	if ok {
+		return id, nil
+	}
+
+	b.types = append(b.types, typ)
+
+	id = TypeID(len(b.types))
+	if int(id) != len(b.types) {
+		return 0, fmt.Errorf("no more type IDs")
+	}
+
+	b.stableIDs[typ] = id
+	return id, nil
+}
+
+// Marshal encodes all types in the Marshaler into BTF wire format.
+//
+// opts may be nil.
+func (b *Builder) Marshal(buf []byte, opts *MarshalOptions) ([]byte, error) {
+	stb := b.strings
+	if stb == nil {
+		// Assume that most types are named. This makes encoding large BTF like
+		// vmlinux a lot cheaper.
+		stb = newStringTableBuilder(len(b.types))
+	} else {
+		// Avoid modifying the Builder's string table.
+		stb = b.strings.Copy()
+	}
+
+	if opts == nil {
+		opts = &MarshalOptions{Order: internal.NativeEndian}
+	}
+
+	// Reserve space for the BTF header.
+	buf = slices.Grow(buf, btfHeaderLen)[:btfHeaderLen]
+
+	w := internal.NewBuffer(buf)
+	defer internal.PutBuffer(w)
+
+	e := encoder{
+		MarshalOptions: *opts,
+		buf:            w,
+		strings:        stb,
+		lastID:         TypeID(len(b.types)),
+		ids:            make(map[Type]TypeID, len(b.types)),
+	}
+
+	// Ensure that types are marshaled in the exact order they were Add()ed.
+	// Otherwise the ID returned from Add() won't match.
+	e.pending.Grow(len(b.types))
+	for _, typ := range b.types {
+		e.pending.Push(typ)
+		e.ids[typ] = b.stableIDs[typ]
+	}
+
+	if err := e.deflatePending(); err != nil {
+		return nil, err
+	}
+
+	length := e.buf.Len()
+	typeLen := uint32(length - btfHeaderLen)
+
+	stringLen := e.strings.Length()
+	buf = e.strings.AppendEncoded(e.buf.Bytes())
+
+	// Fill out the header, and write it out.
+	header := &btfHeader{
+		Magic:     btfMagic,
+		Version:   1,
+		Flags:     0,
+		HdrLen:    uint32(btfHeaderLen),
+		TypeOff:   0,
+		TypeLen:   typeLen,
+		StringOff: typeLen,
+		StringLen: uint32(stringLen),
+	}
+
+	err := binary.Write(sliceWriter(buf[:btfHeaderLen]), e.Order, header)
+	if err != nil {
+		return nil, fmt.Errorf("write header: %v", err)
+	}
+
+	return buf, nil
+}
+
+// addString adds a string to the resulting BTF.
+//
+// Adding the same string multiple times will return the same result.
+//
+// Returns an identifier into the string table or an error if the string
+// contains invalid characters.
+func (b *Builder) addString(str string) (uint32, error) {
+	if b.strings == nil {
+		b.strings = newStringTableBuilder(0)
+	}
+
+	return b.strings.Add(str)
+}
+
+func (e *encoder) allocateID(typ Type) error {
+	id := e.lastID + 1
+	if id < e.lastID {
+		return errors.New("type ID overflow")
+	}
+
+	e.pending.Push(typ)
+	e.ids[typ] = id
+	e.lastID = id
+	return nil
+}
+
+// id returns the ID for the given type or panics with an error.
+func (e *encoder) id(typ Type) TypeID {
+	if _, ok := typ.(*Void); ok {
+		return 0
+	}
+
+	id, ok := e.ids[typ]
+	if !ok {
+		panic(fmt.Errorf("no ID for type %v", typ))
+	}
+
+	return id
+}
+
+func (e *encoder) deflatePending() error {
+	// Declare root outside of the loop to avoid repeated heap allocations.
+	var root Type
+	skip := func(t Type) (skip bool) {
+		if t == root {
+			// Force descending into the current root type even if it already
+			// has an ID. Otherwise we miss children of types that have their
+			// ID pre-allocated via Add.
+			return false
+		}
+
+		_, isVoid := t.(*Void)
+		_, alreadyEncoded := e.ids[t]
+		return isVoid || alreadyEncoded
+	}
+
+	for !e.pending.Empty() {
+		root = e.pending.Shift()
+
+		// Allocate IDs for all children of typ, including transitive dependencies.
+		iter := postorderTraversal(root, skip)
+		for iter.Next() {
+			if iter.Type == root {
+				// The iterator yields root at the end, do not allocate another ID.
+				break
+			}
+
+			if err := e.allocateID(iter.Type); err != nil {
+				return err
+			}
+		}
+
+		if err := e.deflateType(root); err != nil {
+			id := e.ids[root]
+			return fmt.Errorf("deflate %v with ID %d: %w", root, id, err)
+		}
+	}
+
+	return nil
+}
+
+func (e *encoder) deflateType(typ Type) (err error) {
+	defer func() {
+		if r := recover(); r != nil {
+			var ok bool
+			err, ok = r.(error)
+			if !ok {
+				panic(r)
+			}
+		}
+	}()
+
+	var raw rawType
+	raw.NameOff, err = e.strings.Add(typ.TypeName())
+	if err != nil {
+		return err
+	}
+
+	switch v := typ.(type) {
+	case *Void:
+		return errors.New("Void is implicit in BTF wire format")
+
+	case *Int:
+		raw.SetKind(kindInt)
+		raw.SetSize(v.Size)
+
+		var bi btfInt
+		bi.SetEncoding(v.Encoding)
+		// We need to set bits in addition to size, since btf_type_int_is_regular
+		// otherwise flags this as a bitfield.
+		bi.SetBits(byte(v.Size) * 8)
+		raw.data = bi
+
+	case *Pointer:
+		raw.SetKind(kindPointer)
+		raw.SetType(e.id(v.Target))
+
+	case *Array:
+		raw.SetKind(kindArray)
+		raw.data = &btfArray{
+			e.id(v.Type),
+			e.id(v.Index),
+			v.Nelems,
+		}
+
+	case *Struct:
+		raw.SetKind(kindStruct)
+		raw.SetSize(v.Size)
+		raw.data, err = e.convertMembers(&raw.btfType, v.Members)
+
+	case *Union:
+		raw.SetKind(kindUnion)
+		raw.SetSize(v.Size)
+		raw.data, err = e.convertMembers(&raw.btfType, v.Members)
+
+	case *Enum:
+		raw.SetSize(v.size())
+		raw.SetVlen(len(v.Values))
+		raw.SetSigned(v.Signed)
+
+		if v.has64BitValues() {
+			raw.SetKind(kindEnum64)
+			raw.data, err = e.deflateEnum64Values(v.Values)
+		} else {
+			raw.SetKind(kindEnum)
+			raw.data, err = e.deflateEnumValues(v.Values)
+		}
+
+	case *Fwd:
+		raw.SetKind(kindForward)
+		raw.SetFwdKind(v.Kind)
+
+	case *Typedef:
+		raw.SetKind(kindTypedef)
+		raw.SetType(e.id(v.Type))
+
+	case *Volatile:
+		raw.SetKind(kindVolatile)
+		raw.SetType(e.id(v.Type))
+
+	case *Const:
+		raw.SetKind(kindConst)
+		raw.SetType(e.id(v.Type))
+
+	case *Restrict:
+		raw.SetKind(kindRestrict)
+		raw.SetType(e.id(v.Type))
+
+	case *Func:
+		raw.SetKind(kindFunc)
+		raw.SetType(e.id(v.Type))
+		if !e.StripFuncLinkage {
+			raw.SetLinkage(v.Linkage)
+		}
+
+	case *FuncProto:
+		raw.SetKind(kindFuncProto)
+		raw.SetType(e.id(v.Return))
+		raw.SetVlen(len(v.Params))
+		raw.data, err = e.deflateFuncParams(v.Params)
+
+	case *Var:
+		raw.SetKind(kindVar)
+		raw.SetType(e.id(v.Type))
+		raw.data = btfVariable{uint32(v.Linkage)}
+
+	case *Datasec:
+		raw.SetKind(kindDatasec)
+		raw.SetSize(v.Size)
+		raw.SetVlen(len(v.Vars))
+		raw.data = e.deflateVarSecinfos(v.Vars)
+
+	case *Float:
+		raw.SetKind(kindFloat)
+		raw.SetSize(v.Size)
+
+	case *declTag:
+		raw.SetKind(kindDeclTag)
+		raw.SetType(e.id(v.Type))
+		raw.data = &btfDeclTag{uint32(v.Index)}
+		raw.NameOff, err = e.strings.Add(v.Value)
+
+	case *typeTag:
+		raw.SetKind(kindTypeTag)
+		raw.SetType(e.id(v.Type))
+		raw.NameOff, err = e.strings.Add(v.Value)
+
+	default:
+		return fmt.Errorf("don't know how to deflate %T", v)
+	}
+
+	if err != nil {
+		return err
+	}
+
+	return raw.Marshal(e.buf, e.Order)
+}
+
+func (e *encoder) convertMembers(header *btfType, members []Member) ([]btfMember, error) {
+	bms := make([]btfMember, 0, len(members))
+	isBitfield := false
+	for _, member := range members {
+		isBitfield = isBitfield || member.BitfieldSize > 0
+
+		offset := member.Offset
+		if isBitfield {
+			offset = member.BitfieldSize<<24 | (member.Offset & 0xffffff)
+		}
+
+		nameOff, err := e.strings.Add(member.Name)
+		if err != nil {
+			return nil, err
+		}
+
+		bms = append(bms, btfMember{
+			nameOff,
+			e.id(member.Type),
+			uint32(offset),
+		})
+	}
+
+	header.SetVlen(len(members))
+	header.SetBitfield(isBitfield)
+	return bms, nil
+}
+
+func (e *encoder) deflateEnumValues(values []EnumValue) ([]btfEnum, error) {
+	bes := make([]btfEnum, 0, len(values))
+	for _, value := range values {
+		nameOff, err := e.strings.Add(value.Name)
+		if err != nil {
+			return nil, err
+		}
+
+		if value.Value > math.MaxUint32 {
+			return nil, fmt.Errorf("value of enum %q exceeds 32 bits", value.Name)
+		}
+
+		bes = append(bes, btfEnum{
+			nameOff,
+			uint32(value.Value),
+		})
+	}
+
+	return bes, nil
+}
+
+func (e *encoder) deflateEnum64Values(values []EnumValue) ([]btfEnum64, error) {
+	bes := make([]btfEnum64, 0, len(values))
+	for _, value := range values {
+		nameOff, err := e.strings.Add(value.Name)
+		if err != nil {
+			return nil, err
+		}
+
+		bes = append(bes, btfEnum64{
+			nameOff,
+			uint32(value.Value),
+			uint32(value.Value >> 32),
+		})
+	}
+
+	return bes, nil
+}
+
+func (e *encoder) deflateFuncParams(params []FuncParam) ([]btfParam, error) {
+	bps := make([]btfParam, 0, len(params))
+	for _, param := range params {
+		nameOff, err := e.strings.Add(param.Name)
+		if err != nil {
+			return nil, err
+		}
+
+		bps = append(bps, btfParam{
+			nameOff,
+			e.id(param.Type),
+		})
+	}
+	return bps, nil
+}
+
+func (e *encoder) deflateVarSecinfos(vars []VarSecinfo) []btfVarSecinfo {
+	vsis := make([]btfVarSecinfo, 0, len(vars))
+	for _, v := range vars {
+		vsis = append(vsis, btfVarSecinfo{
+			e.id(v.Type),
+			v.Offset,
+			v.Size,
+		})
+	}
+	return vsis
+}
+
+// MarshalMapKV creates a BTF object containing a map key and value.
+//
+// The function is intended for the use of the ebpf package and may be removed
+// at any point in time.
+func MarshalMapKV(key, value Type) (_ *Handle, keyID, valueID TypeID, err error) {
+	var b Builder
+
+	if key != nil {
+		keyID, err = b.Add(key)
+		if err != nil {
+			return nil, 0, 0, fmt.Errorf("add key type: %w", err)
+		}
+	}
+
+	if value != nil {
+		valueID, err = b.Add(value)
+		if err != nil {
+			return nil, 0, 0, fmt.Errorf("add value type: %w", err)
+		}
+	}
+
+	handle, err := NewHandle(&b)
+	if err != nil {
+		// Check for 'full' map BTF support, since kernels between 4.18 and 5.2
+		// already support BTF blobs for maps without Var or Datasec just fine.
+		if err := haveMapBTF(); err != nil {
+			return nil, 0, 0, err
+		}
+	}
+	return handle, keyID, valueID, err
+}

vendor/github.com/cilium/ebpf/btf/strings.go

@@ -6,6 +6,9 @@ import (
 	"errors"
 	"fmt"
 	"io"
+	"strings"
+
+	"golang.org/x/exp/maps"
 )
 
 type stringTable struct {
@@ -88,11 +91,6 @@ func (st *stringTable) lookup(offset uint32) (string, error) {
 	return st.strings[i], nil
 }
 
-func (st *stringTable) Length() int {
-	last := len(st.offsets) - 1
-	return int(st.offsets[last]) + len(st.strings[last]) + 1
-}
-
 func (st *stringTable) Marshal(w io.Writer) error {
 	for _, str := range st.strings {
 		_, err := io.WriteString(w, str)
@@ -107,6 +105,11 @@ func (st *stringTable) Marshal(w io.Writer) error {
 	return nil
 }
 
+// Num returns the number of strings in the table.
+func (st *stringTable) Num() int {
+	return len(st.strings)
+}
+
 // search is a copy of sort.Search specialised for uint32.
 //
 // Licensed under https://go.dev/LICENSE
@@ -126,3 +129,86 @@ func search(ints []uint32, needle uint32) int {
 	// i == j, f(i-1) == false, and f(j) (= f(i)) == true  =>  answer is i.
 	return i
 }
+
+// stringTableBuilder builds BTF string tables.
+type stringTableBuilder struct {
+	length  uint32
+	strings map[string]uint32
+}
+
+// newStringTableBuilder creates a builder with the given capacity.
+//
+// capacity may be zero.
+func newStringTableBuilder(capacity int) *stringTableBuilder {
+	var stb stringTableBuilder
+
+	if capacity == 0 {
+		// Use the runtime's small default size.
+		stb.strings = make(map[string]uint32)
+	} else {
+		stb.strings = make(map[string]uint32, capacity)
+	}
+
+	// Ensure that the empty string is at index 0.
+	stb.append("")
+	return &stb
+}
+
+// Add a string to the table.
+//
+// Adding the same string multiple times will only store it once.
+func (stb *stringTableBuilder) Add(str string) (uint32, error) {
+	if strings.IndexByte(str, 0) != -1 {
+		return 0, fmt.Errorf("string contains null: %q", str)
+	}
+
+	offset, ok := stb.strings[str]
+	if ok {
+		return offset, nil
+	}
+
+	return stb.append(str), nil
+}
+
+func (stb *stringTableBuilder) append(str string) uint32 {
+	offset := stb.length
+	stb.length += uint32(len(str)) + 1
+	stb.strings[str] = offset
+	return offset
+}
+
+// Lookup finds the offset of a string in the table.
+//
+// Returns an error if str hasn't been added yet.
+func (stb *stringTableBuilder) Lookup(str string) (uint32, error) {
+	offset, ok := stb.strings[str]
+	if !ok {
+		return 0, fmt.Errorf("string %q is not in table", str)
+	}
+
+	return offset, nil
+}
+
+// Length returns the length in bytes.
+func (stb *stringTableBuilder) Length() int {
+	return int(stb.length)
+}
+
+// AppendEncoded appends the string table to the end of the provided buffer.
+func (stb *stringTableBuilder) AppendEncoded(buf []byte) []byte {
+	n := len(buf)
+	buf = append(buf, make([]byte, stb.Length())...)
+	strings := buf[n:]
+	for str, offset := range stb.strings {
+		copy(strings[offset:], str)
+	}
+	return buf
+}
+
+// Copy the string table builder.
+func (stb *stringTableBuilder) Copy() *stringTableBuilder {
+	return &stringTableBuilder{
+		stb.length,
+		maps.Clone(stb.strings),
+	}
+}

vendor/github.com/cilium/ebpf/btf/traversal.go

@@ -0,0 +1,141 @@
+package btf
+
+import (
+	"fmt"
+
+	"github.com/cilium/ebpf/internal"
+)
+
+// Functions to traverse a cyclic graph of types. The below was very useful:
+// https://eli.thegreenplace.net/2015/directed-graph-traversal-orderings-and-applications-to-data-flow-analysis/#post-order-and-reverse-post-order
+
+type postorderIterator struct {
+	// Iteration skips types for which this function returns true.
+	skip func(Type) bool
+	// The root type. May be nil if skip(root) is true.
+	root Type
+
+	// Contains types which need to be either walked or yielded.
+	types typeDeque
+	// Contains a boolean whether the type has been walked or not.
+	walked internal.Deque[bool]
+	// The set of types which has been pushed onto types.
+	pushed map[Type]struct{}
+
+	// The current type. Only valid after a call to Next().
+	Type Type
+}
+
+// postorderTraversal iterates all types reachable from root by visiting the
+// leaves of the graph first.
+//
+// Types for which skip returns true are ignored. skip may be nil.
+func postorderTraversal(root Type, skip func(Type) (skip bool)) postorderIterator {
+	// Avoid allocations for the common case of a skipped root.
+	if skip != nil && skip(root) {
+		return postorderIterator{}
+	}
+
+	po := postorderIterator{root: root, skip: skip}
+	walkType(root, po.push)
+
+	return po
+}
+
+func (po *postorderIterator) push(t *Type) {
+	if _, ok := po.pushed[*t]; ok || *t == po.root {
+		return
+	}
+
+	if po.skip != nil && po.skip(*t) {
+		return
+	}
+
+	if po.pushed == nil {
+		// Lazily allocate pushed to avoid an allocation for Types without children.
+		po.pushed = make(map[Type]struct{})
+	}
+
+	po.pushed[*t] = struct{}{}
+	po.types.Push(t)
+	po.walked.Push(false)
+}
+
+// Next returns true if there is another Type to traverse.
+func (po *postorderIterator) Next() bool {
+	for !po.types.Empty() {
+		t := po.types.Pop()
+
+		if !po.walked.Pop() {
+			// Push the type again, so that we re-evaluate it in done state
+			// after all children have been handled.
+			po.types.Push(t)
+			po.walked.Push(true)
+
+			// Add all direct children to todo.
+			walkType(*t, po.push)
+		} else {
+			// We've walked this type previously, so we now know that all
+			// children have been handled.
+			po.Type = *t
+			return true
+		}
+	}
+
+	// Only return root once.
+	po.Type, po.root = po.root, nil
+	return po.Type != nil
+}
+
+// walkType calls fn on each child of typ.
+func walkType(typ Type, fn func(*Type)) {
+	// Explicitly type switch on the most common types to allow the inliner to
+	// do its work. This avoids allocating intermediate slices from walk() on
+	// the heap.
+	switch v := typ.(type) {
+	case *Void, *Int, *Enum, *Fwd, *Float:
+		// No children to traverse.
+	case *Pointer:
+		fn(&v.Target)
+	case *Array:
+		fn(&v.Index)
+		fn(&v.Type)
+	case *Struct:
+		for i := range v.Members {
+			fn(&v.Members[i].Type)
+		}
+	case *Union:
+		for i := range v.Members {
+			fn(&v.Members[i].Type)
+		}
+	case *Typedef:
+		fn(&v.Type)
+	case *Volatile:
+		fn(&v.Type)
+	case *Const:
+		fn(&v.Type)
+	case *Restrict:
+		fn(&v.Type)
+	case *Func:
+		fn(&v.Type)
+	case *FuncProto:
+		fn(&v.Return)
+		for i := range v.Params {
+			fn(&v.Params[i].Type)
+		}
+	case *Var:
+		fn(&v.Type)
+	case *Datasec:
+		for i := range v.Vars {
+			fn(&v.Vars[i].Type)
+		}
+	case *declTag:
+		fn(&v.Type)
+	case *typeTag:
+		fn(&v.Type)
+	case *cycle:
+		// cycle has children, but we ignore them deliberately.
+	default:
+		panic(fmt.Sprintf("don't know how to walk Type %T", v))
+	}
+}

vendor/github.com/cilium/ebpf/btf/types.go

@@ -1,6 +1,7 @@
 package btf
 
 import (
+	"errors"
 	"fmt"
 	"io"
 	"math"
@@ -8,14 +9,27 @@ import (
 	"strings"
 
 	"github.com/cilium/ebpf/asm"
+	"github.com/cilium/ebpf/internal"
+	"github.com/cilium/ebpf/internal/sys"
 )
 
 const maxTypeDepth = 32
 
 // TypeID identifies a type in a BTF section.
-type TypeID uint32
+type TypeID = sys.TypeID
 
 // Type represents a type described by BTF.
+//
+// Identity of Type follows the [Go specification]: two Types are considered
+// equal if they have the same concrete type and the same dynamic value, aka
+// they point at the same location in memory. This means that the following
+// Types are considered distinct even though they have the same "shape".
+//
+//	a := &Int{Size: 1}
+//	b := &Int{Size: 1}
+//	a != b
+//
+// [Go specification]: https://go.dev/ref/spec#Comparison_operators
 type Type interface {
 	// Type can be formatted using the %s and %v verbs. %s outputs only the
 	// identity of the type, without any detail. %v outputs additional detail.
@@ -35,9 +49,7 @@ type Type interface {
 	// Make a copy of the type, without copying Type members.
 	copy() Type
 
-	// Enumerate all nested Types. Repeated calls must visit nested
+	// New implementations must update walkType.
-	// types in the same order.
-	walk(*typeDeque)
 }
 
 var (
@@ -51,20 +63,11 @@ var (
 	_ Type = (*Var)(nil)
 	_ Type = (*Datasec)(nil)
 	_ Type = (*Float)(nil)
+	_ Type = (*declTag)(nil)
+	_ Type = (*typeTag)(nil)
+	_ Type = (*cycle)(nil)
 )
 
-// types is a list of Type.
-//
-// The order determines the ID of a type.
-type types []Type
-
-func (ts types) ByID(id TypeID) (Type, error) {
-	if int(id) > len(ts) {
-		return nil, fmt.Errorf("type ID %d: %w", id, ErrNotFound)
-	}
-	return ts[id], nil
-}
-
 // Void is the unit type of BTF.
 type Void struct{}
 
@@ -72,40 +75,32 @@ func (v *Void) Format(fs fmt.State, verb rune) { formatType(fs, verb, v) }
 func (v *Void) TypeName() string               { return "" }
 func (v *Void) size() uint32                   { return 0 }
 func (v *Void) copy() Type                     { return (*Void)(nil) }
-func (v *Void) walk(*typeDeque)                {}
 
 type IntEncoding byte
 
+// Valid IntEncodings.
+//
+// These may look like they are flags, but they aren't.
 const (
-	Signed IntEncoding = 1 << iota
+	Unsigned IntEncoding = 0
-	Char
+	Signed   IntEncoding = 1
-	Bool
+	Char     IntEncoding = 2
+	Bool     IntEncoding = 4
 )
 
-func (ie IntEncoding) IsSigned() bool {
-	return ie&Signed != 0
-}
-
-func (ie IntEncoding) IsChar() bool {
-	return ie&Char != 0
-}
-
-func (ie IntEncoding) IsBool() bool {
-	return ie&Bool != 0
-}
-
 func (ie IntEncoding) String() string {
-	switch {
+	switch ie {
-	case ie.IsChar() && ie.IsSigned():
+	case Char:
+		// NB: There is no way to determine signedness for char.
 		return "char"
-	case ie.IsChar() && !ie.IsSigned():
+	case Bool:
-		return "uchar"
-	case ie.IsBool():
 		return "bool"
-	case ie.IsSigned():
+	case Signed:
 		return "signed"
-	default:
+	case Unsigned:
 		return "unsigned"
+	default:
+		return fmt.Sprintf("IntEncoding(%d)", byte(ie))
 	}
 }
 
@@ -126,7 +121,6 @@ func (i *Int) Format(fs fmt.State, verb rune) {
 
 func (i *Int) TypeName() string { return i.Name }
 func (i *Int) size() uint32     { return i.Size }
-func (i *Int) walk(*typeDeque)  {}
 func (i *Int) copy() Type {
 	cpy := *i
 	return &cpy
@@ -143,7 +137,6 @@ func (p *Pointer) Format(fs fmt.State, verb rune) {
 
 func (p *Pointer) TypeName() string { return "" }
 func (p *Pointer) size() uint32     { return 8 }
-func (p *Pointer) walk(tdq *typeDeque) { tdq.push(&p.Target) }
 func (p *Pointer) copy() Type {
 	cpy := *p
 	return &cpy
@@ -162,11 +155,6 @@ func (arr *Array) Format(fs fmt.State, verb rune) {
 
 func (arr *Array) TypeName() string { return "" }
 
-func (arr *Array) walk(tdq *typeDeque) {
-	tdq.push(&arr.Index)
-	tdq.push(&arr.Type)
-}
-
 func (arr *Array) copy() Type {
 	cpy := *arr
 	return &cpy
@@ -188,12 +176,6 @@ func (s *Struct) TypeName() string { return s.Name }
 
 func (s *Struct) size() uint32 { return s.Size }
 
-func (s *Struct) walk(tdq *typeDeque) {
-	for i := range s.Members {
-		tdq.push(&s.Members[i].Type)
-	}
-}
-
 func (s *Struct) copy() Type {
 	cpy := *s
 	cpy.Members = copyMembers(s.Members)
@@ -220,12 +202,6 @@ func (u *Union) TypeName() string { return u.Name }
 
 func (u *Union) size() uint32 { return u.Size }
 
-func (u *Union) walk(tdq *typeDeque) {
-	for i := range u.Members {
-		tdq.push(&u.Members[i].Type)
-	}
-}
-
 func (u *Union) copy() Type {
 	cpy := *u
 	cpy.Members = copyMembers(u.Members)
@@ -243,6 +219,7 @@ func copyMembers(orig []Member) []Member {
 }
 
 type composite interface {
+	Type
 	members() []Member
 }
 
@@ -274,6 +251,8 @@ type Enum struct {
 	Name string
 	// Size of the enum value in bytes.
 	Size uint32
+	// True if the values should be interpreted as signed integers.
+	Signed bool
 	Values []EnumValue
 }
 
@@ -288,11 +267,10 @@ func (e *Enum) TypeName() string { return e.Name }
 // Is is not a valid Type
 type EnumValue struct {
 	Name  string
-	Value int32
+	Value uint64
 }
 
 func (e *Enum) size() uint32 { return e.Size }
-func (e *Enum) walk(*typeDeque) {}
 func (e *Enum) copy() Type {
 	cpy := *e
 	cpy.Values = make([]EnumValue, len(e.Values))
@@ -300,6 +278,21 @@ func (e *Enum) copy() Type {
 	return &cpy
 }
 
+// has64BitValues returns true if the Enum contains a value larger than 32 bits.
+// Kernels before 6.0 have enum values that overrun u32 replaced with zeroes.
+//
+// 64-bit enums have their Enum.Size attributes correctly set to 8, but if we
+// use the size attribute as a heuristic during BTF marshaling, we'll emit
+// ENUM64s to kernels that don't support them.
+func (e *Enum) has64BitValues() bool {
+	for _, v := range e.Values {
+		if v.Value > math.MaxUint32 {
+			return true
+		}
+	}
+	return false
+}
+
 // FwdKind is the type of forward declaration.
 type FwdKind int
 
@@ -332,7 +325,6 @@ func (f *Fwd) Format(fs fmt.State, verb rune) {
 
 func (f *Fwd) TypeName() string { return f.Name }
 
-func (f *Fwd) walk(*typeDeque) {}
 func (f *Fwd) copy() Type {
 	cpy := *f
 	return &cpy
@@ -350,7 +342,6 @@ func (td *Typedef) Format(fs fmt.State, verb rune) {
 
 func (td *Typedef) TypeName() string { return td.Name }
 
-func (td *Typedef) walk(tdq *typeDeque) { tdq.push(&td.Type) }
 func (td *Typedef) copy() Type {
 	cpy := *td
 	return &cpy
@@ -368,7 +359,6 @@ func (v *Volatile) Format(fs fmt.State, verb rune) {
 func (v *Volatile) TypeName() string { return "" }
 
 func (v *Volatile) qualify() Type { return v.Type }
-func (v *Volatile) walk(tdq *typeDeque) { tdq.push(&v.Type) }
 func (v *Volatile) copy() Type {
 	cpy := *v
 	return &cpy
@@ -386,7 +376,6 @@ func (c *Const) Format(fs fmt.State, verb rune) {
 func (c *Const) TypeName() string { return "" }
 
 func (c *Const) qualify() Type { return c.Type }
-func (c *Const) walk(tdq *typeDeque) { tdq.push(&c.Type) }
 func (c *Const) copy() Type {
 	cpy := *c
 	return &cpy
@@ -404,7 +393,6 @@ func (r *Restrict) Format(fs fmt.State, verb rune) {
 func (r *Restrict) TypeName() string { return "" }
 
 func (r *Restrict) qualify() Type { return r.Type }
-func (r *Restrict) walk(tdq *typeDeque) { tdq.push(&r.Type) }
 func (r *Restrict) copy() Type {
 	cpy := *r
 	return &cpy
@@ -422,13 +410,18 @@ func FuncMetadata(ins *asm.Instruction) *Func {
 	return fn
 }
 
+// WithFuncMetadata adds a btf.Func to the Metadata of asm.Instruction.
+func WithFuncMetadata(ins asm.Instruction, fn *Func) asm.Instruction {
+	ins.Metadata.Set(funcInfoMeta{}, fn)
+	return ins
+}
+
 func (f *Func) Format(fs fmt.State, verb rune) {
 	formatType(fs, verb, f, f.Linkage, "proto=", f.Type)
 }
 
 func (f *Func) TypeName() string { return f.Name }
 
-func (f *Func) walk(tdq *typeDeque) { tdq.push(&f.Type) }
 func (f *Func) copy() Type {
 	cpy := *f
 	return &cpy
@@ -446,13 +439,6 @@ func (fp *FuncProto) Format(fs fmt.State, verb rune) {
 
 func (fp *FuncProto) TypeName() string { return "" }
 
-func (fp *FuncProto) walk(tdq *typeDeque) {
-	tdq.push(&fp.Return)
-	for i := range fp.Params {
-		tdq.push(&fp.Params[i].Type)
-	}
-}
-
 func (fp *FuncProto) copy() Type {
 	cpy := *fp
 	cpy.Params = make([]FuncParam, len(fp.Params))
@@ -478,7 +464,6 @@ func (v *Var) Format(fs fmt.State, verb rune) {
 
 func (v *Var) TypeName() string { return v.Name }
 
-func (v *Var) walk(tdq *typeDeque) { tdq.push(&v.Type) }
 func (v *Var) copy() Type {
 	cpy := *v
 	return &cpy
@@ -499,12 +484,6 @@ func (ds *Datasec) TypeName() string { return ds.Name }
 
 func (ds *Datasec) size() uint32 { return ds.Size }
 
-func (ds *Datasec) walk(tdq *typeDeque) {
-	for i := range ds.Vars {
-		tdq.push(&ds.Vars[i].Type)
-	}
-}
-
 func (ds *Datasec) copy() Type {
 	cpy := *ds
 	cpy.Vars = make([]VarSecinfo, len(ds.Vars))
@@ -516,6 +495,7 @@ func (ds *Datasec) copy() Type {
 //
 // It is not a valid Type.
 type VarSecinfo struct {
+	// Var or Func.
 	Type   Type
 	Offset uint32
 	Size   uint32
@@ -535,12 +515,48 @@ func (f *Float) Format(fs fmt.State, verb rune) {
 
 func (f *Float) TypeName() string { return f.Name }
 func (f *Float) size() uint32     { return f.Size }
-func (f *Float) walk(*typeDeque)  {}
 func (f *Float) copy() Type {
 	cpy := *f
 	return &cpy
 }
 
+// declTag associates metadata with a declaration.
+type declTag struct {
+	Type  Type
+	Value string
+	// The index this tag refers to in the target type. For composite types,
+	// a value of -1 indicates that the tag refers to the whole type. Otherwise
+	// it indicates which member or argument the tag applies to.
+	Index int
+}
+
+func (dt *declTag) Format(fs fmt.State, verb rune) {
+	formatType(fs, verb, dt, "type=", dt.Type, "value=", dt.Value, "index=", dt.Index)
+}
+
+func (dt *declTag) TypeName() string { return "" }
+func (dt *declTag) copy() Type {
+	cpy := *dt
+	return &cpy
+}
+
+// typeTag associates metadata with a type.
+type typeTag struct {
+	Type  Type
+	Value string
+}
+
+func (tt *typeTag) Format(fs fmt.State, verb rune) {
+	formatType(fs, verb, tt, "type=", tt.Type, "value=", tt.Value)
+}
+
+func (tt *typeTag) TypeName() string { return "" }
+func (tt *typeTag) qualify() Type    { return tt.Type }
+func (tt *typeTag) copy() Type {
+	cpy := *tt
+	return &cpy
+}
+
 // cycle is a type which had to be elided since it exceeded maxTypeDepth.
 type cycle struct {
 	root Type
@@ -549,7 +565,6 @@ type cycle struct {
 func (c *cycle) ID() TypeID                     { return math.MaxUint32 }
 func (c *cycle) Format(fs fmt.State, verb rune) { formatType(fs, verb, c, "root=", c.root) }
 func (c *cycle) TypeName() string               { return "" }
-func (c *cycle) walk(*typeDeque)                {}
 func (c *cycle) copy() Type {
 	cpy := *c
 	return &cpy
@@ -576,8 +591,11 @@ var (
 	_ qualifier = (*Const)(nil)
 	_ qualifier = (*Restrict)(nil)
 	_ qualifier = (*Volatile)(nil)
+	_ qualifier = (*typeTag)(nil)
 )
 
+var errUnsizedType = errors.New("type is unsized")
+
 // Sizeof returns the size of a type in bytes.
 //
 // Returns an error if the size can't be computed.
@@ -612,7 +630,7 @@ func Sizeof(typ Type) (int, error) {
 			continue
 
 		default:
-			return 0, fmt.Errorf("unsized type %T", typ)
+			return 0, fmt.Errorf("type %T: %w", typ, errUnsizedType)
 		}
 
 		if n > 0 && elem > math.MaxInt64/n {
@@ -632,16 +650,33 @@ func Sizeof(typ Type) (int, error) {
 
 // alignof returns the alignment of a type.
 //
-// Currently only supports the subset of types necessary for bitfield relocations.
+// Returns an error if the Type can't be aligned, like an integer with an uneven
+// size. Currently only supports the subset of types necessary for bitfield
+// relocations.
 func alignof(typ Type) (int, error) {
+	var n int
+
 	switch t := UnderlyingType(typ).(type) {
 	case *Enum:
-		return int(t.size()), nil
+		n = int(t.size())
 	case *Int:
-		return int(t.Size), nil
+		n = int(t.Size)
+	case *Array:
+		return alignof(t.Type)
 	default:
 		return 0, fmt.Errorf("can't calculate alignment of %T", t)
 	}
+
+	if !pow(n) {
+		return 0, fmt.Errorf("alignment value %d is not a power of two", n)
+	}
+
+	return n, nil
+}
+
+// pow returns true if n is a power of two.
+func pow(n int) bool {
+	return n != 0 && (n&(n-1)) == 0
 }
 
 // Transformer modifies a given Type and returns the result.
@@ -655,7 +690,7 @@ type Transformer func(Type) Type
 // typ may form a cycle. If transform is not nil, it is called with the
 // to be copied type, and the returned value is copied instead.
 func Copy(typ Type, transform Transformer) Type {
-	copies := make(copier)
+	copies := copier{copies: make(map[Type]Type)}
 	copies.copy(&typ, transform)
 	return typ
 }
@@ -667,7 +702,7 @@ func copyTypes(types []Type, transform Transformer) []Type {
 	result := make([]Type, len(types))
 	copy(result, types)
 
-	copies := make(copier)
+	copies := copier{copies: make(map[Type]Type, len(types))}
 	for i := range result {
 		copies.copy(&result[i], transform)
 	}
@@ -675,13 +710,15 @@ func copyTypes(types []Type, transform Transformer) []Type {
 	return result
 }
 
-type copier map[Type]Type
+type copier struct {
+	copies map[Type]Type
+	work   typeDeque
+}
 
-func (c copier) copy(typ *Type, transform Transformer) {
+func (c *copier) copy(typ *Type, transform Transformer) {
-	var work typeDeque
+	for t := typ; t != nil; t = c.work.Pop() {
-	for t := typ; t != nil; t = work.pop() {
 		// *t is the identity of the type.
-		if cpy := c[*t]; cpy != nil {
+		if cpy := c.copies[*t]; cpy != nil {
 			*t = cpy
 			continue
 		}
@@ -693,108 +730,41 @@ func (c copier) copy(typ *Type, transform Transformer) {
 			cpy = (*t).copy()
 		}
 
-		c[*t] = cpy
+		c.copies[*t] = cpy
 		*t = cpy
 
 		// Mark any nested types for copying.
-		cpy.walk(&work)
+		walkType(cpy, c.work.Push)
 	}
 }
 
-// typeDeque keeps track of pointers to types which still
+type typeDeque = internal.Deque[*Type]
-// need to be visited.
-type typeDeque struct {
-	types       []*Type
-	read, write uint64
-	mask        uint64
-}
-
-func (dq *typeDeque) empty() bool {
-	return dq.read == dq.write
-}
-
-// push adds a type to the stack.
-func (dq *typeDeque) push(t *Type) {
-	if dq.write-dq.read < uint64(len(dq.types)) {
-		dq.types[dq.write&dq.mask] = t
-		dq.write++
-		return
-	}
-
-	new := len(dq.types) * 2
-	if new == 0 {
-		new = 8
-	}
-
-	types := make([]*Type, new)
-	pivot := dq.read & dq.mask
-	n := copy(types, dq.types[pivot:])
-	n += copy(types[n:], dq.types[:pivot])
-	types[n] = t
-
-	dq.types = types
-	dq.mask = uint64(new) - 1
-	dq.read, dq.write = 0, uint64(n+1)
-}
-
-// shift returns the first element or null.
-func (dq *typeDeque) shift() *Type {
-	if dq.empty() {
-		return nil
-	}
-
-	index := dq.read & dq.mask
-	t := dq.types[index]
-	dq.types[index] = nil
-	dq.read++
-	return t
-}
-
-// pop returns the last element or null.
-func (dq *typeDeque) pop() *Type {
-	if dq.empty() {
-		return nil
-	}
-
-	dq.write--
-	index := dq.write & dq.mask
-	t := dq.types[index]
-	dq.types[index] = nil
-	return t
-}
-
-// all returns all elements.
-//
-// The deque is empty after calling this method.
-func (dq *typeDeque) all() []*Type {
-	length := dq.write - dq.read
-	types := make([]*Type, 0, length)
-	for t := dq.shift(); t != nil; t = dq.shift() {
-		types = append(types, t)
-	}
-	return types
-}
 
 // inflateRawTypes takes a list of raw btf types linked via type IDs, and turns
 // it into a graph of Types connected via pointers.
 //
-// If baseTypes are provided, then the raw types are
+// If base is provided, then the raw types are considered to be of a split BTF
-// considered to be of a split BTF (e.g., a kernel module).
+// (e.g., a kernel module).
 //
 // Returns a slice of types indexed by TypeID. Since BTF ignores compilation
 // units, multiple types may share the same name. A Type may form a cyclic graph
 // by pointing at itself.
-func inflateRawTypes(rawTypes []rawType, baseTypes types, rawStrings *stringTable) ([]Type, error) {
+func inflateRawTypes(rawTypes []rawType, rawStrings *stringTable, base *Spec) ([]Type, error) {
 	types := make([]Type, 0, len(rawTypes)+1) // +1 for Void added to base types
 
-	typeIDOffset := TypeID(1) // Void is TypeID(0), so the rest starts from TypeID(1)
-
-	if baseTypes == nil {
 	// Void is defined to always be type ID 0, and is thus omitted from BTF.
 	types = append(types, (*Void)(nil))
-	} else {
+
-		// For split BTF, the next ID is max base BTF type ID + 1
+	firstTypeID := TypeID(0)
-		typeIDOffset = TypeID(len(baseTypes))
+	if base != nil {
+		var err error
+		firstTypeID, err = base.nextTypeID()
+		if err != nil {
+			return nil, err
+		}
+
+		// Split BTF doesn't contain Void.
+		types = types[:0]
 	}
 
 	type fixupDef struct {
@@ -803,39 +773,42 @@ func inflateRawTypes(rawTypes []rawType, baseTypes types, rawStrings *stringTabl
 	}
 
 	var fixups []fixupDef
-	fixup := func(id TypeID, typ *Type) {
+	fixup := func(id TypeID, typ *Type) bool {
-		if id < TypeID(len(baseTypes)) {
+		if id < firstTypeID {
-			*typ = baseTypes[id]
+			if baseType, err := base.TypeByID(id); err == nil {
-			return
+				*typ = baseType
+				return true
+			}
 		}
 
-		idx := id
+		idx := int(id - firstTypeID)
-		if baseTypes != nil {
+		if idx < len(types) {
-			idx = id - TypeID(len(baseTypes))
-		}
-		if idx < TypeID(len(types)) {
 			// We've already inflated this type, fix it up immediately.
 			*typ = types[idx]
-			return
+			return true
 		}
+
 		fixups = append(fixups, fixupDef{id, typ})
+		return false
 	}
 
 	type assertion struct {
+		id   TypeID
 		typ  *Type
 		want reflect.Type
 	}
 
 	var assertions []assertion
-	assert := func(typ *Type, want reflect.Type) error {
+	fixupAndAssert := func(id TypeID, typ *Type, want reflect.Type) error {
-		if *typ != nil {
+		if !fixup(id, typ) {
-			// The type has already been fixed up, check the type immediately.
+			assertions = append(assertions, assertion{id, typ, want})
-			if reflect.TypeOf(*typ) != want {
-				return fmt.Errorf("expected %s, got %T", want, *typ)
-			}
 			return nil
 		}
-		assertions = append(assertions, assertion{typ, want})
+
+		// The type has already been fixed up, check the type immediately.
+		if reflect.TypeOf(*typ) != want {
+			return fmt.Errorf("type ID %d: expected %s, got %T", id, want, *typ)
+		}
 		return nil
 	}
 
@@ -903,12 +876,17 @@ func inflateRawTypes(rawTypes []rawType, baseTypes types, rawStrings *stringTabl
 		return members, nil
 	}
 
-	for i, raw := range rawTypes {
+	var declTags []*declTag
+	for _, raw := range rawTypes {
 		var (
-			id  = typeIDOffset + TypeID(i)
+			id  = firstTypeID + TypeID(len(types))
 			typ Type
 		)
 
+		if id < firstTypeID {
+			return nil, fmt.Errorf("no more type IDs")
+		}
+
 		name, err := rawStrings.Lookup(raw.NameOff)
 		if err != nil {
 			return nil, fmt.Errorf("get name for type id %d: %w", id, err)
@@ -936,14 +914,14 @@ func inflateRawTypes(rawTypes []rawType, baseTypes types, rawStrings *stringTabl
 			typ = arr
 
 		case kindStruct:
-			members, err := convertMembers(raw.data.([]btfMember), raw.KindFlag())
+			members, err := convertMembers(raw.data.([]btfMember), raw.Bitfield())
 			if err != nil {
 				return nil, fmt.Errorf("struct %s (id %d): %w", name, id, err)
 			}
 			typ = &Struct{name, raw.Size(), members}
 
 		case kindUnion:
-			members, err := convertMembers(raw.data.([]btfMember), raw.KindFlag())
+			members, err := convertMembers(raw.data.([]btfMember), raw.Bitfield())
 			if err != nil {
 				return nil, fmt.Errorf("union %s (id %d): %w", name, id, err)
 			}
@@ -952,24 +930,23 @@ func inflateRawTypes(rawTypes []rawType, baseTypes types, rawStrings *stringTabl
 		case kindEnum:
 			rawvals := raw.data.([]btfEnum)
 			vals := make([]EnumValue, 0, len(rawvals))
+			signed := raw.Signed()
 			for i, btfVal := range rawvals {
 				name, err := rawStrings.Lookup(btfVal.NameOff)
 				if err != nil {
 					return nil, fmt.Errorf("get name for enum value %d: %s", i, err)
 				}
-				vals = append(vals, EnumValue{
+				value := uint64(btfVal.Val)
-					Name:  name,
+				if signed {
-					Value: btfVal.Val,
+					// Sign extend values to 64 bit.
-				})
+					value = uint64(int32(btfVal.Val))
 				}
-			typ = &Enum{name, raw.Size(), vals}
+				vals = append(vals, EnumValue{name, value})
+			}
+			typ = &Enum{name, raw.Size(), signed, vals}
 
 		case kindForward:
-			if raw.KindFlag() {
+			typ = &Fwd{name, raw.FwdKind()}
-				typ = &Fwd{name, FwdUnion}
-			} else {
-				typ = &Fwd{name, FwdStruct}
-			}
 
 		case kindTypedef:
 			typedef := &Typedef{name, nil}
@@ -993,8 +970,7 @@ func inflateRawTypes(rawTypes []rawType, baseTypes types, rawStrings *stringTabl
 
 		case kindFunc:
 			fn := &Func{name, nil, raw.Linkage()}
-			fixup(raw.Type(), &fn.Type)
+			if err := fixupAndAssert(raw.Type(), &fn.Type, reflect.TypeOf((*FuncProto)(nil))); err != nil {
-			if err := assert(&fn.Type, reflect.TypeOf((*FuncProto)(nil))); err != nil {
 				return nil, err
 			}
 			typ = fn
@@ -1036,15 +1012,42 @@ func inflateRawTypes(rawTypes []rawType, baseTypes types, rawStrings *stringTabl
 			}
 			for i := range vars {
 				fixup(btfVars[i].Type, &vars[i].Type)
-				if err := assert(&vars[i].Type, reflect.TypeOf((*Var)(nil))); err != nil {
-					return nil, err
 			}
-			}
+			typ = &Datasec{name, raw.Size(), vars}
-			typ = &Datasec{name, raw.SizeType, vars}
 
 		case kindFloat:
 			typ = &Float{name, raw.Size()}
 
+		case kindDeclTag:
+			btfIndex := raw.data.(*btfDeclTag).ComponentIdx
+			if uint64(btfIndex) > math.MaxInt {
+				return nil, fmt.Errorf("type id %d: index exceeds int", id)
+			}
+
+			dt := &declTag{nil, name, int(int32(btfIndex))}
+			fixup(raw.Type(), &dt.Type)
+			typ = dt
+
+			declTags = append(declTags, dt)
+
+		case kindTypeTag:
+			tt := &typeTag{nil, name}
+			fixup(raw.Type(), &tt.Type)
+			typ = tt
+
+		case kindEnum64:
+			rawvals := raw.data.([]btfEnum64)
+			vals := make([]EnumValue, 0, len(rawvals))
+			for i, btfVal := range rawvals {
+				name, err := rawStrings.Lookup(btfVal.NameOff)
+				if err != nil {
+					return nil, fmt.Errorf("get name for enum64 value %d: %s", i, err)
+				}
+				value := (uint64(btfVal.ValHi32) << 32) | uint64(btfVal.ValLo32)
+				vals = append(vals, EnumValue{name, value})
+			}
+			typ = &Enum{name, raw.Size(), raw.Signed(), vals}
+
 		default:
 			return nil, fmt.Errorf("type id %d: unknown kind: %v", id, raw.Kind())
 		}
@@ -1053,19 +1056,20 @@ func inflateRawTypes(rawTypes []rawType, baseTypes types, rawStrings *stringTabl
 	}
 
 	for _, fixup := range fixups {
-		i := int(fixup.id)
+		if fixup.id < firstTypeID {
-		if i >= len(types)+len(baseTypes) {
+			return nil, fmt.Errorf("fixup for base type id %d is not expected", fixup.id)
-			return nil, fmt.Errorf("reference to invalid type id: %d", fixup.id)
-		}
-		if i < len(baseTypes) {
-			return nil, fmt.Errorf("fixup for base type id %d is not expected", i)
 		}
 
-		*fixup.typ = types[i-len(baseTypes)]
+		idx := int(fixup.id - firstTypeID)
+		if idx >= len(types) {
+			return nil, fmt.Errorf("reference to invalid type id: %d", fixup.id)
+		}
+
+		*fixup.typ = types[idx]
 	}
 
 	for _, bitfieldFixup := range bitfieldFixups {
-		if bitfieldFixup.id < TypeID(len(baseTypes)) {
+		if bitfieldFixup.id < firstTypeID {
 			return nil, fmt.Errorf("bitfield fixup from split to base types is not expected")
 		}
 
@@ -1079,7 +1083,29 @@ func inflateRawTypes(rawTypes []rawType, baseTypes types, rawStrings *stringTabl
 
 	for _, assertion := range assertions {
 		if reflect.TypeOf(*assertion.typ) != assertion.want {
-			return nil, fmt.Errorf("expected %s, got %T", assertion.want, *assertion.typ)
+			return nil, fmt.Errorf("type ID %d: expected %s, got %T", assertion.id, assertion.want, *assertion.typ)
+		}
+	}
+
+	for _, dt := range declTags {
+		switch t := dt.Type.(type) {
+		case *Var, *Typedef:
+			if dt.Index != -1 {
+				return nil, fmt.Errorf("type %s: index %d is not -1", dt, dt.Index)
+			}
+
+		case composite:
+			if dt.Index >= len(t.members()) {
+				return nil, fmt.Errorf("type %s: index %d exceeds members of %s", dt, dt.Index, t)
+			}
+
+		case *Func:
+			if dt.Index >= len(t.Type.(*FuncProto).Params) {
+				return nil, fmt.Errorf("type %s: index %d exceeds params of %s", dt, dt.Index, t)
+			}
+
+		default:
+			return nil, fmt.Errorf("type %s: decl tag for type %s is not supported", dt, t)
 		}
 	}
 
@@ -1123,6 +1149,29 @@ func UnderlyingType(typ Type) Type {
 	return &cycle{typ}
 }
 
+// as returns typ if is of type T. Otherwise it peels qualifiers and Typedefs
+// until it finds a T.
+//
+// Returns the zero value and false if there is no T or if the type is nested
+// too deeply.
+func as[T Type](typ Type) (T, bool) {
+	for depth := 0; depth <= maxTypeDepth; depth++ {
+		switch v := (typ).(type) {
+		case T:
+			return v, true
+		case qualifier:
+			typ = v.qualify()
+		case *Typedef:
+			typ = v.Type
+		default:
+			goto notFound
+		}
+	}
+notFound:
+	var zero T
+	return zero, false
+}
+
 type formatState struct {
 	fmt.State
 	depth int
@@ -1145,10 +1194,7 @@ func formatType(f fmt.State, verb rune, t formattableType, extra ...interface{})
 		return
 	}
 
-	// This is the same as %T, but elides the package name. Assumes that
+	_, _ = io.WriteString(f, internal.GoTypeName(t))
-	// formattableType is implemented by a pointer receiver.
-	goTypeName := reflect.TypeOf(t).Elem().Name()
-	_, _ = io.WriteString(f, goTypeName)
 
 	if name := t.TypeName(); name != "" {
 		// Output BTF type name if present.

vendor/github.com/cilium/ebpf/btf/workarounds.go

@@ -0,0 +1,26 @@
+package btf
+
+// datasecResolveWorkaround ensures that certain vars in a Datasec are added
+// to a Spec before the Datasec. This avoids a bug in kernel BTF validation.
+//
+// See https://lore.kernel.org/bpf/20230302123440.1193507-1-lmb@isovalent.com/
+func datasecResolveWorkaround(b *Builder, ds *Datasec) error {
+	for _, vsi := range ds.Vars {
+		v, ok := vsi.Type.(*Var)
+		if !ok {
+			continue
+		}
+
+		switch v.Type.(type) {
+		case *Typedef, *Volatile, *Const, *Restrict, *typeTag:
+			// NB: We must never call Add on a Datasec, otherwise we risk
+			// infinite recursion.
+			_, err := b.Add(v.Type)
+			if err != nil {
+				return err
+			}
+		}
+	}
+
+	return nil
+}

vendor/github.com/cilium/ebpf/collection.go

@@ -9,6 +9,8 @@ import (
 
 	"github.com/cilium/ebpf/asm"
 	"github.com/cilium/ebpf/btf"
+	"github.com/cilium/ebpf/internal"
+	"github.com/cilium/ebpf/internal/kconfig"
 )
 
 // CollectionOptions control loading a collection into the kernel.
@@ -107,6 +109,16 @@ func (cs *CollectionSpec) RewriteMaps(maps map[string]*Map) error {
 	return nil
 }
 
+// MissingConstantsError is returned by [CollectionSpec.RewriteConstants].
+type MissingConstantsError struct {
+	// The constants missing from .rodata.
+	Constants []string
+}
+
+func (m *MissingConstantsError) Error() string {
+	return fmt.Sprintf("some constants are missing from .rodata: %s", strings.Join(m.Constants, ", "))
+}
+
 // RewriteConstants replaces the value of multiple constants.
 //
 // The constant must be defined like so in the C program:
@@ -120,7 +132,7 @@ func (cs *CollectionSpec) RewriteMaps(maps map[string]*Map) error {
 //
 // From Linux 5.5 the verifier will use constants to eliminate dead code.
 //
-// Returns an error if a constant doesn't exist.
+// Returns an error wrapping [MissingConstantsError] if a constant doesn't exist.
 func (cs *CollectionSpec) RewriteConstants(consts map[string]interface{}) error {
 	replaced := make(map[string]bool)
 
@@ -151,6 +163,10 @@ func (cs *CollectionSpec) RewriteConstants(consts map[string]interface{}) error
 				continue
 			}
 
+			if _, ok := v.Type.(*btf.Var); !ok {
+				return fmt.Errorf("section %s: unexpected type %T for variable %s", name, v.Type, vname)
+			}
+
 			if replaced[vname] {
 				return fmt.Errorf("section %s: duplicate variable %s", name, vname)
 			}
@@ -180,7 +196,7 @@ func (cs *CollectionSpec) RewriteConstants(consts map[string]interface{}) error
 	}
 
 	if len(missing) != 0 {
-		return fmt.Errorf("spec is missing one or more constants: %s", strings.Join(missing, ","))
+		return fmt.Errorf("rewrite constants: %w", &MissingConstantsError{Constants: missing})
 	}
 
 	return nil
@@ -386,42 +402,11 @@ func NewCollectionWithOptions(spec *CollectionSpec, opts CollectionOptions) (*Co
 	}, nil
 }
 
-type handleCache struct {
-	btfHandles map[*btf.Spec]*btf.Handle
-}
-
-func newHandleCache() *handleCache {
-	return &handleCache{
-		btfHandles: make(map[*btf.Spec]*btf.Handle),
-	}
-}
-
-func (hc handleCache) btfHandle(spec *btf.Spec) (*btf.Handle, error) {
-	if hc.btfHandles[spec] != nil {
-		return hc.btfHandles[spec], nil
-	}
-
-	handle, err := btf.NewHandle(spec)
-	if err != nil {
-		return nil, err
-	}
-
-	hc.btfHandles[spec] = handle
-	return handle, nil
-}
-
-func (hc handleCache) close() {
-	for _, handle := range hc.btfHandles {
-		handle.Close()
-	}
-}
-
 type collectionLoader struct {
 	coll     *CollectionSpec
 	opts     *CollectionOptions
 	maps     map[string]*Map
 	programs map[string]*Program
-	handles  *handleCache
 }
 
 func newCollectionLoader(coll *CollectionSpec, opts *CollectionOptions) (*collectionLoader, error) {
@@ -436,7 +421,7 @@ func newCollectionLoader(coll *CollectionSpec, opts *CollectionOptions) (*collec
 			return nil, fmt.Errorf("replacement map %s not found in CollectionSpec", name)
 		}
 
-		if err := spec.checkCompatibility(m); err != nil {
+		if err := spec.Compatible(m); err != nil {
 			return nil, fmt.Errorf("using replacement map %s: %w", spec.Name, err)
 		}
 	}
@@ -446,13 +431,11 @@ func newCollectionLoader(coll *CollectionSpec, opts *CollectionOptions) (*collec
 		opts,
 		make(map[string]*Map),
 		make(map[string]*Program),
-		newHandleCache(),
 	}, nil
 }
 
 // close all resources left over in the collectionLoader.
 func (cl *collectionLoader) close() {
-	cl.handles.close()
 	for _, m := range cl.maps {
 		m.Close()
 	}
@@ -471,10 +454,6 @@ func (cl *collectionLoader) loadMap(mapName string) (*Map, error) {
 		return nil, fmt.Errorf("missing map %s", mapName)
 	}
 
-	if mapSpec.BTF != nil && cl.coll.Types != mapSpec.BTF {
-		return nil, fmt.Errorf("map %s: BTF doesn't match collection", mapName)
-	}
-
 	if replaceMap, ok := cl.opts.MapReplacements[mapName]; ok {
 		// Clone the map to avoid closing user's map later on.
 		m, err := replaceMap.Clone()
@@ -486,7 +465,7 @@ func (cl *collectionLoader) loadMap(mapName string) (*Map, error) {
 		return m, nil
 	}
 
-	m, err := newMapWithOptions(mapSpec, cl.opts.Maps, cl.handles)
+	m, err := newMapWithOptions(mapSpec, cl.opts.Maps)
 	if err != nil {
 		return nil, fmt.Errorf("map %s: %w", mapName, err)
 	}
@@ -511,10 +490,6 @@ func (cl *collectionLoader) loadProgram(progName string) (*Program, error) {
 		return nil, fmt.Errorf("cannot load program %s: program type is unspecified", progName)
 	}
 
-	if progSpec.BTF != nil && cl.coll.Types != progSpec.BTF {
-		return nil, fmt.Errorf("program %s: BTF doesn't match collection", progName)
-	}
-
 	progSpec = progSpec.Copy()
 
 	// Rewrite any reference to a valid map in the program's instructions,
@@ -543,7 +518,7 @@ func (cl *collectionLoader) loadProgram(progName string) (*Program, error) {
 		}
 	}
 
-	prog, err := newProgramWithOptions(progSpec, cl.opts.Programs, cl.handles)
+	prog, err := newProgramWithOptions(progSpec, cl.opts.Programs)
 	if err != nil {
 		return nil, fmt.Errorf("program %s: %w", progName, err)
 	}
@@ -559,17 +534,22 @@ func (cl *collectionLoader) populateMaps() error {
 			return fmt.Errorf("missing map spec %s", mapName)
 		}
 
-		mapSpec = mapSpec.Copy()
-
 		// MapSpecs that refer to inner maps or programs within the same
 		// CollectionSpec do so using strings. These strings are used as the key
 		// to look up the respective object in the Maps or Programs fields.
 		// Resolve those references to actual Map or Program resources that
 		// have been loaded into the kernel.
+		if mapSpec.Type.canStoreMap() || mapSpec.Type.canStoreProgram() {
+			mapSpec = mapSpec.Copy()
+
 			for i, kv := range mapSpec.Contents {
-			if objName, ok := kv.Value.(string); ok {
+				objName, ok := kv.Value.(string)
-				switch mapSpec.Type {
+				if !ok {
-				case ProgramArray:
+					continue
+				}
+
+				switch t := mapSpec.Type; {
+				case t.canStoreProgram():
 					// loadProgram is idempotent and could return an existing Program.
 					prog, err := cl.loadProgram(objName)
 					if err != nil {
@@ -577,7 +557,7 @@ func (cl *collectionLoader) populateMaps() error {
 					}
 					mapSpec.Contents[i] = MapKV{kv.Key, prog}
 
-				case ArrayOfMaps, HashOfMaps:
+				case t.canStoreMap():
 					// loadMap is idempotent and could return an existing Map.
 					innerMap, err := cl.loadMap(objName)
 					if err != nil {
@@ -597,6 +577,95 @@ func (cl *collectionLoader) populateMaps() error {
 	return nil
 }
 
+// resolveKconfig resolves all variables declared in .kconfig and populates
+// m.Contents. Does nothing if the given m.Contents is non-empty.
+func resolveKconfig(m *MapSpec) error {
+	ds, ok := m.Value.(*btf.Datasec)
+	if !ok {
+		return errors.New("map value is not a Datasec")
+	}
+
+	type configInfo struct {
+		offset uint32
+		typ    btf.Type
+	}
+
+	configs := make(map[string]configInfo)
+
+	data := make([]byte, ds.Size)
+	for _, vsi := range ds.Vars {
+		v := vsi.Type.(*btf.Var)
+		n := v.TypeName()
+
+		switch n {
+		case "LINUX_KERNEL_VERSION":
+			if integer, ok := v.Type.(*btf.Int); !ok || integer.Size != 4 {
+				return fmt.Errorf("variable %s must be a 32 bits integer, got %s", n, v.Type)
+			}
+
+			kv, err := internal.KernelVersion()
+			if err != nil {
+				return fmt.Errorf("getting kernel version: %w", err)
+			}
+			internal.NativeEndian.PutUint32(data[vsi.Offset:], kv.Kernel())
+
+		case "LINUX_HAS_SYSCALL_WRAPPER":
+			if integer, ok := v.Type.(*btf.Int); !ok || integer.Size != 4 {
+				return fmt.Errorf("variable %s must be a 32 bits integer, got %s", n, v.Type)
+			}
+			var value uint32 = 1
+			if err := haveSyscallWrapper(); errors.Is(err, ErrNotSupported) {
+				value = 0
+			} else if err != nil {
+				return fmt.Errorf("unable to derive a value for LINUX_HAS_SYSCALL_WRAPPER: %w", err)
+			}
+
+			internal.NativeEndian.PutUint32(data[vsi.Offset:], value)
+
+		default: // Catch CONFIG_*.
+			configs[n] = configInfo{
+				offset: vsi.Offset,
+				typ:    v.Type,
+			}
+		}
+	}
+
+	// We only parse kconfig file if a CONFIG_* variable was found.
+	if len(configs) > 0 {
+		f, err := kconfig.Find()
+		if err != nil {
+			return fmt.Errorf("cannot find a kconfig file: %w", err)
+		}
+		defer f.Close()
+
+		filter := make(map[string]struct{}, len(configs))
+		for config := range configs {
+			filter[config] = struct{}{}
+		}
+
+		kernelConfig, err := kconfig.Parse(f, filter)
+		if err != nil {
+			return fmt.Errorf("cannot parse kconfig file: %w", err)
+		}
+
+		for n, info := range configs {
+			value, ok := kernelConfig[n]
+			if !ok {
+				return fmt.Errorf("config option %q does not exists for this kernel", n)
+			}
+
+			err := kconfig.PutValue(data[info.offset:], info.typ, value)
+			if err != nil {
+				return fmt.Errorf("problem adding value for %s: %w", n, err)
+			}
+		}
+	}
+
+	m.Contents = []MapKV{{uint32(0), data}}
+
+	return nil
+}
+
 // LoadCollection reads an object file and creates and loads its declared
 // resources into the kernel.
 //

vendor/github.com/cilium/ebpf/elf_reader.go

@@ -18,6 +18,15 @@ import (
 	"github.com/cilium/ebpf/internal/unix"
 )
 
+type kconfigMetaKey struct{}
+
+type kconfigMeta struct {
+	Map    *MapSpec
+	Offset uint32
+}
+
+type kfuncMeta struct{}
+
 // elfCode is a convenience to reduce the amount of arguments that have to
 // be passed around explicitly. You should treat its contents as immutable.
 type elfCode struct {
@@ -27,6 +36,9 @@ type elfCode struct {
 	version  uint32
 	btf      *btf.Spec
 	extInfo  *btf.ExtInfos
+	maps     map[string]*MapSpec
+	kfuncs   map[string]*btf.Func
+	kconfig  *MapSpec
 }
 
 // LoadCollectionSpec parses an ELF file into a CollectionSpec.
@@ -51,6 +63,12 @@ func LoadCollectionSpecFromReader(rd io.ReaderAt) (*CollectionSpec, error) {
 		return nil, err
 	}
 
+	// Checks if the ELF file is for BPF data.
+	// Old LLVM versions set e_machine to EM_NONE.
+	if f.File.Machine != unix.EM_NONE && f.File.Machine != elf.EM_BPF {
+		return nil, fmt.Errorf("unexpected machine type for BPF ELF: %s", f.File.Machine)
+	}
+
 	var (
 		licenseSection *elf.Section
 		versionSection *elf.Section
@@ -107,6 +125,8 @@ func LoadCollectionSpecFromReader(rd io.ReaderAt) (*CollectionSpec, error) {
 		version:     version,
 		btf:         btfSpec,
 		extInfo:     btfExtInfo,
+		maps:        make(map[string]*MapSpec),
+		kfuncs:      make(map[string]*btf.Func),
 	}
 
 	symbols, err := f.Symbols()
@@ -120,27 +140,33 @@ func LoadCollectionSpecFromReader(rd io.ReaderAt) (*CollectionSpec, error) {
 		return nil, fmt.Errorf("load relocations: %w", err)
 	}
 
-	// Collect all the various ways to define maps.
+	if err := ec.loadMaps(); err != nil {
-	maps := make(map[string]*MapSpec)
-	if err := ec.loadMaps(maps); err != nil {
 		return nil, fmt.Errorf("load maps: %w", err)
 	}
 
-	if err := ec.loadBTFMaps(maps); err != nil {
+	if err := ec.loadBTFMaps(); err != nil {
 		return nil, fmt.Errorf("load BTF maps: %w", err)
 	}
 
-	if err := ec.loadDataSections(maps); err != nil {
+	if err := ec.loadDataSections(); err != nil {
 		return nil, fmt.Errorf("load data sections: %w", err)
 	}
 
+	if err := ec.loadKconfigSection(); err != nil {
+		return nil, fmt.Errorf("load virtual .kconfig section: %w", err)
+	}
+
+	if err := ec.loadKsymsSection(); err != nil {
+		return nil, fmt.Errorf("load virtual .ksyms section: %w", err)
+	}
+
 	// Finally, collect programs and link them.
 	progs, err := ec.loadProgramSections()
 	if err != nil {
 		return nil, fmt.Errorf("load programs: %w", err)
 	}
 
-	return &CollectionSpec{maps, progs, btfSpec, ec.ByteOrder}, nil
+	return &CollectionSpec{ec.maps, progs, btfSpec, ec.ByteOrder}, nil
 }
 
 func loadLicense(sec *elf.Section) (string, error) {
@@ -261,10 +287,6 @@ func (ec *elfCode) loadRelocations(relSections map[elf.SectionIndex]*elf.Section
 				return fmt.Errorf("section %q: reference to %q in section %s: %w", section.Name, rel.Name, rel.Section, ErrNotSupported)
 			}
 
-			if target.Flags&elf.SHF_STRINGS > 0 {
-				return fmt.Errorf("section %q: string is not stack allocated: %w", section.Name, ErrNotSupported)
-			}
-
 			target.references++
 		}
 
@@ -312,7 +334,6 @@ func (ec *elfCode) loadProgramSections() (map[string]*ProgramSpec, error) {
 				KernelVersion: ec.version,
 				Instructions:  insns,
 				ByteOrder:     ec.ByteOrder,
-				BTF:           ec.btf,
 			}
 
 			// Function names must be unique within a single ELF blob.
@@ -565,6 +586,10 @@ func (ec *elfCode) relocateInstruction(ins *asm.Instruction, rel elf.Symbol) err
 			return fmt.Errorf("neither a call nor a load instruction: %v", ins)
 		}
 
+	// The Undefined section is used for 'virtual' symbols that aren't backed by
+	// an ELF section. This includes symbol references from inline asm, forward
+	// function declarations, as well as extern kfunc declarations using __ksym
+	// and extern kconfig variables declared using __kconfig.
 	case undefSection:
 		if bind != elf.STB_GLOBAL {
 			return fmt.Errorf("asm relocation: %s: unsupported binding: %s", name, bind)
@@ -574,7 +599,36 @@ func (ec *elfCode) relocateInstruction(ins *asm.Instruction, rel elf.Symbol) err
 			return fmt.Errorf("asm relocation: %s: unsupported type %s", name, typ)
 		}
 
-		// There is nothing to do here but set ins.Reference.
+		kf := ec.kfuncs[name]
+		switch {
+		// If a Call instruction is found and the datasec has a btf.Func with a Name
+		// that matches the symbol name we mark the instruction as a call to a kfunc.
+		case kf != nil && ins.OpCode.JumpOp() == asm.Call:
+			ins.Metadata.Set(kfuncMeta{}, kf)
+			ins.Src = asm.PseudoKfuncCall
+			ins.Constant = -1
+
+		// If no kconfig map is found, this must be a symbol reference from inline
+		// asm (see testdata/loader.c:asm_relocation()) or a call to a forward
+		// function declaration (see testdata/fwd_decl.c). Don't interfere, These
+		// remain standard symbol references.
+		// extern __kconfig reads are represented as dword loads that need to be
+		// rewritten to pseudo map loads from .kconfig. If the map is present,
+		// require it to contain the symbol to disambiguate between inline asm
+		// relos and kconfigs.
+		case ec.kconfig != nil && ins.OpCode.IsDWordLoad():
+			for _, vsi := range ec.kconfig.Value.(*btf.Datasec).Vars {
+				if vsi.Type.(*btf.Var).Name != rel.Name {
+					continue
+				}
+
+				ins.Src = asm.PseudoMapValue
+				ins.Metadata.Set(kconfigMetaKey{}, &kconfigMeta{ec.kconfig, vsi.Offset})
+				return nil
+			}
+
+			return fmt.Errorf("kconfig %s not found in .kconfig", rel.Name)
+		}
 
 	default:
 		return fmt.Errorf("relocation to %q: %w", target.Name, ErrNotSupported)
@@ -584,7 +638,7 @@ func (ec *elfCode) relocateInstruction(ins *asm.Instruction, rel elf.Symbol) err
 	return nil
 }
 
-func (ec *elfCode) loadMaps(maps map[string]*MapSpec) error {
+func (ec *elfCode) loadMaps() error {
 	for _, sec := range ec.sections {
 		if sec.kind != mapSection {
 			continue
@@ -610,7 +664,7 @@ func (ec *elfCode) loadMaps(maps map[string]*MapSpec) error {
 			}
 
 			mapName := mapSym.Name
-			if maps[mapName] != nil {
+			if ec.maps[mapName] != nil {
 				return fmt.Errorf("section %v: map %v already exists", sec.Name, mapSym)
 			}
 
@@ -644,7 +698,7 @@ func (ec *elfCode) loadMaps(maps map[string]*MapSpec) error {
 				return fmt.Errorf("map %s: %w", mapName, err)
 			}
 
-			maps[mapName] = &spec
+			ec.maps[mapName] = &spec
 		}
 	}
 
@@ -654,7 +708,7 @@ func (ec *elfCode) loadMaps(maps map[string]*MapSpec) error {
 // loadBTFMaps iterates over all ELF sections marked as BTF map sections
 // (like .maps) and parses them into MapSpecs. Dump the .maps section and
 // any relocations with `readelf -x .maps -r <elf_file>`.
-func (ec *elfCode) loadBTFMaps(maps map[string]*MapSpec) error {
+func (ec *elfCode) loadBTFMaps() error {
 	for _, sec := range ec.sections {
 		if sec.kind != btfMapSection {
 			continue
@@ -693,7 +747,7 @@ func (ec *elfCode) loadBTFMaps(maps map[string]*MapSpec) error {
 				return fmt.Errorf("section %v: map %s: initializing BTF map definitions: %w", sec.Name, name, internal.ErrNotSupported)
 			}
 
-			if maps[name] != nil {
+			if ec.maps[name] != nil {
 				return fmt.Errorf("section %v: map %s already exists", sec.Name, name)
 			}
 
@@ -712,7 +766,7 @@ func (ec *elfCode) loadBTFMaps(maps map[string]*MapSpec) error {
 				return fmt.Errorf("map %v: %w", name, err)
 			}
 
-			maps[name] = mapSpec
+			ec.maps[name] = mapSpec
 		}
 
 		// Drain the ELF section reader to make sure all bytes are accounted for
@@ -901,13 +955,6 @@ func mapSpecFromBTF(es *elfSection, vs *btf.VarSecinfo, def *btf.Struct, spec *b
 		}
 	}
 
-	if key == nil {
-		key = &btf.Void{}
-	}
-	if value == nil {
-		value = &btf.Void{}
-	}
-
 	return &MapSpec{
 		Name:       SanitizeName(name, -1),
 		Type:       MapType(mapType),
@@ -917,7 +964,6 @@ func mapSpecFromBTF(es *elfSection, vs *btf.VarSecinfo, def *btf.Struct, spec *b
 		Flags:      flags,
 		Key:        key,
 		Value:      value,
-		BTF:        spec,
 		Pinning:    pinType,
 		InnerMap:   innerMapSpec,
 		Contents:   contents,
@@ -1008,14 +1054,14 @@ func resolveBTFValuesContents(es *elfSection, vs *btf.VarSecinfo, member btf.Mem
 		case elf.STT_OBJECT:
 			contents = append(contents, MapKV{uint32(k), r.Name})
 		default:
-			return nil, fmt.Errorf("unknown relocation type %v", t)
+			return nil, fmt.Errorf("unknown relocation type %v for symbol %s", t, r.Name)
 		}
 	}
 
 	return contents, nil
 }
 
-func (ec *elfCode) loadDataSections(maps map[string]*MapSpec) error {
+func (ec *elfCode) loadDataSections() error {
 	for _, sec := range ec.sections {
 		if sec.kind != dataSection {
 			continue
@@ -1027,6 +1073,17 @@ func (ec *elfCode) loadDataSections(maps map[string]*MapSpec) error {
 			continue
 		}
 
+		mapSpec := &MapSpec{
+			Name:       SanitizeName(sec.Name, -1),
+			Type:       Array,
+			KeySize:    4,
+			ValueSize:  uint32(sec.Size),
+			MaxEntries: 1,
+		}
+
+		switch sec.Type {
+		// Only open the section if we know there's actual data to be read.
+		case elf.SHT_PROGBITS:
 			data, err := sec.Data()
 			if err != nil {
 				return fmt.Errorf("data section %s: can't get contents: %w", sec.Name, err)
@@ -1035,14 +1092,14 @@ func (ec *elfCode) loadDataSections(maps map[string]*MapSpec) error {
 			if uint64(len(data)) > math.MaxUint32 {
 				return fmt.Errorf("data section %s: contents exceed maximum size", sec.Name)
 			}
+			mapSpec.Contents = []MapKV{{uint32(0), data}}
 
-		mapSpec := &MapSpec{
+		case elf.SHT_NOBITS:
-			Name:       SanitizeName(sec.Name, -1),
+			// NOBITS sections like .bss contain only zeroes, and since data sections
-			Type:       Array,
+			// are Arrays, the kernel already preallocates them. Skip reading zeroes
-			KeySize:    4,
+			// from the ELF.
-			ValueSize:  uint32(len(data)),
+		default:
-			MaxEntries: 1,
+			return fmt.Errorf("data section %s: unknown section type %s", sec.Name, sec.Type)
-			Contents:   []MapKV{{uint32(0), data}},
 		}
 
 		// It is possible for a data section to exist without a corresponding BTF Datasec
@@ -1051,23 +1108,78 @@ func (ec *elfCode) loadDataSections(maps map[string]*MapSpec) error {
 			var ds *btf.Datasec
 			if ec.btf.TypeByName(sec.Name, &ds) == nil {
 				// Assign the spec's key and BTF only if the Datasec lookup was successful.
-				mapSpec.BTF = ec.btf
 				mapSpec.Key = &btf.Void{}
 				mapSpec.Value = ds
 			}
 		}
 
-		switch n := sec.Name; {
+		if strings.HasPrefix(sec.Name, ".rodata") {
-		case strings.HasPrefix(n, ".rodata"):
 			mapSpec.Flags = unix.BPF_F_RDONLY_PROG
 			mapSpec.Freeze = true
-		case n == ".bss":
-			// The kernel already zero-initializes the map
-			mapSpec.Contents = nil
 		}
 
-		maps[sec.Name] = mapSpec
+		ec.maps[sec.Name] = mapSpec
 	}
+
+	return nil
+}
+
+// loadKconfigSection handles the 'virtual' Datasec .kconfig that doesn't
+// have a corresponding ELF section and exist purely in BTF.
+func (ec *elfCode) loadKconfigSection() error {
+	if ec.btf == nil {
+		return nil
+	}
+
+	var ds *btf.Datasec
+	err := ec.btf.TypeByName(".kconfig", &ds)
+	if errors.Is(err, btf.ErrNotFound) {
+		return nil
+	}
+	if err != nil {
+		return err
+	}
+
+	if ds.Size == 0 {
+		return errors.New("zero-length .kconfig")
+	}
+
+	ec.kconfig = &MapSpec{
+		Name:       ".kconfig",
+		Type:       Array,
+		KeySize:    uint32(4),
+		ValueSize:  ds.Size,
+		MaxEntries: 1,
+		Flags:      unix.BPF_F_RDONLY_PROG | unix.BPF_F_MMAPABLE,
+		Freeze:     true,
+		Key:        &btf.Int{Size: 4},
+		Value:      ds,
+	}
+
+	return nil
+}
+
+// loadKsymsSection handles the 'virtual' Datasec .ksyms that doesn't
+// have a corresponding ELF section and exist purely in BTF.
+func (ec *elfCode) loadKsymsSection() error {
+	if ec.btf == nil {
+		return nil
+	}
+
+	var ds *btf.Datasec
+	err := ec.btf.TypeByName(".ksyms", &ds)
+	if errors.Is(err, btf.ErrNotFound) {
+		return nil
+	}
+	if err != nil {
+		return err
+	}
+
+	for _, v := range ds.Vars {
+		// we have already checked the .ksyms Datasec to only contain Func Vars.
+		ec.kfuncs[v.Type.TypeName()] = v.Type.(*btf.Func)
+	}
+
 	return nil
 }
 
@@ -1107,9 +1219,13 @@ func getProgType(sectionName string) (ProgramType, AttachType, uint32, string) {
 		{"lsm/", LSM, AttachLSMMac, 0},
 		{"lsm.s/", LSM, AttachLSMMac, unix.BPF_F_SLEEPABLE},
 		{"iter/", Tracing, AttachTraceIter, 0},
+		{"iter.s/", Tracing, AttachTraceIter, unix.BPF_F_SLEEPABLE},
 		{"syscall", Syscall, AttachNone, 0},
+		{"xdp.frags_devmap/", XDP, AttachXDPDevMap, unix.BPF_F_XDP_HAS_FRAGS},
 		{"xdp_devmap/", XDP, AttachXDPDevMap, 0},
+		{"xdp.frags_cpumap/", XDP, AttachXDPCPUMap, unix.BPF_F_XDP_HAS_FRAGS},
 		{"xdp_cpumap/", XDP, AttachXDPCPUMap, 0},
+		{"xdp.frags", XDP, AttachNone, unix.BPF_F_XDP_HAS_FRAGS},
 		{"xdp", XDP, AttachNone, 0},
 		{"perf_event", PerfEvent, AttachNone, 0},
 		{"lwt_in", LWTIn, AttachNone, 0},
@@ -1149,8 +1265,9 @@ func getProgType(sectionName string) (ProgramType, AttachType, uint32, string) {
 		{"cgroup/setsockopt", CGroupSockopt, AttachCGroupSetsockopt, 0},
 		{"struct_ops+", StructOps, AttachNone, 0},
 		{"sk_lookup/", SkLookup, AttachSkLookup, 0},
-
 		{"seccomp", SocketFilter, AttachNone, 0},
+		{"kprobe.multi", Kprobe, AttachTraceKprobeMulti, 0},
+		{"kretprobe.multi", Kprobe, AttachTraceKprobeMulti, 0},
 	}
 
 	for _, t := range types {

vendor/github.com/cilium/ebpf/info.go

@@ -48,7 +48,7 @@ func newMapInfoFromFd(fd *sys.FD) (*MapInfo, error) {
 		info.KeySize,
 		info.ValueSize,
 		info.MaxEntries,
-		info.MapFlags,
+		uint32(info.MapFlags),
 		unix.ByteSliceToString(info.Name[:]),
 	}, nil
 }
@@ -94,6 +94,8 @@ type ProgramInfo struct {
 	// Name as supplied by user space at load time. Available from 4.15.
 	Name string
 
+	createdByUID     uint32
+	haveCreatedByUID bool
 	btf              btf.ID
 	stats            *programStats
 
@@ -130,6 +132,18 @@ func newProgramInfoFromFd(fd *sys.FD) (*ProgramInfo, error) {
 		pi.maps = make([]MapID, info.NrMapIds)
 		info2.NrMapIds = info.NrMapIds
 		info2.MapIds = sys.NewPointer(unsafe.Pointer(&pi.maps[0]))
+	} else if haveProgramInfoMapIDs() == nil {
+		// This program really has no associated maps.
+		pi.maps = make([]MapID, 0)
+	} else {
+		// The kernel doesn't report associated maps.
+		pi.maps = nil
+	}
+
+	// createdByUID and NrMapIds were introduced in the same kernel version.
+	if pi.maps != nil {
+		pi.createdByUID = info.CreatedByUid
+		pi.haveCreatedByUID = true
 	}
 
 	if info.XlatedProgLen > 0 {
@@ -175,6 +189,15 @@ func (pi *ProgramInfo) ID() (ProgramID, bool) {
 	return pi.id, pi.id > 0
 }
 
+// CreatedByUID returns the Uid that created the program.
+//
+// Available from 4.15.
+//
+// The bool return value indicates whether this optional field is available.
+func (pi *ProgramInfo) CreatedByUID() (uint32, bool) {
+	return pi.createdByUID, pi.haveCreatedByUID
+}
+
 // BTFID returns the BTF ID associated with the program.
 //
 // The ID is only valid as long as the associated program is kept alive.
@@ -321,3 +344,30 @@ func EnableStats(which uint32) (io.Closer, error) {
 	}
 	return fd, nil
 }
+
+var haveProgramInfoMapIDs = internal.NewFeatureTest("map IDs in program info", "4.15", func() error {
+	prog, err := progLoad(asm.Instructions{
+		asm.LoadImm(asm.R0, 0, asm.DWord),
+		asm.Return(),
+	}, SocketFilter, "MIT")
+	if err != nil {
+		return err
+	}
+	defer prog.Close()
+
+	err = sys.ObjInfo(prog, &sys.ProgInfo{
+		// NB: Don't need to allocate MapIds since the program isn't using
+		// any maps.
+		NrMapIds: 1,
+	})
+	if errors.Is(err, unix.EINVAL) {
+		// Most likely the syscall doesn't exist.
+		return internal.ErrNotSupported
+	}
+	if errors.Is(err, unix.E2BIG) {
+		// We've hit check_uarg_tail_zero on older kernels.
+		return internal.ErrNotSupported
+	}
+
+	return err
+})

vendor/github.com/cilium/ebpf/internal/align.go

@@ -1,6 +1,8 @@
 package internal
 
+import "golang.org/x/exp/constraints"
+
 // Align returns 'n' updated to 'alignment' boundary.
-func Align(n, alignment int) int {
+func Align[I constraints.Integer](n, alignment I) I {
-	return (int(n) + alignment - 1) / alignment * alignment
+	return (n + alignment - 1) / alignment * alignment
 }

vendor/github.com/cilium/ebpf/internal/buffer.go

@@ -0,0 +1,31 @@
+package internal
+
+import (
+	"bytes"
+	"sync"
+)
+
+var bytesBufferPool = sync.Pool{
+	New: func() interface{} {
+		return new(bytes.Buffer)
+	},
+}
+
+// NewBuffer retrieves a [bytes.Buffer] from a pool an re-initialises it.
+//
+// The returned buffer should be passed to [PutBuffer].
+func NewBuffer(buf []byte) *bytes.Buffer {
+	wr := bytesBufferPool.Get().(*bytes.Buffer)
+	// Reinitialize the Buffer with a new backing slice since it is returned to
+	// the caller by wr.Bytes() below. Pooling is faster despite calling
+	// NewBuffer. The pooled alloc is still reused, it only needs to be zeroed.
+	*wr = *bytes.NewBuffer(buf)
+	return wr
+}
+
+// PutBuffer releases a buffer to the pool.
+func PutBuffer(buf *bytes.Buffer) {
+	// Release reference to the backing buffer.
+	*buf = *bytes.NewBuffer(nil)
+	bytesBufferPool.Put(buf)
+}

vendor/github.com/cilium/ebpf/internal/cpu.go

@@ -4,24 +4,13 @@ import (
 	"fmt"
 	"os"
 	"strings"
-	"sync"
 )
 
-var sysCPU struct {
-	once sync.Once
-	err  error
-	num  int
-}
-
 // PossibleCPUs returns the max number of CPUs a system may possibly have
 // Logical CPU numbers must be of the form 0-n
-func PossibleCPUs() (int, error) {
+var PossibleCPUs = Memoize(func() (int, error) {
-	sysCPU.once.Do(func() {
+	return parseCPUsFromFile("/sys/devices/system/cpu/possible")
-		sysCPU.num, sysCPU.err = parseCPUsFromFile("/sys/devices/system/cpu/possible")
+})
-	})
-
-	return sysCPU.num, sysCPU.err
-}
 
 func parseCPUsFromFile(path string) (int, error) {
 	spec, err := os.ReadFile(path)

vendor/github.com/cilium/ebpf/internal/deque.go

@@ -0,0 +1,91 @@
+package internal
+
+import "math/bits"
+
+// Deque implements a double ended queue.
+type Deque[T any] struct {
+	elems       []T
+	read, write uint64
+	mask        uint64
+}
+
+// Reset clears the contents of the deque while retaining the backing buffer.
+func (dq *Deque[T]) Reset() {
+	var zero T
+
+	for i := dq.read; i < dq.write; i++ {
+		dq.elems[i&dq.mask] = zero
+	}
+
+	dq.read, dq.write = 0, 0
+}
+
+func (dq *Deque[T]) Empty() bool {
+	return dq.read == dq.write
+}
+
+// Push adds an element to the end.
+func (dq *Deque[T]) Push(e T) {
+	dq.Grow(1)
+	dq.elems[dq.write&dq.mask] = e
+	dq.write++
+}
+
+// Shift returns the first element or the zero value.
+func (dq *Deque[T]) Shift() T {
+	var zero T
+
+	if dq.Empty() {
+		return zero
+	}
+
+	index := dq.read & dq.mask
+	t := dq.elems[index]
+	dq.elems[index] = zero
+	dq.read++
+	return t
+}
+
+// Pop returns the last element or the zero value.
+func (dq *Deque[T]) Pop() T {
+	var zero T
+
+	if dq.Empty() {
+		return zero
+	}
+
+	dq.write--
+	index := dq.write & dq.mask
+	t := dq.elems[index]
+	dq.elems[index] = zero
+	return t
+}
+
+// Grow the deque's capacity, if necessary, to guarantee space for another n
+// elements.
+func (dq *Deque[T]) Grow(n int) {
+	have := dq.write - dq.read
+	need := have + uint64(n)
+	if need < have {
+		panic("overflow")
+	}
+	if uint64(len(dq.elems)) >= need {
+		return
+	}
+
+	// Round up to the new power of two which is at least 8.
+	// See https://jameshfisher.com/2018/03/30/round-up-power-2/
+	capacity := 1 << (64 - bits.LeadingZeros64(need-1))
+	if capacity < 8 {
+		capacity = 8
+	}
+
+	elems := make([]T, have, capacity)
+	pivot := dq.read & dq.mask
+	copied := copy(elems, dq.elems[pivot:])
+	copy(elems[copied:], dq.elems[:pivot])
+
+	dq.elems = elems[:capacity]
+	dq.mask = uint64(capacity) - 1
+	dq.read, dq.write = 0, have
+}

vendor/github.com/cilium/ebpf/internal/endian_be.go

@@ -1,5 +1,4 @@
 //go:build armbe || arm64be || mips || mips64 || mips64p32 || ppc64 || s390 || s390x || sparc || sparc64
-// +build armbe arm64be mips mips64 mips64p32 ppc64 s390 s390x sparc sparc64
 
 package internal
 

vendor/github.com/cilium/ebpf/internal/endian_le.go

@@ -1,5 +1,4 @@
-//go:build 386 || amd64 || amd64p32 || arm || arm64 || mipsle || mips64le || mips64p32le || ppc64le || riscv64
+//go:build 386 || amd64 || amd64p32 || arm || arm64 || loong64 || mipsle || mips64le || mips64p32le || ppc64le || riscv64
-// +build 386 amd64 amd64p32 arm arm64 mipsle mips64le mips64p32le ppc64le riscv64
 
 package internal
 

vendor/github.com/cilium/ebpf/internal/errors.go

@@ -7,32 +7,25 @@ import (
 	"strings"
 )
 
-// ErrorWithLog returns an error which includes logs from the kernel verifier.
+// ErrorWithLog wraps err in a VerifierError that includes the parsed verifier
+// log buffer.
 //
 // The default error output is a summary of the full log. The latter can be
 // accessed via VerifierError.Log or by formatting the error, see Format.
-//
+func ErrorWithLog(source string, err error, log []byte, truncated bool) *VerifierError {
-// A set of heuristics is used to determine whether the log has been truncated.
-func ErrorWithLog(err error, log []byte) *VerifierError {
 	const whitespace = "\t\r\v\n "
 
 	// Convert verifier log C string by truncating it on the first 0 byte
 	// and trimming trailing whitespace before interpreting as a Go string.
-	truncated := false
 	if i := bytes.IndexByte(log, 0); i != -1 {
-		if i == len(log)-1 && !bytes.HasSuffix(log[:i], []byte{'\n'}) {
-			// The null byte is at the end of the buffer and it's not preceded
-			// by a newline character. Most likely the buffer was too short.
-			truncated = true
-		}
-
 		log = log[:i]
-	} else if len(log) > 0 {
-		// No null byte? Dodgy!
-		truncated = true
 	}
 
 	log = bytes.Trim(log, whitespace)
+	if len(log) == 0 {
+		return &VerifierError{source, err, nil, truncated}
+	}
+
 	logLines := bytes.Split(log, []byte{'\n'})
 	lines := make([]string, 0, len(logLines))
 	for _, line := range logLines {
@@ -41,13 +34,14 @@ func ErrorWithLog(err error, log []byte) *VerifierError {
 		lines = append(lines, string(bytes.TrimRight(line, whitespace)))
 	}
 
-	return &VerifierError{err, lines, truncated}
+	return &VerifierError{source, err, lines, truncated}
 }
 
 // VerifierError includes information from the eBPF verifier.
 //
 // It summarises the log output, see Format if you want to output the full contents.
 type VerifierError struct {
+	source string
 	// The error which caused this error.
 	Cause error
 	// The verifier output split into lines.
@@ -67,9 +61,12 @@ func (le *VerifierError) Error() string {
 		log = log[:n-1]
 	}
 
+	var b strings.Builder
+	fmt.Fprintf(&b, "%s: %s", le.source, le.Cause.Error())
+
 	n := len(log)
 	if n == 0 {
-		return le.Cause.Error()
+		return b.String()
 	}
 
 	lines := log[n-1:]
@@ -78,14 +75,9 @@ func (le *VerifierError) Error() string {
 		lines = log[n-2:]
 	}
 
-	var b strings.Builder
+	for _, line := range lines {
-	fmt.Fprintf(&b, "%s: ", le.Cause.Error())
-
-	for i, line := range lines {
-		b.WriteString(strings.TrimSpace(line))
-		if i != len(lines)-1 {
 		b.WriteString(": ")
-		}
+		b.WriteString(strings.TrimSpace(line))
 	}
 
 	omitted := len(le.Log) - len(lines)
@@ -143,8 +135,8 @@ func includePreviousLine(line string) bool {
 // Understood verbs are %s and %v, which are equivalent to calling Error(). %v
 // allows outputting additional information using the following flags:
 //
-//     +   Output the first <width> lines, or all lines if no width is given.
+//	%+<width>v: Output the first <width> lines, or all lines if no width is given.
-//     -   Output the last <width> lines, or all lines if no width is given.
+//	%-<width>v: Output the last <width> lines, or all lines if no width is given.
 //
 // Use width to specify how many lines to output. Use the '-' flag to output
 // lines from the end of the log instead of the beginning.
@@ -174,7 +166,7 @@ func (le *VerifierError) Format(f fmt.State, verb rune) {
 			return
 		}
 
-		fmt.Fprintf(f, "%s:", le.Cause.Error())
+		fmt.Fprintf(f, "%s: %s:", le.source, le.Cause.Error())
 
 		omitted := len(le.Log) - n
 		lines := le.Log[:n]

vendor/github.com/cilium/ebpf/internal/feature.go

@@ -31,9 +31,19 @@ func (ufe *UnsupportedFeatureError) Is(target error) bool {
 	return target == ErrNotSupported
 }
 
-type featureTest struct {
+// FeatureTest caches the result of a [FeatureTestFn].
-	sync.RWMutex
+//
-	successful bool
+// Fields should not be modified after creation.
+type FeatureTest struct {
+	// The name of the feature being detected.
+	Name string
+	// Version in in the form Major.Minor[.Patch].
+	Version string
+	// The feature test itself.
+	Fn FeatureTestFn
+
+	mu     sync.RWMutex
+	done   bool
 	result error
 }
 
@@ -47,54 +57,128 @@ type featureTest struct {
 //	err != nil: the test couldn't be executed
 type FeatureTestFn func() error
 
-// FeatureTest wraps a function so that it is run at most once.
+// NewFeatureTest is a convenient way to create a single [FeatureTest].
-//
+func NewFeatureTest(name, version string, fn FeatureTestFn) func() error {
-// name should identify the tested feature, while version must be in the
+	ft := &FeatureTest{
-// form Major.Minor[.Patch].
+		Name:    name,
-//
+		Version: version,
-// Returns an error wrapping ErrNotSupported if the feature is not supported.
+		Fn:      fn,
-func FeatureTest(name, version string, fn FeatureTestFn) func() error {
-	ft := new(featureTest)
-	return func() error {
-		ft.RLock()
-		if ft.successful {
-			defer ft.RUnlock()
-			return ft.result
-		}
-		ft.RUnlock()
-		ft.Lock()
-		defer ft.Unlock()
-		// check one more time on the off
-		// chance that two go routines
-		// were able to call into the write
-		// lock
-		if ft.successful {
-			return ft.result
-		}
-		err := fn()
-		switch {
-		case errors.Is(err, ErrNotSupported):
-			v, err := NewVersion(version)
-			if err != nil {
-				return err
 	}
 
+	return ft.execute
+}
+
+// execute the feature test.
+//
+// The result is cached if the test is conclusive.
+//
+// See [FeatureTestFn] for the meaning of the returned error.
+func (ft *FeatureTest) execute() error {
+	ft.mu.RLock()
+	result, done := ft.result, ft.done
+	ft.mu.RUnlock()
+
+	if done {
+		return result
+	}
+
+	ft.mu.Lock()
+	defer ft.mu.Unlock()
+
+	// The test may have been executed by another caller while we were
+	// waiting to acquire ft.mu.
+	if ft.done {
+		return ft.result
+	}
+
+	err := ft.Fn()
+	if err == nil {
+		ft.done = true
+		return nil
+	}
+
+	if errors.Is(err, ErrNotSupported) {
+		var v Version
+		if ft.Version != "" {
+			v, err = NewVersion(ft.Version)
+			if err != nil {
+				return fmt.Errorf("feature %s: %w", ft.Name, err)
+			}
+		}
+
+		ft.done = true
 		ft.result = &UnsupportedFeatureError{
 			MinimumVersion: v,
-				Name:           name,
+			Name:           ft.Name,
 		}
-			fallthrough
 
-		case err == nil:
+		return ft.result
-			ft.successful = true
+	}
 
-		default:
 	// We couldn't execute the feature test to a point
 	// where it could make a determination.
 	// Don't cache the result, just return it.
-			return fmt.Errorf("detect support for %s: %w", name, err)
+	return fmt.Errorf("detect support for %s: %w", ft.Name, err)
+}
+
+// FeatureMatrix groups multiple related feature tests into a map.
+//
+// Useful when there is a small number of discrete features which are known
+// at compile time.
+//
+// It must not be modified concurrently with calling [FeatureMatrix.Result].
+type FeatureMatrix[K comparable] map[K]*FeatureTest
+
+// Result returns the outcome of the feature test for the given key.
+//
+// It's safe to call this function concurrently.
+func (fm FeatureMatrix[K]) Result(key K) error {
+	ft, ok := fm[key]
+	if !ok {
+		return fmt.Errorf("no feature probe for %v", key)
 	}
 
-		return ft.result
+	return ft.execute()
+}
+
+// FeatureCache caches a potentially unlimited number of feature probes.
+//
+// Useful when there is a high cardinality for a feature test.
+type FeatureCache[K comparable] struct {
+	mu       sync.RWMutex
+	newTest  func(K) *FeatureTest
+	features map[K]*FeatureTest
+}
+
+func NewFeatureCache[K comparable](newTest func(K) *FeatureTest) *FeatureCache[K] {
+	return &FeatureCache[K]{
+		newTest:  newTest,
+		features: make(map[K]*FeatureTest),
 	}
 }
+
+func (fc *FeatureCache[K]) Result(key K) error {
+	// NB: Executing the feature test happens without fc.mu taken.
+	return fc.retrieve(key).execute()
+}
+
+func (fc *FeatureCache[K]) retrieve(key K) *FeatureTest {
+	fc.mu.RLock()
+	ft := fc.features[key]
+	fc.mu.RUnlock()
+
+	if ft != nil {
+		return ft
+	}
+
+	fc.mu.Lock()
+	defer fc.mu.Unlock()
+
+	if ft := fc.features[key]; ft != nil {
+		return ft
+	}
+
+	ft = fc.newTest(key)
+	fc.features[key] = ft
+	return ft
+}

vendor/github.com/cilium/ebpf/internal/io.go

@@ -2,10 +2,14 @@ package internal
 
 import (
 	"bufio"
+	"bytes"
 	"compress/gzip"
 	"errors"
+	"fmt"
 	"io"
 	"os"
+	"path/filepath"
+	"sync"
 )
 
 // NewBufferedSectionReader wraps an io.ReaderAt in an appropriately-sized
@@ -60,3 +64,65 @@ func ReadAllCompressed(file string) ([]byte, error) {
 
 	return io.ReadAll(gz)
 }
+
+// ReadUint64FromFile reads a uint64 from a file.
+//
+// format specifies the contents of the file in fmt.Scanf syntax.
+func ReadUint64FromFile(format string, path ...string) (uint64, error) {
+	filename := filepath.Join(path...)
+	data, err := os.ReadFile(filename)
+	if err != nil {
+		return 0, fmt.Errorf("reading file %q: %w", filename, err)
+	}
+
+	var value uint64
+	n, err := fmt.Fscanf(bytes.NewReader(data), format, &value)
+	if err != nil {
+		return 0, fmt.Errorf("parsing file %q: %w", filename, err)
+	}
+	if n != 1 {
+		return 0, fmt.Errorf("parsing file %q: expected 1 item, got %d", filename, n)
+	}
+
+	return value, nil
+}
+
+type uint64FromFileKey struct {
+	format, path string
+}
+
+var uint64FromFileCache = struct {
+	sync.RWMutex
+	values map[uint64FromFileKey]uint64
+}{
+	values: map[uint64FromFileKey]uint64{},
+}
+
+// ReadUint64FromFileOnce is like readUint64FromFile but memoizes the result.
+func ReadUint64FromFileOnce(format string, path ...string) (uint64, error) {
+	filename := filepath.Join(path...)
+	key := uint64FromFileKey{format, filename}
+
+	uint64FromFileCache.RLock()
+	if value, ok := uint64FromFileCache.values[key]; ok {
+		uint64FromFileCache.RUnlock()
+		return value, nil
+	}
+	uint64FromFileCache.RUnlock()
+
+	value, err := ReadUint64FromFile(format, filename)
+	if err != nil {
+		return 0, err
+	}
+
+	uint64FromFileCache.Lock()
+	defer uint64FromFileCache.Unlock()
+
+	if value, ok := uint64FromFileCache.values[key]; ok {
+		// Someone else got here before us, use what is cached.
+		return value, nil
+	}
+
+	uint64FromFileCache.values[key] = value
+	return value, nil
+}

vendor/github.com/cilium/ebpf/internal/kconfig/kconfig.go

@@ -0,0 +1,267 @@
+package kconfig
+
+import (
+	"bufio"
+	"bytes"
+	"compress/gzip"
+	"fmt"
+	"io"
+	"math"
+	"os"
+	"strconv"
+	"strings"
+
+	"github.com/cilium/ebpf/btf"
+	"github.com/cilium/ebpf/internal"
+)
+
+// Find find a kconfig file on the host.
+// It first reads from /boot/config- of the current running kernel and tries
+// /proc/config.gz if nothing was found in /boot.
+// If none of the file provide a kconfig, it returns an error.
+func Find() (*os.File, error) {
+	kernelRelease, err := internal.KernelRelease()
+	if err != nil {
+		return nil, fmt.Errorf("cannot get kernel release: %w", err)
+	}
+
+	path := "/boot/config-" + kernelRelease
+	f, err := os.Open(path)
+	if err == nil {
+		return f, nil
+	}
+
+	f, err = os.Open("/proc/config.gz")
+	if err == nil {
+		return f, nil
+	}
+
+	return nil, fmt.Errorf("neither %s nor /proc/config.gz provide a kconfig", path)
+}
+
+// Parse parses the kconfig file for which a reader is given.
+// All the CONFIG_* which are in filter and which are set set will be
+// put in the returned map as key with their corresponding value as map value.
+// If filter is nil, no filtering will occur.
+// If the kconfig file is not valid, error will be returned.
+func Parse(source io.ReaderAt, filter map[string]struct{}) (map[string]string, error) {
+	var r io.Reader
+	zr, err := gzip.NewReader(io.NewSectionReader(source, 0, math.MaxInt64))
+	if err != nil {
+		r = io.NewSectionReader(source, 0, math.MaxInt64)
+	} else {
+		// Source is gzip compressed, transparently decompress.
+		r = zr
+	}
+
+	ret := make(map[string]string, len(filter))
+
+	s := bufio.NewScanner(r)
+
+	for s.Scan() {
+		line := s.Bytes()
+		err = processKconfigLine(line, ret, filter)
+		if err != nil {
+			return nil, fmt.Errorf("cannot parse line: %w", err)
+		}
+
+		if filter != nil && len(ret) == len(filter) {
+			break
+		}
+	}
+
+	if err := s.Err(); err != nil {
+		return nil, fmt.Errorf("cannot parse: %w", err)
+	}
+
+	if zr != nil {
+		return ret, zr.Close()
+	}
+
+	return ret, nil
+}
+
+// Golang translation of libbpf bpf_object__process_kconfig_line():
+// https://github.com/libbpf/libbpf/blob/fbd60dbff51c870f5e80a17c4f2fd639eb80af90/src/libbpf.c#L1874
+// It does the same checks but does not put the data inside the BPF map.
+func processKconfigLine(line []byte, m map[string]string, filter map[string]struct{}) error {
+	// Ignore empty lines and "# CONFIG_* is not set".
+	if !bytes.HasPrefix(line, []byte("CONFIG_")) {
+		return nil
+	}
+
+	key, value, found := bytes.Cut(line, []byte{'='})
+	if !found {
+		return fmt.Errorf("line %q does not contain separator '='", line)
+	}
+
+	if len(value) == 0 {
+		return fmt.Errorf("line %q has no value", line)
+	}
+
+	if filter != nil {
+		// NB: map[string(key)] gets special optimisation help from the compiler
+		// and doesn't allocate. Don't turn this into a variable.
+		_, ok := filter[string(key)]
+		if !ok {
+			return nil
+		}
+	}
+
+	// This can seem odd, but libbpf only sets the value the first time the key is
+	// met:
+	// https://github.com/torvalds/linux/blob/0d85b27b0cc6/tools/lib/bpf/libbpf.c#L1906-L1908
+	_, ok := m[string(key)]
+	if !ok {
+		m[string(key)] = string(value)
+	}
+
+	return nil
+}
+
+// PutValue translates the value given as parameter depending on the BTF
+// type, the translated value is then written to the byte array.
+func PutValue(data []byte, typ btf.Type, value string) error {
+	typ = btf.UnderlyingType(typ)
+
+	switch value {
+	case "y", "n", "m":
+		return putValueTri(data, typ, value)
+	default:
+		if strings.HasPrefix(value, `"`) {
+			return putValueString(data, typ, value)
+		}
+		return putValueNumber(data, typ, value)
+	}
+}
+
+// Golang translation of libbpf_tristate enum:
+// https://github.com/libbpf/libbpf/blob/fbd60dbff51c870f5e80a17c4f2fd639eb80af90/src/bpf_helpers.h#L169
+type triState int
+
+const (
+	TriNo     triState = 0
+	TriYes    triState = 1
+	TriModule triState = 2
+)
+
+func putValueTri(data []byte, typ btf.Type, value string) error {
+	switch v := typ.(type) {
+	case *btf.Int:
+		if v.Encoding != btf.Bool {
+			return fmt.Errorf("cannot add tri value, expected btf.Bool, got: %v", v.Encoding)
+		}
+
+		if v.Size != 1 {
+			return fmt.Errorf("cannot add tri value, expected size of 1 byte, got: %d", v.Size)
+		}
+
+		switch value {
+		case "y":
+			data[0] = 1
+		case "n":
+			data[0] = 0
+		default:
+			return fmt.Errorf("cannot use %q for btf.Bool", value)
+		}
+	case *btf.Enum:
+		if v.Name != "libbpf_tristate" {
+			return fmt.Errorf("cannot use enum %q, only libbpf_tristate is supported", v.Name)
+		}
+
+		var tri triState
+		switch value {
+		case "y":
+			tri = TriYes
+		case "m":
+			tri = TriModule
+		case "n":
+			tri = TriNo
+		default:
+			return fmt.Errorf("value %q is not support for libbpf_tristate", value)
+		}
+
+		internal.NativeEndian.PutUint64(data, uint64(tri))
+	default:
+		return fmt.Errorf("cannot add number value, expected btf.Int or btf.Enum, got: %T", v)
+	}
+
+	return nil
+}
+
+func putValueString(data []byte, typ btf.Type, value string) error {
+	array, ok := typ.(*btf.Array)
+	if !ok {
+		return fmt.Errorf("cannot add string value, expected btf.Array, got %T", array)
+	}
+
+	contentType, ok := btf.UnderlyingType(array.Type).(*btf.Int)
+	if !ok {
+		return fmt.Errorf("cannot add string value, expected array of btf.Int, got %T", contentType)
+	}
+
+	// Any Int, which is not bool, of one byte could be used to store char:
+	// https://github.com/torvalds/linux/blob/1a5304fecee5/tools/lib/bpf/libbpf.c#L3637-L3638
+	if contentType.Size != 1 && contentType.Encoding != btf.Bool {
+		return fmt.Errorf("cannot add string value, expected array of btf.Int of size 1, got array of btf.Int of size: %v", contentType.Size)
+	}
+
+	if !strings.HasPrefix(value, `"`) || !strings.HasSuffix(value, `"`) {
+		return fmt.Errorf(`value %q must start and finish with '"'`, value)
+	}
+
+	str := strings.Trim(value, `"`)
+
+	// We need to trim string if the bpf array is smaller.
+	if uint32(len(str)) >= array.Nelems {
+		str = str[:array.Nelems]
+	}
+
+	// Write the string content to .kconfig.
+	copy(data, str)
+
+	return nil
+}
+
+func putValueNumber(data []byte, typ btf.Type, value string) error {
+	integer, ok := typ.(*btf.Int)
+	if !ok {
+		return fmt.Errorf("cannot add number value, expected *btf.Int, got: %T", integer)
+	}
+
+	size := integer.Size
+	sizeInBits := size * 8
+
+	var n uint64
+	var err error
+	if integer.Encoding == btf.Signed {
+		parsed, e := strconv.ParseInt(value, 0, int(sizeInBits))
+
+		n = uint64(parsed)
+		err = e
+	} else {
+		parsed, e := strconv.ParseUint(value, 0, int(sizeInBits))
+
+		n = uint64(parsed)
+		err = e
+	}
+
+	if err != nil {
+		return fmt.Errorf("cannot parse value: %w", err)
+	}
+
+	switch size {
+	case 1:
+		data[0] = byte(n)
+	case 2:
+		internal.NativeEndian.PutUint16(data, uint16(n))
+	case 4:
+		internal.NativeEndian.PutUint32(data, uint32(n))
+	case 8:
+		internal.NativeEndian.PutUint64(data, uint64(n))
+	default:
+		return fmt.Errorf("size (%d) is not valid, expected: 1, 2, 4 or 8", size)
+	}
+
+	return nil
+}

vendor/github.com/cilium/ebpf/internal/memoize.go

@@ -0,0 +1,26 @@
+package internal
+
+import (
+	"sync"
+)
+
+type memoizedFunc[T any] struct {
+	once   sync.Once
+	fn     func() (T, error)
+	result T
+	err    error
+}
+
+func (mf *memoizedFunc[T]) do() (T, error) {
+	mf.once.Do(func() {
+		mf.result, mf.err = mf.fn()
+	})
+	return mf.result, mf.err
+}
+
+// Memoize the result of a function call.
+//
+// fn is only ever called once, even if it returns an error.
+func Memoize[T any](fn func() (T, error)) func() (T, error) {
+	return (&memoizedFunc[T]{fn: fn}).do
+}

vendor/github.com/cilium/ebpf/internal/output.go

@@ -6,6 +6,7 @@ import (
 	"go/format"
 	"go/scanner"
 	"io"
+	"reflect"
 	"strings"
 	"unicode"
 )
@@ -82,3 +83,15 @@ func WriteFormatted(src []byte, out io.Writer) error {
 
 	return nel
 }
+
+// GoTypeName is like %T, but elides the package name.
+//
+// Pointers to a type are peeled off.
+func GoTypeName(t any) string {
+	rT := reflect.TypeOf(t)
+	for rT.Kind() == reflect.Pointer {
+		rT = rT.Elem()
+	}
+	// Doesn't return the correct Name for generic types due to https://github.com/golang/go/issues/55924
+	return rT.Name()
+}

vendor/github.com/cilium/ebpf/internal/pinning.go

@@ -6,15 +6,12 @@ import (
 	"os"
 	"path/filepath"
 	"runtime"
-	"unsafe"
 
 	"github.com/cilium/ebpf/internal/sys"
 	"github.com/cilium/ebpf/internal/unix"
 )
 
 func Pin(currentPath, newPath string, fd *sys.FD) error {
-	const bpfFSType = 0xcafe4a11
-
 	if newPath == "" {
 		return errors.New("given pinning path cannot be empty")
 	}
@@ -22,20 +19,11 @@ func Pin(currentPath, newPath string, fd *sys.FD) error {
 		return nil
 	}
 
-	var statfs unix.Statfs_t
+	fsType, err := FSType(filepath.Dir(newPath))
-	if err := unix.Statfs(filepath.Dir(newPath), &statfs); err != nil {
+	if err != nil {
 		return err
 	}
-
+	if fsType != unix.BPF_FS_MAGIC {
-	fsType := int64(statfs.Type)
-	if unsafe.Sizeof(statfs.Type) == 4 {
-		// We're on a 32 bit arch, where statfs.Type is int32. bpfFSType is a
-		// negative number when interpreted as int32 so we need to cast via
-		// uint32 to avoid sign extension.
-		fsType = int64(uint32(statfs.Type))
-	}
-
-	if fsType != bpfFSType {
 		return fmt.Errorf("%s is not on a bpf filesystem", newPath)
 	}
 
@@ -50,7 +38,7 @@ func Pin(currentPath, newPath string, fd *sys.FD) error {
 
 	// Renameat2 is used instead of os.Rename to disallow the new path replacing
 	// an existing path.
-	err := unix.Renameat2(unix.AT_FDCWD, currentPath, unix.AT_FDCWD, newPath, unix.RENAME_NOREPLACE)
+	err = unix.Renameat2(unix.AT_FDCWD, currentPath, unix.AT_FDCWD, newPath, unix.RENAME_NOREPLACE)
 	if err == nil {
 		// Object is now moved to the new pinning path.
 		return nil

vendor/github.com/cilium/ebpf/internal/platform.go

@@ -0,0 +1,43 @@
+package internal
+
+import (
+	"runtime"
+)
+
+// PlatformPrefix returns the platform-dependent syscall wrapper prefix used by
+// the linux kernel.
+//
+// Based on https://github.com/golang/go/blob/master/src/go/build/syslist.go
+// and https://github.com/libbpf/libbpf/blob/master/src/libbpf.c#L10047
+func PlatformPrefix() string {
+	switch runtime.GOARCH {
+	case "386":
+		return "__ia32_"
+	case "amd64", "amd64p32":
+		return "__x64_"
+
+	case "arm", "armbe":
+		return "__arm_"
+	case "arm64", "arm64be":
+		return "__arm64_"
+
+	case "mips", "mipsle", "mips64", "mips64le", "mips64p32", "mips64p32le":
+		return "__mips_"
+
+	case "s390":
+		return "__s390_"
+	case "s390x":
+		return "__s390x_"
+
+	case "riscv", "riscv64":
+		return "__riscv_"
+
+	case "ppc":
+		return "__powerpc_"
+	case "ppc64", "ppc64le":
+		return "__powerpc64_"
+
+	default:
+		return ""
+	}
+}

vendor/github.com/cilium/ebpf/internal/prog.go

@@ -0,0 +1,11 @@
+package internal
+
+// EmptyBPFContext is the smallest-possible BPF input context to be used for
+// invoking `Program.{Run,Benchmark,Test}`.
+//
+// Programs require a context input buffer of at least 15 bytes. Looking in
+// net/bpf/test_run.c, bpf_test_init() requires that the input is at least
+// ETH_HLEN (14) bytes. As of Linux commit fd18942 ("bpf: Don't redirect packets
+// with invalid pkt_len"), it also requires the skb to be non-empty after
+// removing the Layer 2 header.
+var EmptyBPFContext = make([]byte, 15)

vendor/github.com/cilium/ebpf/internal/statfs.go

@@ -0,0 +1,23 @@
+package internal
+
+import (
+	"unsafe"
+
+	"github.com/cilium/ebpf/internal/unix"
+)
+
+func FSType(path string) (int64, error) {
+	var statfs unix.Statfs_t
+	if err := unix.Statfs(path, &statfs); err != nil {
+		return 0, err
+	}
+
+	fsType := int64(statfs.Type)
+	if unsafe.Sizeof(statfs.Type) == 4 {
+		// We're on a 32 bit arch, where statfs.Type is int32. bpfFSType is a
+		// negative number when interpreted as int32 so we need to cast via
+		// uint32 to avoid sign extension.
+		fsType = int64(uint32(statfs.Type))
+	}
+	return fsType, nil
+}

vendor/github.com/cilium/ebpf/internal/sys/fd.go

@@ -17,11 +17,39 @@ type FD struct {
 }
 
 func newFD(value int) *FD {
+	if onLeakFD != nil {
+		// Attempt to store the caller's stack for the given fd value.
+		// Panic if fds contains an existing stack for the fd.
+		old, exist := fds.LoadOrStore(value, callersFrames())
+		if exist {
+			f := old.(*runtime.Frames)
+			panic(fmt.Sprintf("found existing stack for fd %d:\n%s", value, FormatFrames(f)))
+		}
+	}
+
 	fd := &FD{value}
-	runtime.SetFinalizer(fd, (*FD).Close)
+	runtime.SetFinalizer(fd, (*FD).finalize)
 	return fd
 }
 
+// finalize is set as the FD's runtime finalizer and
+// sends a leak trace before calling FD.Close().
+func (fd *FD) finalize() {
+	if fd.raw < 0 {
+		return
+	}
+
+	// Invoke the fd leak callback. Calls LoadAndDelete to guarantee the callback
+	// is invoked at most once for one sys.FD allocation, runtime.Frames can only
+	// be unwound once.
+	f, ok := fds.LoadAndDelete(fd.Int())
+	if ok && onLeakFD != nil {
+		onLeakFD(f.(*runtime.Frames))
+	}
+
+	_ = fd.Close()
+}
+
 // NewFD wraps a raw fd with a finalizer.
 //
 // You must not use the raw fd after calling this function, since the underlying
@@ -64,15 +92,16 @@ func (fd *FD) Close() error {
 		return nil
 	}
 
-	value := int(fd.raw)
+	return unix.Close(fd.disown())
-	fd.raw = -1
-
-	fd.Forget()
-	return unix.Close(value)
 }
 
-func (fd *FD) Forget() {
+func (fd *FD) disown() int {
+	value := int(fd.raw)
+	fds.Delete(int(value))
+	fd.raw = -1
+
 	runtime.SetFinalizer(fd, nil)
+	return value
 }
 
 func (fd *FD) Dup() (*FD, error) {
@@ -90,7 +119,15 @@ func (fd *FD) Dup() (*FD, error) {
 	return newFD(dup), nil
 }
 
+// File takes ownership of FD and turns it into an [*os.File].
+//
+// You must not use the FD after the call returns.
+//
+// Returns nil if the FD is not valid.
 func (fd *FD) File(name string) *os.File {
-	fd.Forget()
+	if fd.raw < 0 {
-	return os.NewFile(uintptr(fd.raw), name)
+		return nil
+	}
+
+	return os.NewFile(uintptr(fd.disown()), name)
 }

vendor/github.com/cilium/ebpf/internal/sys/fd_trace.go

@@ -0,0 +1,93 @@
+package sys
+
+import (
+	"bytes"
+	"fmt"
+	"runtime"
+	"sync"
+)
+
+// OnLeakFD controls tracing [FD] lifetime to detect resources that are not
+// closed by Close().
+//
+// If fn is not nil, tracing is enabled for all FDs created going forward. fn is
+// invoked for all FDs that are closed by the garbage collector instead of an
+// explicit Close() by a caller. Calling OnLeakFD twice with a non-nil fn
+// (without disabling tracing in the meantime) will cause a panic.
+//
+// If fn is nil, tracing will be disabled. Any FDs that have not been closed are
+// considered to be leaked, fn will be invoked for them, and the process will be
+// terminated.
+//
+// fn will be invoked at most once for every unique sys.FD allocation since a
+// runtime.Frames can only be unwound once.
+func OnLeakFD(fn func(*runtime.Frames)) {
+	// Enable leak tracing if new fn is provided.
+	if fn != nil {
+		if onLeakFD != nil {
+			panic("OnLeakFD called twice with non-nil fn")
+		}
+
+		onLeakFD = fn
+		return
+	}
+
+	// fn is nil past this point.
+
+	if onLeakFD == nil {
+		return
+	}
+
+	// Call onLeakFD for all open fds.
+	if fs := flushFrames(); len(fs) != 0 {
+		for _, f := range fs {
+			onLeakFD(f)
+		}
+	}
+
+	onLeakFD = nil
+}
+
+var onLeakFD func(*runtime.Frames)
+
+// fds is a registry of all file descriptors wrapped into sys.fds that were
+// created while an fd tracer was active.
+var fds sync.Map // map[int]*runtime.Frames
+
+// flushFrames removes all elements from fds and returns them as a slice. This
+// deals with the fact that a runtime.Frames can only be unwound once using
+// Next().
+func flushFrames() []*runtime.Frames {
+	var frames []*runtime.Frames
+	fds.Range(func(key, value any) bool {
+		frames = append(frames, value.(*runtime.Frames))
+		fds.Delete(key)
+		return true
+	})
+	return frames
+}
+
+func callersFrames() *runtime.Frames {
+	c := make([]uintptr, 32)
+
+	// Skip runtime.Callers and this function.
+	i := runtime.Callers(2, c)
+	if i == 0 {
+		return nil
+	}
+
+	return runtime.CallersFrames(c)
+}
+
+// FormatFrames formats a runtime.Frames as a human-readable string.
+func FormatFrames(fs *runtime.Frames) string {
+	var b bytes.Buffer
+	for {
+		f, more := fs.Next()
+		b.WriteString(fmt.Sprintf("\t%s+%#x\n\t\t%s:%d\n", f.Function, f.PC-f.Entry, f.File, f.Line))
+		if !more {
+			break
+		}
+	}
+	return b.String()
+}

vendor/github.com/cilium/ebpf/internal/sys/mapflags_string.go

@@ -0,0 +1,49 @@
+// Code generated by "stringer -type MapFlags"; DO NOT EDIT.
+
+package sys
+
+import "strconv"
+
+func _() {
+	// An "invalid array index" compiler error signifies that the constant values have changed.
+	// Re-run the stringer command to generate them again.
+	var x [1]struct{}
+	_ = x[BPF_F_NO_PREALLOC-1]
+	_ = x[BPF_F_NO_COMMON_LRU-2]
+	_ = x[BPF_F_NUMA_NODE-4]
+	_ = x[BPF_F_RDONLY-8]
+	_ = x[BPF_F_WRONLY-16]
+	_ = x[BPF_F_STACK_BUILD_ID-32]
+	_ = x[BPF_F_ZERO_SEED-64]
+	_ = x[BPF_F_RDONLY_PROG-128]
+	_ = x[BPF_F_WRONLY_PROG-256]
+	_ = x[BPF_F_CLONE-512]
+	_ = x[BPF_F_MMAPABLE-1024]
+	_ = x[BPF_F_PRESERVE_ELEMS-2048]
+	_ = x[BPF_F_INNER_MAP-4096]
+}
+
+const _MapFlags_name = "BPF_F_NO_PREALLOCBPF_F_NO_COMMON_LRUBPF_F_NUMA_NODEBPF_F_RDONLYBPF_F_WRONLYBPF_F_STACK_BUILD_IDBPF_F_ZERO_SEEDBPF_F_RDONLY_PROGBPF_F_WRONLY_PROGBPF_F_CLONEBPF_F_MMAPABLEBPF_F_PRESERVE_ELEMSBPF_F_INNER_MAP"
+
+var _MapFlags_map = map[MapFlags]string{
+	1:    _MapFlags_name[0:17],
+	2:    _MapFlags_name[17:36],
+	4:    _MapFlags_name[36:51],
+	8:    _MapFlags_name[51:63],
+	16:   _MapFlags_name[63:75],
+	32:   _MapFlags_name[75:95],
+	64:   _MapFlags_name[95:110],
+	128:  _MapFlags_name[110:127],
+	256:  _MapFlags_name[127:144],
+	512:  _MapFlags_name[144:155],
+	1024: _MapFlags_name[155:169],
+	2048: _MapFlags_name[169:189],
+	4096: _MapFlags_name[189:204],
+}
+
+func (i MapFlags) String() string {
+	if str, ok := _MapFlags_map[i]; ok {
+		return str
+	}
+	return "MapFlags(" + strconv.FormatInt(int64(i), 10) + ")"
+}

vendor/github.com/cilium/ebpf/internal/sys/ptr.go

@@ -20,7 +20,7 @@ func NewSlicePointer(buf []byte) Pointer {
 	return Pointer{ptr: unsafe.Pointer(&buf[0])}
 }
 
-// NewSlicePointer creates a 64-bit pointer from a byte slice.
+// NewSlicePointerLen creates a 64-bit pointer from a byte slice.
 //
 // Useful to assign both the pointer and the length in one go.
 func NewSlicePointerLen(buf []byte) (Pointer, uint32) {
@@ -36,3 +36,17 @@ func NewStringPointer(str string) Pointer {
 
 	return Pointer{ptr: unsafe.Pointer(p)}
 }
+
+// NewStringSlicePointer allocates an array of Pointers to each string in the
+// given slice of strings and returns a 64-bit pointer to the start of the
+// resulting array.
+//
+// Use this function to pass arrays of strings as syscall arguments.
+func NewStringSlicePointer(strings []string) Pointer {
+	sp := make([]Pointer, 0, len(strings))
+	for _, s := range strings {
+		sp = append(sp, NewStringPointer(s))
+	}
+
+	return Pointer{ptr: unsafe.Pointer(&sp[0])}
+}

vendor/github.com/cilium/ebpf/internal/sys/ptr_32_be.go

@@ -1,5 +1,4 @@
 //go:build armbe || mips || mips64p32
-// +build armbe mips mips64p32
 
 package sys
 

vendor/github.com/cilium/ebpf/internal/sys/ptr_32_le.go

@@ -1,5 +1,4 @@
 //go:build 386 || amd64p32 || arm || mipsle || mips64p32le
-// +build 386 amd64p32 arm mipsle mips64p32le
 
 package sys
 

vendor/github.com/cilium/ebpf/internal/sys/ptr_64.go

@@ -1,5 +1,4 @@
 //go:build !386 && !amd64p32 && !arm && !mipsle && !mips64p32le && !armbe && !mips && !mips64p32
-// +build !386,!amd64p32,!arm,!mipsle,!mips64p32le,!armbe,!mips,!mips64p32
 
 package sys
 

vendor/github.com/cilium/ebpf/internal/sys/signals.go

@@ -0,0 +1,83 @@
+package sys
+
+import (
+	"fmt"
+	"runtime"
+	"unsafe"
+
+	"github.com/cilium/ebpf/internal/unix"
+)
+
+// A sigset containing only SIGPROF.
+var profSet unix.Sigset_t
+
+func init() {
+	// See sigsetAdd for details on the implementation. Open coded here so
+	// that the compiler will check the constant calculations for us.
+	profSet.Val[sigprofBit/wordBits] |= 1 << (sigprofBit % wordBits)
+}
+
+// maskProfilerSignal locks the calling goroutine to its underlying OS thread
+// and adds SIGPROF to the thread's signal mask. This prevents pprof from
+// interrupting expensive syscalls like e.g. BPF_PROG_LOAD.
+//
+// The caller must defer unmaskProfilerSignal() to reverse the operation.
+func maskProfilerSignal() {
+	runtime.LockOSThread()
+
+	if err := unix.PthreadSigmask(unix.SIG_BLOCK, &profSet, nil); err != nil {
+		runtime.UnlockOSThread()
+		panic(fmt.Errorf("masking profiler signal: %w", err))
+	}
+}
+
+// unmaskProfilerSignal removes SIGPROF from the underlying thread's signal
+// mask, allowing it to be interrupted for profiling once again.
+//
+// It also unlocks the current goroutine from its underlying OS thread.
+func unmaskProfilerSignal() {
+	defer runtime.UnlockOSThread()
+
+	if err := unix.PthreadSigmask(unix.SIG_UNBLOCK, &profSet, nil); err != nil {
+		panic(fmt.Errorf("unmasking profiler signal: %w", err))
+	}
+}
+
+const (
+	// Signal is the nth bit in the bitfield.
+	sigprofBit = int(unix.SIGPROF - 1)
+	// The number of bits in one Sigset_t word.
+	wordBits = int(unsafe.Sizeof(unix.Sigset_t{}.Val[0])) * 8
+)
+
+// sigsetAdd adds signal to set.
+//
+// Note: Sigset_t.Val's value type is uint32 or uint64 depending on the arch.
+// This function must be able to deal with both and so must avoid any direct
+// references to u32 or u64 types.
+func sigsetAdd(set *unix.Sigset_t, signal unix.Signal) error {
+	if signal < 1 {
+		return fmt.Errorf("signal %d must be larger than 0", signal)
+	}
+
+	// For amd64, runtime.sigaddset() performs the following operation:
+	// set[(signal-1)/32] |= 1 << ((uint32(signal) - 1) & 31)
+	//
+	// This trick depends on sigset being two u32's, causing a signal in the the
+	// bottom 31 bits to be written to the low word if bit 32 is low, or the high
+	// word if bit 32 is high.
+
+	// Signal is the nth bit in the bitfield.
+	bit := int(signal - 1)
+	// Word within the sigset the bit needs to be written to.
+	word := bit / wordBits
+
+	if word >= len(set.Val) {
+		return fmt.Errorf("signal %d does not fit within unix.Sigset_t", signal)
+	}
+
+	// Write the signal bit into its corresponding word at the corrected offset.
+	set.Val[word] |= 1 << (bit % wordBits)
+
+	return nil
+}

vendor/github.com/cilium/ebpf/internal/sys/syscall.go

@@ -8,10 +8,22 @@ import (
 	"github.com/cilium/ebpf/internal/unix"
 )
 
+// ENOTSUPP is a Linux internal error code that has leaked into UAPI.
+//
+// It is not the same as ENOTSUP or EOPNOTSUPP.
+var ENOTSUPP = syscall.Errno(524)
+
 // BPF wraps SYS_BPF.
 //
 // Any pointers contained in attr must use the Pointer type from this package.
 func BPF(cmd Cmd, attr unsafe.Pointer, size uintptr) (uintptr, error) {
+	// Prevent the Go profiler from repeatedly interrupting the verifier,
+	// which could otherwise lead to a livelock due to receiving EAGAIN.
+	if cmd == BPF_PROG_LOAD || cmd == BPF_PROG_RUN {
+		maskProfilerSignal()
+		defer unmaskProfilerSignal()
+	}
+
 	for {
 		r1, _, errNo := unix.Syscall(unix.SYS_BPF, uintptr(cmd), uintptr(attr), size)
 		runtime.KeepAlive(attr)
@@ -90,12 +102,45 @@ func NewObjName(name string) ObjName {
 	return result
 }
 
+// LogLevel controls the verbosity of the kernel's eBPF program verifier.
+type LogLevel uint32
+
+const (
+	BPF_LOG_LEVEL1 LogLevel = 1 << iota
+	BPF_LOG_LEVEL2
+	BPF_LOG_STATS
+)
+
 // LinkID uniquely identifies a bpf_link.
 type LinkID uint32
 
 // BTFID uniquely identifies a BTF blob loaded into the kernel.
 type BTFID uint32
 
+// TypeID identifies a type in a BTF blob.
+type TypeID uint32
+
+// MapFlags control map behaviour.
+type MapFlags uint32
+
+//go:generate stringer -type MapFlags
+
+const (
+	BPF_F_NO_PREALLOC MapFlags = 1 << iota
+	BPF_F_NO_COMMON_LRU
+	BPF_F_NUMA_NODE
+	BPF_F_RDONLY
+	BPF_F_WRONLY
+	BPF_F_STACK_BUILD_ID
+	BPF_F_ZERO_SEED
+	BPF_F_RDONLY_PROG
+	BPF_F_WRONLY_PROG
+	BPF_F_CLONE
+	BPF_F_MMAPABLE
+	BPF_F_PRESERVE_ELEMS
+	BPF_F_INNER_MAP
+)
+
 // wrappedErrno wraps syscall.Errno to prevent direct comparisons with
 // syscall.E* or unix.E* constants.
 //
@@ -108,6 +153,13 @@ func (we wrappedErrno) Unwrap() error {
 	return we.Errno
 }
 
+func (we wrappedErrno) Error() string {
+	if we.Errno == ENOTSUPP {
+		return "operation not supported"
+	}
+	return we.Errno.Error()
+}
+
 type syscallError struct {
 	error
 	errno syscall.Errno

vendor/github.com/cilium/ebpf/internal/sys/types.go

@@ -6,14 +6,14 @@ import (
 	"unsafe"
 )
 
-type AdjRoomMode int32
+type AdjRoomMode uint32
 
 const (
 	BPF_ADJ_ROOM_NET AdjRoomMode = 0
 	BPF_ADJ_ROOM_MAC AdjRoomMode = 1
 )
 
-type AttachType int32
+type AttachType uint32
 
 const (
 	BPF_CGROUP_INET_INGRESS            AttachType = 0
@@ -62,7 +62,7 @@ const (
 	__MAX_BPF_ATTACH_TYPE              AttachType = 43
 )
 
-type Cmd int32
+type Cmd uint32
 
 const (
 	BPF_MAP_CREATE                  Cmd = 0
@@ -104,7 +104,7 @@ const (
 	BPF_PROG_BIND_MAP               Cmd = 35
 )
 
-type FunctionId int32
+type FunctionId uint32
 
 const (
 	BPF_FUNC_unspec                         FunctionId = 0
@@ -301,17 +301,27 @@ const (
 	BPF_FUNC_copy_from_user_task            FunctionId = 191
 	BPF_FUNC_skb_set_tstamp                 FunctionId = 192
 	BPF_FUNC_ima_file_hash                  FunctionId = 193
-	__BPF_FUNC_MAX_ID                       FunctionId = 194
+	BPF_FUNC_kptr_xchg                      FunctionId = 194
+	BPF_FUNC_map_lookup_percpu_elem         FunctionId = 195
+	BPF_FUNC_skc_to_mptcp_sock              FunctionId = 196
+	BPF_FUNC_dynptr_from_mem                FunctionId = 197
+	BPF_FUNC_ringbuf_reserve_dynptr         FunctionId = 198
+	BPF_FUNC_ringbuf_submit_dynptr          FunctionId = 199
+	BPF_FUNC_ringbuf_discard_dynptr         FunctionId = 200
+	BPF_FUNC_dynptr_read                    FunctionId = 201
+	BPF_FUNC_dynptr_write                   FunctionId = 202
+	BPF_FUNC_dynptr_data                    FunctionId = 203
+	__BPF_FUNC_MAX_ID                       FunctionId = 204
 )
 
-type HdrStartOff int32
+type HdrStartOff uint32
 
 const (
 	BPF_HDR_START_MAC HdrStartOff = 0
 	BPF_HDR_START_NET HdrStartOff = 1
 )
 
-type LinkType int32
+type LinkType uint32
 
 const (
 	BPF_LINK_TYPE_UNSPEC         LinkType = 0
@@ -323,10 +333,11 @@ const (
 	BPF_LINK_TYPE_XDP            LinkType = 6
 	BPF_LINK_TYPE_PERF_EVENT     LinkType = 7
 	BPF_LINK_TYPE_KPROBE_MULTI   LinkType = 8
-	MAX_BPF_LINK_TYPE            LinkType = 9
+	BPF_LINK_TYPE_STRUCT_OPS     LinkType = 9
+	MAX_BPF_LINK_TYPE            LinkType = 10
 )
 
-type MapType int32
+type MapType uint32
 
 const (
 	BPF_MAP_TYPE_UNSPEC                MapType = 0
@@ -362,7 +373,7 @@ const (
 	BPF_MAP_TYPE_BLOOM_FILTER          MapType = 30
 )
 
-type ProgType int32
+type ProgType uint32
 
 const (
 	BPF_PROG_TYPE_UNSPEC                  ProgType = 0
@@ -399,7 +410,7 @@ const (
 	BPF_PROG_TYPE_SYSCALL                 ProgType = 31
 )
 
-type RetCode int32
+type RetCode uint32
 
 const (
 	BPF_OK          RetCode = 0
@@ -408,14 +419,14 @@ const (
 	BPF_LWT_REROUTE RetCode = 128
 )
 
-type SkAction int32
+type SkAction uint32
 
 const (
 	SK_DROP SkAction = 0
 	SK_PASS SkAction = 1
 )
 
-type StackBuildIdStatus int32
+type StackBuildIdStatus uint32
 
 const (
 	BPF_STACK_BUILD_ID_EMPTY StackBuildIdStatus = 0
@@ -423,13 +434,13 @@ const (
 	BPF_STACK_BUILD_ID_IP    StackBuildIdStatus = 2
 )
 
-type StatsType int32
+type StatsType uint32
 
 const (
 	BPF_STATS_RUN_TIME StatsType = 0
 )
 
-type XdpAction int32
+type XdpAction uint32
 
 const (
 	XDP_ABORTED  XdpAction = 0
@@ -474,15 +485,15 @@ type MapInfo struct {
 	KeySize               uint32
 	ValueSize             uint32
 	MaxEntries            uint32
-	MapFlags              uint32
+	MapFlags              MapFlags
 	Name                  ObjName
 	Ifindex               uint32
-	BtfVmlinuxValueTypeId uint32
+	BtfVmlinuxValueTypeId TypeID
 	NetnsDev              uint64
 	NetnsIno              uint64
 	BtfId                 uint32
-	BtfKeyTypeId          uint32
+	BtfKeyTypeId          TypeID
-	BtfValueTypeId        uint32
+	BtfValueTypeId        TypeID
 	_                     [4]byte
 	MapExtra              uint64
 }
@@ -508,7 +519,7 @@ type ProgInfo struct {
 	NrJitedFuncLens      uint32
 	JitedKsyms           uint64
 	JitedFuncLens        uint64
-	BtfId                uint32
+	BtfId                BTFID
 	FuncInfoRecSize      uint32
 	FuncInfo             uint64
 	NrFuncInfo           uint32
@@ -616,7 +627,7 @@ type LinkCreateAttr struct {
 	TargetFd    uint32
 	AttachType  AttachType
 	Flags       uint32
-	TargetBtfId uint32
+	TargetBtfId TypeID
 	_           [28]byte
 }
 
@@ -646,6 +657,26 @@ func LinkCreateIter(attr *LinkCreateIterAttr) (*FD, error) {
 	return NewFD(int(fd))
 }
 
+type LinkCreateKprobeMultiAttr struct {
+	ProgFd           uint32
+	TargetFd         uint32
+	AttachType       AttachType
+	Flags            uint32
+	KprobeMultiFlags uint32
+	Count            uint32
+	Syms             Pointer
+	Addrs            Pointer
+	Cookies          Pointer
+}
+
+func LinkCreateKprobeMulti(attr *LinkCreateKprobeMultiAttr) (*FD, error) {
+	fd, err := BPF(BPF_LINK_CREATE, unsafe.Pointer(attr), unsafe.Sizeof(*attr))
+	if err != nil {
+		return nil, err
+	}
+	return NewFD(int(fd))
+}
+
 type LinkCreatePerfEventAttr struct {
 	ProgFd     uint32
 	TargetFd   uint32
@@ -663,6 +694,25 @@ func LinkCreatePerfEvent(attr *LinkCreatePerfEventAttr) (*FD, error) {
 	return NewFD(int(fd))
 }
 
+type LinkCreateTracingAttr struct {
+	ProgFd      uint32
+	TargetFd    uint32
+	AttachType  AttachType
+	Flags       uint32
+	TargetBtfId BTFID
+	_           [4]byte
+	Cookie      uint64
+	_           [16]byte
+}
+
+func LinkCreateTracing(attr *LinkCreateTracingAttr) (*FD, error) {
+	fd, err := BPF(BPF_LINK_CREATE, unsafe.Pointer(attr), unsafe.Sizeof(*attr))
+	if err != nil {
+		return nil, err
+	}
+	return NewFD(int(fd))
+}
+
 type LinkUpdateAttr struct {
 	LinkFd    uint32
 	NewProgFd uint32
@@ -680,15 +730,15 @@ type MapCreateAttr struct {
 	KeySize               uint32
 	ValueSize             uint32
 	MaxEntries            uint32
-	MapFlags              uint32
+	MapFlags              MapFlags
 	InnerMapFd            uint32
 	NumaNode              uint32
 	MapName               ObjName
 	MapIfindex            uint32
 	BtfFd                 uint32
-	BtfKeyTypeId          uint32
+	BtfKeyTypeId          TypeID
-	BtfValueTypeId        uint32
+	BtfValueTypeId        TypeID
-	BtfVmlinuxValueTypeId uint32
+	BtfVmlinuxValueTypeId TypeID
 	MapExtra              uint64
 }
 
@@ -951,7 +1001,7 @@ type ProgLoadAttr struct {
 	InsnCnt            uint32
 	Insns              Pointer
 	License            Pointer
-	LogLevel           uint32
+	LogLevel           LogLevel
 	LogSize            uint32
 	LogBuf             Pointer
 	KernVersion        uint32
@@ -966,8 +1016,8 @@ type ProgLoadAttr struct {
 	LineInfoRecSize    uint32
 	LineInfo           Pointer
 	LineInfoCnt        uint32
-	AttachBtfId        uint32
+	AttachBtfId        TypeID
-	AttachProgFd       uint32
+	AttachBtfObjFd     uint32
 	CoreReloCnt        uint32
 	FdArray            Pointer
 	CoreRelos          Pointer
@@ -983,6 +1033,21 @@ func ProgLoad(attr *ProgLoadAttr) (*FD, error) {
 	return NewFD(int(fd))
 }
 
+type ProgQueryAttr struct {
+	TargetFd    uint32
+	AttachType  AttachType
+	QueryFlags  uint32
+	AttachFlags uint32
+	ProgIds     Pointer
+	ProgCount   uint32
+	_           [4]byte
+}
+
+func ProgQuery(attr *ProgQueryAttr) error {
+	_, err := BPF(BPF_PROG_QUERY, unsafe.Pointer(attr), unsafe.Sizeof(*attr))
+	return err
+}
+
 type ProgRunAttr struct {
 	ProgFd      uint32
 	Retval      uint32
@@ -1046,7 +1111,7 @@ type RawTracepointLinkInfo struct {
 type TracingLinkInfo struct {
 	AttachType  AttachType
 	TargetObjId uint32
-	TargetBtfId uint32
+	TargetBtfId TypeID
 }
 
 type XDPLinkInfo struct{ Ifindex uint32 }

vendor/github.com/cilium/ebpf/internal/tracefs/kprobe.go

@@ -0,0 +1,359 @@
+package tracefs
+
+import (
+	"crypto/rand"
+	"errors"
+	"fmt"
+	"os"
+	"path/filepath"
+	"runtime"
+	"strings"
+	"syscall"
+
+	"github.com/cilium/ebpf/internal"
+	"github.com/cilium/ebpf/internal/unix"
+)
+
+var (
+	ErrInvalidInput = errors.New("invalid input")
+
+	ErrInvalidMaxActive = errors.New("can only set maxactive on kretprobes")
+)
+
+//go:generate stringer -type=ProbeType -linecomment
+
+type ProbeType uint8
+
+const (
+	Kprobe ProbeType = iota // kprobe
+	Uprobe                  // uprobe
+)
+
+func (pt ProbeType) eventsFile() (*os.File, error) {
+	path, err := sanitizeTracefsPath(fmt.Sprintf("%s_events", pt.String()))
+	if err != nil {
+		return nil, err
+	}
+
+	return os.OpenFile(path, os.O_APPEND|os.O_WRONLY, 0666)
+}
+
+type ProbeArgs struct {
+	Type                         ProbeType
+	Symbol, Group, Path          string
+	Offset, RefCtrOffset, Cookie uint64
+	Pid, RetprobeMaxActive       int
+	Ret                          bool
+}
+
+// RandomGroup generates a pseudorandom string for use as a tracefs group name.
+// Returns an error when the output string would exceed 63 characters (kernel
+// limitation), when rand.Read() fails or when prefix contains characters not
+// allowed by IsValidTraceID.
+func RandomGroup(prefix string) (string, error) {
+	if !validIdentifier(prefix) {
+		return "", fmt.Errorf("prefix '%s' must be alphanumeric or underscore: %w", prefix, ErrInvalidInput)
+	}
+
+	b := make([]byte, 8)
+	if _, err := rand.Read(b); err != nil {
+		return "", fmt.Errorf("reading random bytes: %w", err)
+	}
+
+	group := fmt.Sprintf("%s_%x", prefix, b)
+	if len(group) > 63 {
+		return "", fmt.Errorf("group name '%s' cannot be longer than 63 characters: %w", group, ErrInvalidInput)
+	}
+
+	return group, nil
+}
+
+// validIdentifier implements the equivalent of a regex match
+// against "^[a-zA-Z_][0-9a-zA-Z_]*$".
+//
+// Trace event groups, names and kernel symbols must adhere to this set
+// of characters. Non-empty, first character must not be a number, all
+// characters must be alphanumeric or underscore.
+func validIdentifier(s string) bool {
+	if len(s) < 1 {
+		return false
+	}
+	for i, c := range []byte(s) {
+		switch {
+		case c >= 'a' && c <= 'z':
+		case c >= 'A' && c <= 'Z':
+		case c == '_':
+		case i > 0 && c >= '0' && c <= '9':
+
+		default:
+			return false
+		}
+	}
+
+	return true
+}
+
+func sanitizeTracefsPath(path ...string) (string, error) {
+	base, err := getTracefsPath()
+	if err != nil {
+		return "", err
+	}
+	l := filepath.Join(path...)
+	p := filepath.Join(base, l)
+	if !strings.HasPrefix(p, base) {
+		return "", fmt.Errorf("path '%s' attempts to escape base path '%s': %w", l, base, ErrInvalidInput)
+	}
+	return p, nil
+}
+
+// getTracefsPath will return a correct path to the tracefs mount point.
+// Since kernel 4.1 tracefs should be mounted by default at /sys/kernel/tracing,
+// but may be also be available at /sys/kernel/debug/tracing if debugfs is mounted.
+// The available tracefs paths will depends on distribution choices.
+var getTracefsPath = internal.Memoize(func() (string, error) {
+	for _, p := range []struct {
+		path   string
+		fsType int64
+	}{
+		{"/sys/kernel/tracing", unix.TRACEFS_MAGIC},
+		{"/sys/kernel/debug/tracing", unix.TRACEFS_MAGIC},
+		// RHEL/CentOS
+		{"/sys/kernel/debug/tracing", unix.DEBUGFS_MAGIC},
+	} {
+		if fsType, err := internal.FSType(p.path); err == nil && fsType == p.fsType {
+			return p.path, nil
+		}
+	}
+
+	return "", errors.New("neither debugfs nor tracefs are mounted")
+})
+
+// sanitizeIdentifier replaces every invalid character for the tracefs api with an underscore.
+//
+// It is equivalent to calling regexp.MustCompile("[^a-zA-Z0-9]+").ReplaceAllString("_").
+func sanitizeIdentifier(s string) string {
+	var skip bool
+	return strings.Map(func(c rune) rune {
+		switch {
+		case c >= 'a' && c <= 'z',
+			c >= 'A' && c <= 'Z',
+			c >= '0' && c <= '9':
+			skip = false
+			return c
+
+		case skip:
+			return -1
+
+		default:
+			skip = true
+			return '_'
+		}
+	}, s)
+}
+
+// EventID reads a trace event's ID from tracefs given its group and name.
+// The kernel requires group and name to be alphanumeric or underscore.
+func EventID(group, name string) (uint64, error) {
+	if !validIdentifier(group) {
+		return 0, fmt.Errorf("invalid tracefs group: %q", group)
+	}
+
+	if !validIdentifier(name) {
+		return 0, fmt.Errorf("invalid tracefs name: %q", name)
+	}
+
+	path, err := sanitizeTracefsPath("events", group, name, "id")
+	if err != nil {
+		return 0, err
+	}
+	tid, err := internal.ReadUint64FromFile("%d\n", path)
+	if errors.Is(err, os.ErrNotExist) {
+		return 0, err
+	}
+	if err != nil {
+		return 0, fmt.Errorf("reading trace event ID of %s/%s: %w", group, name, err)
+	}
+
+	return tid, nil
+}
+
+func probePrefix(ret bool, maxActive int) string {
+	if ret {
+		if maxActive > 0 {
+			return fmt.Sprintf("r%d", maxActive)
+		}
+		return "r"
+	}
+	return "p"
+}
+
+// Event represents an entry in a tracefs probe events file.
+type Event struct {
+	typ         ProbeType
+	group, name string
+	// event id allocated by the kernel. 0 if the event has already been removed.
+	id uint64
+}
+
+// NewEvent creates a new ephemeral trace event.
+//
+// Returns os.ErrNotExist if symbol is not a valid
+// kernel symbol, or if it is not traceable with kprobes. Returns os.ErrExist
+// if a probe with the same group and symbol already exists. Returns an error if
+// args.RetprobeMaxActive is used on non kprobe types. Returns ErrNotSupported if
+// the kernel is too old to support kretprobe maxactive.
+func NewEvent(args ProbeArgs) (*Event, error) {
+	// Before attempting to create a trace event through tracefs,
+	// check if an event with the same group and name already exists.
+	// Kernels 4.x and earlier don't return os.ErrExist on writing a duplicate
+	// entry, so we need to rely on reads for detecting uniqueness.
+	eventName := sanitizeIdentifier(args.Symbol)
+	_, err := EventID(args.Group, eventName)
+	if err == nil {
+		return nil, fmt.Errorf("trace event %s/%s: %w", args.Group, eventName, os.ErrExist)
+	}
+	if err != nil && !errors.Is(err, os.ErrNotExist) {
+		return nil, fmt.Errorf("checking trace event %s/%s: %w", args.Group, eventName, err)
+	}
+
+	// Open the kprobe_events file in tracefs.
+	f, err := args.Type.eventsFile()
+	if err != nil {
+		return nil, err
+	}
+	defer f.Close()
+
+	var pe, token string
+	switch args.Type {
+	case Kprobe:
+		// The kprobe_events syntax is as follows (see Documentation/trace/kprobetrace.txt):
+		// p[:[GRP/]EVENT] [MOD:]SYM[+offs]|MEMADDR [FETCHARGS] : Set a probe
+		// r[MAXACTIVE][:[GRP/]EVENT] [MOD:]SYM[+0] [FETCHARGS] : Set a return probe
+		// -:[GRP/]EVENT                                        : Clear a probe
+		//
+		// Some examples:
+		// r:ebpf_1234/r_my_kretprobe nf_conntrack_destroy
+		// p:ebpf_5678/p_my_kprobe __x64_sys_execve
+		//
+		// Leaving the kretprobe's MAXACTIVE set to 0 (or absent) will make the
+		// kernel default to NR_CPUS. This is desired in most eBPF cases since
+		// subsampling or rate limiting logic can be more accurately implemented in
+		// the eBPF program itself.
+		// See Documentation/kprobes.txt for more details.
+		if args.RetprobeMaxActive != 0 && !args.Ret {
+			return nil, ErrInvalidMaxActive
+		}
+		token = KprobeToken(args)
+		pe = fmt.Sprintf("%s:%s/%s %s", probePrefix(args.Ret, args.RetprobeMaxActive), args.Group, eventName, token)
+	case Uprobe:
+		// The uprobe_events syntax is as follows:
+		// p[:[GRP/]EVENT] PATH:OFFSET [FETCHARGS] : Set a probe
+		// r[:[GRP/]EVENT] PATH:OFFSET [FETCHARGS] : Set a return probe
+		// -:[GRP/]EVENT                           : Clear a probe
+		//
+		// Some examples:
+		// r:ebpf_1234/readline /bin/bash:0x12345
+		// p:ebpf_5678/main_mySymbol /bin/mybin:0x12345(0x123)
+		//
+		// See Documentation/trace/uprobetracer.txt for more details.
+		if args.RetprobeMaxActive != 0 {
+			return nil, ErrInvalidMaxActive
+		}
+		token = UprobeToken(args)
+		pe = fmt.Sprintf("%s:%s/%s %s", probePrefix(args.Ret, 0), args.Group, eventName, token)
+	}
+	_, err = f.WriteString(pe)
+
+	// Since commit 97c753e62e6c, ENOENT is correctly returned instead of EINVAL
+	// when trying to create a retprobe for a missing symbol.
+	if errors.Is(err, os.ErrNotExist) {
+		return nil, fmt.Errorf("token %s: not found: %w", token, err)
+	}
+	// Since commit ab105a4fb894, EILSEQ is returned when a kprobe sym+offset is resolved
+	// to an invalid insn boundary. The exact conditions that trigger this error are
+	// arch specific however.
+	if errors.Is(err, syscall.EILSEQ) {
+		return nil, fmt.Errorf("token %s: bad insn boundary: %w", token, os.ErrNotExist)
+	}
+	// ERANGE is returned when the `SYM[+offs]` token is too big and cannot
+	// be resolved.
+	if errors.Is(err, syscall.ERANGE) {
+		return nil, fmt.Errorf("token %s: offset too big: %w", token, os.ErrNotExist)
+	}
+
+	if err != nil {
+		return nil, fmt.Errorf("token %s: writing '%s': %w", token, pe, err)
+	}
+
+	// Get the newly-created trace event's id.
+	tid, err := EventID(args.Group, eventName)
+	if args.RetprobeMaxActive != 0 && errors.Is(err, os.ErrNotExist) {
+		// Kernels < 4.12 don't support maxactive and therefore auto generate
+		// group and event names from the symbol and offset. The symbol is used
+		// without any sanitization.
+		// See https://elixir.bootlin.com/linux/v4.10/source/kernel/trace/trace_kprobe.c#L712
+		event := fmt.Sprintf("kprobes/r_%s_%d", args.Symbol, args.Offset)
+		if err := removeEvent(args.Type, event); err != nil {
+			return nil, fmt.Errorf("failed to remove spurious maxactive event: %s", err)
+		}
+		return nil, fmt.Errorf("create trace event with non-default maxactive: %w", internal.ErrNotSupported)
+	}
+	if err != nil {
+		return nil, fmt.Errorf("get trace event id: %w", err)
+	}
+
+	evt := &Event{args.Type, args.Group, eventName, tid}
+	runtime.SetFinalizer(evt, (*Event).Close)
+	return evt, nil
+}
+
+// Close removes the event from tracefs.
+//
+// Returns os.ErrClosed if the event has already been closed before.
+func (evt *Event) Close() error {
+	if evt.id == 0 {
+		return os.ErrClosed
+	}
+
+	evt.id = 0
+	runtime.SetFinalizer(evt, nil)
+	pe := fmt.Sprintf("%s/%s", evt.group, evt.name)
+	return removeEvent(evt.typ, pe)
+}
+
+func removeEvent(typ ProbeType, pe string) error {
+	f, err := typ.eventsFile()
+	if err != nil {
+		return err
+	}
+	defer f.Close()
+
+	// See [k,u]probe_events syntax above. The probe type does not need to be specified
+	// for removals.
+	if _, err = f.WriteString("-:" + pe); err != nil {
+		return fmt.Errorf("remove event %q from %s: %w", pe, f.Name(), err)
+	}
+
+	return nil
+}
+
+// ID returns the tracefs ID associated with the event.
+func (evt *Event) ID() uint64 {
+	return evt.id
+}
+
+// Group returns the tracefs group used by the event.
+func (evt *Event) Group() string {
+	return evt.group
+}
+
+// KprobeToken creates the SYM[+offs] token for the tracefs api.
+func KprobeToken(args ProbeArgs) string {
+	po := args.Symbol
+
+	if args.Offset != 0 {
+		po += fmt.Sprintf("+%#x", args.Offset)
+	}
+
+	return po
+}

vendor/github.com/cilium/ebpf/internal/tracefs/probetype_string.go

@@ -0,0 +1,24 @@
+// Code generated by "stringer -type=ProbeType -linecomment"; DO NOT EDIT.
+
+package tracefs
+
+import "strconv"
+
+func _() {
+	// An "invalid array index" compiler error signifies that the constant values have changed.
+	// Re-run the stringer command to generate them again.
+	var x [1]struct{}
+	_ = x[Kprobe-0]
+	_ = x[Uprobe-1]
+}
+
+const _ProbeType_name = "kprobeuprobe"
+
+var _ProbeType_index = [...]uint8{0, 6, 12}
+
+func (i ProbeType) String() string {
+	if i >= ProbeType(len(_ProbeType_index)-1) {
+		return "ProbeType(" + strconv.FormatInt(int64(i), 10) + ")"
+	}
+	return _ProbeType_name[_ProbeType_index[i]:_ProbeType_index[i+1]]
+}

vendor/github.com/cilium/ebpf/internal/tracefs/uprobe.go

@@ -0,0 +1,16 @@
+package tracefs
+
+import "fmt"
+
+// UprobeToken creates the PATH:OFFSET(REF_CTR_OFFSET) token for the tracefs api.
+func UprobeToken(args ProbeArgs) string {
+	po := fmt.Sprintf("%s:%#x", args.Path, args.Offset)
+
+	if args.RefCtrOffset != 0 {
+		// This is not documented in Documentation/trace/uprobetracer.txt.
+		// elixir.bootlin.com/linux/v5.15-rc7/source/kernel/trace/trace.c#L5564
+		po += fmt.Sprintf("(%#x)", args.RefCtrOffset)
+	}
+
+	return po
+}

vendor/github.com/cilium/ebpf/internal/unix/doc.go

@@ -0,0 +1,11 @@
+// Package unix re-exports Linux specific parts of golang.org/x/sys/unix.
+//
+// It avoids breaking compilation on other OS by providing stubs as follows:
+//   - Invoking a function always returns an error.
+//   - Errnos have distinct, non-zero values.
+//   - Constants have distinct but meaningless values.
+//   - Types use the same names for members, but may or may not follow the
+//     Linux layout.
+package unix
+
+// Note: please don't add any custom API to this package. Use internal/sys instead.

vendor/github.com/cilium/ebpf/internal/unix/types_linux.go

@@ -1,5 +1,4 @@
 //go:build linux
-// +build linux
 
 package unix
 
@@ -24,9 +23,11 @@ const (
 	E2BIG      = linux.E2BIG
 	EFAULT     = linux.EFAULT
 	EACCES     = linux.EACCES
-	// ENOTSUPP is not the same as ENOTSUP or EOPNOTSUP
+	EILSEQ     = linux.EILSEQ
-	ENOTSUPP = syscall.Errno(0x20c)
+	EOPNOTSUPP = linux.EOPNOTSUPP
+)
 
+const (
 	BPF_F_NO_PREALLOC         = linux.BPF_F_NO_PREALLOC
 	BPF_F_NUMA_NODE           = linux.BPF_F_NUMA_NODE
 	BPF_F_RDONLY              = linux.BPF_F_RDONLY
@@ -34,8 +35,10 @@ const (
 	BPF_F_RDONLY_PROG         = linux.BPF_F_RDONLY_PROG
 	BPF_F_WRONLY_PROG         = linux.BPF_F_WRONLY_PROG
 	BPF_F_SLEEPABLE           = linux.BPF_F_SLEEPABLE
+	BPF_F_XDP_HAS_FRAGS       = linux.BPF_F_XDP_HAS_FRAGS
 	BPF_F_MMAPABLE            = linux.BPF_F_MMAPABLE
 	BPF_F_INNER_MAP           = linux.BPF_F_INNER_MAP
+	BPF_F_KPROBE_MULTI_RETURN = linux.BPF_F_KPROBE_MULTI_RETURN
 	BPF_OBJ_NAME_LEN          = linux.BPF_OBJ_NAME_LEN
 	BPF_TAG_SIZE              = linux.BPF_TAG_SIZE
 	BPF_RINGBUF_BUSY_BIT      = linux.BPF_RINGBUF_BUSY_BIT
@@ -47,9 +50,12 @@ const (
 	EPOLL_CLOEXEC             = linux.EPOLL_CLOEXEC
 	O_CLOEXEC                 = linux.O_CLOEXEC
 	O_NONBLOCK                = linux.O_NONBLOCK
+	PROT_NONE                 = linux.PROT_NONE
 	PROT_READ                 = linux.PROT_READ
 	PROT_WRITE                = linux.PROT_WRITE
+	MAP_ANON                  = linux.MAP_ANON
 	MAP_SHARED                = linux.MAP_SHARED
+	MAP_PRIVATE               = linux.MAP_PRIVATE
 	PERF_ATTR_SIZE_VER1       = linux.PERF_ATTR_SIZE_VER1
 	PERF_TYPE_SOFTWARE        = linux.PERF_TYPE_SOFTWARE
 	PERF_TYPE_TRACEPOINT      = linux.PERF_TYPE_TRACEPOINT
@@ -58,6 +64,7 @@ const (
 	PERF_EVENT_IOC_ENABLE     = linux.PERF_EVENT_IOC_ENABLE
 	PERF_EVENT_IOC_SET_BPF    = linux.PERF_EVENT_IOC_SET_BPF
 	PerfBitWatermark          = linux.PerfBitWatermark
+	PerfBitWriteBackward      = linux.PerfBitWriteBackward
 	PERF_SAMPLE_RAW           = linux.PERF_SAMPLE_RAW
 	PERF_FLAG_FD_CLOEXEC      = linux.PERF_FLAG_FD_CLOEXEC
 	RLIM_INFINITY             = linux.RLIM_INFINITY
@@ -70,129 +77,110 @@ const (
 	SO_ATTACH_BPF             = linux.SO_ATTACH_BPF
 	SO_DETACH_BPF             = linux.SO_DETACH_BPF
 	SOL_SOCKET                = linux.SOL_SOCKET
+	SIGPROF                   = linux.SIGPROF
+	SIG_BLOCK                 = linux.SIG_BLOCK
+	SIG_UNBLOCK               = linux.SIG_UNBLOCK
+	EM_NONE                   = linux.EM_NONE
+	EM_BPF                    = linux.EM_BPF
+	BPF_FS_MAGIC              = linux.BPF_FS_MAGIC
+	TRACEFS_MAGIC             = linux.TRACEFS_MAGIC
+	DEBUGFS_MAGIC             = linux.DEBUGFS_MAGIC
 )
 
-// Statfs_t is a wrapper
 type Statfs_t = linux.Statfs_t
-
 type Stat_t = linux.Stat_t
-
-// Rlimit is a wrapper
 type Rlimit = linux.Rlimit
+type Signal = linux.Signal
+type Sigset_t = linux.Sigset_t
+type PerfEventMmapPage = linux.PerfEventMmapPage
+type EpollEvent = linux.EpollEvent
+type PerfEventAttr = linux.PerfEventAttr
+type Utsname = linux.Utsname
 
-// Syscall is a wrapper
 func Syscall(trap, a1, a2, a3 uintptr) (r1, r2 uintptr, err syscall.Errno) {
 	return linux.Syscall(trap, a1, a2, a3)
 }
 
-// FcntlInt is a wrapper
+func PthreadSigmask(how int, set, oldset *Sigset_t) error {
+	return linux.PthreadSigmask(how, set, oldset)
+}
+
 func FcntlInt(fd uintptr, cmd, arg int) (int, error) {
 	return linux.FcntlInt(fd, cmd, arg)
 }
 
-// IoctlSetInt is a wrapper
 func IoctlSetInt(fd int, req uint, value int) error {
 	return linux.IoctlSetInt(fd, req, value)
 }
 
-// Statfs is a wrapper
 func Statfs(path string, buf *Statfs_t) (err error) {
 	return linux.Statfs(path, buf)
 }
 
-// Close is a wrapper
 func Close(fd int) (err error) {
 	return linux.Close(fd)
 }
 
-// EpollEvent is a wrapper
-type EpollEvent = linux.EpollEvent
-
-// EpollWait is a wrapper
 func EpollWait(epfd int, events []EpollEvent, msec int) (n int, err error) {
 	return linux.EpollWait(epfd, events, msec)
 }
 
-// EpollCtl is a wrapper
 func EpollCtl(epfd int, op int, fd int, event *EpollEvent) (err error) {
 	return linux.EpollCtl(epfd, op, fd, event)
 }
 
-// Eventfd is a wrapper
 func Eventfd(initval uint, flags int) (fd int, err error) {
 	return linux.Eventfd(initval, flags)
 }
 
-// Write is a wrapper
 func Write(fd int, p []byte) (n int, err error) {
 	return linux.Write(fd, p)
 }
 
-// EpollCreate1 is a wrapper
 func EpollCreate1(flag int) (fd int, err error) {
 	return linux.EpollCreate1(flag)
 }
 
-// PerfEventMmapPage is a wrapper
-type PerfEventMmapPage linux.PerfEventMmapPage
-
-// SetNonblock is a wrapper
 func SetNonblock(fd int, nonblocking bool) (err error) {
 	return linux.SetNonblock(fd, nonblocking)
 }
 
-// Mmap is a wrapper
 func Mmap(fd int, offset int64, length int, prot int, flags int) (data []byte, err error) {
 	return linux.Mmap(fd, offset, length, prot, flags)
 }
 
-// Munmap is a wrapper
 func Munmap(b []byte) (err error) {
 	return linux.Munmap(b)
 }
 
-// PerfEventAttr is a wrapper
-type PerfEventAttr = linux.PerfEventAttr
-
-// PerfEventOpen is a wrapper
 func PerfEventOpen(attr *PerfEventAttr, pid int, cpu int, groupFd int, flags int) (fd int, err error) {
 	return linux.PerfEventOpen(attr, pid, cpu, groupFd, flags)
 }
 
-// Utsname is a wrapper
-type Utsname = linux.Utsname
-
-// Uname is a wrapper
 func Uname(buf *Utsname) (err error) {
 	return linux.Uname(buf)
 }
 
-// Getpid is a wrapper
 func Getpid() int {
 	return linux.Getpid()
 }
 
-// Gettid is a wrapper
 func Gettid() int {
 	return linux.Gettid()
 }
 
-// Tgkill is a wrapper
 func Tgkill(tgid int, tid int, sig syscall.Signal) (err error) {
 	return linux.Tgkill(tgid, tid, sig)
 }
 
-// BytePtrFromString is a wrapper
 func BytePtrFromString(s string) (*byte, error) {
 	return linux.BytePtrFromString(s)
 }
 
-// ByteSliceToString is a wrapper
 func ByteSliceToString(s []byte) string {
 	return linux.ByteSliceToString(s)
 }
 
-// Renameat2 is a wrapper
 func Renameat2(olddirfd int, oldpath string, newdirfd int, newpath string, flags uint) error {
 	return linux.Renameat2(olddirfd, oldpath, newdirfd, newpath, flags)
 }
@@ -208,3 +196,7 @@ func Open(path string, mode int, perm uint32) (int, error) {
 func Fstat(fd int, stat *Stat_t) error {
 	return linux.Fstat(fd, stat)
 }
+
+func SetsockoptInt(fd, level, opt, value int) error {
+	return linux.SetsockoptInt(fd, level, opt, value)
+}

vendor/github.com/cilium/ebpf/internal/unix/types_other.go

@@ -1,5 +1,4 @@
 //go:build !linux
-// +build !linux
 
 package unix
 
@@ -11,70 +10,87 @@ import (
 
 var errNonLinux = fmt.Errorf("unsupported platform %s/%s", runtime.GOOS, runtime.GOARCH)
 
+// Errnos are distinct and non-zero.
 const (
-	ENOENT = syscall.ENOENT
+	ENOENT syscall.Errno = iota + 1
-	EEXIST = syscall.EEXIST
+	EEXIST
-	EAGAIN = syscall.EAGAIN
+	EAGAIN
-	ENOSPC = syscall.ENOSPC
+	ENOSPC
-	EINVAL = syscall.EINVAL
+	EINVAL
-	EINTR  = syscall.EINTR
+	EINTR
-	EPERM  = syscall.EPERM
+	EPERM
-	ESRCH  = syscall.ESRCH
+	ESRCH
-	ENODEV = syscall.ENODEV
+	ENODEV
-	EBADF  = syscall.Errno(0)
+	EBADF
-	E2BIG  = syscall.Errno(0)
+	E2BIG
-	EFAULT = syscall.EFAULT
+	EFAULT
-	EACCES = syscall.Errno(0)
+	EACCES
-	// ENOTSUPP is not the same as ENOTSUP or EOPNOTSUP
+	EILSEQ
-	ENOTSUPP = syscall.Errno(0x20c)
+	EOPNOTSUPP
-
+)
-	BPF_F_NO_PREALLOC        = 0
+
-	BPF_F_NUMA_NODE          = 0
+// Constants are distinct to avoid breaking switch statements.
-	BPF_F_RDONLY             = 0
+const (
-	BPF_F_WRONLY             = 0
+	BPF_F_NO_PREALLOC = iota
-	BPF_F_RDONLY_PROG        = 0
+	BPF_F_NUMA_NODE
-	BPF_F_WRONLY_PROG        = 0
+	BPF_F_RDONLY
-	BPF_F_SLEEPABLE          = 0
+	BPF_F_WRONLY
-	BPF_F_MMAPABLE           = 0
+	BPF_F_RDONLY_PROG
-	BPF_F_INNER_MAP          = 0
+	BPF_F_WRONLY_PROG
-	BPF_OBJ_NAME_LEN         = 0x10
+	BPF_F_SLEEPABLE
-	BPF_TAG_SIZE             = 0x8
+	BPF_F_MMAPABLE
-	BPF_RINGBUF_BUSY_BIT     = 0
+	BPF_F_INNER_MAP
-	BPF_RINGBUF_DISCARD_BIT  = 0
+	BPF_F_KPROBE_MULTI_RETURN
-	BPF_RINGBUF_HDR_SZ       = 0
+	BPF_F_XDP_HAS_FRAGS
-	SYS_BPF                  = 321
+	BPF_OBJ_NAME_LEN
-	F_DUPFD_CLOEXEC          = 0x406
+	BPF_TAG_SIZE
-	EPOLLIN                  = 0x1
+	BPF_RINGBUF_BUSY_BIT
-	EPOLL_CTL_ADD            = 0x1
+	BPF_RINGBUF_DISCARD_BIT
-	EPOLL_CLOEXEC            = 0x80000
+	BPF_RINGBUF_HDR_SZ
-	O_CLOEXEC                = 0x80000
+	SYS_BPF
-	O_NONBLOCK               = 0x800
+	F_DUPFD_CLOEXEC
-	PROT_READ                = 0x1
+	EPOLLIN
-	PROT_WRITE               = 0x2
+	EPOLL_CTL_ADD
-	MAP_SHARED               = 0x1
+	EPOLL_CLOEXEC
-	PERF_ATTR_SIZE_VER1      = 0
+	O_CLOEXEC
-	PERF_TYPE_SOFTWARE       = 0x1
+	O_NONBLOCK
-	PERF_TYPE_TRACEPOINT     = 0
+	PROT_NONE
-	PERF_COUNT_SW_BPF_OUTPUT = 0xa
+	PROT_READ
-	PERF_EVENT_IOC_DISABLE   = 0
+	PROT_WRITE
-	PERF_EVENT_IOC_ENABLE    = 0
+	MAP_ANON
-	PERF_EVENT_IOC_SET_BPF   = 0
+	MAP_SHARED
-	PerfBitWatermark         = 0x4000
+	MAP_PRIVATE
-	PERF_SAMPLE_RAW          = 0x400
+	PERF_ATTR_SIZE_VER1
-	PERF_FLAG_FD_CLOEXEC     = 0x8
+	PERF_TYPE_SOFTWARE
-	RLIM_INFINITY            = 0x7fffffffffffffff
+	PERF_TYPE_TRACEPOINT
-	RLIMIT_MEMLOCK           = 8
+	PERF_COUNT_SW_BPF_OUTPUT
-	BPF_STATS_RUN_TIME       = 0
+	PERF_EVENT_IOC_DISABLE
-	PERF_RECORD_LOST         = 2
+	PERF_EVENT_IOC_ENABLE
-	PERF_RECORD_SAMPLE       = 9
+	PERF_EVENT_IOC_SET_BPF
-	AT_FDCWD                 = -0x2
+	PerfBitWatermark
-	RENAME_NOREPLACE         = 0x1
+	PerfBitWriteBackward
-	SO_ATTACH_BPF            = 0x32
+	PERF_SAMPLE_RAW
-	SO_DETACH_BPF            = 0x1b
+	PERF_FLAG_FD_CLOEXEC
-	SOL_SOCKET               = 0x1
+	RLIM_INFINITY
+	RLIMIT_MEMLOCK
+	BPF_STATS_RUN_TIME
+	PERF_RECORD_LOST
+	PERF_RECORD_SAMPLE
+	AT_FDCWD
+	RENAME_NOREPLACE
+	SO_ATTACH_BPF
+	SO_DETACH_BPF
+	SOL_SOCKET
+	SIGPROF
+	SIG_BLOCK
+	SIG_UNBLOCK
+	EM_NONE
+	EM_BPF
+	BPF_FS_MAGIC
+	TRACEFS_MAGIC
+	DEBUGFS_MAGIC
 )
 
-// Statfs_t is a wrapper
 type Statfs_t struct {
 	Type    int64
 	Bsize   int64
@@ -90,72 +106,81 @@ type Statfs_t struct {
 	Spare   [4]int64
 }
 
-type Stat_t struct{}
+type Stat_t struct {
+	Dev     uint64
+	Ino     uint64
+	Nlink   uint64
+	Mode    uint32
+	Uid     uint32
+	Gid     uint32
+	_       int32
+	Rdev    uint64
+	Size    int64
+	Blksize int64
+	Blocks  int64
+}
 
-// Rlimit is a wrapper
 type Rlimit struct {
 	Cur uint64
 	Max uint64
 }
 
-// Syscall is a wrapper
+type Signal int
-func Syscall(trap, a1, a2, a3 uintptr) (r1, r2 uintptr, err syscall.Errno) {
+
-	return 0, 0, syscall.Errno(1)
+type Sigset_t struct {
+	Val [4]uint64
+}
+
+func Syscall(trap, a1, a2, a3 uintptr) (r1, r2 uintptr, err syscall.Errno) {
+	return 0, 0, syscall.ENOTSUP
+}
+
+func PthreadSigmask(how int, set, oldset *Sigset_t) error {
+	return errNonLinux
 }
 
-// FcntlInt is a wrapper
 func FcntlInt(fd uintptr, cmd, arg int) (int, error) {
 	return -1, errNonLinux
 }
 
-// IoctlSetInt is a wrapper
 func IoctlSetInt(fd int, req uint, value int) error {
 	return errNonLinux
 }
 
-// Statfs is a wrapper
 func Statfs(path string, buf *Statfs_t) error {
 	return errNonLinux
 }
 
-// Close is a wrapper
 func Close(fd int) (err error) {
 	return errNonLinux
 }
 
-// EpollEvent is a wrapper
 type EpollEvent struct {
 	Events uint32
 	Fd     int32
 	Pad    int32
 }
 
-// EpollWait is a wrapper
 func EpollWait(epfd int, events []EpollEvent, msec int) (n int, err error) {
 	return 0, errNonLinux
 }
 
-// EpollCtl is a wrapper
 func EpollCtl(epfd int, op int, fd int, event *EpollEvent) (err error) {
 	return errNonLinux
 }
 
-// Eventfd is a wrapper
 func Eventfd(initval uint, flags int) (fd int, err error) {
 	return 0, errNonLinux
 }
 
-// Write is a wrapper
 func Write(fd int, p []byte) (n int, err error) {
 	return 0, errNonLinux
 }
 
-// EpollCreate1 is a wrapper
 func EpollCreate1(flag int) (fd int, err error) {
 	return 0, errNonLinux
 }
 
-// PerfEventMmapPage is a wrapper
 type PerfEventMmapPage struct {
 	Version        uint32
 	Compat_version uint32
@@ -182,22 +207,18 @@ type PerfEventMmapPage struct {
 	Aux_size    uint64
 }
 
-// SetNonblock is a wrapper
 func SetNonblock(fd int, nonblocking bool) (err error) {
 	return errNonLinux
 }
 
-// Mmap is a wrapper
 func Mmap(fd int, offset int64, length int, prot int, flags int) (data []byte, err error) {
 	return []byte{}, errNonLinux
 }
 
-// Munmap is a wrapper
 func Munmap(b []byte) (err error) {
 	return errNonLinux
 }
 
-// PerfEventAttr is a wrapper
 type PerfEventAttr struct {
 	Type               uint32
 	Size               uint32
@@ -219,48 +240,39 @@ type PerfEventAttr struct {
 	Sample_max_stack   uint16
 }
 
-// PerfEventOpen is a wrapper
 func PerfEventOpen(attr *PerfEventAttr, pid int, cpu int, groupFd int, flags int) (fd int, err error) {
 	return 0, errNonLinux
 }
 
-// Utsname is a wrapper
 type Utsname struct {
 	Release [65]byte
 	Version [65]byte
 }
 
-// Uname is a wrapper
 func Uname(buf *Utsname) (err error) {
 	return errNonLinux
 }
 
-// Getpid is a wrapper
 func Getpid() int {
 	return -1
 }
 
-// Gettid is a wrapper
 func Gettid() int {
 	return -1
 }
 
-// Tgkill is a wrapper
 func Tgkill(tgid int, tid int, sig syscall.Signal) (err error) {
 	return errNonLinux
 }
 
-// BytePtrFromString is a wrapper
 func BytePtrFromString(s string) (*byte, error) {
 	return nil, errNonLinux
 }
 
-// ByteSliceToString is a wrapper
 func ByteSliceToString(s []byte) string {
 	return ""
 }
 
-// Renameat2 is a wrapper
 func Renameat2(olddirfd int, oldpath string, newdirfd int, newpath string, flags uint) error {
 	return errNonLinux
 }
@@ -276,3 +288,7 @@ func Open(path string, mode int, perm uint32) (int, error) {
 func Fstat(fd int, stat *Stat_t) error {
 	return errNonLinux
 }
+
+func SetsockoptInt(fd, level, opt, value int) error {
+	return errNonLinux
+}

vendor/github.com/cilium/ebpf/internal/vdso.go

@@ -23,6 +23,9 @@ func vdsoVersion() (uint32, error) {
 	// to the process. Go does not expose that data, so we must read it from procfs.
 	// https://man7.org/linux/man-pages/man3/getauxval.3.html
 	av, err := os.Open("/proc/self/auxv")
+	if errors.Is(err, unix.EACCES) {
+		return 0, fmt.Errorf("opening auxv: %w (process may not be dumpable due to file capabilities)", err)
+	}
 	if err != nil {
 		return 0, fmt.Errorf("opening auxv: %w", err)
 	}
@@ -117,7 +120,7 @@ func vdsoLinuxVersionCode(r io.ReaderAt) (uint32, error) {
 			var name string
 			if n.NameSize > 0 {
 				// Read the note name, aligned to 4 bytes.
-				buf := make([]byte, Align(int(n.NameSize), 4))
+				buf := make([]byte, Align(n.NameSize, 4))
 				if err := binary.Read(sr, hdr.ByteOrder, &buf); err != nil {
 					return 0, fmt.Errorf("reading note name: %w", err)
 				}
@@ -139,7 +142,7 @@ func vdsoLinuxVersionCode(r io.ReaderAt) (uint32, error) {
 				}
 
 				// Discard the note descriptor if it exists but we're not interested in it.
-				if _, err := io.CopyN(io.Discard, sr, int64(Align(int(n.DescSize), 4))); err != nil {
+				if _, err := io.CopyN(io.Discard, sr, int64(Align(n.DescSize, 4))); err != nil {
 					return 0, err
 				}
 			}

vendor/github.com/cilium/ebpf/internal/version.go

@@ -2,7 +2,6 @@ package internal
 
 import (
 	"fmt"
-	"sync"
 
 	"github.com/cilium/ebpf/internal/unix"
 )
@@ -15,14 +14,6 @@ const (
 	MagicKernelVersion = 0xFFFFFFFE
 )
 
-var (
-	kernelVersion = struct {
-		once    sync.Once
-		version Version
-		err     error
-	}{}
-)
-
 // A Version in the form Major.Minor.Patch.
 type Version [3]uint16
 
@@ -88,16 +79,9 @@ func (v Version) Kernel() uint32 {
 }
 
 // KernelVersion returns the version of the currently running kernel.
-func KernelVersion() (Version, error) {
+var KernelVersion = Memoize(func() (Version, error) {
-	kernelVersion.once.Do(func() {
+	return detectKernelVersion()
-		kernelVersion.version, kernelVersion.err = detectKernelVersion()
+})
-	})
-
-	if kernelVersion.err != nil {
-		return Version{}, kernelVersion.err
-	}
-	return kernelVersion.version, nil
-}
 
 // detectKernelVersion returns the version of the running kernel.
 func detectKernelVersion() (Version, error) {

vendor/github.com/cilium/ebpf/link/cgroup.go

@@ -10,10 +10,15 @@ import (
 
 type cgroupAttachFlags uint32
 
-// cgroup attach flags
 const (
+	// Allow programs attached to sub-cgroups to override the verdict of this
+	// program.
 	flagAllowOverride cgroupAttachFlags = 1 << iota
+	// Allow attaching multiple programs to the cgroup. Only works if the cgroup
+	// has zero or more programs attached using the Multi flag. Implies override.
 	flagAllowMulti
+	// Set automatically by progAttachCgroup.Update(). Used for updating a
+	// specific given program attached in multi-mode.
 	flagReplace
 )
 
@@ -27,29 +32,39 @@ type CgroupOptions struct {
 }
 
 // AttachCgroup links a BPF program to a cgroup.
-func AttachCgroup(opts CgroupOptions) (Link, error) {
+//
+// If the running kernel doesn't support bpf_link, attempts to emulate its
+// semantics using the legacy PROG_ATTACH mechanism. If bpf_link is not
+// available, the returned [Link] will not support pinning to bpffs.
+//
+// If you need more control over attachment flags or the attachment mechanism
+// used, look at [RawAttachProgram] and [AttachRawLink] instead.
+func AttachCgroup(opts CgroupOptions) (cg Link, err error) {
 	cgroup, err := os.Open(opts.Path)
 	if err != nil {
 		return nil, fmt.Errorf("can't open cgroup: %s", err)
 	}
-
+	defer func() {
-	clone, err := opts.Program.Clone()
+		if _, ok := cg.(*progAttachCgroup); ok {
-	if err != nil {
+			// Skip closing the cgroup handle if we return a valid progAttachCgroup,
+			// where the handle is retained to implement Update().
+			return
+		}
 		cgroup.Close()
-		return nil, err
+	}()
+
+	cg, err = newLinkCgroup(cgroup, opts.Attach, opts.Program)
+	if err == nil {
+		return cg, nil
 	}
 
-	var cg Link
-	cg, err = newLinkCgroup(cgroup, opts.Attach, clone)
 	if errors.Is(err, ErrNotSupported) {
-		cg, err = newProgAttachCgroup(cgroup, opts.Attach, clone, flagAllowMulti)
+		cg, err = newProgAttachCgroup(cgroup, opts.Attach, opts.Program, flagAllowMulti)
 	}
 	if errors.Is(err, ErrNotSupported) {
-		cg, err = newProgAttachCgroup(cgroup, opts.Attach, clone, flagAllowOverride)
+		cg, err = newProgAttachCgroup(cgroup, opts.Attach, opts.Program, flagAllowOverride)
 	}
 	if err != nil {
-		cgroup.Close()
-		clone.Close()
 		return nil, err
 	}
 
@@ -67,6 +82,8 @@ var _ Link = (*progAttachCgroup)(nil)
 
 func (cg *progAttachCgroup) isLink() {}
 
+// newProgAttachCgroup attaches prog to cgroup using BPF_PROG_ATTACH.
+// cgroup and prog are retained by [progAttachCgroup].
 func newProgAttachCgroup(cgroup *os.File, attach ebpf.AttachType, prog *ebpf.Program, flags cgroupAttachFlags) (*progAttachCgroup, error) {
 	if flags&flagAllowMulti > 0 {
 		if err := haveProgAttachReplace(); err != nil {
@@ -74,17 +91,24 @@ func newProgAttachCgroup(cgroup *os.File, attach ebpf.AttachType, prog *ebpf.Pro
 		}
 	}
 
-	err := RawAttachProgram(RawAttachProgramOptions{
+	// Use a program handle that cannot be closed by the caller.
+	clone, err := prog.Clone()
+	if err != nil {
+		return nil, err
+	}
+
+	err = RawAttachProgram(RawAttachProgramOptions{
 		Target:  int(cgroup.Fd()),
-		Program: prog,
+		Program: clone,
 		Flags:   uint32(flags),
 		Attach:  attach,
 	})
 	if err != nil {
+		clone.Close()
 		return nil, fmt.Errorf("cgroup: %w", err)
 	}
 
-	return &progAttachCgroup{cgroup, prog, attach, flags}, nil
+	return &progAttachCgroup{cgroup, clone, attach, flags}, nil
 }
 
 func (cg *progAttachCgroup) Close() error {
@@ -138,7 +162,7 @@ func (cg *progAttachCgroup) Pin(string) error {
 }
 
 func (cg *progAttachCgroup) Unpin() error {
-	return fmt.Errorf("can't pin cgroup: %w", ErrNotSupported)
+	return fmt.Errorf("can't unpin cgroup: %w", ErrNotSupported)
 }
 
 func (cg *progAttachCgroup) Info() (*Info, error) {
@@ -151,6 +175,7 @@ type linkCgroup struct {
 
 var _ Link = (*linkCgroup)(nil)
 
+// newLinkCgroup attaches prog to cgroup using BPF_LINK_CREATE.
 func newLinkCgroup(cgroup *os.File, attach ebpf.AttachType, prog *ebpf.Program) (*linkCgroup, error) {
 	link, err := AttachRawLink(RawLinkOptions{
 		Target:  int(cgroup.Fd()),

vendor/github.com/cilium/ebpf/link/kprobe.go

@@ -1,42 +1,20 @@
 package link
 
 import (
-	"bytes"
-	"crypto/rand"
 	"errors"
 	"fmt"
 	"os"
-	"path/filepath"
 	"runtime"
 	"strings"
-	"sync"
-	"syscall"
 	"unsafe"
 
 	"github.com/cilium/ebpf"
+	"github.com/cilium/ebpf/internal"
 	"github.com/cilium/ebpf/internal/sys"
+	"github.com/cilium/ebpf/internal/tracefs"
 	"github.com/cilium/ebpf/internal/unix"
 )
 
-var (
-	kprobeEventsPath = filepath.Join(tracefsPath, "kprobe_events")
-
-	kprobeRetprobeBit = struct {
-		once  sync.Once
-		value uint64
-		err   error
-	}{}
-)
-
-type probeType uint8
-
-type probeArgs struct {
-	symbol, group, path          string
-	offset, refCtrOffset, cookie uint64
-	pid                          int
-	ret                          bool
-}
-
 // KprobeOptions defines additional parameters that will be used
 // when loading Kprobes.
 type KprobeOptions struct {
@@ -49,45 +27,23 @@ type KprobeOptions struct {
 	// Can be used to insert kprobes at arbitrary offsets in kernel functions,
 	// e.g. in places where functions have been inlined.
 	Offset uint64
+	// Increase the maximum number of concurrent invocations of a kretprobe.
+	// Required when tracing some long running functions in the kernel.
+	//
+	// Deprecated: this setting forces the use of an outdated kernel API and is not portable
+	// across kernel versions.
+	RetprobeMaxActive int
+	// Prefix used for the event name if the kprobe must be attached using tracefs.
+	// The group name will be formatted as `<prefix>_<randomstr>`.
+	// The default empty string is equivalent to "ebpf" as the prefix.
+	TraceFSPrefix string
 }
 
-const (
+func (ko *KprobeOptions) cookie() uint64 {
-	kprobeType probeType = iota
+	if ko == nil {
-	uprobeType
+		return 0
-)
-
-func (pt probeType) String() string {
-	if pt == kprobeType {
-		return "kprobe"
 	}
-	return "uprobe"
+	return ko.Cookie
-}
-
-func (pt probeType) EventsPath() string {
-	if pt == kprobeType {
-		return kprobeEventsPath
-	}
-	return uprobeEventsPath
-}
-
-func (pt probeType) PerfEventType(ret bool) perfEventType {
-	if pt == kprobeType {
-		if ret {
-			return kretprobeEvent
-		}
-		return kprobeEvent
-	}
-	if ret {
-		return uretprobeEvent
-	}
-	return uprobeEvent
-}
-
-func (pt probeType) RetprobeBit() (uint64, error) {
-	if pt == kprobeType {
-		return kretprobeBit()
-	}
-	return uretprobeBit()
 }
 
 // Kprobe attaches the given eBPF program to a perf event that fires when the
@@ -99,13 +55,17 @@ func (pt probeType) RetprobeBit() (uint64, error) {
 // Losing the reference to the resulting Link (kp) will close the Kprobe
 // and prevent further execution of prog. The Link must be Closed during
 // program shutdown to avoid leaking system resources.
+//
+// If attaching to symbol fails, automatically retries with the running
+// platform's syscall prefix (e.g. __x64_) to support attaching to syscalls
+// in a portable fashion.
 func Kprobe(symbol string, prog *ebpf.Program, opts *KprobeOptions) (Link, error) {
 	k, err := kprobe(symbol, prog, opts, false)
 	if err != nil {
 		return nil, err
 	}
 
-	lnk, err := attachPerfEvent(k, prog)
+	lnk, err := attachPerfEvent(k, prog, opts.cookie())
 	if err != nil {
 		k.Close()
 		return nil, err
@@ -123,13 +83,20 @@ func Kprobe(symbol string, prog *ebpf.Program, opts *KprobeOptions) (Link, error
 // Losing the reference to the resulting Link (kp) will close the Kretprobe
 // and prevent further execution of prog. The Link must be Closed during
 // program shutdown to avoid leaking system resources.
+//
+// If attaching to symbol fails, automatically retries with the running
+// platform's syscall prefix (e.g. __x64_) to support attaching to syscalls
+// in a portable fashion.
+//
+// On kernels 5.10 and earlier, setting a kretprobe on a nonexistent symbol
+// incorrectly returns unix.EINVAL instead of os.ErrNotExist.
 func Kretprobe(symbol string, prog *ebpf.Program, opts *KprobeOptions) (Link, error) {
 	k, err := kprobe(symbol, prog, opts, true)
 	if err != nil {
 		return nil, err
 	}
 
-	lnk, err := attachPerfEvent(k, prog)
+	lnk, err := attachPerfEvent(k, prog, opts.cookie())
 	if err != nil {
 		k.Close()
 		return nil, err
@@ -181,50 +148,51 @@ func kprobe(symbol string, prog *ebpf.Program, opts *KprobeOptions, ret bool) (*
 		return nil, fmt.Errorf("eBPF program type %s is not a Kprobe: %w", prog.Type(), errInvalidInput)
 	}
 
-	args := probeArgs{
+	args := tracefs.ProbeArgs{
-		pid:    perfAllThreads,
+		Type:   tracefs.Kprobe,
-		symbol: symbol,
+		Pid:    perfAllThreads,
-		ret:    ret,
+		Symbol: symbol,
+		Ret:    ret,
 	}
 
 	if opts != nil {
-		args.cookie = opts.Cookie
+		args.RetprobeMaxActive = opts.RetprobeMaxActive
-		args.offset = opts.Offset
+		args.Cookie = opts.Cookie
+		args.Offset = opts.Offset
+		args.Group = opts.TraceFSPrefix
 	}
 
 	// Use kprobe PMU if the kernel has it available.
-	tp, err := pmuKprobe(args)
+	tp, err := pmuProbe(args)
-	if errors.Is(err, os.ErrNotExist) {
+	if errors.Is(err, os.ErrNotExist) || errors.Is(err, unix.EINVAL) {
-		args.symbol = platformPrefix(symbol)
+		if prefix := internal.PlatformPrefix(); prefix != "" {
-		tp, err = pmuKprobe(args)
+			args.Symbol = prefix + symbol
+			tp, err = pmuProbe(args)
+		}
 	}
 	if err == nil {
 		return tp, nil
 	}
 	if err != nil && !errors.Is(err, ErrNotSupported) {
-		return nil, fmt.Errorf("creating perf_kprobe PMU: %w", err)
+		return nil, fmt.Errorf("creating perf_kprobe PMU (arch-specific fallback for %q): %w", symbol, err)
 	}
 
 	// Use tracefs if kprobe PMU is missing.
-	args.symbol = symbol
+	args.Symbol = symbol
-	tp, err = tracefsKprobe(args)
+	tp, err = tracefsProbe(args)
-	if errors.Is(err, os.ErrNotExist) {
+	if errors.Is(err, os.ErrNotExist) || errors.Is(err, unix.EINVAL) {
-		args.symbol = platformPrefix(symbol)
+		if prefix := internal.PlatformPrefix(); prefix != "" {
-		tp, err = tracefsKprobe(args)
+			args.Symbol = prefix + symbol
+			tp, err = tracefsProbe(args)
+		}
 	}
 	if err != nil {
-		return nil, fmt.Errorf("creating trace event '%s' in tracefs: %w", symbol, err)
+		return nil, fmt.Errorf("creating tracefs event (arch-specific fallback for %q): %w", symbol, err)
 	}
 
 	return tp, nil
 }
 
-// pmuKprobe opens a perf event based on the kprobe PMU.
-// Returns os.ErrNotExist if the given symbol does not exist in the kernel.
-func pmuKprobe(args probeArgs) (*perfEvent, error) {
-	return pmuProbe(kprobeType, args)
-}
-
 // pmuProbe opens a perf event based on a Performance Monitoring Unit.
 //
 // Requires at least a 4.17 kernel.
@@ -232,17 +200,25 @@ func pmuKprobe(args probeArgs) (*perfEvent, error) {
 // 33ea4b24277b "perf/core: Implement the 'perf_uprobe' PMU"
 //
 // Returns ErrNotSupported if the kernel doesn't support perf_[k,u]probe PMU
-func pmuProbe(typ probeType, args probeArgs) (*perfEvent, error) {
+func pmuProbe(args tracefs.ProbeArgs) (*perfEvent, error) {
 	// Getting the PMU type will fail if the kernel doesn't support
 	// the perf_[k,u]probe PMU.
-	et, err := getPMUEventType(typ)
+	eventType, err := internal.ReadUint64FromFileOnce("%d\n", "/sys/bus/event_source/devices", args.Type.String(), "type")
+	if errors.Is(err, os.ErrNotExist) {
+		return nil, fmt.Errorf("%s: %w", args.Type, ErrNotSupported)
+	}
 	if err != nil {
 		return nil, err
 	}
 
+	// Use tracefs if we want to set kretprobe's retprobeMaxActive.
+	if args.RetprobeMaxActive != 0 {
+		return nil, fmt.Errorf("pmu probe: non-zero retprobeMaxActive: %w", ErrNotSupported)
+	}
+
 	var config uint64
-	if args.ret {
+	if args.Ret {
-		bit, err := typ.RetprobeBit()
+		bit, err := internal.ReadUint64FromFileOnce("config:%d\n", "/sys/bus/event_source/devices", args.Type.String(), "/format/retprobe")
 		if err != nil {
 			return nil, err
 		}
@@ -252,73 +228,79 @@ func pmuProbe(typ probeType, args probeArgs) (*perfEvent, error) {
 	var (
 		attr  unix.PerfEventAttr
 		sp    unsafe.Pointer
+		token string
 	)
-	switch typ {
+	switch args.Type {
-	case kprobeType:
+	case tracefs.Kprobe:
 		// Create a pointer to a NUL-terminated string for the kernel.
-		sp, err = unsafeStringPtr(args.symbol)
+		sp, err = unsafeStringPtr(args.Symbol)
 		if err != nil {
 			return nil, err
 		}
 
+		token = tracefs.KprobeToken(args)
+
 		attr = unix.PerfEventAttr{
 			// The minimum size required for PMU kprobes is PERF_ATTR_SIZE_VER1,
 			// since it added the config2 (Ext2) field. Use Ext2 as probe_offset.
 			Size:   unix.PERF_ATTR_SIZE_VER1,
-			Type:   uint32(et),          // PMU event type read from sysfs
+			Type:   uint32(eventType),   // PMU event type read from sysfs
 			Ext1:   uint64(uintptr(sp)), // Kernel symbol to trace
-			Ext2:   args.offset,         // Kernel symbol offset
+			Ext2:   args.Offset,         // Kernel symbol offset
 			Config: config,              // Retprobe flag
 		}
-	case uprobeType:
+	case tracefs.Uprobe:
-		sp, err = unsafeStringPtr(args.path)
+		sp, err = unsafeStringPtr(args.Path)
 		if err != nil {
 			return nil, err
 		}
 
-		if args.refCtrOffset != 0 {
+		if args.RefCtrOffset != 0 {
-			config |= args.refCtrOffset << uprobeRefCtrOffsetShift
+			config |= args.RefCtrOffset << uprobeRefCtrOffsetShift
 		}
 
+		token = tracefs.UprobeToken(args)
+
 		attr = unix.PerfEventAttr{
 			// The minimum size required for PMU uprobes is PERF_ATTR_SIZE_VER1,
 			// since it added the config2 (Ext2) field. The Size field controls the
 			// size of the internal buffer the kernel allocates for reading the
 			// perf_event_attr argument from userspace.
 			Size:   unix.PERF_ATTR_SIZE_VER1,
-			Type:   uint32(et),          // PMU event type read from sysfs
+			Type:   uint32(eventType),   // PMU event type read from sysfs
 			Ext1:   uint64(uintptr(sp)), // Uprobe path
-			Ext2:   args.offset,         // Uprobe offset
+			Ext2:   args.Offset,         // Uprobe offset
 			Config: config,              // RefCtrOffset, Retprobe flag
 		}
 	}
 
-	rawFd, err := unix.PerfEventOpen(&attr, args.pid, 0, -1, unix.PERF_FLAG_FD_CLOEXEC)
+	rawFd, err := unix.PerfEventOpen(&attr, args.Pid, 0, -1, unix.PERF_FLAG_FD_CLOEXEC)
 
 	// On some old kernels, kprobe PMU doesn't allow `.` in symbol names and
 	// return -EINVAL. Return ErrNotSupported to allow falling back to tracefs.
 	// https://github.com/torvalds/linux/blob/94710cac0ef4/kernel/trace/trace_kprobe.c#L340-L343
-	if errors.Is(err, unix.EINVAL) && strings.Contains(args.symbol, ".") {
+	if errors.Is(err, unix.EINVAL) && strings.Contains(args.Symbol, ".") {
-		return nil, fmt.Errorf("symbol '%s+%#x': older kernels don't accept dots: %w", args.symbol, args.offset, ErrNotSupported)
+		return nil, fmt.Errorf("token %s: older kernels don't accept dots: %w", token, ErrNotSupported)
 	}
 	// Since commit 97c753e62e6c, ENOENT is correctly returned instead of EINVAL
-	// when trying to create a kretprobe for a missing symbol. Make sure ENOENT
+	// when trying to create a retprobe for a missing symbol.
-	// is returned to the caller.
+	if errors.Is(err, os.ErrNotExist) {
-	if errors.Is(err, os.ErrNotExist) || errors.Is(err, unix.EINVAL) {
+		return nil, fmt.Errorf("token %s: not found: %w", token, err)
-		return nil, fmt.Errorf("symbol '%s+%#x' not found: %w", args.symbol, args.offset, os.ErrNotExist)
 	}
-	// Since commit ab105a4fb894, -EILSEQ is returned when a kprobe sym+offset is resolved
+	// Since commit ab105a4fb894, EILSEQ is returned when a kprobe sym+offset is resolved
-	// to an invalid insn boundary.
+	// to an invalid insn boundary. The exact conditions that trigger this error are
-	if errors.Is(err, syscall.EILSEQ) {
+	// arch specific however.
-		return nil, fmt.Errorf("symbol '%s+%#x' not found (bad insn boundary): %w", args.symbol, args.offset, os.ErrNotExist)
+	if errors.Is(err, unix.EILSEQ) {
+		return nil, fmt.Errorf("token %s: bad insn boundary: %w", token, os.ErrNotExist)
 	}
 	// Since at least commit cb9a19fe4aa51, ENOTSUPP is returned
 	// when attempting to set a uprobe on a trap instruction.
-	if errors.Is(err, unix.ENOTSUPP) {
+	if errors.Is(err, sys.ENOTSUPP) {
-		return nil, fmt.Errorf("failed setting uprobe on offset %#x (possible trap insn): %w", args.offset, err)
+		return nil, fmt.Errorf("token %s: failed setting uprobe on offset %#x (possible trap insn): %w", token, args.Offset, err)
 	}
+
 	if err != nil {
-		return nil, fmt.Errorf("opening perf event: %w", err)
+		return nil, fmt.Errorf("token %s: opening perf event: %w", token, err)
 	}
 
 	// Ensure the string pointer is not collected before PerfEventOpen returns.
@@ -330,18 +312,7 @@ func pmuProbe(typ probeType, args probeArgs) (*perfEvent, error) {
 	}
 
 	// Kernel has perf_[k,u]probe PMU available, initialize perf event.
-	return &perfEvent{
+	return newPerfEvent(fd, nil), nil
-		typ:    typ.PerfEventType(args.ret),
-		name:   args.symbol,
-		pmuID:  et,
-		cookie: args.cookie,
-		fd:     fd,
-	}, nil
-}
-
-// tracefsKprobe creates a Kprobe tracefs entry.
-func tracefsKprobe(args probeArgs) (*perfEvent, error) {
-	return tracefsProbe(kprobeType, args)
 }
 
 // tracefsProbe creates a trace event by writing an entry to <tracefs>/[k,u]probe_events.
@@ -350,219 +321,37 @@ func tracefsKprobe(args probeArgs) (*perfEvent, error) {
 // Path and offset are only set in the case of uprobe(s) and are used to set
 // the executable/library path on the filesystem and the offset where the probe is inserted.
 // A perf event is then opened on the newly-created trace event and returned to the caller.
-func tracefsProbe(typ probeType, args probeArgs) (_ *perfEvent, err error) {
+func tracefsProbe(args tracefs.ProbeArgs) (*perfEvent, error) {
+	groupPrefix := "ebpf"
+	if args.Group != "" {
+		groupPrefix = args.Group
+	}
+
 	// Generate a random string for each trace event we attempt to create.
 	// This value is used as the 'group' token in tracefs to allow creating
 	// multiple kprobe trace events with the same name.
-	group, err := randomGroup("ebpf")
+	group, err := tracefs.RandomGroup(groupPrefix)
 	if err != nil {
 		return nil, fmt.Errorf("randomizing group name: %w", err)
 	}
-	args.group = group
+	args.Group = group
-
-	// Before attempting to create a trace event through tracefs,
-	// check if an event with the same group and name already exists.
-	// Kernels 4.x and earlier don't return os.ErrExist on writing a duplicate
-	// entry, so we need to rely on reads for detecting uniqueness.
-	_, err = getTraceEventID(group, args.symbol)
-	if err == nil {
-		return nil, fmt.Errorf("trace event already exists: %s/%s", group, args.symbol)
-	}
-	if err != nil && !errors.Is(err, os.ErrNotExist) {
-		return nil, fmt.Errorf("checking trace event %s/%s: %w", group, args.symbol, err)
-	}
 
 	// Create the [k,u]probe trace event using tracefs.
-	if err := createTraceFSProbeEvent(typ, args); err != nil {
+	evt, err := tracefs.NewEvent(args)
+	if err != nil {
 		return nil, fmt.Errorf("creating probe entry on tracefs: %w", err)
 	}
-	defer func() {
+
+	// Kprobes are ephemeral tracepoints and share the same perf event type.
+	fd, err := openTracepointPerfEvent(evt.ID(), args.Pid)
 	if err != nil {
 		// Make sure we clean up the created tracefs event when we return error.
 		// If a livepatch handler is already active on the symbol, the write to
 		// tracefs will succeed, a trace event will show up, but creating the
 		// perf event will fail with EBUSY.
-			_ = closeTraceFSProbeEvent(typ, args.group, args.symbol)
+		_ = evt.Close()
-		}
-	}()
-
-	// Get the newly-created trace event's id.
-	tid, err := getTraceEventID(group, args.symbol)
-	if err != nil {
-		return nil, fmt.Errorf("getting trace event id: %w", err)
-	}
-
-	// Kprobes are ephemeral tracepoints and share the same perf event type.
-	fd, err := openTracepointPerfEvent(tid, args.pid)
-	if err != nil {
 		return nil, err
 	}
 
-	return &perfEvent{
+	return newPerfEvent(fd, evt), nil
-		typ:       typ.PerfEventType(args.ret),
-		group:     group,
-		name:      args.symbol,
-		tracefsID: tid,
-		cookie:    args.cookie,
-		fd:        fd,
-	}, nil
-}
-
-// createTraceFSProbeEvent creates a new ephemeral trace event by writing to
-// <tracefs>/[k,u]probe_events. Returns os.ErrNotExist if symbol is not a valid
-// kernel symbol, or if it is not traceable with kprobes. Returns os.ErrExist
-// if a probe with the same group and symbol already exists.
-func createTraceFSProbeEvent(typ probeType, args probeArgs) error {
-	// Open the kprobe_events file in tracefs.
-	f, err := os.OpenFile(typ.EventsPath(), os.O_APPEND|os.O_WRONLY, 0666)
-	if err != nil {
-		return fmt.Errorf("error opening '%s': %w", typ.EventsPath(), err)
-	}
-	defer f.Close()
-
-	var pe, token string
-	switch typ {
-	case kprobeType:
-		// The kprobe_events syntax is as follows (see Documentation/trace/kprobetrace.txt):
-		// p[:[GRP/]EVENT] [MOD:]SYM[+offs]|MEMADDR [FETCHARGS] : Set a probe
-		// r[MAXACTIVE][:[GRP/]EVENT] [MOD:]SYM[+0] [FETCHARGS] : Set a return probe
-		// -:[GRP/]EVENT                                        : Clear a probe
-		//
-		// Some examples:
-		// r:ebpf_1234/r_my_kretprobe nf_conntrack_destroy
-		// p:ebpf_5678/p_my_kprobe __x64_sys_execve
-		//
-		// Leaving the kretprobe's MAXACTIVE set to 0 (or absent) will make the
-		// kernel default to NR_CPUS. This is desired in most eBPF cases since
-		// subsampling or rate limiting logic can be more accurately implemented in
-		// the eBPF program itself.
-		// See Documentation/kprobes.txt for more details.
-		token = kprobeToken(args)
-		pe = fmt.Sprintf("%s:%s/%s %s", probePrefix(args.ret), args.group, sanitizeSymbol(args.symbol), token)
-	case uprobeType:
-		// The uprobe_events syntax is as follows:
-		// p[:[GRP/]EVENT] PATH:OFFSET [FETCHARGS] : Set a probe
-		// r[:[GRP/]EVENT] PATH:OFFSET [FETCHARGS] : Set a return probe
-		// -:[GRP/]EVENT                           : Clear a probe
-		//
-		// Some examples:
-		// r:ebpf_1234/readline /bin/bash:0x12345
-		// p:ebpf_5678/main_mySymbol /bin/mybin:0x12345(0x123)
-		//
-		// See Documentation/trace/uprobetracer.txt for more details.
-		token = uprobeToken(args)
-		pe = fmt.Sprintf("%s:%s/%s %s", probePrefix(args.ret), args.group, args.symbol, token)
-	}
-	_, err = f.WriteString(pe)
-	// Since commit 97c753e62e6c, ENOENT is correctly returned instead of EINVAL
-	// when trying to create a kretprobe for a missing symbol. Make sure ENOENT
-	// is returned to the caller.
-	// EINVAL is also returned on pre-5.2 kernels when the `SYM[+offs]` token
-	// is resolved to an invalid insn boundary.
-	if errors.Is(err, os.ErrNotExist) || errors.Is(err, unix.EINVAL) {
-		return fmt.Errorf("token %s: %w", token, os.ErrNotExist)
-	}
-	// Since commit ab105a4fb894, -EILSEQ is returned when a kprobe sym+offset is resolved
-	// to an invalid insn boundary.
-	if errors.Is(err, syscall.EILSEQ) {
-		return fmt.Errorf("token %s: bad insn boundary: %w", token, os.ErrNotExist)
-	}
-	// ERANGE is returned when the `SYM[+offs]` token is too big and cannot
-	// be resolved.
-	if errors.Is(err, syscall.ERANGE) {
-		return fmt.Errorf("token %s: offset too big: %w", token, os.ErrNotExist)
-	}
-	if err != nil {
-		return fmt.Errorf("writing '%s' to '%s': %w", pe, typ.EventsPath(), err)
-	}
-
-	return nil
-}
-
-// closeTraceFSProbeEvent removes the [k,u]probe with the given type, group and symbol
-// from <tracefs>/[k,u]probe_events.
-func closeTraceFSProbeEvent(typ probeType, group, symbol string) error {
-	f, err := os.OpenFile(typ.EventsPath(), os.O_APPEND|os.O_WRONLY, 0666)
-	if err != nil {
-		return fmt.Errorf("error opening %s: %w", typ.EventsPath(), err)
-	}
-	defer f.Close()
-
-	// See [k,u]probe_events syntax above. The probe type does not need to be specified
-	// for removals.
-	pe := fmt.Sprintf("-:%s/%s", group, sanitizeSymbol(symbol))
-	if _, err = f.WriteString(pe); err != nil {
-		return fmt.Errorf("writing '%s' to '%s': %w", pe, typ.EventsPath(), err)
-	}
-
-	return nil
-}
-
-// randomGroup generates a pseudorandom string for use as a tracefs group name.
-// Returns an error when the output string would exceed 63 characters (kernel
-// limitation), when rand.Read() fails or when prefix contains characters not
-// allowed by isValidTraceID.
-func randomGroup(prefix string) (string, error) {
-	if !isValidTraceID(prefix) {
-		return "", fmt.Errorf("prefix '%s' must be alphanumeric or underscore: %w", prefix, errInvalidInput)
-	}
-
-	b := make([]byte, 8)
-	if _, err := rand.Read(b); err != nil {
-		return "", fmt.Errorf("reading random bytes: %w", err)
-	}
-
-	group := fmt.Sprintf("%s_%x", prefix, b)
-	if len(group) > 63 {
-		return "", fmt.Errorf("group name '%s' cannot be longer than 63 characters: %w", group, errInvalidInput)
-	}
-
-	return group, nil
-}
-
-func probePrefix(ret bool) string {
-	if ret {
-		return "r"
-	}
-	return "p"
-}
-
-// determineRetprobeBit reads a Performance Monitoring Unit's retprobe bit
-// from /sys/bus/event_source/devices/<pmu>/format/retprobe.
-func determineRetprobeBit(typ probeType) (uint64, error) {
-	p := filepath.Join("/sys/bus/event_source/devices/", typ.String(), "/format/retprobe")
-
-	data, err := os.ReadFile(p)
-	if err != nil {
-		return 0, err
-	}
-
-	var rp uint64
-	n, err := fmt.Sscanf(string(bytes.TrimSpace(data)), "config:%d", &rp)
-	if err != nil {
-		return 0, fmt.Errorf("parse retprobe bit: %w", err)
-	}
-	if n != 1 {
-		return 0, fmt.Errorf("parse retprobe bit: expected 1 item, got %d", n)
-	}
-
-	return rp, nil
-}
-
-func kretprobeBit() (uint64, error) {
-	kprobeRetprobeBit.once.Do(func() {
-		kprobeRetprobeBit.value, kprobeRetprobeBit.err = determineRetprobeBit(kprobeType)
-	})
-	return kprobeRetprobeBit.value, kprobeRetprobeBit.err
-}
-
-// kprobeToken creates the SYM[+offs] token for the tracefs api.
-func kprobeToken(args probeArgs) string {
-	po := args.symbol
-
-	if args.offset != 0 {
-		po += fmt.Sprintf("+%#x", args.offset)
-	}
-
-	return po
 }

vendor/github.com/cilium/ebpf/link/kprobe_multi.go

@@ -0,0 +1,180 @@
+package link
+
+import (
+	"errors"
+	"fmt"
+	"os"
+	"unsafe"
+
+	"github.com/cilium/ebpf"
+	"github.com/cilium/ebpf/asm"
+	"github.com/cilium/ebpf/internal"
+	"github.com/cilium/ebpf/internal/sys"
+	"github.com/cilium/ebpf/internal/unix"
+)
+
+// KprobeMultiOptions defines additional parameters that will be used
+// when opening a KprobeMulti Link.
+type KprobeMultiOptions struct {
+	// Symbols takes a list of kernel symbol names to attach an ebpf program to.
+	//
+	// Mutually exclusive with Addresses.
+	Symbols []string
+
+	// Addresses takes a list of kernel symbol addresses in case they can not
+	// be referred to by name.
+	//
+	// Note that only start addresses can be specified, since the fprobe API
+	// limits the attach point to the function entry or return.
+	//
+	// Mutually exclusive with Symbols.
+	Addresses []uintptr
+
+	// Cookies specifies arbitrary values that can be fetched from an eBPF
+	// program via `bpf_get_attach_cookie()`.
+	//
+	// If set, its length should be equal to the length of Symbols or Addresses.
+	// Each Cookie is assigned to the Symbol or Address specified at the
+	// corresponding slice index.
+	Cookies []uint64
+}
+
+// KprobeMulti attaches the given eBPF program to the entry point of a given set
+// of kernel symbols.
+//
+// The difference with Kprobe() is that multi-kprobe accomplishes this in a
+// single system call, making it significantly faster than attaching many
+// probes one at a time.
+//
+// Requires at least Linux 5.18.
+func KprobeMulti(prog *ebpf.Program, opts KprobeMultiOptions) (Link, error) {
+	return kprobeMulti(prog, opts, 0)
+}
+
+// KretprobeMulti attaches the given eBPF program to the return point of a given
+// set of kernel symbols.
+//
+// The difference with Kretprobe() is that multi-kprobe accomplishes this in a
+// single system call, making it significantly faster than attaching many
+// probes one at a time.
+//
+// Requires at least Linux 5.18.
+func KretprobeMulti(prog *ebpf.Program, opts KprobeMultiOptions) (Link, error) {
+	return kprobeMulti(prog, opts, unix.BPF_F_KPROBE_MULTI_RETURN)
+}
+
+func kprobeMulti(prog *ebpf.Program, opts KprobeMultiOptions, flags uint32) (Link, error) {
+	if prog == nil {
+		return nil, errors.New("cannot attach a nil program")
+	}
+
+	syms := uint32(len(opts.Symbols))
+	addrs := uint32(len(opts.Addresses))
+	cookies := uint32(len(opts.Cookies))
+
+	if syms == 0 && addrs == 0 {
+		return nil, fmt.Errorf("one of Symbols or Addresses is required: %w", errInvalidInput)
+	}
+	if syms != 0 && addrs != 0 {
+		return nil, fmt.Errorf("Symbols and Addresses are mutually exclusive: %w", errInvalidInput)
+	}
+	if cookies > 0 && cookies != syms && cookies != addrs {
+		return nil, fmt.Errorf("Cookies must be exactly Symbols or Addresses in length: %w", errInvalidInput)
+	}
+
+	if err := haveBPFLinkKprobeMulti(); err != nil {
+		return nil, err
+	}
+
+	attr := &sys.LinkCreateKprobeMultiAttr{
+		ProgFd:           uint32(prog.FD()),
+		AttachType:       sys.BPF_TRACE_KPROBE_MULTI,
+		KprobeMultiFlags: flags,
+	}
+
+	switch {
+	case syms != 0:
+		attr.Count = syms
+		attr.Syms = sys.NewStringSlicePointer(opts.Symbols)
+
+	case addrs != 0:
+		attr.Count = addrs
+		attr.Addrs = sys.NewPointer(unsafe.Pointer(&opts.Addresses[0]))
+	}
+
+	if cookies != 0 {
+		attr.Cookies = sys.NewPointer(unsafe.Pointer(&opts.Cookies[0]))
+	}
+
+	fd, err := sys.LinkCreateKprobeMulti(attr)
+	if errors.Is(err, unix.ESRCH) {
+		return nil, fmt.Errorf("couldn't find one or more symbols: %w", os.ErrNotExist)
+	}
+	if errors.Is(err, unix.EINVAL) {
+		return nil, fmt.Errorf("%w (missing kernel symbol or prog's AttachType not AttachTraceKprobeMulti?)", err)
+	}
+	if err != nil {
+		return nil, err
+	}
+
+	return &kprobeMultiLink{RawLink{fd, ""}}, nil
+}
+
+type kprobeMultiLink struct {
+	RawLink
+}
+
+var _ Link = (*kprobeMultiLink)(nil)
+
+func (kml *kprobeMultiLink) Update(prog *ebpf.Program) error {
+	return fmt.Errorf("update kprobe_multi: %w", ErrNotSupported)
+}
+
+func (kml *kprobeMultiLink) Pin(string) error {
+	return fmt.Errorf("pin kprobe_multi: %w", ErrNotSupported)
+}
+
+func (kml *kprobeMultiLink) Unpin() error {
+	return fmt.Errorf("unpin kprobe_multi: %w", ErrNotSupported)
+}
+
+var haveBPFLinkKprobeMulti = internal.NewFeatureTest("bpf_link_kprobe_multi", "5.18", func() error {
+	prog, err := ebpf.NewProgram(&ebpf.ProgramSpec{
+		Name: "probe_kpm_link",
+		Type: ebpf.Kprobe,
+		Instructions: asm.Instructions{
+			asm.Mov.Imm(asm.R0, 0),
+			asm.Return(),
+		},
+		AttachType: ebpf.AttachTraceKprobeMulti,
+		License:    "MIT",
+	})
+	if errors.Is(err, unix.E2BIG) {
+		// Kernel doesn't support AttachType field.
+		return internal.ErrNotSupported
+	}
+	if err != nil {
+		return err
+	}
+	defer prog.Close()
+
+	fd, err := sys.LinkCreateKprobeMulti(&sys.LinkCreateKprobeMultiAttr{
+		ProgFd:     uint32(prog.FD()),
+		AttachType: sys.BPF_TRACE_KPROBE_MULTI,
+		Count:      1,
+		Syms:       sys.NewStringSlicePointer([]string{"vprintk"}),
+	})
+	switch {
+	case errors.Is(err, unix.EINVAL):
+		return internal.ErrNotSupported
+	// If CONFIG_FPROBE isn't set.
+	case errors.Is(err, unix.EOPNOTSUPP):
+		return internal.ErrNotSupported
+	case err != nil:
+		return err
+	}
+
+	fd.Close()
+
+	return nil
+})

vendor/github.com/cilium/ebpf/link/link.go

@@ -46,6 +46,18 @@ type Link interface {
 	isLink()
 }
 
+// NewLinkFromFD creates a link from a raw fd.
+//
+// You should not use fd after calling this function.
+func NewLinkFromFD(fd int) (Link, error) {
+	sysFD, err := sys.NewFD(fd)
+	if err != nil {
+		return nil, err
+	}
+
+	return wrapRawLink(&RawLink{fd: sysFD})
+}
+
 // LoadPinnedLink loads a link that was persisted into a bpffs.
 func LoadPinnedLink(fileName string, opts *ebpf.LoadPinOptions) (Link, error) {
 	raw, err := loadPinnedRawLink(fileName, opts)
@@ -59,10 +71,15 @@ func LoadPinnedLink(fileName string, opts *ebpf.LoadPinOptions) (Link, error) {
 // wrap a RawLink in a more specific type if possible.
 //
 // The function takes ownership of raw and closes it on error.
-func wrapRawLink(raw *RawLink) (Link, error) {
+func wrapRawLink(raw *RawLink) (_ Link, err error) {
-	info, err := raw.Info()
+	defer func() {
 		if err != nil {
 			raw.Close()
+		}
+	}()
+
+	info, err := raw.Info()
+	if err != nil {
 		return nil, err
 	}
 
@@ -77,6 +94,10 @@ func wrapRawLink(raw *RawLink) (Link, error) {
 		return &Iter{*raw}, nil
 	case NetNsType:
 		return &NetNsLink{*raw}, nil
+	case KprobeMultiType:
+		return &kprobeMultiLink{*raw}, nil
+	case PerfEventType:
+		return nil, fmt.Errorf("recovering perf event fd: %w", ErrNotSupported)
 	default:
 		return raw, nil
 	}
@@ -172,12 +193,12 @@ func AttachRawLink(opts RawLinkOptions) (*RawLink, error) {
 		TargetFd:    uint32(opts.Target),
 		ProgFd:      uint32(progFd),
 		AttachType:  sys.AttachType(opts.Attach),
-		TargetBtfId: uint32(opts.BTF),
+		TargetBtfId: opts.BTF,
 		Flags:       opts.Flags,
 	}
 	fd, err := sys.LinkCreate(&attr)
 	if err != nil {
-		return nil, fmt.Errorf("can't create link: %s", err)
+		return nil, fmt.Errorf("create link: %w", err)
 	}
 
 	return &RawLink{fd, ""}, nil
@@ -230,6 +251,11 @@ func (l *RawLink) Unpin() error {
 	return nil
 }
 
+// IsPinned returns true if the Link has a non-empty pinned path.
+func (l *RawLink) IsPinned() bool {
+	return l.pinnedPath != ""
+}
+
 // Update implements the Link interface.
 func (l *RawLink) Update(new *ebpf.Program) error {
 	return l.UpdateArgs(RawLinkUpdateOptions{
@@ -280,27 +306,24 @@ func (l *RawLink) Info() (*Info, error) {
 	switch info.Type {
 	case CgroupType:
 		extra = &CgroupInfo{}
-	case IterType:
-		// not supported
 	case NetNsType:
 		extra = &NetNsInfo{}
-	case RawTracepointType:
-		// not supported
 	case TracingType:
 		extra = &TracingInfo{}
 	case XDPType:
 		extra = &XDPInfo{}
-	case PerfEventType:
+	case RawTracepointType, IterType,
-		// no extra
+		PerfEventType, KprobeMultiType:
+		// Extra metadata not supported.
 	default:
 		return nil, fmt.Errorf("unknown link info type: %d", info.Type)
 	}
 
-	if info.Type != RawTracepointType && info.Type != IterType && info.Type != PerfEventType {
+	if extra != nil {
 		buf := bytes.NewReader(info.Extra[:])
 		err := binary.Read(buf, internal.NativeEndian, extra)
 		if err != nil {
-			return nil, fmt.Errorf("can not read extra link info: %w", err)
+			return nil, fmt.Errorf("cannot read extra link info: %w", err)
 		}
 	}
 

vendor/github.com/cilium/ebpf/link/perf_event.go

@@ -1,20 +1,16 @@
 package link
 
 import (
-	"bytes"
 	"errors"
 	"fmt"
-	"os"
-	"path/filepath"
 	"runtime"
-	"strconv"
-	"strings"
 	"unsafe"
 
 	"github.com/cilium/ebpf"
 	"github.com/cilium/ebpf/asm"
 	"github.com/cilium/ebpf/internal"
 	"github.com/cilium/ebpf/internal/sys"
+	"github.com/cilium/ebpf/internal/tracefs"
 	"github.com/cilium/ebpf/internal/unix"
 )
 
@@ -42,67 +38,41 @@ import (
 //   stops any further invocations of the attached eBPF program.
 
 var (
-	tracefsPath = "/sys/kernel/debug/tracing"
+	errInvalidInput = tracefs.ErrInvalidInput
-
-	errInvalidInput = errors.New("invalid input")
 )
 
 const (
 	perfAllThreads = -1
 )
 
-type perfEventType uint8
-
-const (
-	tracepointEvent perfEventType = iota
-	kprobeEvent
-	kretprobeEvent
-	uprobeEvent
-	uretprobeEvent
-)
-
 // A perfEvent represents a perf event kernel object. Exactly one eBPF program
 // can be attached to it. It is created based on a tracefs trace event or a
 // Performance Monitoring Unit (PMU).
 type perfEvent struct {
-	// The event type determines the types of programs that can be attached.
+	// Trace event backing this perfEvent. May be nil.
-	typ perfEventType
+	tracefsEvent *tracefs.Event
-
-	// Group and name of the tracepoint/kprobe/uprobe.
-	group string
-	name  string
-
-	// PMU event ID read from sysfs. Valid IDs are non-zero.
-	pmuID uint64
-	// ID of the trace event read from tracefs. Valid IDs are non-zero.
-	tracefsID uint64
-
-	// User provided arbitrary value.
-	cookie uint64
 
 	// This is the perf event FD.
 	fd *sys.FD
 }
 
+func newPerfEvent(fd *sys.FD, event *tracefs.Event) *perfEvent {
+	pe := &perfEvent{event, fd}
+	// Both event and fd have their own finalizer, but we want to
+	// guarantee that they are closed in a certain order.
+	runtime.SetFinalizer(pe, (*perfEvent).Close)
+	return pe
+}
+
 func (pe *perfEvent) Close() error {
+	runtime.SetFinalizer(pe, nil)
+
 	if err := pe.fd.Close(); err != nil {
 		return fmt.Errorf("closing perf event fd: %w", err)
 	}
 
-	switch pe.typ {
+	if pe.tracefsEvent != nil {
-	case kprobeEvent, kretprobeEvent:
+		return pe.tracefsEvent.Close()
-		// Clean up kprobe tracefs entry.
-		if pe.tracefsID != 0 {
-			return closeTraceFSProbeEvent(kprobeType, pe.group, pe.name)
-		}
-	case uprobeEvent, uretprobeEvent:
-		// Clean up uprobe tracefs entry.
-		if pe.tracefsID != 0 {
-			return closeTraceFSProbeEvent(uprobeType, pe.group, pe.name)
-		}
-	case tracepointEvent:
-		// Tracepoint trace events don't hold any extra resources.
-		return nil
 	}
 
 	return nil
@@ -136,10 +106,14 @@ func (pl *perfEventLink) Unpin() error {
 }
 
 func (pl *perfEventLink) Close() error {
-	if err := pl.pe.Close(); err != nil {
+	if err := pl.fd.Close(); err != nil {
-		return fmt.Errorf("perf event link close: %w", err)
+		return fmt.Errorf("perf link close: %w", err)
 	}
-	return pl.fd.Close()
+
+	if err := pl.pe.Close(); err != nil {
+		return fmt.Errorf("perf event close: %w", err)
+	}
+	return nil
 }
 
 func (pl *perfEventLink) Update(prog *ebpf.Program) error {
@@ -183,7 +157,7 @@ func (pi *perfEventIoctl) Info() (*Info, error) {
 // attach the given eBPF prog to the perf event stored in pe.
 // pe must contain a valid perf event fd.
 // prog's type must match the program type stored in pe.
-func attachPerfEvent(pe *perfEvent, prog *ebpf.Program) (Link, error) {
+func attachPerfEvent(pe *perfEvent, prog *ebpf.Program, cookie uint64) (Link, error) {
 	if prog == nil {
 		return nil, errors.New("cannot attach a nil program")
 	}
@@ -191,30 +165,18 @@ func attachPerfEvent(pe *perfEvent, prog *ebpf.Program) (Link, error) {
 		return nil, fmt.Errorf("invalid program: %w", sys.ErrClosedFd)
 	}
 
-	switch pe.typ {
+	if err := haveBPFLinkPerfEvent(); err == nil {
-	case kprobeEvent, kretprobeEvent, uprobeEvent, uretprobeEvent:
+		return attachPerfEventLink(pe, prog, cookie)
-		if t := prog.Type(); t != ebpf.Kprobe {
-			return nil, fmt.Errorf("invalid program type (expected %s): %s", ebpf.Kprobe, t)
-		}
-	case tracepointEvent:
-		if t := prog.Type(); t != ebpf.TracePoint {
-			return nil, fmt.Errorf("invalid program type (expected %s): %s", ebpf.TracePoint, t)
-		}
-	default:
-		return nil, fmt.Errorf("unknown perf event type: %d", pe.typ)
 	}
 
-	if err := haveBPFLinkPerfEvent(); err == nil {
+	if cookie != 0 {
-		return attachPerfEventLink(pe, prog)
+		return nil, fmt.Errorf("cookies are not supported: %w", ErrNotSupported)
 	}
+
 	return attachPerfEventIoctl(pe, prog)
 }
 
 func attachPerfEventIoctl(pe *perfEvent, prog *ebpf.Program) (*perfEventIoctl, error) {
-	if pe.cookie != 0 {
-		return nil, fmt.Errorf("cookies are not supported: %w", ErrNotSupported)
-	}
-
 	// Assign the eBPF program to the perf event.
 	err := unix.IoctlSetInt(pe.fd.Int(), unix.PERF_EVENT_IOC_SET_BPF, prog.FD())
 	if err != nil {
@@ -226,32 +188,24 @@ func attachPerfEventIoctl(pe *perfEvent, prog *ebpf.Program) (*perfEventIoctl, e
 		return nil, fmt.Errorf("enable perf event: %s", err)
 	}
 
-	pi := &perfEventIoctl{pe}
+	return &perfEventIoctl{pe}, nil
-
-	// Close the perf event when its reference is lost to avoid leaking system resources.
-	runtime.SetFinalizer(pi, (*perfEventIoctl).Close)
-	return pi, nil
 }
 
 // Use the bpf api to attach the perf event (BPF_LINK_TYPE_PERF_EVENT, 5.15+).
 //
 // https://github.com/torvalds/linux/commit/b89fbfbb854c9afc3047e8273cc3a694650b802e
-func attachPerfEventLink(pe *perfEvent, prog *ebpf.Program) (*perfEventLink, error) {
+func attachPerfEventLink(pe *perfEvent, prog *ebpf.Program, cookie uint64) (*perfEventLink, error) {
 	fd, err := sys.LinkCreatePerfEvent(&sys.LinkCreatePerfEventAttr{
 		ProgFd:     uint32(prog.FD()),
 		TargetFd:   pe.fd.Uint(),
 		AttachType: sys.BPF_PERF_EVENT,
-		BpfCookie:  pe.cookie,
+		BpfCookie:  cookie,
 	})
 	if err != nil {
 		return nil, fmt.Errorf("cannot create bpf perf link: %v", err)
 	}
 
-	pl := &perfEventLink{RawLink{fd: fd}, pe}
+	return &perfEventLink{RawLink{fd: fd}, pe}, nil
-
-	// Close the perf event when its reference is lost to avoid leaking system resources.
-	runtime.SetFinalizer(pl, (*perfEventLink).Close)
-	return pl, nil
 }
 
 // unsafeStringPtr returns an unsafe.Pointer to a NUL-terminated copy of str.
@@ -263,40 +217,6 @@ func unsafeStringPtr(str string) (unsafe.Pointer, error) {
 	return unsafe.Pointer(p), nil
 }
 
-// getTraceEventID reads a trace event's ID from tracefs given its group and name.
-// The kernel requires group and name to be alphanumeric or underscore.
-//
-// name automatically has its invalid symbols converted to underscores so the caller
-// can pass a raw symbol name, e.g. a kernel symbol containing dots.
-func getTraceEventID(group, name string) (uint64, error) {
-	name = sanitizeSymbol(name)
-	tid, err := uint64FromFile(tracefsPath, "events", group, name, "id")
-	if errors.Is(err, os.ErrNotExist) {
-		return 0, fmt.Errorf("trace event %s/%s: %w", group, name, os.ErrNotExist)
-	}
-	if err != nil {
-		return 0, fmt.Errorf("reading trace event ID of %s/%s: %w", group, name, err)
-	}
-
-	return tid, nil
-}
-
-// getPMUEventType reads a Performance Monitoring Unit's type (numeric identifier)
-// from /sys/bus/event_source/devices/<pmu>/type.
-//
-// Returns ErrNotSupported if the pmu type is not supported.
-func getPMUEventType(typ probeType) (uint64, error) {
-	et, err := uint64FromFile("/sys/bus/event_source/devices", typ.String(), "type")
-	if errors.Is(err, os.ErrNotExist) {
-		return 0, fmt.Errorf("pmu type %s: %w", typ, ErrNotSupported)
-	}
-	if err != nil {
-		return 0, fmt.Errorf("reading pmu type %s: %w", typ, err)
-	}
-
-	return et, nil
-}
-
 // openTracepointPerfEvent opens a tracepoint-type perf event. System-wide
 // [k,u]probes created by writing to <tracefs>/[k,u]probe_events are tracepoints
 // behind the scenes, and can be attached to using these perf events.
@@ -317,30 +237,11 @@ func openTracepointPerfEvent(tid uint64, pid int) (*sys.FD, error) {
 	return sys.NewFD(fd)
 }
 
-// uint64FromFile reads a uint64 from a file. All elements of path are sanitized
-// and joined onto base. Returns error if base no longer prefixes the path after
-// joining all components.
-func uint64FromFile(base string, path ...string) (uint64, error) {
-	l := filepath.Join(path...)
-	p := filepath.Join(base, l)
-	if !strings.HasPrefix(p, base) {
-		return 0, fmt.Errorf("path '%s' attempts to escape base path '%s': %w", l, base, errInvalidInput)
-	}
-
-	data, err := os.ReadFile(p)
-	if err != nil {
-		return 0, fmt.Errorf("reading file %s: %w", p, err)
-	}
-
-	et := bytes.TrimSpace(data)
-	return strconv.ParseUint(string(et), 10, 64)
-}
-
 // Probe BPF perf link.
 //
 // https://elixir.bootlin.com/linux/v5.16.8/source/kernel/bpf/syscall.c#L4307
 // https://github.com/torvalds/linux/commit/b89fbfbb854c9afc3047e8273cc3a694650b802e
-var haveBPFLinkPerfEvent = internal.FeatureTest("bpf_link_perf_event", "5.15", func() error {
+var haveBPFLinkPerfEvent = internal.NewFeatureTest("bpf_link_perf_event", "5.15", func() error {
 	prog, err := ebpf.NewProgram(&ebpf.ProgramSpec{
 		Name: "probe_bpf_perf_link",
 		Type: ebpf.Kprobe,
@@ -367,28 +268,3 @@ var haveBPFLinkPerfEvent = internal.FeatureTest("bpf_link_perf_event", "5.15", f
 	}
 	return err
 })
-
-// isValidTraceID implements the equivalent of a regex match
-// against "^[a-zA-Z_][0-9a-zA-Z_]*$".
-//
-// Trace event groups, names and kernel symbols must adhere to this set
-// of characters. Non-empty, first character must not be a number, all
-// characters must be alphanumeric or underscore.
-func isValidTraceID(s string) bool {
-	if len(s) < 1 {
-		return false
-	}
-	for i, c := range []byte(s) {
-		switch {
-		case c >= 'a' && c <= 'z':
-		case c >= 'A' && c <= 'Z':
-		case c == '_':
-		case i > 0 && c >= '0' && c <= '9':
-
-		default:
-			return false
-		}
-	}
-
-	return true
-}

vendor/github.com/cilium/ebpf/link/platform.go

@@ -1,25 +0,0 @@
-package link
-
-import (
-	"fmt"
-	"runtime"
-)
-
-func platformPrefix(symbol string) string {
-
-	prefix := runtime.GOARCH
-
-	// per https://github.com/golang/go/blob/master/src/go/build/syslist.go
-	switch prefix {
-	case "386":
-		prefix = "ia32"
-	case "amd64", "amd64p32":
-		prefix = "x64"
-	case "arm64", "arm64be":
-		prefix = "arm64"
-	default:
-		return symbol
-	}
-
-	return fmt.Sprintf("__%s_%s", prefix, symbol)
-}

vendor/github.com/cilium/ebpf/link/query.go

@@ -0,0 +1,63 @@
+package link
+
+import (
+	"fmt"
+	"os"
+	"unsafe"
+
+	"github.com/cilium/ebpf"
+	"github.com/cilium/ebpf/internal/sys"
+)
+
+// QueryOptions defines additional parameters when querying for programs.
+type QueryOptions struct {
+	// Path can be a path to a cgroup, netns or LIRC2 device
+	Path string
+	// Attach specifies the AttachType of the programs queried for
+	Attach ebpf.AttachType
+	// QueryFlags are flags for BPF_PROG_QUERY, e.g. BPF_F_QUERY_EFFECTIVE
+	QueryFlags uint32
+}
+
+// QueryPrograms retrieves ProgramIDs associated with the AttachType.
+//
+// Returns (nil, nil) if there are no programs attached to the queried kernel
+// resource. Calling QueryPrograms on a kernel missing PROG_QUERY will result in
+// ErrNotSupported.
+func QueryPrograms(opts QueryOptions) ([]ebpf.ProgramID, error) {
+	if haveProgQuery() != nil {
+		return nil, fmt.Errorf("can't query program IDs: %w", ErrNotSupported)
+	}
+
+	f, err := os.Open(opts.Path)
+	if err != nil {
+		return nil, fmt.Errorf("can't open file: %s", err)
+	}
+	defer f.Close()
+
+	// query the number of programs to allocate correct slice size
+	attr := sys.ProgQueryAttr{
+		TargetFd:   uint32(f.Fd()),
+		AttachType: sys.AttachType(opts.Attach),
+		QueryFlags: opts.QueryFlags,
+	}
+	if err := sys.ProgQuery(&attr); err != nil {
+		return nil, fmt.Errorf("can't query program count: %w", err)
+	}
+
+	// return nil if no progs are attached
+	if attr.ProgCount == 0 {
+		return nil, nil
+	}
+
+	// we have at least one prog, so we query again
+	progIds := make([]ebpf.ProgramID, attr.ProgCount)
+	attr.ProgIds = sys.NewPointer(unsafe.Pointer(&progIds[0]))
+	attr.ProgCount = uint32(len(progIds))
+	if err := sys.ProgQuery(&attr); err != nil {
+		return nil, fmt.Errorf("can't query program IDs: %w", err)
+	}
+
+	return progIds, nil
+
+}

vendor/github.com/cilium/ebpf/link/socket_filter.go

@@ -15,7 +15,7 @@ func AttachSocketFilter(conn syscall.Conn, program *ebpf.Program) error {
 	}
 	var ssoErr error
 	err = rawConn.Control(func(fd uintptr) {
-		ssoErr = syscall.SetsockoptInt(int(fd), unix.SOL_SOCKET, unix.SO_ATTACH_BPF, program.FD())
+		ssoErr = unix.SetsockoptInt(int(fd), unix.SOL_SOCKET, unix.SO_ATTACH_BPF, program.FD())
 	})
 	if ssoErr != nil {
 		return ssoErr
@@ -31,7 +31,7 @@ func DetachSocketFilter(conn syscall.Conn) error {
 	}
 	var ssoErr error
 	err = rawConn.Control(func(fd uintptr) {
-		ssoErr = syscall.SetsockoptInt(int(fd), unix.SOL_SOCKET, unix.SO_DETACH_BPF, 0)
+		ssoErr = unix.SetsockoptInt(int(fd), unix.SOL_SOCKET, unix.SO_DETACH_BPF, 0)
 	})
 	if ssoErr != nil {
 		return ssoErr

vendor/github.com/cilium/ebpf/link/syscalls.go

@@ -23,9 +23,10 @@ const (
 	NetNsType         = sys.BPF_LINK_TYPE_NETNS
 	XDPType           = sys.BPF_LINK_TYPE_XDP
 	PerfEventType     = sys.BPF_LINK_TYPE_PERF_EVENT
+	KprobeMultiType   = sys.BPF_LINK_TYPE_KPROBE_MULTI
 )
 
-var haveProgAttach = internal.FeatureTest("BPF_PROG_ATTACH", "4.10", func() error {
+var haveProgAttach = internal.NewFeatureTest("BPF_PROG_ATTACH", "4.10", func() error {
 	prog, err := ebpf.NewProgram(&ebpf.ProgramSpec{
 		Type:    ebpf.CGroupSKB,
 		License: "MIT",
@@ -45,7 +46,7 @@ var haveProgAttach = internal.FeatureTest("BPF_PROG_ATTACH", "4.10", func() erro
 	return nil
 })
 
-var haveProgAttachReplace = internal.FeatureTest("BPF_PROG_ATTACH atomic replacement", "5.5", func() error {
+var haveProgAttachReplace = internal.NewFeatureTest("BPF_PROG_ATTACH atomic replacement of MULTI progs", "5.5", func() error {
 	if err := haveProgAttach(); err != nil {
 		return err
 	}
@@ -85,7 +86,7 @@ var haveProgAttachReplace = internal.FeatureTest("BPF_PROG_ATTACH atomic replace
 	return err
 })
 
-var haveBPFLink = internal.FeatureTest("bpf_link", "5.7", func() error {
+var haveBPFLink = internal.NewFeatureTest("bpf_link", "5.7", func() error {
 	attr := sys.LinkCreateAttr{
 		// This is a hopefully invalid file descriptor, which triggers EBADF.
 		TargetFd:   ^uint32(0),
@@ -101,3 +102,22 @@ var haveBPFLink = internal.FeatureTest("bpf_link", "5.7", func() error {
 	}
 	return err
 })
+
+var haveProgQuery = internal.NewFeatureTest("BPF_PROG_QUERY", "4.15", func() error {
+	attr := sys.ProgQueryAttr{
+		// We rely on this being checked during the syscall.
+		// With an otherwise correct payload we expect EBADF here
+		// as an indication that the feature is present.
+		TargetFd:   ^uint32(0),
+		AttachType: sys.AttachType(ebpf.AttachCGroupInetIngress),
+	}
+
+	err := sys.ProgQuery(&attr)
+	if errors.Is(err, unix.EINVAL) {
+		return internal.ErrNotSupported
+	}
+	if errors.Is(err, unix.EBADF) {
+		return nil
+	}
+	return err
+})

vendor/github.com/cilium/ebpf/link/tracepoint.go

@@ -4,6 +4,7 @@ import (
 	"fmt"
 
 	"github.com/cilium/ebpf"
+	"github.com/cilium/ebpf/internal/tracefs"
 )
 
 // TracepointOptions defines additional parameters that will be used
@@ -17,7 +18,7 @@ type TracepointOptions struct {
 }
 
 // Tracepoint attaches the given eBPF program to the tracepoint with the given
-// group and name. See /sys/kernel/debug/tracing/events to find available
+// group and name. See /sys/kernel/tracing/events to find available
 // tracepoints. The top-level directory is the group, the event's subdirectory
 // is the name. Example:
 //
@@ -36,14 +37,11 @@ func Tracepoint(group, name string, prog *ebpf.Program, opts *TracepointOptions)
 	if prog == nil {
 		return nil, fmt.Errorf("prog cannot be nil: %w", errInvalidInput)
 	}
-	if !isValidTraceID(group) || !isValidTraceID(name) {
-		return nil, fmt.Errorf("group and name '%s/%s' must be alphanumeric or underscore: %w", group, name, errInvalidInput)
-	}
 	if prog.Type() != ebpf.TracePoint {
 		return nil, fmt.Errorf("eBPF program type %s is not a Tracepoint: %w", prog.Type(), errInvalidInput)
 	}
 
-	tid, err := getTraceEventID(group, name)
+	tid, err := tracefs.EventID(group, name)
 	if err != nil {
 		return nil, err
 	}
@@ -58,16 +56,9 @@ func Tracepoint(group, name string, prog *ebpf.Program, opts *TracepointOptions)
 		cookie = opts.Cookie
 	}
 
-	pe := &perfEvent{
+	pe := newPerfEvent(fd, nil)
-		typ:       tracepointEvent,
-		group:     group,
-		name:      name,
-		tracefsID: tid,
-		cookie:    cookie,
-		fd:        fd,
-	}
 
-	lnk, err := attachPerfEvent(pe, prog)
+	lnk, err := attachPerfEvent(pe, prog, cookie)
 	if err != nil {
 		pe.Close()
 		return nil, err

vendor/github.com/cilium/ebpf/link/tracing.go

@@ -1,11 +1,13 @@
 package link
 
 import (
+	"errors"
 	"fmt"
 
 	"github.com/cilium/ebpf"
 	"github.com/cilium/ebpf/btf"
 	"github.com/cilium/ebpf/internal/sys"
+	"github.com/cilium/ebpf/internal/unix"
 )
 
 type tracing struct {
@@ -70,6 +72,10 @@ func AttachFreplace(targetProg *ebpf.Program, name string, prog *ebpf.Program) (
 		Attach:  ebpf.AttachNone,
 		BTF:     typeID,
 	})
+	if errors.Is(err, sys.ENOTSUPP) {
+		// This may be returned by bpf_tracing_prog_attach via bpf_arch_text_poke.
+		return nil, fmt.Errorf("create raw tracepoint: %w", ErrNotSupported)
+	}
 	if err != nil {
 		return nil, err
 	}
@@ -82,25 +88,71 @@ type TracingOptions struct {
 	// AttachTraceFEntry/AttachTraceFExit/AttachModifyReturn or
 	// AttachTraceRawTp.
 	Program *ebpf.Program
+	// Program attach type. Can be one of:
+	// 	- AttachTraceFEntry
+	// 	- AttachTraceFExit
+	// 	- AttachModifyReturn
+	// 	- AttachTraceRawTp
+	// This field is optional.
+	AttachType ebpf.AttachType
+	// Arbitrary value that can be fetched from an eBPF program
+	// via `bpf_get_attach_cookie()`.
+	Cookie uint64
 }
 
 type LSMOptions struct {
 	// Program must be of type LSM with attach type
 	// AttachLSMMac.
 	Program *ebpf.Program
+	// Arbitrary value that can be fetched from an eBPF program
+	// via `bpf_get_attach_cookie()`.
+	Cookie uint64
 }
 
 // attachBTFID links all BPF program types (Tracing/LSM) that they attach to a btf_id.
-func attachBTFID(program *ebpf.Program) (Link, error) {
+func attachBTFID(program *ebpf.Program, at ebpf.AttachType, cookie uint64) (Link, error) {
 	if program.FD() < 0 {
 		return nil, fmt.Errorf("invalid program %w", sys.ErrClosedFd)
 	}
 
-	fd, err := sys.RawTracepointOpen(&sys.RawTracepointOpenAttr{
+	var (
+		fd  *sys.FD
+		err error
+	)
+	switch at {
+	case ebpf.AttachTraceFEntry, ebpf.AttachTraceFExit, ebpf.AttachTraceRawTp,
+		ebpf.AttachModifyReturn, ebpf.AttachLSMMac:
+		// Attach via BPF link
+		fd, err = sys.LinkCreateTracing(&sys.LinkCreateTracingAttr{
+			ProgFd:     uint32(program.FD()),
+			AttachType: sys.AttachType(at),
+			Cookie:     cookie,
+		})
+		if err == nil {
+			break
+		}
+		if !errors.Is(err, unix.EINVAL) && !errors.Is(err, sys.ENOTSUPP) {
+			return nil, fmt.Errorf("create tracing link: %w", err)
+		}
+		fallthrough
+	case ebpf.AttachNone:
+		// Attach via RawTracepointOpen
+		if cookie > 0 {
+			return nil, fmt.Errorf("create raw tracepoint with cookie: %w", ErrNotSupported)
+		}
+
+		fd, err = sys.RawTracepointOpen(&sys.RawTracepointOpenAttr{
 			ProgFd: uint32(program.FD()),
 		})
+		if errors.Is(err, sys.ENOTSUPP) {
+			// This may be returned by bpf_tracing_prog_attach via bpf_arch_text_poke.
+			return nil, fmt.Errorf("create raw tracepoint: %w", ErrNotSupported)
+		}
 		if err != nil {
-		return nil, err
+			return nil, fmt.Errorf("create raw tracepoint: %w", err)
+		}
+	default:
+		return nil, fmt.Errorf("invalid attach type: %s", at.String())
 	}
 
 	raw := RawLink{fd: fd}
@@ -115,8 +167,7 @@ func attachBTFID(program *ebpf.Program) (Link, error) {
 		// a raw_tracepoint link. Other types return a tracing link.
 		return &rawTracepoint{raw}, nil
 	}
-
+	return &tracing{raw}, nil
-	return &tracing{RawLink: RawLink{fd: fd}}, nil
 }
 
 // AttachTracing links a tracing (fentry/fexit/fmod_ret) BPF program or
@@ -127,7 +178,14 @@ func AttachTracing(opts TracingOptions) (Link, error) {
 		return nil, fmt.Errorf("invalid program type %s, expected Tracing", t)
 	}
 
-	return attachBTFID(opts.Program)
+	switch opts.AttachType {
+	case ebpf.AttachTraceFEntry, ebpf.AttachTraceFExit, ebpf.AttachModifyReturn,
+		ebpf.AttachTraceRawTp, ebpf.AttachNone:
+	default:
+		return nil, fmt.Errorf("invalid attach type: %s", opts.AttachType.String())
+	}
+
+	return attachBTFID(opts.Program, opts.AttachType, opts.Cookie)
 }
 
 // AttachLSM links a Linux security module (LSM) BPF Program to a BPF
@@ -137,5 +195,5 @@ func AttachLSM(opts LSMOptions) (Link, error) {
 		return nil, fmt.Errorf("invalid program type %s, expected LSM", t)
 	}
 
-	return attachBTFID(opts.Program)
+	return attachBTFID(opts.Program, ebpf.AttachLSMMac, opts.Cookie)
 }

vendor/github.com/cilium/ebpf/link/uprobe.go

@@ -5,27 +5,18 @@ import (
 	"errors"
 	"fmt"
 	"os"
-	"path/filepath"
-	"strings"
 	"sync"
 
 	"github.com/cilium/ebpf"
 	"github.com/cilium/ebpf/internal"
+	"github.com/cilium/ebpf/internal/tracefs"
 )
 
 var (
-	uprobeEventsPath = filepath.Join(tracefsPath, "uprobe_events")
-
-	uprobeRetprobeBit = struct {
-		once  sync.Once
-		value uint64
-		err   error
-	}{}
-
 	uprobeRefCtrOffsetPMUPath = "/sys/bus/event_source/devices/uprobe/format/ref_ctr_offset"
 	// elixir.bootlin.com/linux/v5.15-rc7/source/kernel/events/core.c#L9799
 	uprobeRefCtrOffsetShift = 32
-	haveRefCtrOffsetPMU     = internal.FeatureTest("RefCtrOffsetPMU", "4.20", func() error {
+	haveRefCtrOffsetPMU     = internal.NewFeatureTest("RefCtrOffsetPMU", "4.20", func() error {
 		_, err := os.Stat(uprobeRefCtrOffsetPMUPath)
 		if err != nil {
 			return internal.ErrNotSupported
@@ -44,6 +35,8 @@ type Executable struct {
 	path string
 	// Parsed ELF and dynamic symbols' addresses.
 	addresses map[string]uint64
+	// Keep track of symbol table lazy load.
+	addressesOnce sync.Once
 }
 
 // UprobeOptions defines additional parameters that will be used
@@ -77,6 +70,17 @@ type UprobeOptions struct {
 	//
 	// Needs kernel 5.15+.
 	Cookie uint64
+	// Prefix used for the event name if the uprobe must be attached using tracefs.
+	// The group name will be formatted as `<prefix>_<randomstr>`.
+	// The default empty string is equivalent to "ebpf" as the prefix.
+	TraceFSPrefix string
+}
+
+func (uo *UprobeOptions) cookie() uint64 {
+	if uo == nil {
+		return 0
+	}
+	return uo.Cookie
 }
 
 // To open a new Executable, use:
@@ -89,32 +93,21 @@ func OpenExecutable(path string) (*Executable, error) {
 		return nil, fmt.Errorf("path cannot be empty")
 	}
 
-	f, err := os.Open(path)
+	f, err := internal.OpenSafeELFFile(path)
-	if err != nil {
-		return nil, fmt.Errorf("open file '%s': %w", path, err)
-	}
-	defer f.Close()
-
-	se, err := internal.NewSafeELFFile(f)
 	if err != nil {
 		return nil, fmt.Errorf("parse ELF file: %w", err)
 	}
+	defer f.Close()
 
-	if se.Type != elf.ET_EXEC && se.Type != elf.ET_DYN {
+	if f.Type != elf.ET_EXEC && f.Type != elf.ET_DYN {
 		// ELF is not an executable or a shared object.
 		return nil, errors.New("the given file is not an executable or a shared object")
 	}
 
-	ex := Executable{
+	return &Executable{
 		path:      path,
 		addresses: make(map[string]uint64),
-	}
+	}, nil
-
-	if err := ex.load(se); err != nil {
-		return nil, err
-	}
-
-	return &ex, nil
 }
 
 func (ex *Executable) load(f *internal.SafeELFFile) error {
@@ -171,6 +164,22 @@ func (ex *Executable) address(symbol string, opts *UprobeOptions) (uint64, error
 		return opts.Address + opts.Offset, nil
 	}
 
+	var err error
+	ex.addressesOnce.Do(func() {
+		var f *internal.SafeELFFile
+		f, err = internal.OpenSafeELFFile(ex.path)
+		if err != nil {
+			err = fmt.Errorf("parse ELF file: %w", err)
+			return
+		}
+		defer f.Close()
+
+		err = ex.load(f)
+	})
+	if err != nil {
+		return 0, fmt.Errorf("lazy load symbols: %w", err)
+	}
+
 	address, ok := ex.addresses[symbol]
 	if !ok {
 		return 0, fmt.Errorf("symbol %s: %w", symbol, ErrNoSymbol)
@@ -216,7 +225,7 @@ func (ex *Executable) Uprobe(symbol string, prog *ebpf.Program, opts *UprobeOpti
 		return nil, err
 	}
 
-	lnk, err := attachPerfEvent(u, prog)
+	lnk, err := attachPerfEvent(u, prog, opts.cookie())
 	if err != nil {
 		u.Close()
 		return nil, err
@@ -250,7 +259,7 @@ func (ex *Executable) Uretprobe(symbol string, prog *ebpf.Program, opts *UprobeO
 		return nil, err
 	}
 
-	lnk, err := attachPerfEvent(u, prog)
+	lnk, err := attachPerfEvent(u, prog, opts.cookie())
 	if err != nil {
 		u.Close()
 		return nil, err
@@ -288,18 +297,20 @@ func (ex *Executable) uprobe(symbol string, prog *ebpf.Program, opts *UprobeOpti
 		}
 	}
 
-	args := probeArgs{
+	args := tracefs.ProbeArgs{
-		symbol:       symbol,
+		Type:         tracefs.Uprobe,
-		path:         ex.path,
+		Symbol:       symbol,
-		offset:       offset,
+		Path:         ex.path,
-		pid:          pid,
+		Offset:       offset,
-		refCtrOffset: opts.RefCtrOffset,
+		Pid:          pid,
-		ret:          ret,
+		RefCtrOffset: opts.RefCtrOffset,
-		cookie:       opts.Cookie,
+		Ret:          ret,
+		Cookie:       opts.Cookie,
+		Group:        opts.TraceFSPrefix,
 	}
 
 	// Use uprobe PMU if the kernel has it available.
-	tp, err := pmuUprobe(args)
+	tp, err := pmuProbe(args)
 	if err == nil {
 		return tp, nil
 	}
@@ -308,66 +319,10 @@ func (ex *Executable) uprobe(symbol string, prog *ebpf.Program, opts *UprobeOpti
 	}
 
 	// Use tracefs if uprobe PMU is missing.
-	args.symbol = sanitizeSymbol(symbol)
+	tp, err = tracefsProbe(args)
-	tp, err = tracefsUprobe(args)
 	if err != nil {
 		return nil, fmt.Errorf("creating trace event '%s:%s' in tracefs: %w", ex.path, symbol, err)
 	}
 
 	return tp, nil
 }
-
-// pmuUprobe opens a perf event based on the uprobe PMU.
-func pmuUprobe(args probeArgs) (*perfEvent, error) {
-	return pmuProbe(uprobeType, args)
-}
-
-// tracefsUprobe creates a Uprobe tracefs entry.
-func tracefsUprobe(args probeArgs) (*perfEvent, error) {
-	return tracefsProbe(uprobeType, args)
-}
-
-// sanitizeSymbol replaces every invalid character for the tracefs api with an underscore.
-// It is equivalent to calling regexp.MustCompile("[^a-zA-Z0-9]+").ReplaceAllString("_").
-func sanitizeSymbol(s string) string {
-	var b strings.Builder
-	b.Grow(len(s))
-	var skip bool
-	for _, c := range []byte(s) {
-		switch {
-		case c >= 'a' && c <= 'z',
-			c >= 'A' && c <= 'Z',
-			c >= '0' && c <= '9':
-			skip = false
-			b.WriteByte(c)
-
-		default:
-			if !skip {
-				b.WriteByte('_')
-				skip = true
-			}
-		}
-	}
-
-	return b.String()
-}
-
-// uprobeToken creates the PATH:OFFSET(REF_CTR_OFFSET) token for the tracefs api.
-func uprobeToken(args probeArgs) string {
-	po := fmt.Sprintf("%s:%#x", args.path, args.offset)
-
-	if args.refCtrOffset != 0 {
-		// This is not documented in Documentation/trace/uprobetracer.txt.
-		// elixir.bootlin.com/linux/v5.15-rc7/source/kernel/trace/trace.c#L5564
-		po += fmt.Sprintf("(%#x)", args.refCtrOffset)
-	}
-
-	return po
-}
-
-func uretprobeBit() (uint64, error) {
-	uprobeRetprobeBit.once.Do(func() {
-		uprobeRetprobeBit.value, uprobeRetprobeBit.err = determineRetprobeBit(uprobeType)
-	})
-	return uprobeRetprobeBit.value, uprobeRetprobeBit.err
-}

vendor/github.com/cilium/ebpf/linker.go

@@ -1,14 +1,51 @@
 package ebpf
 
 import (
+	"encoding/binary"
 	"errors"
 	"fmt"
-	"sync"
+	"io"
+	"math"
 
 	"github.com/cilium/ebpf/asm"
 	"github.com/cilium/ebpf/btf"
+	"github.com/cilium/ebpf/internal"
 )
 
+// handles stores handle objects to avoid gc cleanup
+type handles []*btf.Handle
+
+func (hs *handles) add(h *btf.Handle) (int, error) {
+	if h == nil {
+		return 0, nil
+	}
+
+	if len(*hs) == math.MaxInt16 {
+		return 0, fmt.Errorf("can't add more than %d module FDs to fdArray", math.MaxInt16)
+	}
+
+	*hs = append(*hs, h)
+
+	// return length of slice so that indexes start at 1
+	return len(*hs), nil
+}
+
+func (hs handles) fdArray() []int32 {
+	// first element of fda is reserved as no module can be indexed with 0
+	fda := []int32{0}
+	for _, h := range hs {
+		fda = append(fda, int32(h.FD()))
+	}
+
+	return fda
+}
+
+func (hs handles) close() {
+	for _, h := range hs {
+		h.Close()
+	}
+}
+
 // splitSymbols splits insns into subsections delimited by Symbol Instructions.
 // insns cannot be empty and must start with a Symbol Instruction.
 //
@@ -67,7 +104,7 @@ func hasFunctionReferences(insns asm.Instructions) bool {
 //
 // Passing a nil target will relocate against the running kernel. insns are
 // modified in place.
-func applyRelocations(insns asm.Instructions, local, target *btf.Spec) error {
+func applyRelocations(insns asm.Instructions, target *btf.Spec, bo binary.ByteOrder) error {
 	var relos []*btf.CORERelocation
 	var reloInsns []*asm.Instruction
 	iter := insns.Iterate()
@@ -82,19 +119,18 @@ func applyRelocations(insns asm.Instructions, local, target *btf.Spec) error {
 		return nil
 	}
 
-	target, err := maybeLoadKernelBTF(target)
+	if bo == nil {
-	if err != nil {
+		bo = internal.NativeEndian
-		return err
 	}
 
-	fixups, err := btf.CORERelocate(local, target, relos)
+	fixups, err := btf.CORERelocate(relos, target, bo)
 	if err != nil {
 		return err
 	}
 
 	for i, fixup := range fixups {
 		if err := fixup.Apply(reloInsns[i]); err != nil {
-			return fmt.Errorf("apply fixup %s: %w", &fixup, err)
+			return fmt.Errorf("fixup for %s: %w", relos[i], err)
 		}
 	}
 
@@ -181,8 +217,9 @@ func fixupAndValidate(insns asm.Instructions) error {
 		ins := iter.Ins
 
 		// Map load was tagged with a Reference, but does not contain a Map pointer.
-		if ins.IsLoadFromMap() && ins.Reference() != "" && ins.Map() == nil {
+		needsMap := ins.Reference() != "" || ins.Metadata.Get(kconfigMetaKey{}) != nil
-			return fmt.Errorf("instruction %d: map %s: %w", iter.Index, ins.Reference(), asm.ErrUnsatisfiedMapReference)
+		if ins.IsLoadFromMap() && needsMap && ins.Map() == nil {
+			return fmt.Errorf("instruction %d: %w", iter.Index, asm.ErrUnsatisfiedMapReference)
 		}
 
 		fixupProbeReadKernel(ins)
@@ -191,6 +228,88 @@ func fixupAndValidate(insns asm.Instructions) error {
 	return nil
 }
 
+// fixupKfuncs loops over all instructions in search for kfunc calls.
+// If at least one is found, the current kernels BTF and module BTFis are searched to set Instruction.Constant
+// and Instruction.Offset to the correct values.
+func fixupKfuncs(insns asm.Instructions) (handles, error) {
+	iter := insns.Iterate()
+	for iter.Next() {
+		ins := iter.Ins
+		if ins.IsKfuncCall() {
+			goto fixups
+		}
+	}
+
+	return nil, nil
+
+fixups:
+	// only load the kernel spec if we found at least one kfunc call
+	kernelSpec, err := btf.LoadKernelSpec()
+	if err != nil {
+		return nil, err
+	}
+
+	fdArray := make(handles, 0)
+	for {
+		ins := iter.Ins
+
+		if !ins.IsKfuncCall() {
+			if !iter.Next() {
+				// break loop if this was the last instruction in the stream.
+				break
+			}
+			continue
+		}
+
+		// check meta, if no meta return err
+		kfm, _ := ins.Metadata.Get(kfuncMeta{}).(*btf.Func)
+		if kfm == nil {
+			return nil, fmt.Errorf("kfunc call has no kfuncMeta")
+		}
+
+		target := btf.Type((*btf.Func)(nil))
+		spec, module, err := findTargetInKernel(kernelSpec, kfm.Name, &target)
+		if errors.Is(err, btf.ErrNotFound) {
+			return nil, fmt.Errorf("kfunc %q: %w", kfm.Name, ErrNotSupported)
+		}
+		if err != nil {
+			return nil, err
+		}
+
+		if err := btf.CheckTypeCompatibility(kfm.Type, target.(*btf.Func).Type); err != nil {
+			return nil, &incompatibleKfuncError{kfm.Name, err}
+		}
+
+		id, err := spec.TypeID(target)
+		if err != nil {
+			return nil, err
+		}
+
+		idx, err := fdArray.add(module)
+		if err != nil {
+			return nil, err
+		}
+
+		ins.Constant = int64(id)
+		ins.Offset = int16(idx)
+
+		if !iter.Next() {
+			break
+		}
+	}
+
+	return fdArray, nil
+}
+
+type incompatibleKfuncError struct {
+	name string
+	err  error
+}
+
+func (ike *incompatibleKfuncError) Error() string {
+	return fmt.Sprintf("kfunc %q: %s", ike.name, ike.err)
+}
+
 // fixupProbeReadKernel replaces calls to bpf_probe_read_{kernel,user}(_str)
 // with bpf_probe_read(_str) on kernels that don't support it yet.
 func fixupProbeReadKernel(ins *asm.Instruction) {
@@ -211,28 +330,62 @@ func fixupProbeReadKernel(ins *asm.Instruction) {
 	}
 }
 
-var kernelBTF struct {
+// resolveKconfigReferences creates and populates a .kconfig map if necessary.
-	sync.Mutex
-	spec *btf.Spec
-}
-
-// maybeLoadKernelBTF loads the current kernel's BTF if spec is nil, otherwise
-// it returns spec unchanged.
 //
-// The kernel BTF is cached for the lifetime of the process.
+// Returns a nil Map and no error if no references exist.
-func maybeLoadKernelBTF(spec *btf.Spec) (*btf.Spec, error) {
+func resolveKconfigReferences(insns asm.Instructions) (_ *Map, err error) {
-	if spec != nil {
+	closeOnError := func(c io.Closer) {
-		return spec, nil
+		if err != nil {
+			c.Close()
+		}
 	}
 
-	kernelBTF.Lock()
+	var spec *MapSpec
-	defer kernelBTF.Unlock()
+	iter := insns.Iterate()
-
+	for iter.Next() {
-	if kernelBTF.spec != nil {
+		meta, _ := iter.Ins.Metadata.Get(kconfigMetaKey{}).(*kconfigMeta)
-		return kernelBTF.spec, nil
+		if meta != nil {
+			spec = meta.Map
+			break
+		}
 	}
 
-	var err error
+	if spec == nil {
-	kernelBTF.spec, err = btf.LoadKernelSpec()
+		return nil, nil
-	return kernelBTF.spec, err
+	}
+
+	cpy := spec.Copy()
+	if err := resolveKconfig(cpy); err != nil {
+		return nil, err
+	}
+
+	kconfig, err := NewMap(cpy)
+	if err != nil {
+		return nil, err
+	}
+	defer closeOnError(kconfig)
+
+	// Resolve all instructions which load from .kconfig map with actual map
+	// and offset inside it.
+	iter = insns.Iterate()
+	for iter.Next() {
+		meta, _ := iter.Ins.Metadata.Get(kconfigMetaKey{}).(*kconfigMeta)
+		if meta == nil {
+			continue
+		}
+
+		if meta.Map != spec {
+			return nil, fmt.Errorf("instruction %d: reference to multiple .kconfig maps is not allowed", iter.Index)
+		}
+
+		if err := iter.Ins.AssociateMap(kconfig); err != nil {
+			return nil, fmt.Errorf("instruction %d: %w", iter.Index, err)
+		}
+
+		// Encode a map read at the offset of the var in the datasec.
+		iter.Ins.Constant = int64(uint64(meta.Offset) << 32)
+		iter.Ins.Metadata.Set(kconfigMetaKey{}, nil)
+	}
+
+	return kconfig, nil
 }

vendor/github.com/cilium/ebpf/map.go

@@ -6,6 +6,7 @@ import (
 	"fmt"
 	"io"
 	"math/rand"
+	"os"
 	"path/filepath"
 	"reflect"
 	"time"
@@ -77,9 +78,6 @@ type MapSpec struct {
 
 	// The key and value type of this map. May be nil.
 	Key, Value btf.Type
-
-	// The BTF associated with this map.
-	BTF *btf.Spec
 }
 
 func (ms *MapSpec) String() string {
@@ -104,12 +102,6 @@ func (ms *MapSpec) Copy() *MapSpec {
 	return &cpy
 }
 
-// hasBTF returns true if the MapSpec has a valid BTF spec and if its
-// map type supports associated BTF metadata in the kernel.
-func (ms *MapSpec) hasBTF() bool {
-	return ms.BTF != nil && ms.Type.hasBTF()
-}
-
 func (ms *MapSpec) clampPerfEventArraySize() error {
 	if ms.Type != PerfEventArray {
 		return nil
@@ -158,7 +150,11 @@ type MapKV struct {
 	Value interface{}
 }
 
-func (ms *MapSpec) checkCompatibility(m *Map) error {
+// Compatible returns nil if an existing map may be used instead of creating
+// one from the spec.
+//
+// Returns an error wrapping [ErrMapIncompatible] otherwise.
+func (ms *MapSpec) Compatible(m *Map) error {
 	switch {
 	case m.typ != ms.Type:
 		return fmt.Errorf("expected type %v, got %v: %w", ms.Type, m.typ, ErrMapIncompatible)
@@ -173,7 +169,10 @@ func (ms *MapSpec) checkCompatibility(m *Map) error {
 		m.maxEntries != ms.MaxEntries:
 		return fmt.Errorf("expected max entries %v, got %v: %w", ms.MaxEntries, m.maxEntries, ErrMapIncompatible)
 
-	case m.flags != ms.Flags:
+	// BPF_F_RDONLY_PROG is set unconditionally for devmaps. Explicitly allow
+	// this mismatch.
+	case !((ms.Type == DevMap || ms.Type == DevMapHash) && m.flags^ms.Flags == unix.BPF_F_RDONLY_PROG) &&
+		m.flags != ms.Flags:
 		return fmt.Errorf("expected flags %v, got %v: %w", ms.Flags, m.flags, ErrMapIncompatible)
 	}
 	return nil
@@ -241,10 +240,7 @@ func NewMap(spec *MapSpec) (*Map, error) {
 //
 // May return an error wrapping ErrMapIncompatible.
 func NewMapWithOptions(spec *MapSpec, opts MapOptions) (*Map, error) {
-	handles := newHandleCache()
+	m, err := newMapWithOptions(spec, opts)
-	defer handles.close()
-
-	m, err := newMapWithOptions(spec, opts, handles)
 	if err != nil {
 		return nil, fmt.Errorf("creating map: %w", err)
 	}
@@ -257,7 +253,7 @@ func NewMapWithOptions(spec *MapSpec, opts MapOptions) (*Map, error) {
 	return m, nil
 }
 
-func newMapWithOptions(spec *MapSpec, opts MapOptions, handles *handleCache) (_ *Map, err error) {
+func newMapWithOptions(spec *MapSpec, opts MapOptions) (_ *Map, err error) {
 	closeOnError := func(c io.Closer) {
 		if err != nil {
 			c.Close()
@@ -284,7 +280,7 @@ func newMapWithOptions(spec *MapSpec, opts MapOptions, handles *handleCache) (_
 		}
 		defer closeOnError(m)
 
-		if err := spec.checkCompatibility(m); err != nil {
+		if err := spec.Compatible(m); err != nil {
 			return nil, fmt.Errorf("use pinned map %s: %w", spec.Name, err)
 		}
 
@@ -307,7 +303,7 @@ func newMapWithOptions(spec *MapSpec, opts MapOptions, handles *handleCache) (_
 			return nil, errors.New("inner maps cannot be pinned")
 		}
 
-		template, err := spec.InnerMap.createMap(nil, opts, handles)
+		template, err := spec.InnerMap.createMap(nil, opts)
 		if err != nil {
 			return nil, fmt.Errorf("inner map: %w", err)
 		}
@@ -319,7 +315,7 @@ func newMapWithOptions(spec *MapSpec, opts MapOptions, handles *handleCache) (_
 		innerFd = template.fd
 	}
 
-	m, err := spec.createMap(innerFd, opts, handles)
+	m, err := spec.createMap(innerFd, opts)
 	if err != nil {
 		return nil, err
 	}
@@ -328,7 +324,7 @@ func newMapWithOptions(spec *MapSpec, opts MapOptions, handles *handleCache) (_
 	if spec.Pinning == PinByName {
 		path := filepath.Join(opts.PinPath, spec.Name)
 		if err := m.Pin(path); err != nil {
-			return nil, fmt.Errorf("pin map: %w", err)
+			return nil, fmt.Errorf("pin map to %s: %w", path, err)
 		}
 	}
 
@@ -337,15 +333,13 @@ func newMapWithOptions(spec *MapSpec, opts MapOptions, handles *handleCache) (_
 
 // createMap validates the spec's properties and creates the map in the kernel
 // using the given opts. It does not populate or freeze the map.
-func (spec *MapSpec) createMap(inner *sys.FD, opts MapOptions, handles *handleCache) (_ *Map, err error) {
+func (spec *MapSpec) createMap(inner *sys.FD, opts MapOptions) (_ *Map, err error) {
 	closeOnError := func(closer io.Closer) {
 		if err != nil {
 			closer.Close()
 		}
 	}
 
-	spec = spec.Copy()
-
 	// Kernels 4.13 through 5.4 used a struct bpf_map_def that contained
 	// additional 'inner_map_idx' and later 'numa_node' fields.
 	// In order to support loading these definitions, tolerate the presence of
@@ -365,17 +359,21 @@ func (spec *MapSpec) createMap(inner *sys.FD, opts MapOptions, handles *handleCa
 		if spec.ValueSize != 0 && spec.ValueSize != 4 {
 			return nil, errors.New("ValueSize must be zero or four for map of map")
 		}
+
+		spec = spec.Copy()
 		spec.ValueSize = 4
 
 	case PerfEventArray:
 		if spec.KeySize != 0 && spec.KeySize != 4 {
 			return nil, errors.New("KeySize must be zero or four for perf event array")
 		}
-		spec.KeySize = 4
 
 		if spec.ValueSize != 0 && spec.ValueSize != 4 {
 			return nil, errors.New("ValueSize must be zero or four for perf event array")
 		}
+
+		spec = spec.Copy()
+		spec.KeySize = 4
 		spec.ValueSize = 4
 
 		if spec.MaxEntries == 0 {
@@ -413,7 +411,7 @@ func (spec *MapSpec) createMap(inner *sys.FD, opts MapOptions, handles *handleCa
 		KeySize:    spec.KeySize,
 		ValueSize:  spec.ValueSize,
 		MaxEntries: spec.MaxEntries,
-		MapFlags:   spec.Flags,
+		MapFlags:   sys.MapFlags(spec.Flags),
 		NumaNode:   spec.NumaNode,
 	}
 
@@ -425,40 +423,43 @@ func (spec *MapSpec) createMap(inner *sys.FD, opts MapOptions, handles *handleCa
 		attr.MapName = sys.NewObjName(spec.Name)
 	}
 
-	if spec.hasBTF() {
+	if spec.Key != nil || spec.Value != nil {
-		handle, err := handles.btfHandle(spec.BTF)
+		handle, keyTypeID, valueTypeID, err := btf.MarshalMapKV(spec.Key, spec.Value)
 		if err != nil && !errors.Is(err, btf.ErrNotSupported) {
 			return nil, fmt.Errorf("load BTF: %w", err)
 		}
 
 		if handle != nil {
-			keyTypeID, err := spec.BTF.TypeID(spec.Key)
+			defer handle.Close()
-			if err != nil {
-				return nil, err
-			}
-
-			valueTypeID, err := spec.BTF.TypeID(spec.Value)
-			if err != nil {
-				return nil, err
-			}
 
+			// Use BTF k/v during map creation.
 			attr.BtfFd = uint32(handle.FD())
-			attr.BtfKeyTypeId = uint32(keyTypeID)
+			attr.BtfKeyTypeId = keyTypeID
-			attr.BtfValueTypeId = uint32(valueTypeID)
+			attr.BtfValueTypeId = valueTypeID
 		}
 	}
 
 	fd, err := sys.MapCreate(&attr)
+	// Some map types don't support BTF k/v in earlier kernel versions.
+	// Remove BTF metadata and retry map creation.
+	if (errors.Is(err, sys.ENOTSUPP) || errors.Is(err, unix.EINVAL)) && attr.BtfFd != 0 {
+		attr.BtfFd, attr.BtfKeyTypeId, attr.BtfValueTypeId = 0, 0, 0
+		fd, err = sys.MapCreate(&attr)
+	}
+
 	if err != nil {
 		if errors.Is(err, unix.EPERM) {
 			return nil, fmt.Errorf("map create: %w (MEMLOCK may be too low, consider rlimit.RemoveMemlock)", err)
 		}
-		if !spec.hasBTF() {
-			return nil, fmt.Errorf("map create without BTF: %w", err)
-		}
 		if errors.Is(err, unix.EINVAL) && attr.MaxEntries == 0 {
 			return nil, fmt.Errorf("map create: %w (MaxEntries may be incorrectly set to zero)", err)
 		}
+		if errors.Is(err, unix.EINVAL) && spec.Type == UnspecifiedMap {
+			return nil, fmt.Errorf("map create: cannot use type %s", UnspecifiedMap)
+		}
+		if attr.BtfFd == 0 {
+			return nil, fmt.Errorf("map create: %w (without BTF k/v)", err)
+		}
 		return nil, fmt.Errorf("map create: %w", err)
 	}
 	defer closeOnError(fd)
@@ -495,7 +496,7 @@ func newMap(fd *sys.FD, name string, typ MapType, keySize, valueSize, maxEntries
 		return nil, err
 	}
 
-	m.fullValueSize = internal.Align(int(valueSize), 8) * possibleCPUs
+	m.fullValueSize = int(internal.Align(valueSize, 8)) * possibleCPUs
 	return m, nil
 }
 
@@ -549,12 +550,7 @@ const LookupLock MapLookupFlags = 4
 //
 // Returns an error if the key doesn't exist, see ErrKeyNotExist.
 func (m *Map) Lookup(key, valueOut interface{}) error {
-	valuePtr, valueBytes := makeBuffer(valueOut, m.fullValueSize)
+	return m.LookupWithFlags(key, valueOut, 0)
-	if err := m.lookup(key, valuePtr, 0); err != nil {
-		return err
-	}
-
-	return m.unmarshalValue(valueOut, valueBytes)
 }
 
 // LookupWithFlags retrieves a value from a Map with flags.
@@ -568,6 +564,10 @@ func (m *Map) Lookup(key, valueOut interface{}) error {
 //
 // Returns an error if the key doesn't exist, see ErrKeyNotExist.
 func (m *Map) LookupWithFlags(key, valueOut interface{}, flags MapLookupFlags) error {
+	if m.typ.hasPerCPUValue() {
+		return m.lookupPerCPU(key, valueOut, flags)
+	}
+
 	valuePtr, valueBytes := makeBuffer(valueOut, m.fullValueSize)
 	if err := m.lookup(key, valuePtr, flags); err != nil {
 		return err
@@ -580,7 +580,7 @@ func (m *Map) LookupWithFlags(key, valueOut interface{}, flags MapLookupFlags) e
 //
 // Returns ErrKeyNotExist if the key doesn't exist.
 func (m *Map) LookupAndDelete(key, valueOut interface{}) error {
-	return m.lookupAndDelete(key, valueOut, 0)
+	return m.LookupAndDeleteWithFlags(key, valueOut, 0)
 }
 
 // LookupAndDeleteWithFlags retrieves and deletes a value from a Map.
@@ -591,7 +591,15 @@ func (m *Map) LookupAndDelete(key, valueOut interface{}) error {
 //
 // Returns ErrKeyNotExist if the key doesn't exist.
 func (m *Map) LookupAndDeleteWithFlags(key, valueOut interface{}, flags MapLookupFlags) error {
-	return m.lookupAndDelete(key, valueOut, flags)
+	if m.typ.hasPerCPUValue() {
+		return m.lookupAndDeletePerCPU(key, valueOut, flags)
+	}
+
+	valuePtr, valueBytes := makeBuffer(valueOut, m.fullValueSize)
+	if err := m.lookupAndDelete(key, valuePtr, flags); err != nil {
+		return err
+	}
+	return m.unmarshalValue(valueOut, valueBytes)
 }
 
 // LookupBytes gets a value from Map.
@@ -609,6 +617,14 @@ func (m *Map) LookupBytes(key interface{}) ([]byte, error) {
 	return valueBytes, err
 }
 
+func (m *Map) lookupPerCPU(key, valueOut any, flags MapLookupFlags) error {
+	valueBytes := make([]byte, m.fullValueSize)
+	if err := m.lookup(key, sys.NewSlicePointer(valueBytes), flags); err != nil {
+		return err
+	}
+	return unmarshalPerCPUValue(valueOut, int(m.valueSize), valueBytes)
+}
+
 func (m *Map) lookup(key interface{}, valueOut sys.Pointer, flags MapLookupFlags) error {
 	keyPtr, err := m.marshalKey(key)
 	if err != nil {
@@ -628,9 +644,15 @@ func (m *Map) lookup(key interface{}, valueOut sys.Pointer, flags MapLookupFlags
 	return nil
 }
 
-func (m *Map) lookupAndDelete(key, valueOut interface{}, flags MapLookupFlags) error {
+func (m *Map) lookupAndDeletePerCPU(key, valueOut any, flags MapLookupFlags) error {
-	valuePtr, valueBytes := makeBuffer(valueOut, m.fullValueSize)
+	valueBytes := make([]byte, m.fullValueSize)
+	if err := m.lookupAndDelete(key, sys.NewSlicePointer(valueBytes), flags); err != nil {
+		return err
+	}
+	return unmarshalPerCPUValue(valueOut, int(m.valueSize), valueBytes)
+}
 
+func (m *Map) lookupAndDelete(key any, valuePtr sys.Pointer, flags MapLookupFlags) error {
 	keyPtr, err := m.marshalKey(key)
 	if err != nil {
 		return fmt.Errorf("can't marshal key: %w", err)
@@ -647,7 +669,7 @@ func (m *Map) lookupAndDelete(key, valueOut interface{}, flags MapLookupFlags) e
 		return fmt.Errorf("lookup and delete: %w", wrapMapError(err))
 	}
 
-	return m.unmarshalValue(valueOut, valueBytes)
+	return nil
 }
 
 // MapUpdateFlags controls the behaviour of the Map.Update call.
@@ -674,15 +696,32 @@ func (m *Map) Put(key, value interface{}) error {
 }
 
 // Update changes the value of a key.
-func (m *Map) Update(key, value interface{}, flags MapUpdateFlags) error {
+func (m *Map) Update(key, value any, flags MapUpdateFlags) error {
-	keyPtr, err := m.marshalKey(key)
+	if m.typ.hasPerCPUValue() {
-	if err != nil {
+		return m.updatePerCPU(key, value, flags)
-		return fmt.Errorf("can't marshal key: %w", err)
 	}
 
 	valuePtr, err := m.marshalValue(value)
 	if err != nil {
-		return fmt.Errorf("can't marshal value: %w", err)
+		return fmt.Errorf("marshal value: %w", err)
+	}
+
+	return m.update(key, valuePtr, flags)
+}
+
+func (m *Map) updatePerCPU(key, value any, flags MapUpdateFlags) error {
+	valuePtr, err := marshalPerCPUValue(value, int(m.valueSize))
+	if err != nil {
+		return fmt.Errorf("marshal value: %w", err)
+	}
+
+	return m.update(key, valuePtr, flags)
+}
+
+func (m *Map) update(key any, valuePtr sys.Pointer, flags MapUpdateFlags) error {
+	keyPtr, err := m.marshalKey(key)
+	if err != nil {
+		return fmt.Errorf("marshal key: %w", err)
 	}
 
 	attr := sys.MapUpdateElemAttr{
@@ -798,12 +837,22 @@ func (m *Map) nextKey(key interface{}, nextKeyOut sys.Pointer) error {
 	return nil
 }
 
+var mmapProtectedPage = internal.Memoize(func() ([]byte, error) {
+	return unix.Mmap(-1, 0, os.Getpagesize(), unix.PROT_NONE, unix.MAP_ANON|unix.MAP_SHARED)
+})
+
 // guessNonExistentKey attempts to perform a map lookup that returns ENOENT.
 // This is necessary on kernels before 4.4.132, since those don't support
 // iterating maps from the start by providing an invalid key pointer.
 func (m *Map) guessNonExistentKey() ([]byte, error) {
-	// Provide an invalid value pointer to prevent a copy on the kernel side.
+	// Map a protected page and use that as the value pointer. This saves some
-	valuePtr := sys.NewPointer(unsafe.Pointer(^uintptr(0)))
+	// work copying out the value, which we're not interested in.
+	page, err := mmapProtectedPage()
+	if err != nil {
+		return nil, err
+	}
+	valuePtr := sys.NewSlicePointer(page)
+
 	randKey := make([]byte, int(m.keySize))
 
 	for i := 0; i < 4; i++ {
@@ -1095,7 +1144,8 @@ func (m *Map) Clone() (*Map, error) {
 // the new path already exists. Re-pinning across filesystems is not supported.
 // You can Clone a map to pin it to a different path.
 //
-// This requires bpffs to be mounted above fileName. See https://docs.cilium.io/en/k8s-doc/admin/#admin-mount-bpffs
+// This requires bpffs to be mounted above fileName.
+// See https://docs.cilium.io/en/stable/network/kubernetes/configuration/#mounting-bpffs-with-systemd
 func (m *Map) Pin(fileName string) error {
 	if err := internal.Pin(m.pinnedPath, fileName, m.fd); err != nil {
 		return err
@@ -1180,10 +1230,6 @@ func (m *Map) unmarshalKey(data interface{}, buf []byte) error {
 }
 
 func (m *Map) marshalValue(data interface{}) (sys.Pointer, error) {
-	if m.typ.hasPerCPUValue() {
-		return marshalPerCPUValue(data, int(m.valueSize))
-	}
-
 	var (
 		buf []byte
 		err error
@@ -1316,8 +1362,7 @@ func marshalMap(m *Map, length int) ([]byte, error) {
 // See Map.Iterate.
 type MapIterator struct {
 	target            *Map
-	prevKey           interface{}
+	curKey            []byte
-	prevBytes         []byte
 	count, maxEntries uint32
 	done              bool
 	err               error
@@ -1327,7 +1372,6 @@ func newMapIterator(target *Map) *MapIterator {
 	return &MapIterator{
 		target:     target,
 		maxEntries: target.maxEntries,
-		prevBytes:  make([]byte, target.keySize),
 	}
 }
 
@@ -1349,26 +1393,35 @@ func (mi *MapIterator) Next(keyOut, valueOut interface{}) bool {
 	// For array-like maps NextKeyBytes returns nil only on after maxEntries
 	// iterations.
 	for mi.count <= mi.maxEntries {
-		var nextBytes []byte
+		var nextKey []byte
-		nextBytes, mi.err = mi.target.NextKeyBytes(mi.prevKey)
+		if mi.curKey == nil {
+			// Pass nil interface to NextKeyBytes to make sure the Map's first key
+			// is returned. If we pass an uninitialized []byte instead, it'll see a
+			// non-nil interface and try to marshal it.
+			nextKey, mi.err = mi.target.NextKeyBytes(nil)
+
+			mi.curKey = make([]byte, mi.target.keySize)
+		} else {
+			nextKey, mi.err = mi.target.NextKeyBytes(mi.curKey)
+		}
 		if mi.err != nil {
+			mi.err = fmt.Errorf("get next key: %w", mi.err)
 			return false
 		}
 
-		if nextBytes == nil {
+		if nextKey == nil {
 			mi.done = true
 			return false
 		}
 
-		// The user can get access to nextBytes since unmarshalBytes
+		// The user can get access to nextKey since unmarshalBytes
 		// does not copy when unmarshaling into a []byte.
 		// Make a copy to prevent accidental corruption of
 		// iterator state.
-		copy(mi.prevBytes, nextBytes)
+		copy(mi.curKey, nextKey)
-		mi.prevKey = mi.prevBytes
 
 		mi.count++
-		mi.err = mi.target.Lookup(nextBytes, valueOut)
+		mi.err = mi.target.Lookup(nextKey, valueOut)
 		if errors.Is(mi.err, ErrKeyNotExist) {
 			// Even though the key should be valid, we couldn't look up
 			// its value. If we're iterating a hash map this is probably
@@ -1381,10 +1434,11 @@ func (mi *MapIterator) Next(keyOut, valueOut interface{}) bool {
 			continue
 		}
 		if mi.err != nil {
+			mi.err = fmt.Errorf("look up next key: %w", mi.err)
 			return false
 		}
 
-		mi.err = mi.target.unmarshalKey(keyOut, nextBytes)
+		mi.err = mi.target.unmarshalKey(keyOut, nextKey)
 		return mi.err == nil
 	}
 

vendor/github.com/cilium/ebpf/marshalers.go

@@ -57,8 +57,10 @@ func marshalBytes(data interface{}, length int) (buf []byte, err error) {
 	case Map, *Map, Program, *Program:
 		err = fmt.Errorf("can't marshal %T", value)
 	default:
-		var wr bytes.Buffer
+		wr := internal.NewBuffer(make([]byte, 0, length))
-		err = binary.Write(&wr, internal.NativeEndian, value)
+		defer internal.PutBuffer(wr)
+
+		err = binary.Write(wr, internal.NativeEndian, value)
 		if err != nil {
 			err = fmt.Errorf("encoding %T: %v", value, err)
 		}

vendor/github.com/cilium/ebpf/prog.go

@@ -10,6 +10,7 @@ import (
 	"runtime"
 	"strings"
 	"time"
+	"unsafe"
 
 	"github.com/cilium/ebpf/asm"
 	"github.com/cilium/ebpf/btf"
@@ -35,16 +36,44 @@ const (
 // verifier log.
 const DefaultVerifierLogSize = 64 * 1024
 
+// maxVerifierLogSize is the maximum size of verifier log buffer the kernel
+// will accept before returning EINVAL.
+const maxVerifierLogSize = math.MaxUint32 >> 2
+
 // ProgramOptions control loading a program into the kernel.
 type ProgramOptions struct {
-	// Controls the detail emitted by the kernel verifier. Set to non-zero
+	// Bitmap controlling the detail emitted by the kernel's eBPF verifier log.
-	// to enable logging.
+	// LogLevel-type values can be ORed together to request specific kinds of
-	LogLevel uint32
+	// verifier output. See the documentation on [ebpf.LogLevel] for details.
-	// Controls the output buffer size for the verifier. Defaults to
+	//
-	// DefaultVerifierLogSize.
+	//  opts.LogLevel = (ebpf.LogLevelBranch | ebpf.LogLevelStats)
+	//
+	// If left to its default value, the program will first be loaded without
+	// verifier output enabled. Upon error, the program load will be repeated
+	// with LogLevelBranch and the given (or default) LogSize value.
+	//
+	// Setting this to a non-zero value will unconditionally enable the verifier
+	// log, populating the [ebpf.Program.VerifierLog] field on successful loads
+	// and including detailed verifier errors if the program is rejected. This
+	// will always allocate an output buffer, but will result in only a single
+	// attempt at loading the program.
+	LogLevel LogLevel
+
+	// Controls the output buffer size for the verifier log, in bytes. See the
+	// documentation on ProgramOptions.LogLevel for details about how this value
+	// is used.
+	//
+	// If this value is set too low to fit the verifier log, the resulting
+	// [ebpf.VerifierError]'s Truncated flag will be true, and the error string
+	// will also contain a hint to that effect.
+	//
+	// Defaults to DefaultVerifierLogSize.
 	LogSize int
-	// Type information used for CO-RE relocations and when attaching to
+
-	// kernel functions.
+	// Disables the verifier log completely, regardless of other options.
+	LogDisabled bool
+
+	// Type information used for CO-RE relocations.
 	//
 	// This is useful in environments where the kernel BTF is not available
 	// (containers) or where it is in a non-standard location. Defaults to
@@ -74,7 +103,7 @@ type ProgramSpec struct {
 	// The program to attach to. Must be provided manually.
 	AttachTarget *Program
 
-	// The name of the ELF section this program orininated from.
+	// The name of the ELF section this program originated from.
 	SectionName string
 
 	Instructions asm.Instructions
@@ -95,11 +124,6 @@ type ProgramSpec struct {
 	// detect this value automatically.
 	KernelVersion uint32
 
-	// The BTF associated with this program. Changing Instructions
-	// will most likely invalidate the contained data, and may
-	// result in errors when attempting to load it into the kernel.
-	BTF *btf.Spec
-
 	// The byte order this program was compiled for, may be nil.
 	ByteOrder binary.ByteOrder
 }
@@ -123,6 +147,10 @@ func (ps *ProgramSpec) Tag() (string, error) {
 	return ps.Instructions.Tag(internal.NativeEndian)
 }
 
+// VerifierError is returned by [NewProgram] and [NewProgramWithOptions] if a
+// program is rejected by the verifier.
+//
+// Use [errors.As] to access the error.
 type VerifierError = internal.VerifierError
 
 // Program represents BPF program loaded into the kernel.
@@ -141,7 +169,10 @@ type Program struct {
 
 // NewProgram creates a new Program.
 //
-// See NewProgramWithOptions for details.
+// See [NewProgramWithOptions] for details.
+//
+// Returns a [VerifierError] containing the full verifier log if the program is
+// rejected by the kernel.
 func NewProgram(spec *ProgramSpec) (*Program, error) {
 	return NewProgramWithOptions(spec, ProgramOptions{})
 }
@@ -151,24 +182,21 @@ func NewProgram(spec *ProgramSpec) (*Program, error) {
 // Loading a program for the first time will perform
 // feature detection by loading small, temporary programs.
 //
-// Returns an error wrapping VerifierError if the program or its BTF is rejected
+// Returns a [VerifierError] containing the full verifier log if the program is
-// by the kernel.
+// rejected by the kernel.
 func NewProgramWithOptions(spec *ProgramSpec, opts ProgramOptions) (*Program, error) {
 	if spec == nil {
 		return nil, errors.New("can't load a program from a nil spec")
 	}
 
-	handles := newHandleCache()
+	prog, err := newProgramWithOptions(spec, opts)
-	defer handles.close()
-
-	prog, err := newProgramWithOptions(spec, opts, handles)
 	if errors.Is(err, asm.ErrUnsatisfiedMapReference) {
 		return nil, fmt.Errorf("cannot load program without loading its whole collection: %w", err)
 	}
 	return prog, err
 }
 
-func newProgramWithOptions(spec *ProgramSpec, opts ProgramOptions, handles *handleCache) (*Program, error) {
+func newProgramWithOptions(spec *ProgramSpec, opts ProgramOptions) (*Program, error) {
 	if len(spec.Instructions) == 0 {
 		return nil, errors.New("instructions cannot be empty")
 	}
@@ -181,6 +209,10 @@ func newProgramWithOptions(spec *ProgramSpec, opts ProgramOptions, handles *hand
 		return nil, fmt.Errorf("can't load %s program on %s", spec.ByteOrder, internal.NativeEndian)
 	}
 
+	if opts.LogSize < 0 {
+		return nil, errors.New("ProgramOptions.LogSize must be a positive value; disable verifier logs using ProgramOptions.LogDisabled")
+	}
+
 	// Kernels before 5.0 (6c4fc209fcf9 "bpf: remove useless version check for prog load")
 	// require the version field to be set to the value of the KERNEL_VERSION
 	// macro for kprobe-type programs.
@@ -206,30 +238,17 @@ func newProgramWithOptions(spec *ProgramSpec, opts ProgramOptions, handles *hand
 		attr.ProgName = sys.NewObjName(spec.Name)
 	}
 
-	kernelTypes := opts.KernelTypes
-
 	insns := make(asm.Instructions, len(spec.Instructions))
 	copy(insns, spec.Instructions)
 
-	var btfDisabled bool
+	handle, fib, lib, err := btf.MarshalExtInfos(insns)
-	if spec.BTF != nil {
+	if err != nil && !errors.Is(err, btf.ErrNotSupported) {
-		if err := applyRelocations(insns, spec.BTF, kernelTypes); err != nil {
+		return nil, fmt.Errorf("load ext_infos: %w", err)
-			return nil, fmt.Errorf("apply CO-RE relocations: %w", err)
 	}
-
-		handle, err := handles.btfHandle(spec.BTF)
-		btfDisabled = errors.Is(err, btf.ErrNotSupported)
-		if err != nil && !btfDisabled {
-			return nil, fmt.Errorf("load BTF: %w", err)
-		}
-
 	if handle != nil {
-			attr.ProgBtfFd = uint32(handle.FD())
+		defer handle.Close()
 
-			fib, lib, err := btf.MarshalExtInfos(insns, spec.BTF.TypeID)
+		attr.ProgBtfFd = uint32(handle.FD())
-			if err != nil {
-				return nil, err
-			}
 
 		attr.FuncInfoRecSize = btf.FuncInfoSize
 		attr.FuncInfoCnt = uint32(len(fib)) / btf.FuncInfoSize
@@ -239,14 +258,34 @@ func newProgramWithOptions(spec *ProgramSpec, opts ProgramOptions, handles *hand
 		attr.LineInfoCnt = uint32(len(lib)) / btf.LineInfoSize
 		attr.LineInfo = sys.NewSlicePointer(lib)
 	}
+
+	if err := applyRelocations(insns, opts.KernelTypes, spec.ByteOrder); err != nil {
+		return nil, fmt.Errorf("apply CO-RE relocations: %w", err)
 	}
 
+	kconfig, err := resolveKconfigReferences(insns)
+	if err != nil {
+		return nil, fmt.Errorf("resolve .kconfig: %w", err)
+	}
+	defer kconfig.Close()
+
 	if err := fixupAndValidate(insns); err != nil {
 		return nil, err
 	}
 
+	handles, err := fixupKfuncs(insns)
+	if err != nil {
+		return nil, fmt.Errorf("fixing up kfuncs: %w", err)
+	}
+	defer handles.close()
+
+	if len(handles) > 0 {
+		fdArray := handles.fdArray()
+		attr.FdArray = sys.NewPointer(unsafe.Pointer(&fdArray[0]))
+	}
+
 	buf := bytes.NewBuffer(make([]byte, 0, insns.Size()))
-	err := insns.Marshal(buf, internal.NativeEndian)
+	err = insns.Marshal(buf, internal.NativeEndian)
 	if err != nil {
 		return nil, err
 	}
@@ -261,28 +300,32 @@ func newProgramWithOptions(spec *ProgramSpec, opts ProgramOptions, handles *hand
 			return nil, fmt.Errorf("attach %s/%s: %w", spec.Type, spec.AttachType, err)
 		}
 
-		attr.AttachBtfId = uint32(targetID)
+		attr.AttachBtfId = targetID
-		attr.AttachProgFd = uint32(spec.AttachTarget.FD())
+		attr.AttachBtfObjFd = uint32(spec.AttachTarget.FD())
 		defer runtime.KeepAlive(spec.AttachTarget)
 	} else if spec.AttachTo != "" {
-		targetID, err := findTargetInKernel(kernelTypes, spec.AttachTo, spec.Type, spec.AttachType)
+		module, targetID, err := findProgramTargetInKernel(spec.AttachTo, spec.Type, spec.AttachType)
 		if err != nil && !errors.Is(err, errUnrecognizedAttachType) {
 			// We ignore errUnrecognizedAttachType since AttachTo may be non-empty
 			// for programs that don't attach anywhere.
 			return nil, fmt.Errorf("attach %s/%s: %w", spec.Type, spec.AttachType, err)
 		}
 
-		attr.AttachBtfId = uint32(targetID)
+		attr.AttachBtfId = targetID
+		if module != nil {
+			attr.AttachBtfObjFd = uint32(module.FD())
+			defer module.Close()
+		}
 	}
 
-	logSize := DefaultVerifierLogSize
+	if opts.LogSize == 0 {
-	if opts.LogSize > 0 {
+		opts.LogSize = DefaultVerifierLogSize
-		logSize = opts.LogSize
 	}
 
+	// The caller requested a specific verifier log level. Set up the log buffer.
 	var logBuf []byte
-	if opts.LogLevel > 0 {
+	if !opts.LogDisabled && opts.LogLevel != 0 {
-		logBuf = make([]byte, logSize)
+		logBuf = make([]byte, opts.LogSize)
 		attr.LogLevel = opts.LogLevel
 		attr.LogSize = uint32(len(logBuf))
 		attr.LogBuf = sys.NewSlicePointer(logBuf)
@@ -293,13 +336,19 @@ func newProgramWithOptions(spec *ProgramSpec, opts ProgramOptions, handles *hand
 		return &Program{unix.ByteSliceToString(logBuf), fd, spec.Name, "", spec.Type}, nil
 	}
 
-	if opts.LogLevel == 0 && opts.LogSize >= 0 {
+	// An error occurred loading the program, but the caller did not explicitly
-		// Re-run with the verifier enabled to get better error messages.
+	// enable the verifier log. Re-run with branch-level verifier logs enabled to
-		logBuf = make([]byte, logSize)
+	// obtain more info. Preserve the original error to return it to the caller.
-		attr.LogLevel = 1
+	// An undersized log buffer will result in ENOSPC regardless of the underlying
+	// cause.
+	var err2 error
+	if !opts.LogDisabled && opts.LogLevel == 0 {
+		logBuf = make([]byte, opts.LogSize)
+		attr.LogLevel = LogLevelBranch
 		attr.LogSize = uint32(len(logBuf))
 		attr.LogBuf = sys.NewSlicePointer(logBuf)
-		_, _ = sys.ProgLoad(attr)
+
+		_, err2 = sys.ProgLoad(attr)
 	}
 
 	switch {
@@ -318,13 +367,14 @@ func newProgramWithOptions(spec *ProgramSpec, opts ProgramOptions, handles *hand
 				return nil, fmt.Errorf("load program: %w", err)
 			}
 		}
+
+		if opts.LogSize > maxVerifierLogSize {
+			return nil, fmt.Errorf("load program: %w (ProgramOptions.LogSize exceeds maximum value of %d)", err, maxVerifierLogSize)
+		}
 	}
 
-	err = internal.ErrorWithLog(err, logBuf)
+	truncated := errors.Is(err, unix.ENOSPC) || errors.Is(err2, unix.ENOSPC)
-	if btfDisabled {
+	return nil, internal.ErrorWithLog("load program", err, logBuf, truncated)
-		return nil, fmt.Errorf("load program: %w (BTF disabled)", err)
-	}
-	return nil, fmt.Errorf("load program: %w", err)
 }
 
 // NewProgramFromFD creates a program from a raw fd.
@@ -362,7 +412,7 @@ func newProgramFromFD(fd *sys.FD) (*Program, error) {
 		return nil, fmt.Errorf("discover program type: %w", err)
 	}
 
-	return &Program{"", fd, "", "", info.Type}, nil
+	return &Program{"", fd, info.Name, "", info.Type}, nil
 }
 
 func (p *Program) String() string {
@@ -433,7 +483,8 @@ func (p *Program) Clone() (*Program, error) {
 // Calling Pin on a previously pinned program will overwrite the path, except when
 // the new path already exists. Re-pinning across filesystems is not supported.
 //
-// This requires bpffs to be mounted above fileName. See https://docs.cilium.io/en/k8s-doc/admin/#admin-mount-bpffs
+// This requires bpffs to be mounted above fileName.
+// See https://docs.cilium.io/en/stable/network/kubernetes/configuration/#mounting-bpffs-with-systemd
 func (p *Program) Pin(fileName string) error {
 	if err := internal.Pin(p.pinnedPath, fileName, p.fd); err != nil {
 		return err
@@ -474,6 +525,9 @@ func (p *Program) Close() error {
 // Various options for Run'ing a Program
 type RunOptions struct {
 	// Program's data input. Required field.
+	//
+	// The kernel expects at least 14 bytes input for an ethernet header for
+	// XDP and SKB programs.
 	Data []byte
 	// Program's data after Program has run. Caller must allocate. Optional field.
 	DataOut []byte
@@ -481,7 +535,10 @@ type RunOptions struct {
 	Context interface{}
 	// Program's context after Program has run. Must be a pointer or slice. Optional field.
 	ContextOut interface{}
-	// Number of times to run Program. Optional field. Defaults to 1.
+	// Minimum number of times to run Program. Optional field. Defaults to 1.
+	//
+	// The program may be executed more often than this due to interruptions, e.g.
+	// when runtime.AllThreadsSyscall is invoked.
 	Repeat uint32
 	// Optional flags.
 	Flags uint32
@@ -490,6 +547,8 @@ type RunOptions struct {
 	CPU uint32
 	// Called whenever the syscall is interrupted, and should be set to testing.B.ResetTimer
 	// or similar. Typically used during benchmarking. Optional field.
+	//
+	// Deprecated: use [testing.B.ReportMetric] with unit "ns/op" instead.
 	Reset func()
 }
 
@@ -517,9 +576,9 @@ func (p *Program) Test(in []byte) (uint32, []byte, error) {
 		Repeat:  1,
 	}
 
-	ret, _, err := p.testRun(&opts)
+	ret, _, err := p.run(&opts)
 	if err != nil {
-		return ret, nil, fmt.Errorf("can't test program: %w", err)
+		return ret, nil, fmt.Errorf("test program: %w", err)
 	}
 	return ret, opts.DataOut, nil
 }
@@ -528,9 +587,9 @@ func (p *Program) Test(in []byte) (uint32, []byte, error) {
 //
 // Note: the same restrictions from Test apply.
 func (p *Program) Run(opts *RunOptions) (uint32, error) {
-	ret, _, err := p.testRun(opts)
+	ret, _, err := p.run(opts)
 	if err != nil {
-		return ret, fmt.Errorf("can't test program: %w", err)
+		return ret, fmt.Errorf("run program: %w", err)
 	}
 	return ret, nil
 }
@@ -542,9 +601,6 @@ func (p *Program) Run(opts *RunOptions) (uint32, error) {
 // run or an error. reset is called whenever the benchmark syscall is
 // interrupted, and should be set to testing.B.ResetTimer or similar.
 //
-// Note: profiling a call to this function will skew it's results, see
-// https://github.com/cilium/ebpf/issues/24
-//
 // This function requires at least Linux 4.12.
 func (p *Program) Benchmark(in []byte, repeat int, reset func()) (uint32, time.Duration, error) {
 	if uint(repeat) > math.MaxUint32 {
@@ -557,14 +613,14 @@ func (p *Program) Benchmark(in []byte, repeat int, reset func()) (uint32, time.D
 		Reset:  reset,
 	}
 
-	ret, total, err := p.testRun(&opts)
+	ret, total, err := p.run(&opts)
 	if err != nil {
-		return ret, total, fmt.Errorf("can't benchmark program: %w", err)
+		return ret, total, fmt.Errorf("benchmark program: %w", err)
 	}
 	return ret, total, nil
 }
 
-var haveProgTestRun = internal.FeatureTest("BPF_PROG_TEST_RUN", "4.12", func() error {
+var haveProgRun = internal.NewFeatureTest("BPF_PROG_RUN", "4.12", func() error {
 	prog, err := NewProgram(&ProgramSpec{
 		// SocketFilter does not require privileges on newer kernels.
 		Type: SocketFilter,
@@ -580,8 +636,7 @@ var haveProgTestRun = internal.FeatureTest("BPF_PROG_TEST_RUN", "4.12", func() e
 	}
 	defer prog.Close()
 
-	// Programs require at least 14 bytes input
+	in := internal.EmptyBPFContext
-	in := make([]byte, 14)
 	attr := sys.ProgRunAttr{
 		ProgFd:     uint32(prog.FD()),
 		DataSizeIn: uint32(len(in)),
@@ -599,7 +654,7 @@ var haveProgTestRun = internal.FeatureTest("BPF_PROG_TEST_RUN", "4.12", func() e
 		// We know that PROG_TEST_RUN is supported if we get EINTR.
 		return nil
 
-	case errors.Is(err, unix.ENOTSUPP):
+	case errors.Is(err, sys.ENOTSUPP):
 		// The first PROG_TEST_RUN patches shipped in 4.12 didn't include
 		// a test runner for SocketFilter. ENOTSUPP means PROG_TEST_RUN is
 		// supported, but not for the program type used in the probe.
@@ -609,12 +664,12 @@ var haveProgTestRun = internal.FeatureTest("BPF_PROG_TEST_RUN", "4.12", func() e
 	return err
 })
 
-func (p *Program) testRun(opts *RunOptions) (uint32, time.Duration, error) {
+func (p *Program) run(opts *RunOptions) (uint32, time.Duration, error) {
 	if uint(len(opts.Data)) > math.MaxUint32 {
 		return 0, 0, fmt.Errorf("input is too long")
 	}
 
-	if err := haveProgTestRun(); err != nil {
+	if err := haveProgRun(); err != nil {
 		return 0, 0, err
 	}
 
@@ -647,24 +702,45 @@ func (p *Program) testRun(opts *RunOptions) (uint32, time.Duration, error) {
 		Cpu:         opts.CPU,
 	}
 
+	if attr.Repeat == 0 {
+		attr.Repeat = 1
+	}
+
+retry:
 	for {
 		err := sys.ProgRun(&attr)
 		if err == nil {
-			break
+			break retry
 		}
 
 		if errors.Is(err, unix.EINTR) {
+			if attr.Repeat == 1 {
+				// Older kernels check whether enough repetitions have been
+				// executed only after checking for pending signals.
+				//
+				//     run signal? done? run ...
+				//
+				// As a result we can get EINTR for repeat==1 even though
+				// the program was run exactly once. Treat this as a
+				// successful run instead.
+				//
+				// Since commit 607b9cc92bd7 ("bpf: Consolidate shared test timing code")
+				// the conditions are reversed:
+				//     run done? signal? ...
+				break retry
+			}
+
 			if opts.Reset != nil {
 				opts.Reset()
 			}
-			continue
+			continue retry
 		}
 
-		if errors.Is(err, unix.ENOTSUPP) {
+		if errors.Is(err, sys.ENOTSUPP) {
-			return 0, 0, fmt.Errorf("kernel doesn't support testing program type %s: %w", p.Type(), ErrNotSupported)
+			return 0, 0, fmt.Errorf("kernel doesn't support running %s: %w", p.Type(), ErrNotSupported)
 		}
 
-		return 0, 0, fmt.Errorf("can't run test: %w", err)
+		return 0, 0, err
 	}
 
 	if opts.DataOut != nil {
@@ -726,7 +802,14 @@ func LoadPinnedProgram(fileName string, opts *LoadPinOptions) (*Program, error)
 		return nil, fmt.Errorf("info for %s: %w", fileName, err)
 	}
 
-	return &Program{"", fd, filepath.Base(fileName), fileName, info.Type}, nil
+	var progName string
+	if haveObjName() == nil {
+		progName = info.Name
+	} else {
+		progName = filepath.Base(fileName)
+	}
+
+	return &Program{"", fd, progName, fileName, info.Type}, nil
 }
 
 // SanitizeName replaces all invalid characters in name with replacement.
@@ -770,11 +853,15 @@ var errUnrecognizedAttachType = errors.New("unrecognized attach type")
 
 // find an attach target type in the kernel.
 //
-// spec may be nil and defaults to the canonical kernel BTF. name together with
+// name, progType and attachType determine which type we need to attach to.
-// progType and attachType determine which type we need to attach to.
 //
-// Returns errUnrecognizedAttachType.
+// The attach target may be in a loaded kernel module.
-func findTargetInKernel(spec *btf.Spec, name string, progType ProgramType, attachType AttachType) (btf.TypeID, error) {
+// In that case the returned handle will be non-nil.
+// The caller is responsible for closing the handle.
+//
+// Returns errUnrecognizedAttachType if the combination of progType and attachType
+// is not recognised.
+func findProgramTargetInKernel(name string, progType ProgramType, attachType AttachType) (*btf.Handle, btf.TypeID, error) {
 	type match struct {
 		p ProgramType
 		a AttachType
@@ -782,59 +869,123 @@ func findTargetInKernel(spec *btf.Spec, name string, progType ProgramType, attac
 
 	var (
 		typeName, featureName string
-		isBTFTypeFunc         = true
+		target                btf.Type
 	)
 
 	switch (match{progType, attachType}) {
 	case match{LSM, AttachLSMMac}:
 		typeName = "bpf_lsm_" + name
 		featureName = name + " LSM hook"
+		target = (*btf.Func)(nil)
 	case match{Tracing, AttachTraceIter}:
 		typeName = "bpf_iter_" + name
 		featureName = name + " iterator"
+		target = (*btf.Func)(nil)
 	case match{Tracing, AttachTraceFEntry}:
 		typeName = name
 		featureName = fmt.Sprintf("fentry %s", name)
+		target = (*btf.Func)(nil)
 	case match{Tracing, AttachTraceFExit}:
 		typeName = name
 		featureName = fmt.Sprintf("fexit %s", name)
+		target = (*btf.Func)(nil)
 	case match{Tracing, AttachModifyReturn}:
 		typeName = name
 		featureName = fmt.Sprintf("fmod_ret %s", name)
+		target = (*btf.Func)(nil)
 	case match{Tracing, AttachTraceRawTp}:
 		typeName = fmt.Sprintf("btf_trace_%s", name)
 		featureName = fmt.Sprintf("raw_tp %s", name)
-		isBTFTypeFunc = false
+		target = (*btf.Typedef)(nil)
 	default:
-		return 0, errUnrecognizedAttachType
+		return nil, 0, errUnrecognizedAttachType
 	}
 
-	spec, err := maybeLoadKernelBTF(spec)
+	spec, err := btf.LoadKernelSpec()
 	if err != nil {
-		return 0, fmt.Errorf("load kernel spec: %w", err)
+		return nil, 0, fmt.Errorf("load kernel spec: %w", err)
 	}
 
-	var target btf.Type
+	spec, module, err := findTargetInKernel(spec, typeName, &target)
-	if isBTFTypeFunc {
-		var targetFunc *btf.Func
-		err = spec.TypeByName(typeName, &targetFunc)
-		target = targetFunc
-	} else {
-		var targetTypedef *btf.Typedef
-		err = spec.TypeByName(typeName, &targetTypedef)
-		target = targetTypedef
-	}
-
-	if err != nil {
 	if errors.Is(err, btf.ErrNotFound) {
-			return 0, &internal.UnsupportedFeatureError{
+		return nil, 0, &internal.UnsupportedFeatureError{Name: featureName}
-				Name: featureName,
 	}
+	// See cilium/ebpf#894. Until we can disambiguate between equally-named kernel
+	// symbols, we should explicitly refuse program loads. They will not reliably
+	// do what the caller intended.
+	if errors.Is(err, btf.ErrMultipleMatches) {
+		return nil, 0, fmt.Errorf("attaching to ambiguous kernel symbol is not supported: %w", err)
 	}
-		return 0, fmt.Errorf("find target for %s: %w", featureName, err)
+	if err != nil {
+		return nil, 0, fmt.Errorf("find target for %s: %w", featureName, err)
 	}
 
-	return spec.TypeID(target)
+	id, err := spec.TypeID(target)
+	return module, id, err
+}
+
+// findTargetInKernel attempts to find a named type in the current kernel.
+//
+// target will point at the found type after a successful call. Searches both
+// vmlinux and any loaded modules.
+//
+// Returns a non-nil handle if the type was found in a module, or btf.ErrNotFound
+// if the type wasn't found at all.
+func findTargetInKernel(kernelSpec *btf.Spec, typeName string, target *btf.Type) (*btf.Spec, *btf.Handle, error) {
+	err := kernelSpec.TypeByName(typeName, target)
+	if errors.Is(err, btf.ErrNotFound) {
+		spec, module, err := findTargetInModule(kernelSpec, typeName, target)
+		if err != nil {
+			return nil, nil, fmt.Errorf("find target in modules: %w", err)
+		}
+		return spec, module, nil
+	}
+	if err != nil {
+		return nil, nil, fmt.Errorf("find target in vmlinux: %w", err)
+	}
+	return kernelSpec, nil, err
+}
+
+// findTargetInModule attempts to find a named type in any loaded module.
+//
+// base must contain the kernel's types and is used to parse kmod BTF. Modules
+// are searched in the order they were loaded.
+//
+// Returns btf.ErrNotFound if the target can't be found in any module.
+func findTargetInModule(base *btf.Spec, typeName string, target *btf.Type) (*btf.Spec, *btf.Handle, error) {
+	it := new(btf.HandleIterator)
+	defer it.Handle.Close()
+
+	for it.Next() {
+		info, err := it.Handle.Info()
+		if err != nil {
+			return nil, nil, fmt.Errorf("get info for BTF ID %d: %w", it.ID, err)
+		}
+
+		if !info.IsModule() {
+			continue
+		}
+
+		spec, err := it.Handle.Spec(base)
+		if err != nil {
+			return nil, nil, fmt.Errorf("parse types for module %s: %w", info.Name, err)
+		}
+
+		err = spec.TypeByName(typeName, target)
+		if errors.Is(err, btf.ErrNotFound) {
+			continue
+		}
+		if err != nil {
+			return nil, nil, fmt.Errorf("lookup type in module %s: %w", info.Name, err)
+		}
+
+		return spec, it.Take(), nil
+	}
+	if err := it.Err(); err != nil {
+		return nil, nil, fmt.Errorf("iterate modules: %w", err)
+	}
+
+	return nil, nil, btf.ErrNotFound
 }
 
 // find an attach target type in a program.

vendor/github.com/cilium/ebpf/run-tests.sh

@@ -6,6 +6,8 @@
 #     $ ./run-tests.sh 5.4
 #     Run a subset of tests:
 #     $ ./run-tests.sh 5.4 ./link
+#     Run using a local kernel image
+#     $ ./run-tests.sh /path/to/bzImage
 
 set -euo pipefail
 
@@ -95,38 +97,45 @@ elif [[ "${1:-}" = "--exec-test" ]]; then
   exit $rc # this return code is "swallowed" by qemu
 fi
 
-readonly kernel_version="${1:-}"
+if [[ -z "${1:-}" ]]; then
-if [[ -z "${kernel_version}" ]]; then
+  echo "Expecting kernel version or path as first argument"
-  echo "Expecting kernel version as first argument"
   exit 1
 fi
-shift
 
-readonly kernel="linux-${kernel_version}.bz"
-readonly selftests="linux-${kernel_version}-selftests-bpf.tgz"
 readonly input="$(mktemp -d)"
 readonly tmp_dir="${TMPDIR:-/tmp}"
-readonly branch="${BRANCH:-master}"
 
 fetch() {
     echo Fetching "${1}"
     pushd "${tmp_dir}" > /dev/null
-    curl -s -L -O --fail --etag-compare "${1}.etag" --etag-save "${1}.etag" "https://github.com/cilium/ci-kernels/raw/${branch}/${1}"
+    curl --no-progress-meter -L -O --fail --etag-compare "${1}.etag" --etag-save "${1}.etag" "https://github.com/cilium/ci-kernels/raw/${BRANCH:-master}/${1}"
     local ret=$?
     popd > /dev/null
     return $ret
 }
 
-fetch "${kernel}"
+if [[ -f "${1}" ]]; then
-cp "${tmp_dir}/${kernel}" "${input}/bzImage"
+  readonly kernel="${1}"
+  cp "${1}" "${input}/bzImage"
+else
+# LINUX_VERSION_CODE test compares this to discovered value.
+  export KERNEL_VERSION="${1}"
 
-if fetch "${selftests}"; then
+  readonly kernel="linux-${1}.bz"
+  readonly selftests="linux-${1}-selftests-bpf.tgz"
+
+  fetch "${kernel}"
+  cp "${tmp_dir}/${kernel}" "${input}/bzImage"
+
+  if fetch "${selftests}"; then
     echo "Decompressing selftests"
     mkdir "${input}/bpf"
     tar --strip-components=4 -xf "${tmp_dir}/${selftests}" -C "${input}/bpf"
-else
+  else
     echo "No selftests found, disabling"
+  fi
 fi
+shift
 
 args=(-short -coverpkg=./... -coverprofile=coverage.out -count 1 ./...)
 if (( $# > 0 )); then
@@ -135,11 +144,9 @@ fi
 
 export GOFLAGS=-mod=readonly
 export CGO_ENABLED=0
-# LINUX_VERSION_CODE test compares this to discovered value.
-export KERNEL_VERSION="${kernel_version}"
 
-echo Testing on "${kernel_version}"
+echo Testing on "${kernel}"
 go test -exec "$script --exec-vm $input" "${args[@]}"
-echo "Test successful on ${kernel_version}"
+echo "Test successful on ${kernel}"
 
 rm -r "${input}"

vendor/github.com/cilium/ebpf/syscalls.go

@@ -4,13 +4,25 @@ import (
 	"bytes"
 	"errors"
 	"fmt"
+	"os"
+	"runtime"
 
 	"github.com/cilium/ebpf/asm"
 	"github.com/cilium/ebpf/internal"
 	"github.com/cilium/ebpf/internal/sys"
+	"github.com/cilium/ebpf/internal/tracefs"
 	"github.com/cilium/ebpf/internal/unix"
 )
 
+var (
+	// pre-allocating these here since they may
+	// get called in hot code paths and cause
+	// unnecessary memory allocations
+	sysErrKeyNotExist  = sys.Error(ErrKeyNotExist, unix.ENOENT)
+	sysErrKeyExist     = sys.Error(ErrKeyExist, unix.EEXIST)
+	sysErrNotSupported = sys.Error(ErrNotSupported, sys.ENOTSUPP)
+)
+
 // invalidBPFObjNameChar returns true if char may not appear in
 // a BPF object name.
 func invalidBPFObjNameChar(char rune) bool {
@@ -47,7 +59,7 @@ func progLoad(insns asm.Instructions, typ ProgramType, license string) (*sys.FD,
 	})
 }
 
-var haveNestedMaps = internal.FeatureTest("nested maps", "4.12", func() error {
+var haveNestedMaps = internal.NewFeatureTest("nested maps", "4.12", func() error {
 	_, err := sys.MapCreate(&sys.MapCreateAttr{
 		MapType:    sys.MapType(ArrayOfMaps),
 		KeySize:    4,
@@ -65,7 +77,7 @@ var haveNestedMaps = internal.FeatureTest("nested maps", "4.12", func() error {
 	return err
 })
 
-var haveMapMutabilityModifiers = internal.FeatureTest("read- and write-only maps", "5.2", func() error {
+var haveMapMutabilityModifiers = internal.NewFeatureTest("read- and write-only maps", "5.2", func() error {
 	// This checks BPF_F_RDONLY_PROG and BPF_F_WRONLY_PROG. Since
 	// BPF_MAP_FREEZE appeared in 5.2 as well we don't do a separate check.
 	m, err := sys.MapCreate(&sys.MapCreateAttr{
@@ -82,7 +94,7 @@ var haveMapMutabilityModifiers = internal.FeatureTest("read- and write-only maps
 	return nil
 })
 
-var haveMmapableMaps = internal.FeatureTest("mmapable maps", "5.5", func() error {
+var haveMmapableMaps = internal.NewFeatureTest("mmapable maps", "5.5", func() error {
 	// This checks BPF_F_MMAPABLE, which appeared in 5.5 for array maps.
 	m, err := sys.MapCreate(&sys.MapCreateAttr{
 		MapType:    sys.MapType(Array),
@@ -98,7 +110,7 @@ var haveMmapableMaps = internal.FeatureTest("mmapable maps", "5.5", func() error
 	return nil
 })
 
-var haveInnerMaps = internal.FeatureTest("inner maps", "5.10", func() error {
+var haveInnerMaps = internal.NewFeatureTest("inner maps", "5.10", func() error {
 	// This checks BPF_F_INNER_MAP, which appeared in 5.10.
 	m, err := sys.MapCreate(&sys.MapCreateAttr{
 		MapType:    sys.MapType(Array),
@@ -114,7 +126,7 @@ var haveInnerMaps = internal.FeatureTest("inner maps", "5.10", func() error {
 	return nil
 })
 
-var haveNoPreallocMaps = internal.FeatureTest("prealloc maps", "4.6", func() error {
+var haveNoPreallocMaps = internal.NewFeatureTest("prealloc maps", "4.6", func() error {
 	// This checks BPF_F_NO_PREALLOC, which appeared in 4.6.
 	m, err := sys.MapCreate(&sys.MapCreateAttr{
 		MapType:    sys.MapType(Hash),
@@ -136,15 +148,15 @@ func wrapMapError(err error) error {
 	}
 
 	if errors.Is(err, unix.ENOENT) {
-		return sys.Error(ErrKeyNotExist, unix.ENOENT)
+		return sysErrKeyNotExist
 	}
 
 	if errors.Is(err, unix.EEXIST) {
-		return sys.Error(ErrKeyExist, unix.EEXIST)
+		return sysErrKeyExist
 	}
 
-	if errors.Is(err, unix.ENOTSUPP) {
+	if errors.Is(err, sys.ENOTSUPP) {
-		return sys.Error(ErrNotSupported, unix.ENOTSUPP)
+		return sysErrNotSupported
 	}
 
 	if errors.Is(err, unix.E2BIG) {
@@ -154,7 +166,7 @@ func wrapMapError(err error) error {
 	return err
 }
 
-var haveObjName = internal.FeatureTest("object names", "4.15", func() error {
+var haveObjName = internal.NewFeatureTest("object names", "4.15", func() error {
 	attr := sys.MapCreateAttr{
 		MapType:    sys.MapType(Array),
 		KeySize:    4,
@@ -172,7 +184,7 @@ var haveObjName = internal.FeatureTest("object names", "4.15", func() error {
 	return nil
 })
 
-var objNameAllowsDot = internal.FeatureTest("dot in object names", "5.2", func() error {
+var objNameAllowsDot = internal.NewFeatureTest("dot in object names", "5.2", func() error {
 	if err := haveObjName(); err != nil {
 		return err
 	}
@@ -194,7 +206,7 @@ var objNameAllowsDot = internal.FeatureTest("dot in object names", "5.2", func()
 	return nil
 })
 
-var haveBatchAPI = internal.FeatureTest("map batch api", "5.6", func() error {
+var haveBatchAPI = internal.NewFeatureTest("map batch api", "5.6", func() error {
 	var maxEntries uint32 = 2
 	attr := sys.MapCreateAttr{
 		MapType:    sys.MapType(Hash),
@@ -226,7 +238,7 @@ var haveBatchAPI = internal.FeatureTest("map batch api", "5.6", func() error {
 	return nil
 })
 
-var haveProbeReadKernel = internal.FeatureTest("bpf_probe_read_kernel", "5.5", func() error {
+var haveProbeReadKernel = internal.NewFeatureTest("bpf_probe_read_kernel", "5.5", func() error {
 	insns := asm.Instructions{
 		asm.Mov.Reg(asm.R1, asm.R10),
 		asm.Add.Imm(asm.R1, -8),
@@ -244,7 +256,7 @@ var haveProbeReadKernel = internal.FeatureTest("bpf_probe_read_kernel", "5.5", f
 	return nil
 })
 
-var haveBPFToBPFCalls = internal.FeatureTest("bpf2bpf calls", "4.16", func() error {
+var haveBPFToBPFCalls = internal.NewFeatureTest("bpf2bpf calls", "4.16", func() error {
 	insns := asm.Instructions{
 		asm.Call.Label("prog2").WithSymbol("prog1"),
 		asm.Return(),
@@ -262,3 +274,32 @@ var haveBPFToBPFCalls = internal.FeatureTest("bpf2bpf calls", "4.16", func() err
 	_ = fd.Close()
 	return nil
 })
+
+var haveSyscallWrapper = internal.NewFeatureTest("syscall wrapper", "4.17", func() error {
+	prefix := internal.PlatformPrefix()
+	if prefix == "" {
+		return fmt.Errorf("unable to find the platform prefix for (%s)", runtime.GOARCH)
+	}
+
+	args := tracefs.ProbeArgs{
+		Type:   tracefs.Kprobe,
+		Symbol: prefix + "sys_bpf",
+		Pid:    -1,
+	}
+
+	var err error
+	args.Group, err = tracefs.RandomGroup("ebpf_probe")
+	if err != nil {
+		return err
+	}
+
+	evt, err := tracefs.NewEvent(args)
+	if errors.Is(err, os.ErrNotExist) {
+		return internal.ErrNotSupported
+	}
+	if err != nil {
+		return err
+	}
+
+	return evt.Close()
+})

vendor/github.com/cilium/ebpf/types.go

@@ -1,6 +1,7 @@
 package ebpf
 
 import (
+	"github.com/cilium/ebpf/internal/sys"
 	"github.com/cilium/ebpf/internal/unix"
 )
 
@@ -10,11 +11,6 @@ import (
 // that will be initialized in the kernel.
 type MapType uint32
 
-// Max returns the latest supported MapType.
-func (MapType) Max() MapType {
-	return maxMapType - 1
-}
-
 // All the various map types that can be created
 const (
 	UnspecifiedMap MapType = iota
@@ -99,8 +95,6 @@ const (
 	InodeStorage
 	// TaskStorage - Specialized local storage map for task_struct.
 	TaskStorage
-	// maxMapType - Bound enum of MapTypes, has to be last in enum.
-	maxMapType
 )
 
 // hasPerCPUValue returns true if the Map stores a value per CPU.
@@ -120,25 +114,9 @@ func (mt MapType) canStoreProgram() bool {
 	return mt == ProgramArray
 }
 
-// hasBTF returns true if the map type supports BTF key/value metadata.
-func (mt MapType) hasBTF() bool {
-	switch mt {
-	case PerfEventArray, CGroupArray, StackTrace, ArrayOfMaps, HashOfMaps, DevMap,
-		DevMapHash, CPUMap, XSKMap, SockMap, SockHash, Queue, Stack, RingBuf:
-		return false
-	default:
-		return true
-	}
-}
-
 // ProgramType of the eBPF program
 type ProgramType uint32
 
-// Max return the latest supported ProgramType.
-func (ProgramType) Max() ProgramType {
-	return maxProgramType - 1
-}
-
 // eBPF program types
 const (
 	UnspecifiedProgram ProgramType = iota
@@ -173,7 +151,6 @@ const (
 	LSM
 	SkLookup
 	Syscall
-	maxProgramType
 )
 
 // AttachType of the eBPF program, needed to differentiate allowed context accesses in
@@ -229,6 +206,7 @@ const (
 	AttachSkReuseportSelect
 	AttachSkReuseportSelectOrMigrate
 	AttachPerfEvent
+	AttachTraceKprobeMulti
 )
 
 // AttachFlags of the eBPF program used in BPF_PROG_ATTACH command
@@ -282,3 +260,20 @@ type BatchOptions struct {
 	ElemFlags uint64
 	Flags     uint64
 }
+
+// LogLevel controls the verbosity of the kernel's eBPF program verifier.
+// These constants can be used for the ProgramOptions.LogLevel field.
+type LogLevel = sys.LogLevel
+
+const (
+	// Print verifier state at branch points.
+	LogLevelBranch = sys.BPF_LOG_LEVEL1
+
+	// Print verifier state for every instruction.
+	// Available since Linux v5.2.
+	LogLevelInstruction = sys.BPF_LOG_LEVEL2
+
+	// Print verifier errors and stats at the end of the verification process.
+	// Available since Linux v5.2.
+	LogLevelStats = sys.BPF_LOG_STATS
+)

vendor/github.com/cilium/ebpf/types_string.go

@@ -38,12 +38,11 @@ func _() {
 	_ = x[RingBuf-27]
 	_ = x[InodeStorage-28]
 	_ = x[TaskStorage-29]
-	_ = x[maxMapType-30]
 }
 
-const _MapType_name = "UnspecifiedMapHashArrayProgramArrayPerfEventArrayPerCPUHashPerCPUArrayStackTraceCGroupArrayLRUHashLRUCPUHashLPMTrieArrayOfMapsHashOfMapsDevMapSockMapCPUMapXSKMapSockHashCGroupStorageReusePortSockArrayPerCPUCGroupStorageQueueStackSkStorageDevMapHashStructOpsMapRingBufInodeStorageTaskStoragemaxMapType"
+const _MapType_name = "UnspecifiedMapHashArrayProgramArrayPerfEventArrayPerCPUHashPerCPUArrayStackTraceCGroupArrayLRUHashLRUCPUHashLPMTrieArrayOfMapsHashOfMapsDevMapSockMapCPUMapXSKMapSockHashCGroupStorageReusePortSockArrayPerCPUCGroupStorageQueueStackSkStorageDevMapHashStructOpsMapRingBufInodeStorageTaskStorage"
 
-var _MapType_index = [...]uint16{0, 14, 18, 23, 35, 49, 59, 70, 80, 91, 98, 108, 115, 126, 136, 142, 149, 155, 161, 169, 182, 200, 219, 224, 229, 238, 248, 260, 267, 279, 290, 300}
+var _MapType_index = [...]uint16{0, 14, 18, 23, 35, 49, 59, 70, 80, 91, 98, 108, 115, 126, 136, 142, 149, 155, 161, 169, 182, 200, 219, 224, 229, 238, 248, 260, 267, 279, 290}
 
 func (i MapType) String() string {
 	if i >= MapType(len(_MapType_index)-1) {
@@ -87,12 +86,11 @@ func _() {
 	_ = x[LSM-29]
 	_ = x[SkLookup-30]
 	_ = x[Syscall-31]
-	_ = x[maxProgramType-32]
 }
 
-const _ProgramType_name = "UnspecifiedProgramSocketFilterKprobeSchedCLSSchedACTTracePointXDPPerfEventCGroupSKBCGroupSockLWTInLWTOutLWTXmitSockOpsSkSKBCGroupDeviceSkMsgRawTracepointCGroupSockAddrLWTSeg6LocalLircMode2SkReuseportFlowDissectorCGroupSysctlRawTracepointWritableCGroupSockoptTracingStructOpsExtensionLSMSkLookupSyscallmaxProgramType"
+const _ProgramType_name = "UnspecifiedProgramSocketFilterKprobeSchedCLSSchedACTTracePointXDPPerfEventCGroupSKBCGroupSockLWTInLWTOutLWTXmitSockOpsSkSKBCGroupDeviceSkMsgRawTracepointCGroupSockAddrLWTSeg6LocalLircMode2SkReuseportFlowDissectorCGroupSysctlRawTracepointWritableCGroupSockoptTracingStructOpsExtensionLSMSkLookupSyscall"
 
-var _ProgramType_index = [...]uint16{0, 18, 30, 36, 44, 52, 62, 65, 74, 83, 93, 98, 104, 111, 118, 123, 135, 140, 153, 167, 179, 188, 199, 212, 224, 245, 258, 265, 274, 283, 286, 294, 301, 315}
+var _ProgramType_index = [...]uint16{0, 18, 30, 36, 44, 52, 62, 65, 74, 83, 93, 98, 104, 111, 118, 123, 135, 140, 153, 167, 179, 188, 199, 212, 224, 245, 258, 265, 274, 283, 286, 294, 301}
 
 func (i ProgramType) String() string {
 	if i >= ProgramType(len(_ProgramType_index)-1) {

vendor/github.com/containerd/cgroups/v3/Makefile

@@ -13,14 +13,17 @@
 #   limitations under the License.
 
 PACKAGES=$(shell go list ./... | grep -v /vendor/)
+GO_TAGS=$(if $(GO_BUILDTAGS),-tags "$(strip $(GO_BUILDTAGS))",)
+GO ?= go
+GO_BUILD_FLAGS ?=
 
 all: cgutil
-	go build -v
+	$(GO) build -v $(GO_TAGS)
 
 cgutil:
-	cd cmd/cgctl && go build -v
+	cd cmd/cgctl && $(GO) build $(GO_BUILD_FLAGS) -v $(GO_TAGS)
 
 proto:
 	protobuild --quiet ${PACKAGES}
 	# Keep them Go-idiomatic and backward-compatible with the gogo/protobuf era.
-	go-fix-acronym -w -a '(Cpu|Tcp|Rss)' $(shell find cgroup1/stats/ cgroup2/stats/ -name '*.pb.go')
+	go-fix-acronym -w -a '(Cpu|Tcp|Rss|Psi)' $(shell find cgroup1/stats/ cgroup2/stats/ -name '*.pb.go')

vendor/github.com/containerd/cgroups/v3/cgroup1/memory.go

@@ -454,6 +454,9 @@ func getOomControlValue(mem *specs.LinuxMemory) *int64 {
 	if mem.DisableOOMKiller != nil && *mem.DisableOOMKiller {
 		i := int64(1)
 		return &i
+	} else if mem.DisableOOMKiller != nil && !*mem.DisableOOMKiller {
+		i := int64(0)
+		return &i
 	}
 	return nil
 }

vendor/github.com/containerd/cgroups/v3/cgroup2/manager.go

@@ -21,6 +21,7 @@ import (
 	"context"
 	"errors"
 	"fmt"
+	"io/fs"
 	"math"
 	"os"
 	"path/filepath"
@@ -478,9 +479,9 @@ func (c *Manager) Delete() error {
 	return remove(c.path)
 }
 
-func (c *Manager) Procs(recursive bool) ([]uint64, error) {
+func (c *Manager) getTasks(recursive bool, tType string) ([]uint64, error) {
-	var processes []uint64
+	var tasks []uint64
-	err := filepath.Walk(c.path, func(p string, info os.FileInfo, err error) error {
+	err := filepath.Walk(c.path, func(p string, info fs.FileInfo, err error) error {
 		if err != nil {
 			return err
 		}
@@ -491,17 +492,25 @@ func (c *Manager) Procs(recursive bool) ([]uint64, error) {
 			return filepath.SkipDir
 		}
 		_, name := filepath.Split(p)
-		if name != cgroupProcs {
+		if name != tType {
 			return nil
 		}
-		procs, err := parseCgroupProcsFile(p)
+		curTasks, err := parseCgroupTasksFile(p)
 		if err != nil {
 			return err
 		}
-		processes = append(processes, procs...)
+		tasks = append(tasks, curTasks...)
 		return nil
 	})
-	return processes, err
+	return tasks, err
+}
+
+func (c *Manager) Procs(recursive bool) ([]uint64, error) {
+	return c.getTasks(recursive, cgroupProcs)
+}
+
+func (c *Manager) Threads(recursive bool) ([]uint64, error) {
+	return c.getTasks(recursive, cgroupThreads)
 }
 
 func (c *Manager) MoveTo(destination *Manager) error {
@@ -559,6 +568,7 @@ func (c *Manager) Stat() (*stats.Metrics, error) {
 		NrPeriods:     out["nr_periods"],
 		NrThrottled:   out["nr_throttled"],
 		ThrottledUsec: out["throttled_usec"],
+		PSI:           getStatPSIFromFile(filepath.Join(c.path, "cpu.pressure")),
 	}
 	metrics.Memory = &stats.MemoryStat{
 		Anon:                  out["anon"],
@@ -594,8 +604,11 @@ func (c *Manager) Stat() (*stats.Metrics, error) {
 		ThpCollapseAlloc:      out["thp_collapse_alloc"],
 		Usage:                 getStatFileContentUint64(filepath.Join(c.path, "memory.current")),
 		UsageLimit:            getStatFileContentUint64(filepath.Join(c.path, "memory.max")),
+		MaxUsage:              getStatFileContentUint64(filepath.Join(c.path, "memory.peak")),
 		SwapUsage:             getStatFileContentUint64(filepath.Join(c.path, "memory.swap.current")),
 		SwapLimit:             getStatFileContentUint64(filepath.Join(c.path, "memory.swap.max")),
+		SwapMaxUsage:          getStatFileContentUint64(filepath.Join(c.path, "memory.swap.peak")),
+		PSI:                   getStatPSIFromFile(filepath.Join(c.path, "memory.pressure")),
 	}
 	if len(memoryEvents) > 0 {
 		metrics.MemoryEvents = &stats.MemoryEvents{
@@ -606,7 +619,10 @@ func (c *Manager) Stat() (*stats.Metrics, error) {
 			OomKill: memoryEvents["oom_kill"],
 		}
 	}
-	metrics.Io = &stats.IOStat{Usage: readIoStats(c.path)}
+	metrics.Io = &stats.IOStat{
+		Usage: readIoStats(c.path),
+		PSI:   getStatPSIFromFile(filepath.Join(c.path, "io.pressure")),
+	}
 	metrics.Rdma = &stats.RdmaStat{
 		Current: rdmaStats(filepath.Join(c.path, "rdma.current")),
 		Limit:   rdmaStats(filepath.Join(c.path, "rdma.max")),
@@ -870,14 +886,7 @@ func NewSystemd(slice, group string, pid int, resources *Resources) (*Manager, e
 			newSystemdProperty("TasksMax", uint64(resources.Pids.Max)))
 	}
 
-	statusChan := make(chan string, 1)
+	if err := startUnit(conn, group, properties, pid == -1); err != nil {
-	if _, err := conn.StartTransientUnitContext(ctx, group, "replace", properties, statusChan); err == nil {
-		select {
-		case <-statusChan:
-		case <-time.After(time.Second):
-			logrus.Warnf("Timed out while waiting for StartTransientUnit(%s) completion signal from dbus. Continuing...", group)
-		}
-	} else if !isUnitExists(err) {
 		return &Manager{}, err
 	}
 
@@ -886,6 +895,60 @@ func NewSystemd(slice, group string, pid int, resources *Resources) (*Manager, e
 	}, nil
 }
 
+func startUnit(conn *systemdDbus.Conn, group string, properties []systemdDbus.Property, ignoreExists bool) error {
+	ctx := context.TODO()
+
+	statusChan := make(chan string, 1)
+	defer close(statusChan)
+
+	retry := true
+	started := false
+
+	for !started {
+		if _, err := conn.StartTransientUnitContext(ctx, group, "replace", properties, statusChan); err != nil {
+			if !isUnitExists(err) {
+				return err
+			}
+
+			if ignoreExists {
+				return nil
+			}
+
+			if retry {
+				retry = false
+				// When a unit of the same name already exists, it may be a leftover failed unit.
+				// If we reset it once, systemd can try to remove it.
+				attemptFailedUnitReset(conn, group)
+				continue
+			}
+
+			return err
+		} else {
+			started = true
+		}
+	}
+
+	select {
+	case s := <-statusChan:
+		if s != "done" {
+			attemptFailedUnitReset(conn, group)
+			return fmt.Errorf("error creating systemd unit `%s`: got `%s`", group, s)
+		}
+	case <-time.After(30 * time.Second):
+		logrus.Warnf("Timed out while waiting for StartTransientUnit(%s) completion signal from dbus. Continuing...", group)
+	}
+
+	return nil
+}
+
+func attemptFailedUnitReset(conn *systemdDbus.Conn, group string) {
+	err := conn.ResetFailedUnitContext(context.TODO(), group)
+
+	if err != nil {
+		logrus.Warnf("Unable to reset failed unit: %v", err)
+	}
+}
+
 func LoadSystemd(slice, group string) (*Manager, error) {
 	if slice == "" {
 		slice = defaultSlice

vendor/github.com/containerd/cgroups/v3/cgroup2/stats/metrics.pb.go

@@ -115,6 +115,132 @@ func (x *Metrics) GetMemoryEvents() *MemoryEvents {
 	return nil
 }
 
+type PSIData struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+
+	Avg10  float64 `protobuf:"fixed64,1,opt,name=avg10,proto3" json:"avg10,omitempty"`
+	Avg60  float64 `protobuf:"fixed64,2,opt,name=avg60,proto3" json:"avg60,omitempty"`
+	Avg300 float64 `protobuf:"fixed64,3,opt,name=avg300,proto3" json:"avg300,omitempty"`
+	Total  uint64  `protobuf:"varint,4,opt,name=total,proto3" json:"total,omitempty"`
+}
+
+func (x *PSIData) Reset() {
+	*x = PSIData{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[1]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *PSIData) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*PSIData) ProtoMessage() {}
+
+func (x *PSIData) ProtoReflect() protoreflect.Message {
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[1]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use PSIData.ProtoReflect.Descriptor instead.
+func (*PSIData) Descriptor() ([]byte, []int) {
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{1}
+}
+
+func (x *PSIData) GetAvg10() float64 {
+	if x != nil {
+		return x.Avg10
+	}
+	return 0
+}
+
+func (x *PSIData) GetAvg60() float64 {
+	if x != nil {
+		return x.Avg60
+	}
+	return 0
+}
+
+func (x *PSIData) GetAvg300() float64 {
+	if x != nil {
+		return x.Avg300
+	}
+	return 0
+}
+
+func (x *PSIData) GetTotal() uint64 {
+	if x != nil {
+		return x.Total
+	}
+	return 0
+}
+
+type PSIStats struct {
+	state         protoimpl.MessageState
+	sizeCache     protoimpl.SizeCache
+	unknownFields protoimpl.UnknownFields
+
+	Some *PSIData `protobuf:"bytes,1,opt,name=some,proto3" json:"some,omitempty"`
+	Full *PSIData `protobuf:"bytes,2,opt,name=full,proto3" json:"full,omitempty"`
+}
+
+func (x *PSIStats) Reset() {
+	*x = PSIStats{}
+	if protoimpl.UnsafeEnabled {
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[2]
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		ms.StoreMessageInfo(mi)
+	}
+}
+
+func (x *PSIStats) String() string {
+	return protoimpl.X.MessageStringOf(x)
+}
+
+func (*PSIStats) ProtoMessage() {}
+
+func (x *PSIStats) ProtoReflect() protoreflect.Message {
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[2]
+	if protoimpl.UnsafeEnabled && x != nil {
+		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
+		if ms.LoadMessageInfo() == nil {
+			ms.StoreMessageInfo(mi)
+		}
+		return ms
+	}
+	return mi.MessageOf(x)
+}
+
+// Deprecated: Use PSIStats.ProtoReflect.Descriptor instead.
+func (*PSIStats) Descriptor() ([]byte, []int) {
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{2}
+}
+
+func (x *PSIStats) GetSome() *PSIData {
+	if x != nil {
+		return x.Some
+	}
+	return nil
+}
+
+func (x *PSIStats) GetFull() *PSIData {
+	if x != nil {
+		return x.Full
+	}
+	return nil
+}
+
 type PidsStat struct {
 	state         protoimpl.MessageState
 	sizeCache     protoimpl.SizeCache
@@ -127,7 +253,7 @@ type PidsStat struct {
 func (x *PidsStat) Reset() {
 	*x = PidsStat{}
 	if protoimpl.UnsafeEnabled {
-		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[1]
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[3]
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		ms.StoreMessageInfo(mi)
 	}
@@ -140,7 +266,7 @@ func (x *PidsStat) String() string {
 func (*PidsStat) ProtoMessage() {}
 
 func (x *PidsStat) ProtoReflect() protoreflect.Message {
-	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[1]
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[3]
 	if protoimpl.UnsafeEnabled && x != nil {
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		if ms.LoadMessageInfo() == nil {
@@ -153,7 +279,7 @@ func (x *PidsStat) ProtoReflect() protoreflect.Message {
 
 // Deprecated: Use PidsStat.ProtoReflect.Descriptor instead.
 func (*PidsStat) Descriptor() ([]byte, []int) {
-	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{1}
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{3}
 }
 
 func (x *PidsStat) GetCurrent() uint64 {
@@ -181,12 +307,13 @@ type CPUStat struct {
 	NrPeriods     uint64    `protobuf:"varint,4,opt,name=nr_periods,json=nrPeriods,proto3" json:"nr_periods,omitempty"`
 	NrThrottled   uint64    `protobuf:"varint,5,opt,name=nr_throttled,json=nrThrottled,proto3" json:"nr_throttled,omitempty"`
 	ThrottledUsec uint64    `protobuf:"varint,6,opt,name=throttled_usec,json=throttledUsec,proto3" json:"throttled_usec,omitempty"`
+	PSI           *PSIStats `protobuf:"bytes,7,opt,name=psi,proto3" json:"psi,omitempty"`
 }
 
 func (x *CPUStat) Reset() {
 	*x = CPUStat{}
 	if protoimpl.UnsafeEnabled {
-		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[2]
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[4]
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		ms.StoreMessageInfo(mi)
 	}
@@ -199,7 +326,7 @@ func (x *CPUStat) String() string {
 func (*CPUStat) ProtoMessage() {}
 
 func (x *CPUStat) ProtoReflect() protoreflect.Message {
-	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[2]
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[4]
 	if protoimpl.UnsafeEnabled && x != nil {
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		if ms.LoadMessageInfo() == nil {
@@ -212,7 +339,7 @@ func (x *CPUStat) ProtoReflect() protoreflect.Message {
 
 // Deprecated: Use CPUStat.ProtoReflect.Descriptor instead.
 func (*CPUStat) Descriptor() ([]byte, []int) {
-	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{2}
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{4}
 }
 
 func (x *CPUStat) GetUsageUsec() uint64 {
@@ -257,6 +384,13 @@ func (x *CPUStat) GetThrottledUsec() uint64 {
 	return 0
 }
 
+func (x *CPUStat) GetPSI() *PSIStats {
+	if x != nil {
+		return x.PSI
+	}
+	return nil
+}
+
 type MemoryStat struct {
 	state         protoimpl.MessageState
 	sizeCache     protoimpl.SizeCache
@@ -297,12 +431,15 @@ type MemoryStat struct {
 	UsageLimit            uint64    `protobuf:"varint,33,opt,name=usage_limit,json=usageLimit,proto3" json:"usage_limit,omitempty"`
 	SwapUsage             uint64    `protobuf:"varint,34,opt,name=swap_usage,json=swapUsage,proto3" json:"swap_usage,omitempty"`
 	SwapLimit             uint64    `protobuf:"varint,35,opt,name=swap_limit,json=swapLimit,proto3" json:"swap_limit,omitempty"`
+	MaxUsage              uint64    `protobuf:"varint,36,opt,name=max_usage,json=maxUsage,proto3" json:"max_usage,omitempty"`
+	SwapMaxUsage          uint64    `protobuf:"varint,37,opt,name=swap_max_usage,json=swapMaxUsage,proto3" json:"swap_max_usage,omitempty"`
+	PSI                   *PSIStats `protobuf:"bytes,38,opt,name=psi,proto3" json:"psi,omitempty"`
 }
 
 func (x *MemoryStat) Reset() {
 	*x = MemoryStat{}
 	if protoimpl.UnsafeEnabled {
-		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[3]
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[5]
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		ms.StoreMessageInfo(mi)
 	}
@@ -315,7 +452,7 @@ func (x *MemoryStat) String() string {
 func (*MemoryStat) ProtoMessage() {}
 
 func (x *MemoryStat) ProtoReflect() protoreflect.Message {
-	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[3]
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[5]
 	if protoimpl.UnsafeEnabled && x != nil {
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		if ms.LoadMessageInfo() == nil {
@@ -328,7 +465,7 @@ func (x *MemoryStat) ProtoReflect() protoreflect.Message {
 
 // Deprecated: Use MemoryStat.ProtoReflect.Descriptor instead.
 func (*MemoryStat) Descriptor() ([]byte, []int) {
-	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{3}
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{5}
 }
 
 func (x *MemoryStat) GetAnon() uint64 {
@@ -576,6 +713,27 @@ func (x *MemoryStat) GetSwapLimit() uint64 {
 	return 0
 }
 
+func (x *MemoryStat) GetMaxUsage() uint64 {
+	if x != nil {
+		return x.MaxUsage
+	}
+	return 0
+}
+
+func (x *MemoryStat) GetSwapMaxUsage() uint64 {
+	if x != nil {
+		return x.SwapMaxUsage
+	}
+	return 0
+}
+
+func (x *MemoryStat) GetPSI() *PSIStats {
+	if x != nil {
+		return x.PSI
+	}
+	return nil
+}
+
 type MemoryEvents struct {
 	state         protoimpl.MessageState
 	sizeCache     protoimpl.SizeCache
@@ -591,7 +749,7 @@ type MemoryEvents struct {
 func (x *MemoryEvents) Reset() {
 	*x = MemoryEvents{}
 	if protoimpl.UnsafeEnabled {
-		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[4]
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[6]
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		ms.StoreMessageInfo(mi)
 	}
@@ -604,7 +762,7 @@ func (x *MemoryEvents) String() string {
 func (*MemoryEvents) ProtoMessage() {}
 
 func (x *MemoryEvents) ProtoReflect() protoreflect.Message {
-	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[4]
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[6]
 	if protoimpl.UnsafeEnabled && x != nil {
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		if ms.LoadMessageInfo() == nil {
@@ -617,7 +775,7 @@ func (x *MemoryEvents) ProtoReflect() protoreflect.Message {
 
 // Deprecated: Use MemoryEvents.ProtoReflect.Descriptor instead.
 func (*MemoryEvents) Descriptor() ([]byte, []int) {
-	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{4}
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{6}
 }
 
 func (x *MemoryEvents) GetLow() uint64 {
@@ -667,7 +825,7 @@ type RdmaStat struct {
 func (x *RdmaStat) Reset() {
 	*x = RdmaStat{}
 	if protoimpl.UnsafeEnabled {
-		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[5]
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[7]
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		ms.StoreMessageInfo(mi)
 	}
@@ -680,7 +838,7 @@ func (x *RdmaStat) String() string {
 func (*RdmaStat) ProtoMessage() {}
 
 func (x *RdmaStat) ProtoReflect() protoreflect.Message {
-	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[5]
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[7]
 	if protoimpl.UnsafeEnabled && x != nil {
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		if ms.LoadMessageInfo() == nil {
@@ -693,7 +851,7 @@ func (x *RdmaStat) ProtoReflect() protoreflect.Message {
 
 // Deprecated: Use RdmaStat.ProtoReflect.Descriptor instead.
 func (*RdmaStat) Descriptor() ([]byte, []int) {
-	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{5}
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{7}
 }
 
 func (x *RdmaStat) GetCurrent() []*RdmaEntry {
@@ -723,7 +881,7 @@ type RdmaEntry struct {
 func (x *RdmaEntry) Reset() {
 	*x = RdmaEntry{}
 	if protoimpl.UnsafeEnabled {
-		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[6]
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[8]
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		ms.StoreMessageInfo(mi)
 	}
@@ -736,7 +894,7 @@ func (x *RdmaEntry) String() string {
 func (*RdmaEntry) ProtoMessage() {}
 
 func (x *RdmaEntry) ProtoReflect() protoreflect.Message {
-	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[6]
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[8]
 	if protoimpl.UnsafeEnabled && x != nil {
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		if ms.LoadMessageInfo() == nil {
@@ -749,7 +907,7 @@ func (x *RdmaEntry) ProtoReflect() protoreflect.Message {
 
 // Deprecated: Use RdmaEntry.ProtoReflect.Descriptor instead.
 func (*RdmaEntry) Descriptor() ([]byte, []int) {
-	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{6}
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{8}
 }
 
 func (x *RdmaEntry) GetDevice() string {
@@ -779,12 +937,13 @@ type IOStat struct {
 	unknownFields protoimpl.UnknownFields
 
 	Usage []*IOEntry `protobuf:"bytes,1,rep,name=usage,proto3" json:"usage,omitempty"`
+	PSI   *PSIStats  `protobuf:"bytes,2,opt,name=psi,proto3" json:"psi,omitempty"`
 }
 
 func (x *IOStat) Reset() {
 	*x = IOStat{}
 	if protoimpl.UnsafeEnabled {
-		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[7]
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[9]
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		ms.StoreMessageInfo(mi)
 	}
@@ -797,7 +956,7 @@ func (x *IOStat) String() string {
 func (*IOStat) ProtoMessage() {}
 
 func (x *IOStat) ProtoReflect() protoreflect.Message {
-	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[7]
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[9]
 	if protoimpl.UnsafeEnabled && x != nil {
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		if ms.LoadMessageInfo() == nil {
@@ -810,7 +969,7 @@ func (x *IOStat) ProtoReflect() protoreflect.Message {
 
 // Deprecated: Use IOStat.ProtoReflect.Descriptor instead.
 func (*IOStat) Descriptor() ([]byte, []int) {
-	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{7}
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{9}
 }
 
 func (x *IOStat) GetUsage() []*IOEntry {
@@ -820,6 +979,13 @@ func (x *IOStat) GetUsage() []*IOEntry {
 	return nil
 }
 
+func (x *IOStat) GetPSI() *PSIStats {
+	if x != nil {
+		return x.PSI
+	}
+	return nil
+}
+
 type IOEntry struct {
 	state         protoimpl.MessageState
 	sizeCache     protoimpl.SizeCache
@@ -836,7 +1002,7 @@ type IOEntry struct {
 func (x *IOEntry) Reset() {
 	*x = IOEntry{}
 	if protoimpl.UnsafeEnabled {
-		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[8]
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[10]
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		ms.StoreMessageInfo(mi)
 	}
@@ -849,7 +1015,7 @@ func (x *IOEntry) String() string {
 func (*IOEntry) ProtoMessage() {}
 
 func (x *IOEntry) ProtoReflect() protoreflect.Message {
-	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[8]
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[10]
 	if protoimpl.UnsafeEnabled && x != nil {
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		if ms.LoadMessageInfo() == nil {
@@ -862,7 +1028,7 @@ func (x *IOEntry) ProtoReflect() protoreflect.Message {
 
 // Deprecated: Use IOEntry.ProtoReflect.Descriptor instead.
 func (*IOEntry) Descriptor() ([]byte, []int) {
-	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{8}
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{10}
 }
 
 func (x *IOEntry) GetMajor() uint64 {
@@ -920,7 +1086,7 @@ type HugeTlbStat struct {
 func (x *HugeTlbStat) Reset() {
 	*x = HugeTlbStat{}
 	if protoimpl.UnsafeEnabled {
-		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[9]
+		mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[11]
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		ms.StoreMessageInfo(mi)
 	}
@@ -933,7 +1099,7 @@ func (x *HugeTlbStat) String() string {
 func (*HugeTlbStat) ProtoMessage() {}
 
 func (x *HugeTlbStat) ProtoReflect() protoreflect.Message {
-	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[9]
+	mi := &file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[11]
 	if protoimpl.UnsafeEnabled && x != nil {
 		ms := protoimpl.X.MessageStateOf(protoimpl.Pointer(x))
 		if ms.LoadMessageInfo() == nil {
@@ -946,7 +1112,7 @@ func (x *HugeTlbStat) ProtoReflect() protoreflect.Message {
 
 // Deprecated: Use HugeTlbStat.ProtoReflect.Descriptor instead.
 func (*HugeTlbStat) Descriptor() ([]byte, []int) {
-	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{9}
+	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(), []int{11}
 }
 
 func (x *HugeTlbStat) GetCurrent() uint64 {
@@ -1005,141 +1171,169 @@ var file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDesc = []b
 	0x2e, 0x69, 0x6f, 0x2e, 0x63, 0x6f, 0x6e, 0x74, 0x61, 0x69, 0x6e, 0x65, 0x72, 0x64, 0x2e, 0x63,
 	0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2e, 0x76, 0x32, 0x2e, 0x4d, 0x65, 0x6d, 0x6f, 0x72, 0x79,
 	0x45, 0x76, 0x65, 0x6e, 0x74, 0x73, 0x52, 0x0c, 0x6d, 0x65, 0x6d, 0x6f, 0x72, 0x79, 0x45, 0x76,
-	0x65, 0x6e, 0x74, 0x73, 0x22, 0x3a, 0x0a, 0x08, 0x50, 0x69, 0x64, 0x73, 0x53, 0x74, 0x61, 0x74,
+	0x65, 0x6e, 0x74, 0x73, 0x22, 0x63, 0x0a, 0x07, 0x50, 0x53, 0x49, 0x44, 0x61, 0x74, 0x61, 0x12,
-	0x12, 0x18, 0x0a, 0x07, 0x63, 0x75, 0x72, 0x72, 0x65, 0x6e, 0x74, 0x18, 0x01, 0x20, 0x01, 0x28,
+	0x14, 0x0a, 0x05, 0x61, 0x76, 0x67, 0x31, 0x30, 0x18, 0x01, 0x20, 0x01, 0x28, 0x01, 0x52, 0x05,
-	0x04, 0x52, 0x07, 0x63, 0x75, 0x72, 0x72, 0x65, 0x6e, 0x74, 0x12, 0x14, 0x0a, 0x05, 0x6c, 0x69,
+	0x61, 0x76, 0x67, 0x31, 0x30, 0x12, 0x14, 0x0a, 0x05, 0x61, 0x76, 0x67, 0x36, 0x30, 0x18, 0x02,
-	0x6d, 0x69, 0x74, 0x18, 0x02, 0x20, 0x01, 0x28, 0x04, 0x52, 0x05, 0x6c, 0x69, 0x6d, 0x69, 0x74,
+	0x20, 0x01, 0x28, 0x01, 0x52, 0x05, 0x61, 0x76, 0x67, 0x36, 0x30, 0x12, 0x16, 0x0a, 0x06, 0x61,
-	0x22, 0xcf, 0x01, 0x0a, 0x07, 0x43, 0x50, 0x55, 0x53, 0x74, 0x61, 0x74, 0x12, 0x1d, 0x0a, 0x0a,
+	0x76, 0x67, 0x33, 0x30, 0x30, 0x18, 0x03, 0x20, 0x01, 0x28, 0x01, 0x52, 0x06, 0x61, 0x76, 0x67,
-	0x75, 0x73, 0x61, 0x67, 0x65, 0x5f, 0x75, 0x73, 0x65, 0x63, 0x18, 0x01, 0x20, 0x01, 0x28, 0x04,
+	0x33, 0x30, 0x30, 0x12, 0x14, 0x0a, 0x05, 0x74, 0x6f, 0x74, 0x61, 0x6c, 0x18, 0x04, 0x20, 0x01,
-	0x52, 0x09, 0x75, 0x73, 0x61, 0x67, 0x65, 0x55, 0x73, 0x65, 0x63, 0x12, 0x1b, 0x0a, 0x09, 0x75,
+	0x28, 0x04, 0x52, 0x05, 0x74, 0x6f, 0x74, 0x61, 0x6c, 0x22, 0x78, 0x0a, 0x08, 0x50, 0x53, 0x49,
-	0x73, 0x65, 0x72, 0x5f, 0x75, 0x73, 0x65, 0x63, 0x18, 0x02, 0x20, 0x01, 0x28, 0x04, 0x52, 0x08,
+	0x53, 0x74, 0x61, 0x74, 0x73, 0x12, 0x35, 0x0a, 0x04, 0x73, 0x6f, 0x6d, 0x65, 0x18, 0x01, 0x20,
-	0x75, 0x73, 0x65, 0x72, 0x55, 0x73, 0x65, 0x63, 0x12, 0x1f, 0x0a, 0x0b, 0x73, 0x79, 0x73, 0x74,
+	0x01, 0x28, 0x0b, 0x32, 0x21, 0x2e, 0x69, 0x6f, 0x2e, 0x63, 0x6f, 0x6e, 0x74, 0x61, 0x69, 0x6e,
-	0x65, 0x6d, 0x5f, 0x75, 0x73, 0x65, 0x63, 0x18, 0x03, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x73,
+	0x65, 0x72, 0x64, 0x2e, 0x63, 0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2e, 0x76, 0x32, 0x2e, 0x50,
-	0x79, 0x73, 0x74, 0x65, 0x6d, 0x55, 0x73, 0x65, 0x63, 0x12, 0x1d, 0x0a, 0x0a, 0x6e, 0x72, 0x5f,
+	0x53, 0x49, 0x44, 0x61, 0x74, 0x61, 0x52, 0x04, 0x73, 0x6f, 0x6d, 0x65, 0x12, 0x35, 0x0a, 0x04,
-	0x70, 0x65, 0x72, 0x69, 0x6f, 0x64, 0x73, 0x18, 0x04, 0x20, 0x01, 0x28, 0x04, 0x52, 0x09, 0x6e,
+	0x66, 0x75, 0x6c, 0x6c, 0x18, 0x02, 0x20, 0x01, 0x28, 0x0b, 0x32, 0x21, 0x2e, 0x69, 0x6f, 0x2e,
-	0x72, 0x50, 0x65, 0x72, 0x69, 0x6f, 0x64, 0x73, 0x12, 0x21, 0x0a, 0x0c, 0x6e, 0x72, 0x5f, 0x74,
+	0x63, 0x6f, 0x6e, 0x74, 0x61, 0x69, 0x6e, 0x65, 0x72, 0x64, 0x2e, 0x63, 0x67, 0x72, 0x6f, 0x75,
-	0x68, 0x72, 0x6f, 0x74, 0x74, 0x6c, 0x65, 0x64, 0x18, 0x05, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0b,
+	0x70, 0x73, 0x2e, 0x76, 0x32, 0x2e, 0x50, 0x53, 0x49, 0x44, 0x61, 0x74, 0x61, 0x52, 0x04, 0x66,
-	0x6e, 0x72, 0x54, 0x68, 0x72, 0x6f, 0x74, 0x74, 0x6c, 0x65, 0x64, 0x12, 0x25, 0x0a, 0x0e, 0x74,
+	0x75, 0x6c, 0x6c, 0x22, 0x3a, 0x0a, 0x08, 0x50, 0x69, 0x64, 0x73, 0x53, 0x74, 0x61, 0x74, 0x12,
-	0x68, 0x72, 0x6f, 0x74, 0x74, 0x6c, 0x65, 0x64, 0x5f, 0x75, 0x73, 0x65, 0x63, 0x18, 0x06, 0x20,
+	0x18, 0x0a, 0x07, 0x63, 0x75, 0x72, 0x72, 0x65, 0x6e, 0x74, 0x18, 0x01, 0x20, 0x01, 0x28, 0x04,
-	0x01, 0x28, 0x04, 0x52, 0x0d, 0x74, 0x68, 0x72, 0x6f, 0x74, 0x74, 0x6c, 0x65, 0x64, 0x55, 0x73,
+	0x52, 0x07, 0x63, 0x75, 0x72, 0x72, 0x65, 0x6e, 0x74, 0x12, 0x14, 0x0a, 0x05, 0x6c, 0x69, 0x6d,
-	0x65, 0x63, 0x22, 0x8f, 0x09, 0x0a, 0x0a, 0x4d, 0x65, 0x6d, 0x6f, 0x72, 0x79, 0x53, 0x74, 0x61,
+	0x69, 0x74, 0x18, 0x02, 0x20, 0x01, 0x28, 0x04, 0x52, 0x05, 0x6c, 0x69, 0x6d, 0x69, 0x74, 0x22,
-	0x74, 0x12, 0x12, 0x0a, 0x04, 0x61, 0x6e, 0x6f, 0x6e, 0x18, 0x01, 0x20, 0x01, 0x28, 0x04, 0x52,
+	0x85, 0x02, 0x0a, 0x07, 0x43, 0x50, 0x55, 0x53, 0x74, 0x61, 0x74, 0x12, 0x1d, 0x0a, 0x0a, 0x75,
-	0x04, 0x61, 0x6e, 0x6f, 0x6e, 0x12, 0x12, 0x0a, 0x04, 0x66, 0x69, 0x6c, 0x65, 0x18, 0x02, 0x20,
+	0x73, 0x61, 0x67, 0x65, 0x5f, 0x75, 0x73, 0x65, 0x63, 0x18, 0x01, 0x20, 0x01, 0x28, 0x04, 0x52,
-	0x01, 0x28, 0x04, 0x52, 0x04, 0x66, 0x69, 0x6c, 0x65, 0x12, 0x21, 0x0a, 0x0c, 0x6b, 0x65, 0x72,
+	0x09, 0x75, 0x73, 0x61, 0x67, 0x65, 0x55, 0x73, 0x65, 0x63, 0x12, 0x1b, 0x0a, 0x09, 0x75, 0x73,
-	0x6e, 0x65, 0x6c, 0x5f, 0x73, 0x74, 0x61, 0x63, 0x6b, 0x18, 0x03, 0x20, 0x01, 0x28, 0x04, 0x52,
+	0x65, 0x72, 0x5f, 0x75, 0x73, 0x65, 0x63, 0x18, 0x02, 0x20, 0x01, 0x28, 0x04, 0x52, 0x08, 0x75,
-	0x0b, 0x6b, 0x65, 0x72, 0x6e, 0x65, 0x6c, 0x53, 0x74, 0x61, 0x63, 0x6b, 0x12, 0x12, 0x0a, 0x04,
+	0x73, 0x65, 0x72, 0x55, 0x73, 0x65, 0x63, 0x12, 0x1f, 0x0a, 0x0b, 0x73, 0x79, 0x73, 0x74, 0x65,
-	0x73, 0x6c, 0x61, 0x62, 0x18, 0x04, 0x20, 0x01, 0x28, 0x04, 0x52, 0x04, 0x73, 0x6c, 0x61, 0x62,
+	0x6d, 0x5f, 0x75, 0x73, 0x65, 0x63, 0x18, 0x03, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x73, 0x79,
-	0x12, 0x12, 0x0a, 0x04, 0x73, 0x6f, 0x63, 0x6b, 0x18, 0x05, 0x20, 0x01, 0x28, 0x04, 0x52, 0x04,
+	0x73, 0x74, 0x65, 0x6d, 0x55, 0x73, 0x65, 0x63, 0x12, 0x1d, 0x0a, 0x0a, 0x6e, 0x72, 0x5f, 0x70,
-	0x73, 0x6f, 0x63, 0x6b, 0x12, 0x14, 0x0a, 0x05, 0x73, 0x68, 0x6d, 0x65, 0x6d, 0x18, 0x06, 0x20,
+	0x65, 0x72, 0x69, 0x6f, 0x64, 0x73, 0x18, 0x04, 0x20, 0x01, 0x28, 0x04, 0x52, 0x09, 0x6e, 0x72,
-	0x01, 0x28, 0x04, 0x52, 0x05, 0x73, 0x68, 0x6d, 0x65, 0x6d, 0x12, 0x1f, 0x0a, 0x0b, 0x66, 0x69,
+	0x50, 0x65, 0x72, 0x69, 0x6f, 0x64, 0x73, 0x12, 0x21, 0x0a, 0x0c, 0x6e, 0x72, 0x5f, 0x74, 0x68,
-	0x6c, 0x65, 0x5f, 0x6d, 0x61, 0x70, 0x70, 0x65, 0x64, 0x18, 0x07, 0x20, 0x01, 0x28, 0x04, 0x52,
+	0x72, 0x6f, 0x74, 0x74, 0x6c, 0x65, 0x64, 0x18, 0x05, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0b, 0x6e,
-	0x0a, 0x66, 0x69, 0x6c, 0x65, 0x4d, 0x61, 0x70, 0x70, 0x65, 0x64, 0x12, 0x1d, 0x0a, 0x0a, 0x66,
+	0x72, 0x54, 0x68, 0x72, 0x6f, 0x74, 0x74, 0x6c, 0x65, 0x64, 0x12, 0x25, 0x0a, 0x0e, 0x74, 0x68,
-	0x69, 0x6c, 0x65, 0x5f, 0x64, 0x69, 0x72, 0x74, 0x79, 0x18, 0x08, 0x20, 0x01, 0x28, 0x04, 0x52,
+	0x72, 0x6f, 0x74, 0x74, 0x6c, 0x65, 0x64, 0x5f, 0x75, 0x73, 0x65, 0x63, 0x18, 0x06, 0x20, 0x01,
-	0x09, 0x66, 0x69, 0x6c, 0x65, 0x44, 0x69, 0x72, 0x74, 0x79, 0x12, 0x25, 0x0a, 0x0e, 0x66, 0x69,
+	0x28, 0x04, 0x52, 0x0d, 0x74, 0x68, 0x72, 0x6f, 0x74, 0x74, 0x6c, 0x65, 0x64, 0x55, 0x73, 0x65,
-	0x6c, 0x65, 0x5f, 0x77, 0x72, 0x69, 0x74, 0x65, 0x62, 0x61, 0x63, 0x6b, 0x18, 0x09, 0x20, 0x01,
+	0x63, 0x12, 0x34, 0x0a, 0x03, 0x70, 0x73, 0x69, 0x18, 0x07, 0x20, 0x01, 0x28, 0x0b, 0x32, 0x22,
-	0x28, 0x04, 0x52, 0x0d, 0x66, 0x69, 0x6c, 0x65, 0x57, 0x72, 0x69, 0x74, 0x65, 0x62, 0x61, 0x63,
+	0x2e, 0x69, 0x6f, 0x2e, 0x63, 0x6f, 0x6e, 0x74, 0x61, 0x69, 0x6e, 0x65, 0x72, 0x64, 0x2e, 0x63,
-	0x6b, 0x12, 0x19, 0x0a, 0x08, 0x61, 0x6e, 0x6f, 0x6e, 0x5f, 0x74, 0x68, 0x70, 0x18, 0x0a, 0x20,
+	0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2e, 0x76, 0x32, 0x2e, 0x50, 0x53, 0x49, 0x53, 0x74, 0x61,
-	0x01, 0x28, 0x04, 0x52, 0x07, 0x61, 0x6e, 0x6f, 0x6e, 0x54, 0x68, 0x70, 0x12, 0x23, 0x0a, 0x0d,
+	0x74, 0x73, 0x52, 0x03, 0x70, 0x73, 0x69, 0x22, 0x88, 0x0a, 0x0a, 0x0a, 0x4d, 0x65, 0x6d, 0x6f,
-	0x69, 0x6e, 0x61, 0x63, 0x74, 0x69, 0x76, 0x65, 0x5f, 0x61, 0x6e, 0x6f, 0x6e, 0x18, 0x0b, 0x20,
+	0x72, 0x79, 0x53, 0x74, 0x61, 0x74, 0x12, 0x12, 0x0a, 0x04, 0x61, 0x6e, 0x6f, 0x6e, 0x18, 0x01,
-	0x01, 0x28, 0x04, 0x52, 0x0c, 0x69, 0x6e, 0x61, 0x63, 0x74, 0x69, 0x76, 0x65, 0x41, 0x6e, 0x6f,
+	0x20, 0x01, 0x28, 0x04, 0x52, 0x04, 0x61, 0x6e, 0x6f, 0x6e, 0x12, 0x12, 0x0a, 0x04, 0x66, 0x69,
-	0x6e, 0x12, 0x1f, 0x0a, 0x0b, 0x61, 0x63, 0x74, 0x69, 0x76, 0x65, 0x5f, 0x61, 0x6e, 0x6f, 0x6e,
+	0x6c, 0x65, 0x18, 0x02, 0x20, 0x01, 0x28, 0x04, 0x52, 0x04, 0x66, 0x69, 0x6c, 0x65, 0x12, 0x21,
-	0x18, 0x0c, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x61, 0x63, 0x74, 0x69, 0x76, 0x65, 0x41, 0x6e,
+	0x0a, 0x0c, 0x6b, 0x65, 0x72, 0x6e, 0x65, 0x6c, 0x5f, 0x73, 0x74, 0x61, 0x63, 0x6b, 0x18, 0x03,
-	0x6f, 0x6e, 0x12, 0x23, 0x0a, 0x0d, 0x69, 0x6e, 0x61, 0x63, 0x74, 0x69, 0x76, 0x65, 0x5f, 0x66,
+	0x20, 0x01, 0x28, 0x04, 0x52, 0x0b, 0x6b, 0x65, 0x72, 0x6e, 0x65, 0x6c, 0x53, 0x74, 0x61, 0x63,
-	0x69, 0x6c, 0x65, 0x18, 0x0d, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0c, 0x69, 0x6e, 0x61, 0x63, 0x74,
+	0x6b, 0x12, 0x12, 0x0a, 0x04, 0x73, 0x6c, 0x61, 0x62, 0x18, 0x04, 0x20, 0x01, 0x28, 0x04, 0x52,
-	0x69, 0x76, 0x65, 0x46, 0x69, 0x6c, 0x65, 0x12, 0x1f, 0x0a, 0x0b, 0x61, 0x63, 0x74, 0x69, 0x76,
+	0x04, 0x73, 0x6c, 0x61, 0x62, 0x12, 0x12, 0x0a, 0x04, 0x73, 0x6f, 0x63, 0x6b, 0x18, 0x05, 0x20,
-	0x65, 0x5f, 0x66, 0x69, 0x6c, 0x65, 0x18, 0x0e, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x61, 0x63,
+	0x01, 0x28, 0x04, 0x52, 0x04, 0x73, 0x6f, 0x63, 0x6b, 0x12, 0x14, 0x0a, 0x05, 0x73, 0x68, 0x6d,
-	0x74, 0x69, 0x76, 0x65, 0x46, 0x69, 0x6c, 0x65, 0x12, 0x20, 0x0a, 0x0b, 0x75, 0x6e, 0x65, 0x76,
+	0x65, 0x6d, 0x18, 0x06, 0x20, 0x01, 0x28, 0x04, 0x52, 0x05, 0x73, 0x68, 0x6d, 0x65, 0x6d, 0x12,
-	0x69, 0x63, 0x74, 0x61, 0x62, 0x6c, 0x65, 0x18, 0x0f, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0b, 0x75,
+	0x1f, 0x0a, 0x0b, 0x66, 0x69, 0x6c, 0x65, 0x5f, 0x6d, 0x61, 0x70, 0x70, 0x65, 0x64, 0x18, 0x07,
-	0x6e, 0x65, 0x76, 0x69, 0x63, 0x74, 0x61, 0x62, 0x6c, 0x65, 0x12, 0x29, 0x0a, 0x10, 0x73, 0x6c,
+	0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x66, 0x69, 0x6c, 0x65, 0x4d, 0x61, 0x70, 0x70, 0x65, 0x64,
-	0x61, 0x62, 0x5f, 0x72, 0x65, 0x63, 0x6c, 0x61, 0x69, 0x6d, 0x61, 0x62, 0x6c, 0x65, 0x18, 0x10,
+	0x12, 0x1d, 0x0a, 0x0a, 0x66, 0x69, 0x6c, 0x65, 0x5f, 0x64, 0x69, 0x72, 0x74, 0x79, 0x18, 0x08,
-	0x20, 0x01, 0x28, 0x04, 0x52, 0x0f, 0x73, 0x6c, 0x61, 0x62, 0x52, 0x65, 0x63, 0x6c, 0x61, 0x69,
+	0x20, 0x01, 0x28, 0x04, 0x52, 0x09, 0x66, 0x69, 0x6c, 0x65, 0x44, 0x69, 0x72, 0x74, 0x79, 0x12,
-	0x6d, 0x61, 0x62, 0x6c, 0x65, 0x12, 0x2d, 0x0a, 0x12, 0x73, 0x6c, 0x61, 0x62, 0x5f, 0x75, 0x6e,
+	0x25, 0x0a, 0x0e, 0x66, 0x69, 0x6c, 0x65, 0x5f, 0x77, 0x72, 0x69, 0x74, 0x65, 0x62, 0x61, 0x63,
-	0x72, 0x65, 0x63, 0x6c, 0x61, 0x69, 0x6d, 0x61, 0x62, 0x6c, 0x65, 0x18, 0x11, 0x20, 0x01, 0x28,
+	0x6b, 0x18, 0x09, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0d, 0x66, 0x69, 0x6c, 0x65, 0x57, 0x72, 0x69,
-	0x04, 0x52, 0x11, 0x73, 0x6c, 0x61, 0x62, 0x55, 0x6e, 0x72, 0x65, 0x63, 0x6c, 0x61, 0x69, 0x6d,
+	0x74, 0x65, 0x62, 0x61, 0x63, 0x6b, 0x12, 0x19, 0x0a, 0x08, 0x61, 0x6e, 0x6f, 0x6e, 0x5f, 0x74,
-	0x61, 0x62, 0x6c, 0x65, 0x12, 0x18, 0x0a, 0x07, 0x70, 0x67, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x18,
+	0x68, 0x70, 0x18, 0x0a, 0x20, 0x01, 0x28, 0x04, 0x52, 0x07, 0x61, 0x6e, 0x6f, 0x6e, 0x54, 0x68,
-	0x12, 0x20, 0x01, 0x28, 0x04, 0x52, 0x07, 0x70, 0x67, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x12, 0x1e,
+	0x70, 0x12, 0x23, 0x0a, 0x0d, 0x69, 0x6e, 0x61, 0x63, 0x74, 0x69, 0x76, 0x65, 0x5f, 0x61, 0x6e,
-	0x0a, 0x0a, 0x70, 0x67, 0x6d, 0x61, 0x6a, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x18, 0x13, 0x20, 0x01,
+	0x6f, 0x6e, 0x18, 0x0b, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0c, 0x69, 0x6e, 0x61, 0x63, 0x74, 0x69,
-	0x28, 0x04, 0x52, 0x0a, 0x70, 0x67, 0x6d, 0x61, 0x6a, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x12, 0x2d,
+	0x76, 0x65, 0x41, 0x6e, 0x6f, 0x6e, 0x12, 0x1f, 0x0a, 0x0b, 0x61, 0x63, 0x74, 0x69, 0x76, 0x65,
-	0x0a, 0x12, 0x77, 0x6f, 0x72, 0x6b, 0x69, 0x6e, 0x67, 0x73, 0x65, 0x74, 0x5f, 0x72, 0x65, 0x66,
+	0x5f, 0x61, 0x6e, 0x6f, 0x6e, 0x18, 0x0c, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x61, 0x63, 0x74,
-	0x61, 0x75, 0x6c, 0x74, 0x18, 0x14, 0x20, 0x01, 0x28, 0x04, 0x52, 0x11, 0x77, 0x6f, 0x72, 0x6b,
+	0x69, 0x76, 0x65, 0x41, 0x6e, 0x6f, 0x6e, 0x12, 0x23, 0x0a, 0x0d, 0x69, 0x6e, 0x61, 0x63, 0x74,
-	0x69, 0x6e, 0x67, 0x73, 0x65, 0x74, 0x52, 0x65, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x12, 0x2f, 0x0a,
+	0x69, 0x76, 0x65, 0x5f, 0x66, 0x69, 0x6c, 0x65, 0x18, 0x0d, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0c,
-	0x13, 0x77, 0x6f, 0x72, 0x6b, 0x69, 0x6e, 0x67, 0x73, 0x65, 0x74, 0x5f, 0x61, 0x63, 0x74, 0x69,
+	0x69, 0x6e, 0x61, 0x63, 0x74, 0x69, 0x76, 0x65, 0x46, 0x69, 0x6c, 0x65, 0x12, 0x1f, 0x0a, 0x0b,
-	0x76, 0x61, 0x74, 0x65, 0x18, 0x15, 0x20, 0x01, 0x28, 0x04, 0x52, 0x12, 0x77, 0x6f, 0x72, 0x6b,
+	0x61, 0x63, 0x74, 0x69, 0x76, 0x65, 0x5f, 0x66, 0x69, 0x6c, 0x65, 0x18, 0x0e, 0x20, 0x01, 0x28,
-	0x69, 0x6e, 0x67, 0x73, 0x65, 0x74, 0x41, 0x63, 0x74, 0x69, 0x76, 0x61, 0x74, 0x65, 0x12, 0x35,
+	0x04, 0x52, 0x0a, 0x61, 0x63, 0x74, 0x69, 0x76, 0x65, 0x46, 0x69, 0x6c, 0x65, 0x12, 0x20, 0x0a,
-	0x0a, 0x16, 0x77, 0x6f, 0x72, 0x6b, 0x69, 0x6e, 0x67, 0x73, 0x65, 0x74, 0x5f, 0x6e, 0x6f, 0x64,
+	0x0b, 0x75, 0x6e, 0x65, 0x76, 0x69, 0x63, 0x74, 0x61, 0x62, 0x6c, 0x65, 0x18, 0x0f, 0x20, 0x01,
-	0x65, 0x72, 0x65, 0x63, 0x6c, 0x61, 0x69, 0x6d, 0x18, 0x16, 0x20, 0x01, 0x28, 0x04, 0x52, 0x15,
+	0x28, 0x04, 0x52, 0x0b, 0x75, 0x6e, 0x65, 0x76, 0x69, 0x63, 0x74, 0x61, 0x62, 0x6c, 0x65, 0x12,
-	0x77, 0x6f, 0x72, 0x6b, 0x69, 0x6e, 0x67, 0x73, 0x65, 0x74, 0x4e, 0x6f, 0x64, 0x65, 0x72, 0x65,
+	0x29, 0x0a, 0x10, 0x73, 0x6c, 0x61, 0x62, 0x5f, 0x72, 0x65, 0x63, 0x6c, 0x61, 0x69, 0x6d, 0x61,
-	0x63, 0x6c, 0x61, 0x69, 0x6d, 0x12, 0x1a, 0x0a, 0x08, 0x70, 0x67, 0x72, 0x65, 0x66, 0x69, 0x6c,
+	0x62, 0x6c, 0x65, 0x18, 0x10, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0f, 0x73, 0x6c, 0x61, 0x62, 0x52,
-	0x6c, 0x18, 0x17, 0x20, 0x01, 0x28, 0x04, 0x52, 0x08, 0x70, 0x67, 0x72, 0x65, 0x66, 0x69, 0x6c,
+	0x65, 0x63, 0x6c, 0x61, 0x69, 0x6d, 0x61, 0x62, 0x6c, 0x65, 0x12, 0x2d, 0x0a, 0x12, 0x73, 0x6c,
-	0x6c, 0x12, 0x16, 0x0a, 0x06, 0x70, 0x67, 0x73, 0x63, 0x61, 0x6e, 0x18, 0x18, 0x20, 0x01, 0x28,
+	0x61, 0x62, 0x5f, 0x75, 0x6e, 0x72, 0x65, 0x63, 0x6c, 0x61, 0x69, 0x6d, 0x61, 0x62, 0x6c, 0x65,
-	0x04, 0x52, 0x06, 0x70, 0x67, 0x73, 0x63, 0x61, 0x6e, 0x12, 0x18, 0x0a, 0x07, 0x70, 0x67, 0x73,
+	0x18, 0x11, 0x20, 0x01, 0x28, 0x04, 0x52, 0x11, 0x73, 0x6c, 0x61, 0x62, 0x55, 0x6e, 0x72, 0x65,
-	0x74, 0x65, 0x61, 0x6c, 0x18, 0x19, 0x20, 0x01, 0x28, 0x04, 0x52, 0x07, 0x70, 0x67, 0x73, 0x74,
+	0x63, 0x6c, 0x61, 0x69, 0x6d, 0x61, 0x62, 0x6c, 0x65, 0x12, 0x18, 0x0a, 0x07, 0x70, 0x67, 0x66,
-	0x65, 0x61, 0x6c, 0x12, 0x1e, 0x0a, 0x0a, 0x70, 0x67, 0x61, 0x63, 0x74, 0x69, 0x76, 0x61, 0x74,
+	0x61, 0x75, 0x6c, 0x74, 0x18, 0x12, 0x20, 0x01, 0x28, 0x04, 0x52, 0x07, 0x70, 0x67, 0x66, 0x61,
-	0x65, 0x18, 0x1a, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x70, 0x67, 0x61, 0x63, 0x74, 0x69, 0x76,
+	0x75, 0x6c, 0x74, 0x12, 0x1e, 0x0a, 0x0a, 0x70, 0x67, 0x6d, 0x61, 0x6a, 0x66, 0x61, 0x75, 0x6c,
-	0x61, 0x74, 0x65, 0x12, 0x22, 0x0a, 0x0c, 0x70, 0x67, 0x64, 0x65, 0x61, 0x63, 0x74, 0x69, 0x76,
+	0x74, 0x18, 0x13, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x70, 0x67, 0x6d, 0x61, 0x6a, 0x66, 0x61,
-	0x61, 0x74, 0x65, 0x18, 0x1b, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0c, 0x70, 0x67, 0x64, 0x65, 0x61,
+	0x75, 0x6c, 0x74, 0x12, 0x2d, 0x0a, 0x12, 0x77, 0x6f, 0x72, 0x6b, 0x69, 0x6e, 0x67, 0x73, 0x65,
-	0x63, 0x74, 0x69, 0x76, 0x61, 0x74, 0x65, 0x12, 0x1e, 0x0a, 0x0a, 0x70, 0x67, 0x6c, 0x61, 0x7a,
+	0x74, 0x5f, 0x72, 0x65, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x18, 0x14, 0x20, 0x01, 0x28, 0x04, 0x52,
-	0x79, 0x66, 0x72, 0x65, 0x65, 0x18, 0x1c, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x70, 0x67, 0x6c,
+	0x11, 0x77, 0x6f, 0x72, 0x6b, 0x69, 0x6e, 0x67, 0x73, 0x65, 0x74, 0x52, 0x65, 0x66, 0x61, 0x75,
-	0x61, 0x7a, 0x79, 0x66, 0x72, 0x65, 0x65, 0x12, 0x20, 0x0a, 0x0b, 0x70, 0x67, 0x6c, 0x61, 0x7a,
+	0x6c, 0x74, 0x12, 0x2f, 0x0a, 0x13, 0x77, 0x6f, 0x72, 0x6b, 0x69, 0x6e, 0x67, 0x73, 0x65, 0x74,
-	0x79, 0x66, 0x72, 0x65, 0x65, 0x64, 0x18, 0x1d, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0b, 0x70, 0x67,
+	0x5f, 0x61, 0x63, 0x74, 0x69, 0x76, 0x61, 0x74, 0x65, 0x18, 0x15, 0x20, 0x01, 0x28, 0x04, 0x52,
-	0x6c, 0x61, 0x7a, 0x79, 0x66, 0x72, 0x65, 0x65, 0x64, 0x12, 0x26, 0x0a, 0x0f, 0x74, 0x68, 0x70,
+	0x12, 0x77, 0x6f, 0x72, 0x6b, 0x69, 0x6e, 0x67, 0x73, 0x65, 0x74, 0x41, 0x63, 0x74, 0x69, 0x76,
-	0x5f, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x5f, 0x61, 0x6c, 0x6c, 0x6f, 0x63, 0x18, 0x1e, 0x20, 0x01,
+	0x61, 0x74, 0x65, 0x12, 0x35, 0x0a, 0x16, 0x77, 0x6f, 0x72, 0x6b, 0x69, 0x6e, 0x67, 0x73, 0x65,
-	0x28, 0x04, 0x52, 0x0d, 0x74, 0x68, 0x70, 0x46, 0x61, 0x75, 0x6c, 0x74, 0x41, 0x6c, 0x6c, 0x6f,
+	0x74, 0x5f, 0x6e, 0x6f, 0x64, 0x65, 0x72, 0x65, 0x63, 0x6c, 0x61, 0x69, 0x6d, 0x18, 0x16, 0x20,
-	0x63, 0x12, 0x2c, 0x0a, 0x12, 0x74, 0x68, 0x70, 0x5f, 0x63, 0x6f, 0x6c, 0x6c, 0x61, 0x70, 0x73,
+	0x01, 0x28, 0x04, 0x52, 0x15, 0x77, 0x6f, 0x72, 0x6b, 0x69, 0x6e, 0x67, 0x73, 0x65, 0x74, 0x4e,
-	0x65, 0x5f, 0x61, 0x6c, 0x6c, 0x6f, 0x63, 0x18, 0x1f, 0x20, 0x01, 0x28, 0x04, 0x52, 0x10, 0x74,
+	0x6f, 0x64, 0x65, 0x72, 0x65, 0x63, 0x6c, 0x61, 0x69, 0x6d, 0x12, 0x1a, 0x0a, 0x08, 0x70, 0x67,
-	0x68, 0x70, 0x43, 0x6f, 0x6c, 0x6c, 0x61, 0x70, 0x73, 0x65, 0x41, 0x6c, 0x6c, 0x6f, 0x63, 0x12,
+	0x72, 0x65, 0x66, 0x69, 0x6c, 0x6c, 0x18, 0x17, 0x20, 0x01, 0x28, 0x04, 0x52, 0x08, 0x70, 0x67,
-	0x14, 0x0a, 0x05, 0x75, 0x73, 0x61, 0x67, 0x65, 0x18, 0x20, 0x20, 0x01, 0x28, 0x04, 0x52, 0x05,
+	0x72, 0x65, 0x66, 0x69, 0x6c, 0x6c, 0x12, 0x16, 0x0a, 0x06, 0x70, 0x67, 0x73, 0x63, 0x61, 0x6e,
-	0x75, 0x73, 0x61, 0x67, 0x65, 0x12, 0x1f, 0x0a, 0x0b, 0x75, 0x73, 0x61, 0x67, 0x65, 0x5f, 0x6c,
+	0x18, 0x18, 0x20, 0x01, 0x28, 0x04, 0x52, 0x06, 0x70, 0x67, 0x73, 0x63, 0x61, 0x6e, 0x12, 0x18,
-	0x69, 0x6d, 0x69, 0x74, 0x18, 0x21, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x75, 0x73, 0x61, 0x67,
+	0x0a, 0x07, 0x70, 0x67, 0x73, 0x74, 0x65, 0x61, 0x6c, 0x18, 0x19, 0x20, 0x01, 0x28, 0x04, 0x52,
-	0x65, 0x4c, 0x69, 0x6d, 0x69, 0x74, 0x12, 0x1d, 0x0a, 0x0a, 0x73, 0x77, 0x61, 0x70, 0x5f, 0x75,
+	0x07, 0x70, 0x67, 0x73, 0x74, 0x65, 0x61, 0x6c, 0x12, 0x1e, 0x0a, 0x0a, 0x70, 0x67, 0x61, 0x63,
-	0x73, 0x61, 0x67, 0x65, 0x18, 0x22, 0x20, 0x01, 0x28, 0x04, 0x52, 0x09, 0x73, 0x77, 0x61, 0x70,
+	0x74, 0x69, 0x76, 0x61, 0x74, 0x65, 0x18, 0x1a, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0a, 0x70, 0x67,
-	0x55, 0x73, 0x61, 0x67, 0x65, 0x12, 0x1d, 0x0a, 0x0a, 0x73, 0x77, 0x61, 0x70, 0x5f, 0x6c, 0x69,
+	0x61, 0x63, 0x74, 0x69, 0x76, 0x61, 0x74, 0x65, 0x12, 0x22, 0x0a, 0x0c, 0x70, 0x67, 0x64, 0x65,
-	0x6d, 0x69, 0x74, 0x18, 0x23, 0x20, 0x01, 0x28, 0x04, 0x52, 0x09, 0x73, 0x77, 0x61, 0x70, 0x4c,
+	0x61, 0x63, 0x74, 0x69, 0x76, 0x61, 0x74, 0x65, 0x18, 0x1b, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0c,
-	0x69, 0x6d, 0x69, 0x74, 0x22, 0x73, 0x0a, 0x0c, 0x4d, 0x65, 0x6d, 0x6f, 0x72, 0x79, 0x45, 0x76,
+	0x70, 0x67, 0x64, 0x65, 0x61, 0x63, 0x74, 0x69, 0x76, 0x61, 0x74, 0x65, 0x12, 0x1e, 0x0a, 0x0a,
-	0x65, 0x6e, 0x74, 0x73, 0x12, 0x10, 0x0a, 0x03, 0x6c, 0x6f, 0x77, 0x18, 0x01, 0x20, 0x01, 0x28,
+	0x70, 0x67, 0x6c, 0x61, 0x7a, 0x79, 0x66, 0x72, 0x65, 0x65, 0x18, 0x1c, 0x20, 0x01, 0x28, 0x04,
-	0x04, 0x52, 0x03, 0x6c, 0x6f, 0x77, 0x12, 0x12, 0x0a, 0x04, 0x68, 0x69, 0x67, 0x68, 0x18, 0x02,
+	0x52, 0x0a, 0x70, 0x67, 0x6c, 0x61, 0x7a, 0x79, 0x66, 0x72, 0x65, 0x65, 0x12, 0x20, 0x0a, 0x0b,
-	0x20, 0x01, 0x28, 0x04, 0x52, 0x04, 0x68, 0x69, 0x67, 0x68, 0x12, 0x10, 0x0a, 0x03, 0x6d, 0x61,
+	0x70, 0x67, 0x6c, 0x61, 0x7a, 0x79, 0x66, 0x72, 0x65, 0x65, 0x64, 0x18, 0x1d, 0x20, 0x01, 0x28,
-	0x78, 0x18, 0x03, 0x20, 0x01, 0x28, 0x04, 0x52, 0x03, 0x6d, 0x61, 0x78, 0x12, 0x10, 0x0a, 0x03,
+	0x04, 0x52, 0x0b, 0x70, 0x67, 0x6c, 0x61, 0x7a, 0x79, 0x66, 0x72, 0x65, 0x65, 0x64, 0x12, 0x26,
-	0x6f, 0x6f, 0x6d, 0x18, 0x04, 0x20, 0x01, 0x28, 0x04, 0x52, 0x03, 0x6f, 0x6f, 0x6d, 0x12, 0x19,
+	0x0a, 0x0f, 0x74, 0x68, 0x70, 0x5f, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x5f, 0x61, 0x6c, 0x6c, 0x6f,
-	0x0a, 0x08, 0x6f, 0x6f, 0x6d, 0x5f, 0x6b, 0x69, 0x6c, 0x6c, 0x18, 0x05, 0x20, 0x01, 0x28, 0x04,
+	0x63, 0x18, 0x1e, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0d, 0x74, 0x68, 0x70, 0x46, 0x61, 0x75, 0x6c,
-	0x52, 0x07, 0x6f, 0x6f, 0x6d, 0x4b, 0x69, 0x6c, 0x6c, 0x22, 0x84, 0x01, 0x0a, 0x08, 0x52, 0x64,
+	0x74, 0x41, 0x6c, 0x6c, 0x6f, 0x63, 0x12, 0x2c, 0x0a, 0x12, 0x74, 0x68, 0x70, 0x5f, 0x63, 0x6f,
-	0x6d, 0x61, 0x53, 0x74, 0x61, 0x74, 0x12, 0x3d, 0x0a, 0x07, 0x63, 0x75, 0x72, 0x72, 0x65, 0x6e,
+	0x6c, 0x6c, 0x61, 0x70, 0x73, 0x65, 0x5f, 0x61, 0x6c, 0x6c, 0x6f, 0x63, 0x18, 0x1f, 0x20, 0x01,
-	0x74, 0x18, 0x01, 0x20, 0x03, 0x28, 0x0b, 0x32, 0x23, 0x2e, 0x69, 0x6f, 0x2e, 0x63, 0x6f, 0x6e,
+	0x28, 0x04, 0x52, 0x10, 0x74, 0x68, 0x70, 0x43, 0x6f, 0x6c, 0x6c, 0x61, 0x70, 0x73, 0x65, 0x41,
-	0x74, 0x61, 0x69, 0x6e, 0x65, 0x72, 0x64, 0x2e, 0x63, 0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2e,
+	0x6c, 0x6c, 0x6f, 0x63, 0x12, 0x14, 0x0a, 0x05, 0x75, 0x73, 0x61, 0x67, 0x65, 0x18, 0x20, 0x20,
-	0x76, 0x32, 0x2e, 0x52, 0x64, 0x6d, 0x61, 0x45, 0x6e, 0x74, 0x72, 0x79, 0x52, 0x07, 0x63, 0x75,
+	0x01, 0x28, 0x04, 0x52, 0x05, 0x75, 0x73, 0x61, 0x67, 0x65, 0x12, 0x1f, 0x0a, 0x0b, 0x75, 0x73,
-	0x72, 0x72, 0x65, 0x6e, 0x74, 0x12, 0x39, 0x0a, 0x05, 0x6c, 0x69, 0x6d, 0x69, 0x74, 0x18, 0x02,
+	0x61, 0x67, 0x65, 0x5f, 0x6c, 0x69, 0x6d, 0x69, 0x74, 0x18, 0x21, 0x20, 0x01, 0x28, 0x04, 0x52,
-	0x20, 0x03, 0x28, 0x0b, 0x32, 0x23, 0x2e, 0x69, 0x6f, 0x2e, 0x63, 0x6f, 0x6e, 0x74, 0x61, 0x69,
+	0x0a, 0x75, 0x73, 0x61, 0x67, 0x65, 0x4c, 0x69, 0x6d, 0x69, 0x74, 0x12, 0x1d, 0x0a, 0x0a, 0x73,
-	0x6e, 0x65, 0x72, 0x64, 0x2e, 0x63, 0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2e, 0x76, 0x32, 0x2e,
+	0x77, 0x61, 0x70, 0x5f, 0x75, 0x73, 0x61, 0x67, 0x65, 0x18, 0x22, 0x20, 0x01, 0x28, 0x04, 0x52,
-	0x52, 0x64, 0x6d, 0x61, 0x45, 0x6e, 0x74, 0x72, 0x79, 0x52, 0x05, 0x6c, 0x69, 0x6d, 0x69, 0x74,
+	0x09, 0x73, 0x77, 0x61, 0x70, 0x55, 0x73, 0x61, 0x67, 0x65, 0x12, 0x1d, 0x0a, 0x0a, 0x73, 0x77,
-	0x22, 0x65, 0x0a, 0x09, 0x52, 0x64, 0x6d, 0x61, 0x45, 0x6e, 0x74, 0x72, 0x79, 0x12, 0x16, 0x0a,
+	0x61, 0x70, 0x5f, 0x6c, 0x69, 0x6d, 0x69, 0x74, 0x18, 0x23, 0x20, 0x01, 0x28, 0x04, 0x52, 0x09,
-	0x06, 0x64, 0x65, 0x76, 0x69, 0x63, 0x65, 0x18, 0x01, 0x20, 0x01, 0x28, 0x09, 0x52, 0x06, 0x64,
+	0x73, 0x77, 0x61, 0x70, 0x4c, 0x69, 0x6d, 0x69, 0x74, 0x12, 0x1b, 0x0a, 0x09, 0x6d, 0x61, 0x78,
-	0x65, 0x76, 0x69, 0x63, 0x65, 0x12, 0x1f, 0x0a, 0x0b, 0x68, 0x63, 0x61, 0x5f, 0x68, 0x61, 0x6e,
+	0x5f, 0x75, 0x73, 0x61, 0x67, 0x65, 0x18, 0x24, 0x20, 0x01, 0x28, 0x04, 0x52, 0x08, 0x6d, 0x61,
-	0x64, 0x6c, 0x65, 0x73, 0x18, 0x02, 0x20, 0x01, 0x28, 0x0d, 0x52, 0x0a, 0x68, 0x63, 0x61, 0x48,
+	0x78, 0x55, 0x73, 0x61, 0x67, 0x65, 0x12, 0x24, 0x0a, 0x0e, 0x73, 0x77, 0x61, 0x70, 0x5f, 0x6d,
-	0x61, 0x6e, 0x64, 0x6c, 0x65, 0x73, 0x12, 0x1f, 0x0a, 0x0b, 0x68, 0x63, 0x61, 0x5f, 0x6f, 0x62,
+	0x61, 0x78, 0x5f, 0x75, 0x73, 0x61, 0x67, 0x65, 0x18, 0x25, 0x20, 0x01, 0x28, 0x04, 0x52, 0x0c,
-	0x6a, 0x65, 0x63, 0x74, 0x73, 0x18, 0x03, 0x20, 0x01, 0x28, 0x0d, 0x52, 0x0a, 0x68, 0x63, 0x61,
+	0x73, 0x77, 0x61, 0x70, 0x4d, 0x61, 0x78, 0x55, 0x73, 0x61, 0x67, 0x65, 0x12, 0x34, 0x0a, 0x03,
-	0x4f, 0x62, 0x6a, 0x65, 0x63, 0x74, 0x73, 0x22, 0x41, 0x0a, 0x06, 0x49, 0x4f, 0x53, 0x74, 0x61,
+	0x70, 0x73, 0x69, 0x18, 0x26, 0x20, 0x01, 0x28, 0x0b, 0x32, 0x22, 0x2e, 0x69, 0x6f, 0x2e, 0x63,
-	0x74, 0x12, 0x37, 0x0a, 0x05, 0x75, 0x73, 0x61, 0x67, 0x65, 0x18, 0x01, 0x20, 0x03, 0x28, 0x0b,
+	0x6f, 0x6e, 0x74, 0x61, 0x69, 0x6e, 0x65, 0x72, 0x64, 0x2e, 0x63, 0x67, 0x72, 0x6f, 0x75, 0x70,
-	0x32, 0x21, 0x2e, 0x69, 0x6f, 0x2e, 0x63, 0x6f, 0x6e, 0x74, 0x61, 0x69, 0x6e, 0x65, 0x72, 0x64,
+	0x73, 0x2e, 0x76, 0x32, 0x2e, 0x50, 0x53, 0x49, 0x53, 0x74, 0x61, 0x74, 0x73, 0x52, 0x03, 0x70,
-	0x2e, 0x63, 0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2e, 0x76, 0x32, 0x2e, 0x49, 0x4f, 0x45, 0x6e,
+	0x73, 0x69, 0x22, 0x73, 0x0a, 0x0c, 0x4d, 0x65, 0x6d, 0x6f, 0x72, 0x79, 0x45, 0x76, 0x65, 0x6e,
-	0x74, 0x72, 0x79, 0x52, 0x05, 0x75, 0x73, 0x61, 0x67, 0x65, 0x22, 0x8d, 0x01, 0x0a, 0x07, 0x49,
+	0x74, 0x73, 0x12, 0x10, 0x0a, 0x03, 0x6c, 0x6f, 0x77, 0x18, 0x01, 0x20, 0x01, 0x28, 0x04, 0x52,
-	0x4f, 0x45, 0x6e, 0x74, 0x72, 0x79, 0x12, 0x14, 0x0a, 0x05, 0x6d, 0x61, 0x6a, 0x6f, 0x72, 0x18,
+	0x03, 0x6c, 0x6f, 0x77, 0x12, 0x12, 0x0a, 0x04, 0x68, 0x69, 0x67, 0x68, 0x18, 0x02, 0x20, 0x01,
-	0x01, 0x20, 0x01, 0x28, 0x04, 0x52, 0x05, 0x6d, 0x61, 0x6a, 0x6f, 0x72, 0x12, 0x14, 0x0a, 0x05,
+	0x28, 0x04, 0x52, 0x04, 0x68, 0x69, 0x67, 0x68, 0x12, 0x10, 0x0a, 0x03, 0x6d, 0x61, 0x78, 0x18,
-	0x6d, 0x69, 0x6e, 0x6f, 0x72, 0x18, 0x02, 0x20, 0x01, 0x28, 0x04, 0x52, 0x05, 0x6d, 0x69, 0x6e,
+	0x03, 0x20, 0x01, 0x28, 0x04, 0x52, 0x03, 0x6d, 0x61, 0x78, 0x12, 0x10, 0x0a, 0x03, 0x6f, 0x6f,
-	0x6f, 0x72, 0x12, 0x16, 0x0a, 0x06, 0x72, 0x62, 0x79, 0x74, 0x65, 0x73, 0x18, 0x03, 0x20, 0x01,
+	0x6d, 0x18, 0x04, 0x20, 0x01, 0x28, 0x04, 0x52, 0x03, 0x6f, 0x6f, 0x6d, 0x12, 0x19, 0x0a, 0x08,
-	0x28, 0x04, 0x52, 0x06, 0x72, 0x62, 0x79, 0x74, 0x65, 0x73, 0x12, 0x16, 0x0a, 0x06, 0x77, 0x62,
+	0x6f, 0x6f, 0x6d, 0x5f, 0x6b, 0x69, 0x6c, 0x6c, 0x18, 0x05, 0x20, 0x01, 0x28, 0x04, 0x52, 0x07,
-	0x79, 0x74, 0x65, 0x73, 0x18, 0x04, 0x20, 0x01, 0x28, 0x04, 0x52, 0x06, 0x77, 0x62, 0x79, 0x74,
+	0x6f, 0x6f, 0x6d, 0x4b, 0x69, 0x6c, 0x6c, 0x22, 0x84, 0x01, 0x0a, 0x08, 0x52, 0x64, 0x6d, 0x61,
-	0x65, 0x73, 0x12, 0x12, 0x0a, 0x04, 0x72, 0x69, 0x6f, 0x73, 0x18, 0x05, 0x20, 0x01, 0x28, 0x04,
+	0x53, 0x74, 0x61, 0x74, 0x12, 0x3d, 0x0a, 0x07, 0x63, 0x75, 0x72, 0x72, 0x65, 0x6e, 0x74, 0x18,
-	0x52, 0x04, 0x72, 0x69, 0x6f, 0x73, 0x12, 0x12, 0x0a, 0x04, 0x77, 0x69, 0x6f, 0x73, 0x18, 0x06,
+	0x01, 0x20, 0x03, 0x28, 0x0b, 0x32, 0x23, 0x2e, 0x69, 0x6f, 0x2e, 0x63, 0x6f, 0x6e, 0x74, 0x61,
-	0x20, 0x01, 0x28, 0x04, 0x52, 0x04, 0x77, 0x69, 0x6f, 0x73, 0x22, 0x55, 0x0a, 0x0b, 0x48, 0x75,
+	0x69, 0x6e, 0x65, 0x72, 0x64, 0x2e, 0x63, 0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2e, 0x76, 0x32,
-	0x67, 0x65, 0x54, 0x6c, 0x62, 0x53, 0x74, 0x61, 0x74, 0x12, 0x18, 0x0a, 0x07, 0x63, 0x75, 0x72,
+	0x2e, 0x52, 0x64, 0x6d, 0x61, 0x45, 0x6e, 0x74, 0x72, 0x79, 0x52, 0x07, 0x63, 0x75, 0x72, 0x72,
-	0x72, 0x65, 0x6e, 0x74, 0x18, 0x01, 0x20, 0x01, 0x28, 0x04, 0x52, 0x07, 0x63, 0x75, 0x72, 0x72,
+	0x65, 0x6e, 0x74, 0x12, 0x39, 0x0a, 0x05, 0x6c, 0x69, 0x6d, 0x69, 0x74, 0x18, 0x02, 0x20, 0x03,
-	0x65, 0x6e, 0x74, 0x12, 0x10, 0x0a, 0x03, 0x6d, 0x61, 0x78, 0x18, 0x02, 0x20, 0x01, 0x28, 0x04,
+	0x28, 0x0b, 0x32, 0x23, 0x2e, 0x69, 0x6f, 0x2e, 0x63, 0x6f, 0x6e, 0x74, 0x61, 0x69, 0x6e, 0x65,
-	0x52, 0x03, 0x6d, 0x61, 0x78, 0x12, 0x1a, 0x0a, 0x08, 0x70, 0x61, 0x67, 0x65, 0x73, 0x69, 0x7a,
+	0x72, 0x64, 0x2e, 0x63, 0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2e, 0x76, 0x32, 0x2e, 0x52, 0x64,
-	0x65, 0x18, 0x03, 0x20, 0x01, 0x28, 0x09, 0x52, 0x08, 0x70, 0x61, 0x67, 0x65, 0x73, 0x69, 0x7a,
+	0x6d, 0x61, 0x45, 0x6e, 0x74, 0x72, 0x79, 0x52, 0x05, 0x6c, 0x69, 0x6d, 0x69, 0x74, 0x22, 0x65,
-	0x65, 0x42, 0x2d, 0x5a, 0x2b, 0x67, 0x69, 0x74, 0x68, 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, 0x2f,
+	0x0a, 0x09, 0x52, 0x64, 0x6d, 0x61, 0x45, 0x6e, 0x74, 0x72, 0x79, 0x12, 0x16, 0x0a, 0x06, 0x64,
-	0x63, 0x6f, 0x6e, 0x74, 0x61, 0x69, 0x6e, 0x65, 0x72, 0x64, 0x2f, 0x63, 0x67, 0x72, 0x6f, 0x75,
+	0x65, 0x76, 0x69, 0x63, 0x65, 0x18, 0x01, 0x20, 0x01, 0x28, 0x09, 0x52, 0x06, 0x64, 0x65, 0x76,
-	0x70, 0x73, 0x2f, 0x63, 0x67, 0x72, 0x6f, 0x75, 0x70, 0x32, 0x2f, 0x73, 0x74, 0x61, 0x74, 0x73,
+	0x69, 0x63, 0x65, 0x12, 0x1f, 0x0a, 0x0b, 0x68, 0x63, 0x61, 0x5f, 0x68, 0x61, 0x6e, 0x64, 0x6c,
-	0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33,
+	0x65, 0x73, 0x18, 0x02, 0x20, 0x01, 0x28, 0x0d, 0x52, 0x0a, 0x68, 0x63, 0x61, 0x48, 0x61, 0x6e,
+	0x64, 0x6c, 0x65, 0x73, 0x12, 0x1f, 0x0a, 0x0b, 0x68, 0x63, 0x61, 0x5f, 0x6f, 0x62, 0x6a, 0x65,
+	0x63, 0x74, 0x73, 0x18, 0x03, 0x20, 0x01, 0x28, 0x0d, 0x52, 0x0a, 0x68, 0x63, 0x61, 0x4f, 0x62,
+	0x6a, 0x65, 0x63, 0x74, 0x73, 0x22, 0x77, 0x0a, 0x06, 0x49, 0x4f, 0x53, 0x74, 0x61, 0x74, 0x12,
+	0x37, 0x0a, 0x05, 0x75, 0x73, 0x61, 0x67, 0x65, 0x18, 0x01, 0x20, 0x03, 0x28, 0x0b, 0x32, 0x21,
+	0x2e, 0x69, 0x6f, 0x2e, 0x63, 0x6f, 0x6e, 0x74, 0x61, 0x69, 0x6e, 0x65, 0x72, 0x64, 0x2e, 0x63,
+	0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2e, 0x76, 0x32, 0x2e, 0x49, 0x4f, 0x45, 0x6e, 0x74, 0x72,
+	0x79, 0x52, 0x05, 0x75, 0x73, 0x61, 0x67, 0x65, 0x12, 0x34, 0x0a, 0x03, 0x70, 0x73, 0x69, 0x18,
+	0x02, 0x20, 0x01, 0x28, 0x0b, 0x32, 0x22, 0x2e, 0x69, 0x6f, 0x2e, 0x63, 0x6f, 0x6e, 0x74, 0x61,
+	0x69, 0x6e, 0x65, 0x72, 0x64, 0x2e, 0x63, 0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2e, 0x76, 0x32,
+	0x2e, 0x50, 0x53, 0x49, 0x53, 0x74, 0x61, 0x74, 0x73, 0x52, 0x03, 0x70, 0x73, 0x69, 0x22, 0x8d,
+	0x01, 0x0a, 0x07, 0x49, 0x4f, 0x45, 0x6e, 0x74, 0x72, 0x79, 0x12, 0x14, 0x0a, 0x05, 0x6d, 0x61,
+	0x6a, 0x6f, 0x72, 0x18, 0x01, 0x20, 0x01, 0x28, 0x04, 0x52, 0x05, 0x6d, 0x61, 0x6a, 0x6f, 0x72,
+	0x12, 0x14, 0x0a, 0x05, 0x6d, 0x69, 0x6e, 0x6f, 0x72, 0x18, 0x02, 0x20, 0x01, 0x28, 0x04, 0x52,
+	0x05, 0x6d, 0x69, 0x6e, 0x6f, 0x72, 0x12, 0x16, 0x0a, 0x06, 0x72, 0x62, 0x79, 0x74, 0x65, 0x73,
+	0x18, 0x03, 0x20, 0x01, 0x28, 0x04, 0x52, 0x06, 0x72, 0x62, 0x79, 0x74, 0x65, 0x73, 0x12, 0x16,
+	0x0a, 0x06, 0x77, 0x62, 0x79, 0x74, 0x65, 0x73, 0x18, 0x04, 0x20, 0x01, 0x28, 0x04, 0x52, 0x06,
+	0x77, 0x62, 0x79, 0x74, 0x65, 0x73, 0x12, 0x12, 0x0a, 0x04, 0x72, 0x69, 0x6f, 0x73, 0x18, 0x05,
+	0x20, 0x01, 0x28, 0x04, 0x52, 0x04, 0x72, 0x69, 0x6f, 0x73, 0x12, 0x12, 0x0a, 0x04, 0x77, 0x69,
+	0x6f, 0x73, 0x18, 0x06, 0x20, 0x01, 0x28, 0x04, 0x52, 0x04, 0x77, 0x69, 0x6f, 0x73, 0x22, 0x55,
+	0x0a, 0x0b, 0x48, 0x75, 0x67, 0x65, 0x54, 0x6c, 0x62, 0x53, 0x74, 0x61, 0x74, 0x12, 0x18, 0x0a,
+	0x07, 0x63, 0x75, 0x72, 0x72, 0x65, 0x6e, 0x74, 0x18, 0x01, 0x20, 0x01, 0x28, 0x04, 0x52, 0x07,
+	0x63, 0x75, 0x72, 0x72, 0x65, 0x6e, 0x74, 0x12, 0x10, 0x0a, 0x03, 0x6d, 0x61, 0x78, 0x18, 0x02,
+	0x20, 0x01, 0x28, 0x04, 0x52, 0x03, 0x6d, 0x61, 0x78, 0x12, 0x1a, 0x0a, 0x08, 0x70, 0x61, 0x67,
+	0x65, 0x73, 0x69, 0x7a, 0x65, 0x18, 0x03, 0x20, 0x01, 0x28, 0x09, 0x52, 0x08, 0x70, 0x61, 0x67,
+	0x65, 0x73, 0x69, 0x7a, 0x65, 0x42, 0x2d, 0x5a, 0x2b, 0x67, 0x69, 0x74, 0x68, 0x75, 0x62, 0x2e,
+	0x63, 0x6f, 0x6d, 0x2f, 0x63, 0x6f, 0x6e, 0x74, 0x61, 0x69, 0x6e, 0x65, 0x72, 0x64, 0x2f, 0x63,
+	0x67, 0x72, 0x6f, 0x75, 0x70, 0x73, 0x2f, 0x63, 0x67, 0x72, 0x6f, 0x75, 0x70, 0x32, 0x2f, 0x73,
+	0x74, 0x61, 0x74, 0x73, 0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33,
 }
 
 var (
@@ -1154,35 +1348,42 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescGZIP(
 	return file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDescData
 }
 
-var file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes = make([]protoimpl.MessageInfo, 10)
+var file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes = make([]protoimpl.MessageInfo, 12)
 var file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_goTypes = []interface{}{
 	(*Metrics)(nil),      // 0: io.containerd.cgroups.v2.Metrics
-	(*PidsStat)(nil),     // 1: io.containerd.cgroups.v2.PidsStat
+	(*PSIData)(nil),      // 1: io.containerd.cgroups.v2.PSIData
-	(*CPUStat)(nil),      // 2: io.containerd.cgroups.v2.CPUStat
+	(*PSIStats)(nil),     // 2: io.containerd.cgroups.v2.PSIStats
-	(*MemoryStat)(nil),   // 3: io.containerd.cgroups.v2.MemoryStat
+	(*PidsStat)(nil),     // 3: io.containerd.cgroups.v2.PidsStat
-	(*MemoryEvents)(nil), // 4: io.containerd.cgroups.v2.MemoryEvents
+	(*CPUStat)(nil),      // 4: io.containerd.cgroups.v2.CPUStat
-	(*RdmaStat)(nil),     // 5: io.containerd.cgroups.v2.RdmaStat
+	(*MemoryStat)(nil),   // 5: io.containerd.cgroups.v2.MemoryStat
-	(*RdmaEntry)(nil),    // 6: io.containerd.cgroups.v2.RdmaEntry
+	(*MemoryEvents)(nil), // 6: io.containerd.cgroups.v2.MemoryEvents
-	(*IOStat)(nil),       // 7: io.containerd.cgroups.v2.IOStat
+	(*RdmaStat)(nil),     // 7: io.containerd.cgroups.v2.RdmaStat
-	(*IOEntry)(nil),      // 8: io.containerd.cgroups.v2.IOEntry
+	(*RdmaEntry)(nil),    // 8: io.containerd.cgroups.v2.RdmaEntry
-	(*HugeTlbStat)(nil),  // 9: io.containerd.cgroups.v2.HugeTlbStat
+	(*IOStat)(nil),       // 9: io.containerd.cgroups.v2.IOStat
+	(*IOEntry)(nil),      // 10: io.containerd.cgroups.v2.IOEntry
+	(*HugeTlbStat)(nil),  // 11: io.containerd.cgroups.v2.HugeTlbStat
 }
 var file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_depIdxs = []int32{
-	1,  // 0: io.containerd.cgroups.v2.Metrics.pids:type_name -> io.containerd.cgroups.v2.PidsStat
+	3,  // 0: io.containerd.cgroups.v2.Metrics.pids:type_name -> io.containerd.cgroups.v2.PidsStat
-	2,  // 1: io.containerd.cgroups.v2.Metrics.cpu:type_name -> io.containerd.cgroups.v2.CPUStat
+	4,  // 1: io.containerd.cgroups.v2.Metrics.cpu:type_name -> io.containerd.cgroups.v2.CPUStat
-	3,  // 2: io.containerd.cgroups.v2.Metrics.memory:type_name -> io.containerd.cgroups.v2.MemoryStat
+	5,  // 2: io.containerd.cgroups.v2.Metrics.memory:type_name -> io.containerd.cgroups.v2.MemoryStat
-	5,  // 3: io.containerd.cgroups.v2.Metrics.rdma:type_name -> io.containerd.cgroups.v2.RdmaStat
+	7,  // 3: io.containerd.cgroups.v2.Metrics.rdma:type_name -> io.containerd.cgroups.v2.RdmaStat
-	7,  // 4: io.containerd.cgroups.v2.Metrics.io:type_name -> io.containerd.cgroups.v2.IOStat
+	9,  // 4: io.containerd.cgroups.v2.Metrics.io:type_name -> io.containerd.cgroups.v2.IOStat
-	9,  // 5: io.containerd.cgroups.v2.Metrics.hugetlb:type_name -> io.containerd.cgroups.v2.HugeTlbStat
+	11, // 5: io.containerd.cgroups.v2.Metrics.hugetlb:type_name -> io.containerd.cgroups.v2.HugeTlbStat
-	4,  // 6: io.containerd.cgroups.v2.Metrics.memory_events:type_name -> io.containerd.cgroups.v2.MemoryEvents
+	6,  // 6: io.containerd.cgroups.v2.Metrics.memory_events:type_name -> io.containerd.cgroups.v2.MemoryEvents
-	6,  // 7: io.containerd.cgroups.v2.RdmaStat.current:type_name -> io.containerd.cgroups.v2.RdmaEntry
+	1,  // 7: io.containerd.cgroups.v2.PSIStats.some:type_name -> io.containerd.cgroups.v2.PSIData
-	6,  // 8: io.containerd.cgroups.v2.RdmaStat.limit:type_name -> io.containerd.cgroups.v2.RdmaEntry
+	1,  // 8: io.containerd.cgroups.v2.PSIStats.full:type_name -> io.containerd.cgroups.v2.PSIData
-	8,  // 9: io.containerd.cgroups.v2.IOStat.usage:type_name -> io.containerd.cgroups.v2.IOEntry
+	2,  // 9: io.containerd.cgroups.v2.CPUStat.psi:type_name -> io.containerd.cgroups.v2.PSIStats
-	10, // [10:10] is the sub-list for method output_type
+	2,  // 10: io.containerd.cgroups.v2.MemoryStat.psi:type_name -> io.containerd.cgroups.v2.PSIStats
-	10, // [10:10] is the sub-list for method input_type
+	8,  // 11: io.containerd.cgroups.v2.RdmaStat.current:type_name -> io.containerd.cgroups.v2.RdmaEntry
-	10, // [10:10] is the sub-list for extension type_name
+	8,  // 12: io.containerd.cgroups.v2.RdmaStat.limit:type_name -> io.containerd.cgroups.v2.RdmaEntry
-	10, // [10:10] is the sub-list for extension extendee
+	10, // 13: io.containerd.cgroups.v2.IOStat.usage:type_name -> io.containerd.cgroups.v2.IOEntry
-	0,  // [0:10] is the sub-list for field type_name
+	2,  // 14: io.containerd.cgroups.v2.IOStat.psi:type_name -> io.containerd.cgroups.v2.PSIStats
+	15, // [15:15] is the sub-list for method output_type
+	15, // [15:15] is the sub-list for method input_type
+	15, // [15:15] is the sub-list for extension type_name
+	15, // [15:15] is the sub-list for extension extendee
+	0,  // [0:15] is the sub-list for field type_name
 }
 
 func init() { file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() }
@@ -1204,7 +1405,7 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() {
 			}
 		}
 		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[1].Exporter = func(v interface{}, i int) interface{} {
-			switch v := v.(*PidsStat); i {
+			switch v := v.(*PSIData); i {
 			case 0:
 				return &v.state
 			case 1:
@@ -1216,7 +1417,7 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() {
 			}
 		}
 		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[2].Exporter = func(v interface{}, i int) interface{} {
-			switch v := v.(*CPUStat); i {
+			switch v := v.(*PSIStats); i {
 			case 0:
 				return &v.state
 			case 1:
@@ -1228,7 +1429,7 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() {
 			}
 		}
 		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[3].Exporter = func(v interface{}, i int) interface{} {
-			switch v := v.(*MemoryStat); i {
+			switch v := v.(*PidsStat); i {
 			case 0:
 				return &v.state
 			case 1:
@@ -1240,7 +1441,7 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() {
 			}
 		}
 		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[4].Exporter = func(v interface{}, i int) interface{} {
-			switch v := v.(*MemoryEvents); i {
+			switch v := v.(*CPUStat); i {
 			case 0:
 				return &v.state
 			case 1:
@@ -1252,7 +1453,7 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() {
 			}
 		}
 		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[5].Exporter = func(v interface{}, i int) interface{} {
-			switch v := v.(*RdmaStat); i {
+			switch v := v.(*MemoryStat); i {
 			case 0:
 				return &v.state
 			case 1:
@@ -1264,7 +1465,7 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() {
 			}
 		}
 		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[6].Exporter = func(v interface{}, i int) interface{} {
-			switch v := v.(*RdmaEntry); i {
+			switch v := v.(*MemoryEvents); i {
 			case 0:
 				return &v.state
 			case 1:
@@ -1276,7 +1477,7 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() {
 			}
 		}
 		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[7].Exporter = func(v interface{}, i int) interface{} {
-			switch v := v.(*IOStat); i {
+			switch v := v.(*RdmaStat); i {
 			case 0:
 				return &v.state
 			case 1:
@@ -1288,7 +1489,7 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() {
 			}
 		}
 		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[8].Exporter = func(v interface{}, i int) interface{} {
-			switch v := v.(*IOEntry); i {
+			switch v := v.(*RdmaEntry); i {
 			case 0:
 				return &v.state
 			case 1:
@@ -1300,6 +1501,30 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() {
 			}
 		}
 		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[9].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*IOStat); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[10].Exporter = func(v interface{}, i int) interface{} {
+			switch v := v.(*IOEntry); i {
+			case 0:
+				return &v.state
+			case 1:
+				return &v.sizeCache
+			case 2:
+				return &v.unknownFields
+			default:
+				return nil
+			}
+		}
+		file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_msgTypes[11].Exporter = func(v interface{}, i int) interface{} {
 			switch v := v.(*HugeTlbStat); i {
 			case 0:
 				return &v.state
@@ -1318,7 +1543,7 @@ func file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_init() {
 			GoPackagePath: reflect.TypeOf(x{}).PkgPath(),
 			RawDescriptor: file_github_com_containerd_cgroups_cgroup2_stats_metrics_proto_rawDesc,
 			NumEnums:      0,
-			NumMessages:   10,
+			NumMessages:   12,
 			NumExtensions: 0,
 			NumServices:   0,
 		},

vendor/github.com/containerd/cgroups/v3/cgroup2/stats/metrics.pb.txt

@@ -60,6 +60,56 @@ file {
       json_name: "memoryEvents"
     }
   }
+  message_type {
+    name: "PSIData"
+    field {
+      name: "avg10"
+      number: 1
+      label: LABEL_OPTIONAL
+      type: TYPE_DOUBLE
+      json_name: "avg10"
+    }
+    field {
+      name: "avg60"
+      number: 2
+      label: LABEL_OPTIONAL
+      type: TYPE_DOUBLE
+      json_name: "avg60"
+    }
+    field {
+      name: "avg300"
+      number: 3
+      label: LABEL_OPTIONAL
+      type: TYPE_DOUBLE
+      json_name: "avg300"
+    }
+    field {
+      name: "total"
+      number: 4
+      label: LABEL_OPTIONAL
+      type: TYPE_UINT64
+      json_name: "total"
+    }
+  }
+  message_type {
+    name: "PSIStats"
+    field {
+      name: "some"
+      number: 1
+      label: LABEL_OPTIONAL
+      type: TYPE_MESSAGE
+      type_name: ".io.containerd.cgroups.v2.PSIData"
+      json_name: "some"
+    }
+    field {
+      name: "full"
+      number: 2
+      label: LABEL_OPTIONAL
+      type: TYPE_MESSAGE
+      type_name: ".io.containerd.cgroups.v2.PSIData"
+      json_name: "full"
+    }
+  }
   message_type {
     name: "PidsStat"
     field {
@@ -121,6 +171,14 @@ file {
       type: TYPE_UINT64
       json_name: "throttledUsec"
     }
+    field {
+      name: "psi"
+      number: 7
+      label: LABEL_OPTIONAL
+      type: TYPE_MESSAGE
+      type_name: ".io.containerd.cgroups.v2.PSIStats"
+      json_name: "psi"
+    }
   }
   message_type {
     name: "MemoryStat"
@@ -369,6 +427,28 @@ file {
       type: TYPE_UINT64
       json_name: "swapLimit"
     }
+    field {
+      name: "max_usage"
+      number: 36
+      label: LABEL_OPTIONAL
+      type: TYPE_UINT64
+      json_name: "maxUsage"
+    }
+    field {
+      name: "swap_max_usage"
+      number: 37
+      label: LABEL_OPTIONAL
+      type: TYPE_UINT64
+      json_name: "swapMaxUsage"
+    }
+    field {
+      name: "psi"
+      number: 38
+      label: LABEL_OPTIONAL
+      type: TYPE_MESSAGE
+      type_name: ".io.containerd.cgroups.v2.PSIStats"
+      json_name: "psi"
+    }
   }
   message_type {
     name: "MemoryEvents"
@@ -461,6 +541,14 @@ file {
       type_name: ".io.containerd.cgroups.v2.IOEntry"
       json_name: "usage"
     }
+    field {
+      name: "psi"
+      number: 2
+      label: LABEL_OPTIONAL
+      type: TYPE_MESSAGE
+      type_name: ".io.containerd.cgroups.v2.PSIStats"
+      json_name: "psi"
+    }
   }
   message_type {
     name: "IOEntry"

vendor/github.com/containerd/cgroups/v3/cgroup2/stats/metrics.proto

@@ -14,6 +14,18 @@ message Metrics {
 	MemoryEvents memory_events = 8;
 }
 
+message PSIData {
+	double avg10 = 1;
+	double avg60 = 2;
+	double avg300 = 3;
+	uint64 total = 4;
+}
+
+message PSIStats {
+	PSIData some = 1;
+	PSIData full = 2;
+}
+
 message PidsStat {
 	uint64 current = 1;
 	uint64 limit = 2;
@@ -26,6 +38,7 @@ message CPUStat {
 	uint64 nr_periods = 4;
 	uint64 nr_throttled = 5;
 	uint64 throttled_usec = 6;
+	PSIStats psi = 7;
 }
 
 message MemoryStat {
@@ -64,6 +77,9 @@ message MemoryStat {
 	uint64 usage_limit = 33;
 	uint64 swap_usage = 34;
 	uint64 swap_limit = 35;
+	uint64 max_usage = 36;
+	uint64 swap_max_usage = 37;
+	PSIStats psi = 38;
 }
 
 message MemoryEvents {
@@ -87,6 +103,7 @@ message RdmaEntry {
 
 message IOStat {
 	repeated IOEntry usage = 1;
+	PSIStats psi = 2;
 }
 
 message IOEntry {

vendor/github.com/containerd/cgroups/v3/cgroup2/utils.go

@@ -68,8 +68,9 @@ func remove(path string) error {
 	return fmt.Errorf("cgroups: unable to remove path %q: %w", path, err)
 }
 
-// parseCgroupProcsFile parses /sys/fs/cgroup/$GROUPPATH/cgroup.procs
+// parseCgroupTasksFile parses /sys/fs/cgroup/$GROUPPATH/cgroup.procs or
-func parseCgroupProcsFile(path string) ([]uint64, error) {
+// /sys/fs/cgroup/$GROUPPATH/cgroup.threads
+func parseCgroupTasksFile(path string) ([]uint64, error) {
 	f, err := os.Open(path)
 	if err != nil {
 		return nil, err
@@ -176,6 +177,10 @@ func ToResources(spec *specs.LinuxResources) *Resources {
 		resources.Memory = &Memory{}
 		if swap := mem.Swap; swap != nil {
 			resources.Memory.Swap = swap
+			if l := mem.Limit; l != nil {
+				reduce := *swap - *l
+				resources.Memory.Swap = &reduce
+			}
 		}
 		if l := mem.Limit; l != nil {
 			resources.Memory.Max = l
@@ -479,6 +484,74 @@ func getHugePageSizeFromFilenames(fileNames []string) ([]string, error) {
 	return pageSizes, warn
 }
 
+func getStatPSIFromFile(path string) *stats.PSIStats {
+	f, err := os.Open(path)
+	if err != nil {
+		return nil
+	}
+	defer f.Close()
+
+	psistats := &stats.PSIStats{}
+	sc := bufio.NewScanner(f)
+	for sc.Scan() {
+		parts := strings.Fields(sc.Text())
+		var pv *stats.PSIData
+		switch parts[0] {
+		case "some":
+			psistats.Some = &stats.PSIData{}
+			pv = psistats.Some
+		case "full":
+			psistats.Full = &stats.PSIData{}
+			pv = psistats.Full
+		}
+		if pv != nil {
+			err = parsePSIData(parts[1:], pv)
+			if err != nil {
+				logrus.Errorf("failed to read file %s: %v", path, err)
+				return nil
+			}
+		}
+	}
+
+	if err := sc.Err(); err != nil {
+		logrus.Errorf("unable to parse PSI data: %v", err)
+		return nil
+	}
+	return psistats
+}
+
+func parsePSIData(psi []string, data *stats.PSIData) error {
+	for _, f := range psi {
+		kv := strings.SplitN(f, "=", 2)
+		if len(kv) != 2 {
+			return fmt.Errorf("invalid PSI data: %q", f)
+		}
+		var pv *float64
+		switch kv[0] {
+		case "avg10":
+			pv = &data.Avg10
+		case "avg60":
+			pv = &data.Avg60
+		case "avg300":
+			pv = &data.Avg300
+		case "total":
+			v, err := strconv.ParseUint(kv[1], 10, 64)
+			if err != nil {
+				return fmt.Errorf("invalid %s PSI value: %w", kv[0], err)
+			}
+			data.Total = v
+		}
+		if pv != nil {
+			v, err := strconv.ParseFloat(kv[1], 64)
+			if err != nil {
+				return fmt.Errorf("invalid %s PSI value: %w", kv[0], err)
+			}
+			*pv = v
+		}
+	}
+	return nil
+}
+
 func getSubreaper() (int, error) {
 	var i uintptr
 	if err := unix.Prctl(unix.PR_GET_CHILD_SUBREAPER, uintptr(unsafe.Pointer(&i)), 0, 0, 0); err != nil {

vendor/golang.org/x/exp/LICENSE

@@ -0,0 +1,27 @@
+Copyright (c) 2009 The Go Authors. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   * Redistributions of source code must retain the above copyright
+notice, this list of conditions and the following disclaimer.
+   * Redistributions in binary form must reproduce the above
+copyright notice, this list of conditions and the following disclaimer
+in the documentation and/or other materials provided with the
+distribution.
+   * Neither the name of Google Inc. nor the names of its
+contributors may be used to endorse or promote products derived from
+this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

vendor/golang.org/x/exp/PATENTS

@@ -0,0 +1,22 @@
+Additional IP Rights Grant (Patents)
+
+"This implementation" means the copyrightable works distributed by
+Google as part of the Go project.
+
+Google hereby grants to You a perpetual, worldwide, non-exclusive,
+no-charge, royalty-free, irrevocable (except as stated in this section)
+patent license to make, have made, use, offer to sell, sell, import,
+transfer and otherwise run, modify and propagate the contents of this
+implementation of Go, where such license applies only to those patent
+claims, both currently owned or controlled by Google and acquired in
+the future, licensable by Google that are necessarily infringed by this
+implementation of Go.  This grant does not include claims that would be
+infringed only as a consequence of further modification of this
+implementation.  If you or your agent or exclusive licensee institute or
+order or agree to the institution of patent litigation against any
+entity (including a cross-claim or counterclaim in a lawsuit) alleging
+that this implementation of Go or any code incorporated within this
+implementation of Go constitutes direct or contributory patent
+infringement, or inducement of patent infringement, then any patent
+rights granted to you under this License for this implementation of Go
+shall terminate as of the date such litigation is filed.

vendor/golang.org/x/exp/constraints/constraints.go

@@ -0,0 +1,50 @@
+// Copyright 2021 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// Package constraints defines a set of useful constraints to be used
+// with type parameters.
+package constraints
+
+// Signed is a constraint that permits any signed integer type.
+// If future releases of Go add new predeclared signed integer types,
+// this constraint will be modified to include them.
+type Signed interface {
+	~int | ~int8 | ~int16 | ~int32 | ~int64
+}
+
+// Unsigned is a constraint that permits any unsigned integer type.
+// If future releases of Go add new predeclared unsigned integer types,
+// this constraint will be modified to include them.
+type Unsigned interface {
+	~uint | ~uint8 | ~uint16 | ~uint32 | ~uint64 | ~uintptr
+}
+
+// Integer is a constraint that permits any integer type.
+// If future releases of Go add new predeclared integer types,
+// this constraint will be modified to include them.
+type Integer interface {
+	Signed | Unsigned
+}
+
+// Float is a constraint that permits any floating-point type.
+// If future releases of Go add new predeclared floating-point types,
+// this constraint will be modified to include them.
+type Float interface {
+	~float32 | ~float64
+}
+
+// Complex is a constraint that permits any complex numeric type.
+// If future releases of Go add new predeclared complex numeric types,
+// this constraint will be modified to include them.
+type Complex interface {
+	~complex64 | ~complex128
+}
+
+// Ordered is a constraint that permits any ordered type: any type
+// that supports the operators < <= >= >.
+// If future releases of Go add new ordered types,
+// this constraint will be modified to include them.
+type Ordered interface {
+	Integer | Float | ~string
+}

vendor/golang.org/x/exp/maps/maps.go

@@ -0,0 +1,94 @@
+// Copyright 2021 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// Package maps defines various functions useful with maps of any type.
+package maps
+
+// Keys returns the keys of the map m.
+// The keys will be in an indeterminate order.
+func Keys[M ~map[K]V, K comparable, V any](m M) []K {
+	r := make([]K, 0, len(m))
+	for k := range m {
+		r = append(r, k)
+	}
+	return r
+}
+
+// Values returns the values of the map m.
+// The values will be in an indeterminate order.
+func Values[M ~map[K]V, K comparable, V any](m M) []V {
+	r := make([]V, 0, len(m))
+	for _, v := range m {
+		r = append(r, v)
+	}
+	return r
+}
+
+// Equal reports whether two maps contain the same key/value pairs.
+// Values are compared using ==.
+func Equal[M1, M2 ~map[K]V, K, V comparable](m1 M1, m2 M2) bool {
+	if len(m1) != len(m2) {
+		return false
+	}
+	for k, v1 := range m1 {
+		if v2, ok := m2[k]; !ok || v1 != v2 {
+			return false
+		}
+	}
+	return true
+}
+
+// EqualFunc is like Equal, but compares values using eq.
+// Keys are still compared with ==.
+func EqualFunc[M1 ~map[K]V1, M2 ~map[K]V2, K comparable, V1, V2 any](m1 M1, m2 M2, eq func(V1, V2) bool) bool {
+	if len(m1) != len(m2) {
+		return false
+	}
+	for k, v1 := range m1 {
+		if v2, ok := m2[k]; !ok || !eq(v1, v2) {
+			return false
+		}
+	}
+	return true
+}
+
+// Clear removes all entries from m, leaving it empty.
+func Clear[M ~map[K]V, K comparable, V any](m M) {
+	for k := range m {
+		delete(m, k)
+	}
+}
+
+// Clone returns a copy of m.  This is a shallow clone:
+// the new keys and values are set using ordinary assignment.
+func Clone[M ~map[K]V, K comparable, V any](m M) M {
+	// Preserve nil in case it matters.
+	if m == nil {
+		return nil
+	}
+	r := make(M, len(m))
+	for k, v := range m {
+		r[k] = v
+	}
+	return r
+}
+
+// Copy copies all key/value pairs in src adding them to dst.
+// When a key in src is already present in dst,
+// the value in dst will be overwritten by the value associated
+// with the key in src.
+func Copy[M1 ~map[K]V, M2 ~map[K]V, K comparable, V any](dst M1, src M2) {
+	for k, v := range src {
+		dst[k] = v
+	}
+}
+
+// DeleteFunc deletes any key/value pairs from m for which del returns true.
+func DeleteFunc[M ~map[K]V, K comparable, V any](m M, del func(K, V) bool) {
+	for k, v := range m {
+		if del(k, v) {
+			delete(m, k)
+		}
+	}
+}

vendor/golang.org/x/exp/slices/cmp.go

@@ -0,0 +1,44 @@
+// Copyright 2023 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package slices
+
+import "golang.org/x/exp/constraints"
+
+// min is a version of the predeclared function from the Go 1.21 release.
+func min[T constraints.Ordered](a, b T) T {
+	if a < b || isNaN(a) {
+		return a
+	}
+	return b
+}
+
+// max is a version of the predeclared function from the Go 1.21 release.
+func max[T constraints.Ordered](a, b T) T {
+	if a > b || isNaN(a) {
+		return a
+	}
+	return b
+}
+
+// cmpLess is a copy of cmp.Less from the Go 1.21 release.
+func cmpLess[T constraints.Ordered](x, y T) bool {
+	return (isNaN(x) && !isNaN(y)) || x < y
+}
+
+// cmpCompare is a copy of cmp.Compare from the Go 1.21 release.
+func cmpCompare[T constraints.Ordered](x, y T) int {
+	xNaN := isNaN(x)
+	yNaN := isNaN(y)
+	if xNaN && yNaN {
+		return 0
+	}
+	if xNaN || x < y {
+		return -1
+	}
+	if yNaN || x > y {
+		return +1
+	}
+	return 0
+}

vendor/golang.org/x/exp/slices/slices.go

@@ -0,0 +1,499 @@
+// Copyright 2021 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// Package slices defines various functions useful with slices of any type.
+package slices
+
+import (
+	"unsafe"
+
+	"golang.org/x/exp/constraints"
+)
+
+// Equal reports whether two slices are equal: the same length and all
+// elements equal. If the lengths are different, Equal returns false.
+// Otherwise, the elements are compared in increasing index order, and the
+// comparison stops at the first unequal pair.
+// Floating point NaNs are not considered equal.
+func Equal[S ~[]E, E comparable](s1, s2 S) bool {
+	if len(s1) != len(s2) {
+		return false
+	}
+	for i := range s1 {
+		if s1[i] != s2[i] {
+			return false
+		}
+	}
+	return true
+}
+
+// EqualFunc reports whether two slices are equal using an equality
+// function on each pair of elements. If the lengths are different,
+// EqualFunc returns false. Otherwise, the elements are compared in
+// increasing index order, and the comparison stops at the first index
+// for which eq returns false.
+func EqualFunc[S1 ~[]E1, S2 ~[]E2, E1, E2 any](s1 S1, s2 S2, eq func(E1, E2) bool) bool {
+	if len(s1) != len(s2) {
+		return false
+	}
+	for i, v1 := range s1 {
+		v2 := s2[i]
+		if !eq(v1, v2) {
+			return false
+		}
+	}
+	return true
+}
+
+// Compare compares the elements of s1 and s2, using [cmp.Compare] on each pair
+// of elements. The elements are compared sequentially, starting at index 0,
+// until one element is not equal to the other.
+// The result of comparing the first non-matching elements is returned.
+// If both slices are equal until one of them ends, the shorter slice is
+// considered less than the longer one.
+// The result is 0 if s1 == s2, -1 if s1 < s2, and +1 if s1 > s2.
+func Compare[S ~[]E, E constraints.Ordered](s1, s2 S) int {
+	for i, v1 := range s1 {
+		if i >= len(s2) {
+			return +1
+		}
+		v2 := s2[i]
+		if c := cmpCompare(v1, v2); c != 0 {
+			return c
+		}
+	}
+	if len(s1) < len(s2) {
+		return -1
+	}
+	return 0
+}
+
+// CompareFunc is like [Compare] but uses a custom comparison function on each
+// pair of elements.
+// The result is the first non-zero result of cmp; if cmp always
+// returns 0 the result is 0 if len(s1) == len(s2), -1 if len(s1) < len(s2),
+// and +1 if len(s1) > len(s2).
+func CompareFunc[S1 ~[]E1, S2 ~[]E2, E1, E2 any](s1 S1, s2 S2, cmp func(E1, E2) int) int {
+	for i, v1 := range s1 {
+		if i >= len(s2) {
+			return +1
+		}
+		v2 := s2[i]
+		if c := cmp(v1, v2); c != 0 {
+			return c
+		}
+	}
+	if len(s1) < len(s2) {
+		return -1
+	}
+	return 0
+}
+
+// Index returns the index of the first occurrence of v in s,
+// or -1 if not present.
+func Index[S ~[]E, E comparable](s S, v E) int {
+	for i := range s {
+		if v == s[i] {
+			return i
+		}
+	}
+	return -1
+}
+
+// IndexFunc returns the first index i satisfying f(s[i]),
+// or -1 if none do.
+func IndexFunc[S ~[]E, E any](s S, f func(E) bool) int {
+	for i := range s {
+		if f(s[i]) {
+			return i
+		}
+	}
+	return -1
+}
+
+// Contains reports whether v is present in s.
+func Contains[S ~[]E, E comparable](s S, v E) bool {
+	return Index(s, v) >= 0
+}
+
+// ContainsFunc reports whether at least one
+// element e of s satisfies f(e).
+func ContainsFunc[S ~[]E, E any](s S, f func(E) bool) bool {
+	return IndexFunc(s, f) >= 0
+}
+
+// Insert inserts the values v... into s at index i,
+// returning the modified slice.
+// The elements at s[i:] are shifted up to make room.
+// In the returned slice r, r[i] == v[0],
+// and r[i+len(v)] == value originally at r[i].
+// Insert panics if i is out of range.
+// This function is O(len(s) + len(v)).
+func Insert[S ~[]E, E any](s S, i int, v ...E) S {
+	m := len(v)
+	if m == 0 {
+		return s
+	}
+	n := len(s)
+	if i == n {
+		return append(s, v...)
+	}
+	if n+m > cap(s) {
+		// Use append rather than make so that we bump the size of
+		// the slice up to the next storage class.
+		// This is what Grow does but we don't call Grow because
+		// that might copy the values twice.
+		s2 := append(s[:i], make(S, n+m-i)...)
+		copy(s2[i:], v)
+		copy(s2[i+m:], s[i:])
+		return s2
+	}
+	s = s[:n+m]
+
+	// before:
+	// s: aaaaaaaabbbbccccccccdddd
+	//            ^   ^       ^   ^
+	//            i  i+m      n  n+m
+	// after:
+	// s: aaaaaaaavvvvbbbbcccccccc
+	//            ^   ^       ^   ^
+	//            i  i+m      n  n+m
+	//
+	// a are the values that don't move in s.
+	// v are the values copied in from v.
+	// b and c are the values from s that are shifted up in index.
+	// d are the values that get overwritten, never to be seen again.
+
+	if !overlaps(v, s[i+m:]) {
+		// Easy case - v does not overlap either the c or d regions.
+		// (It might be in some of a or b, or elsewhere entirely.)
+		// The data we copy up doesn't write to v at all, so just do it.
+
+		copy(s[i+m:], s[i:])
+
+		// Now we have
+		// s: aaaaaaaabbbbbbbbcccccccc
+		//            ^   ^       ^   ^
+		//            i  i+m      n  n+m
+		// Note the b values are duplicated.
+
+		copy(s[i:], v)
+
+		// Now we have
+		// s: aaaaaaaavvvvbbbbcccccccc
+		//            ^   ^       ^   ^
+		//            i  i+m      n  n+m
+		// That's the result we want.
+		return s
+	}
+
+	// The hard case - v overlaps c or d. We can't just shift up
+	// the data because we'd move or clobber the values we're trying
+	// to insert.
+	// So instead, write v on top of d, then rotate.
+	copy(s[n:], v)
+
+	// Now we have
+	// s: aaaaaaaabbbbccccccccvvvv
+	//            ^   ^       ^   ^
+	//            i  i+m      n  n+m
+
+	rotateRight(s[i:], m)
+
+	// Now we have
+	// s: aaaaaaaavvvvbbbbcccccccc
+	//            ^   ^       ^   ^
+	//            i  i+m      n  n+m
+	// That's the result we want.
+	return s
+}
+
+// Delete removes the elements s[i:j] from s, returning the modified slice.
+// Delete panics if s[i:j] is not a valid slice of s.
+// Delete is O(len(s)-j), so if many items must be deleted, it is better to
+// make a single call deleting them all together than to delete one at a time.
+// Delete might not modify the elements s[len(s)-(j-i):len(s)]. If those
+// elements contain pointers you might consider zeroing those elements so that
+// objects they reference can be garbage collected.
+func Delete[S ~[]E, E any](s S, i, j int) S {
+	_ = s[i:j] // bounds check
+
+	return append(s[:i], s[j:]...)
+}
+
+// DeleteFunc removes any elements from s for which del returns true,
+// returning the modified slice.
+// When DeleteFunc removes m elements, it might not modify the elements
+// s[len(s)-m:len(s)]. If those elements contain pointers you might consider
+// zeroing those elements so that objects they reference can be garbage
+// collected.
+func DeleteFunc[S ~[]E, E any](s S, del func(E) bool) S {
+	i := IndexFunc(s, del)
+	if i == -1 {
+		return s
+	}
+	// Don't start copying elements until we find one to delete.
+	for j := i + 1; j < len(s); j++ {
+		if v := s[j]; !del(v) {
+			s[i] = v
+			i++
+		}
+	}
+	return s[:i]
+}
+
+// Replace replaces the elements s[i:j] by the given v, and returns the
+// modified slice. Replace panics if s[i:j] is not a valid slice of s.
+func Replace[S ~[]E, E any](s S, i, j int, v ...E) S {
+	_ = s[i:j] // verify that i:j is a valid subslice
+
+	if i == j {
+		return Insert(s, i, v...)
+	}
+	if j == len(s) {
+		return append(s[:i], v...)
+	}
+
+	tot := len(s[:i]) + len(v) + len(s[j:])
+	if tot > cap(s) {
+		// Too big to fit, allocate and copy over.
+		s2 := append(s[:i], make(S, tot-i)...) // See Insert
+		copy(s2[i:], v)
+		copy(s2[i+len(v):], s[j:])
+		return s2
+	}
+
+	r := s[:tot]
+
+	if i+len(v) <= j {
+		// Easy, as v fits in the deleted portion.
+		copy(r[i:], v)
+		if i+len(v) != j {
+			copy(r[i+len(v):], s[j:])
+		}
+		return r
+	}
+
+	// We are expanding (v is bigger than j-i).
+	// The situation is something like this:
+	// (example has i=4,j=8,len(s)=16,len(v)=6)
+	// s: aaaaxxxxbbbbbbbbyy
+	//        ^   ^       ^ ^
+	//        i   j  len(s) tot
+	// a: prefix of s
+	// x: deleted range
+	// b: more of s
+	// y: area to expand into
+
+	if !overlaps(r[i+len(v):], v) {
+		// Easy, as v is not clobbered by the first copy.
+		copy(r[i+len(v):], s[j:])
+		copy(r[i:], v)
+		return r
+	}
+
+	// This is a situation where we don't have a single place to which
+	// we can copy v. Parts of it need to go to two different places.
+	// We want to copy the prefix of v into y and the suffix into x, then
+	// rotate |y| spots to the right.
+	//
+	//        v[2:]      v[:2]
+	//         |           |
+	// s: aaaavvvvbbbbbbbbvv
+	//        ^   ^       ^ ^
+	//        i   j  len(s) tot
+	//
+	// If either of those two destinations don't alias v, then we're good.
+	y := len(v) - (j - i) // length of y portion
+
+	if !overlaps(r[i:j], v) {
+		copy(r[i:j], v[y:])
+		copy(r[len(s):], v[:y])
+		rotateRight(r[i:], y)
+		return r
+	}
+	if !overlaps(r[len(s):], v) {
+		copy(r[len(s):], v[:y])
+		copy(r[i:j], v[y:])
+		rotateRight(r[i:], y)
+		return r
+	}
+
+	// Now we know that v overlaps both x and y.
+	// That means that the entirety of b is *inside* v.
+	// So we don't need to preserve b at all; instead we
+	// can copy v first, then copy the b part of v out of
+	// v to the right destination.
+	k := startIdx(v, s[j:])
+	copy(r[i:], v)
+	copy(r[i+len(v):], r[i+k:])
+	return r
+}
+
+// Clone returns a copy of the slice.
+// The elements are copied using assignment, so this is a shallow clone.
+func Clone[S ~[]E, E any](s S) S {
+	// Preserve nil in case it matters.
+	if s == nil {
+		return nil
+	}
+	return append(S([]E{}), s...)
+}
+
+// Compact replaces consecutive runs of equal elements with a single copy.
+// This is like the uniq command found on Unix.
+// Compact modifies the contents of the slice s and returns the modified slice,
+// which may have a smaller length.
+// When Compact discards m elements in total, it might not modify the elements
+// s[len(s)-m:len(s)]. If those elements contain pointers you might consider
+// zeroing those elements so that objects they reference can be garbage collected.
+func Compact[S ~[]E, E comparable](s S) S {
+	if len(s) < 2 {
+		return s
+	}
+	i := 1
+	for k := 1; k < len(s); k++ {
+		if s[k] != s[k-1] {
+			if i != k {
+				s[i] = s[k]
+			}
+			i++
+		}
+	}
+	return s[:i]
+}
+
+// CompactFunc is like [Compact] but uses an equality function to compare elements.
+// For runs of elements that compare equal, CompactFunc keeps the first one.
+func CompactFunc[S ~[]E, E any](s S, eq func(E, E) bool) S {
+	if len(s) < 2 {
+		return s
+	}
+	i := 1
+	for k := 1; k < len(s); k++ {
+		if !eq(s[k], s[k-1]) {
+			if i != k {
+				s[i] = s[k]
+			}
+			i++
+		}
+	}
+	return s[:i]
+}
+
+// Grow increases the slice's capacity, if necessary, to guarantee space for
+// another n elements. After Grow(n), at least n elements can be appended
+// to the slice without another allocation. If n is negative or too large to
+// allocate the memory, Grow panics.
+func Grow[S ~[]E, E any](s S, n int) S {
+	if n < 0 {
+		panic("cannot be negative")
+	}
+	if n -= cap(s) - len(s); n > 0 {
+		// TODO(https://go.dev/issue/53888): Make using []E instead of S
+		// to workaround a compiler bug where the runtime.growslice optimization
+		// does not take effect. Revert when the compiler is fixed.
+		s = append([]E(s)[:cap(s)], make([]E, n)...)[:len(s)]
+	}
+	return s
+}
+
+// Clip removes unused capacity from the slice, returning s[:len(s):len(s)].
+func Clip[S ~[]E, E any](s S) S {
+	return s[:len(s):len(s)]
+}
+
+// Rotation algorithm explanation:
+//
+// rotate left by 2
+// start with
+//   0123456789
+// split up like this
+//   01 234567 89
+// swap first 2 and last 2
+//   89 234567 01
+// join first parts
+//   89234567 01
+// recursively rotate first left part by 2
+//   23456789 01
+// join at the end
+//   2345678901
+//
+// rotate left by 8
+// start with
+//   0123456789
+// split up like this
+//   01 234567 89
+// swap first 2 and last 2
+//   89 234567 01
+// join last parts
+//   89 23456701
+// recursively rotate second part left by 6
+//   89 01234567
+// join at the end
+//   8901234567
+
+// TODO: There are other rotate algorithms.
+// This algorithm has the desirable property that it moves each element exactly twice.
+// The triple-reverse algorithm is simpler and more cache friendly, but takes more writes.
+// The follow-cycles algorithm can be 1-write but it is not very cache friendly.
+
+// rotateLeft rotates b left by n spaces.
+// s_final[i] = s_orig[i+r], wrapping around.
+func rotateLeft[E any](s []E, r int) {
+	for r != 0 && r != len(s) {
+		if r*2 <= len(s) {
+			swap(s[:r], s[len(s)-r:])
+			s = s[:len(s)-r]
+		} else {
+			swap(s[:len(s)-r], s[r:])
+			s, r = s[len(s)-r:], r*2-len(s)
+		}
+	}
+}
+func rotateRight[E any](s []E, r int) {
+	rotateLeft(s, len(s)-r)
+}
+
+// swap swaps the contents of x and y. x and y must be equal length and disjoint.
+func swap[E any](x, y []E) {
+	for i := 0; i < len(x); i++ {
+		x[i], y[i] = y[i], x[i]
+	}
+}
+
+// overlaps reports whether the memory ranges a[0:len(a)] and b[0:len(b)] overlap.
+func overlaps[E any](a, b []E) bool {
+	if len(a) == 0 || len(b) == 0 {
+		return false
+	}
+	elemSize := unsafe.Sizeof(a[0])
+	if elemSize == 0 {
+		return false
+	}
+	// TODO: use a runtime/unsafe facility once one becomes available. See issue 12445.
+	// Also see crypto/internal/alias/alias.go:AnyOverlap
+	return uintptr(unsafe.Pointer(&a[0])) <= uintptr(unsafe.Pointer(&b[len(b)-1]))+(elemSize-1) &&
+		uintptr(unsafe.Pointer(&b[0])) <= uintptr(unsafe.Pointer(&a[len(a)-1]))+(elemSize-1)
+}
+
+// startIdx returns the index in haystack where the needle starts.
+// prerequisite: the needle must be aliased entirely inside the haystack.
+func startIdx[E any](haystack, needle []E) int {
+	p := &needle[0]
+	for i := range haystack {
+		if p == &haystack[i] {
+			return i
+		}
+	}
+	// TODO: what if the overlap is by a non-integral number of Es?
+	panic("needle not found")
+}
+
+// Reverse reverses the elements of the slice in place.
+func Reverse[S ~[]E, E any](s S) {
+	for i, j := 0, len(s)-1; i < j; i, j = i+1, j-1 {
+		s[i], s[j] = s[j], s[i]
+	}
+}

vendor/golang.org/x/exp/slices/sort.go

@@ -0,0 +1,195 @@
+// Copyright 2022 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+//go:generate go run $GOROOT/src/sort/gen_sort_variants.go -exp
+
+package slices
+
+import (
+	"math/bits"
+
+	"golang.org/x/exp/constraints"
+)
+
+// Sort sorts a slice of any ordered type in ascending order.
+// When sorting floating-point numbers, NaNs are ordered before other values.
+func Sort[S ~[]E, E constraints.Ordered](x S) {
+	n := len(x)
+	pdqsortOrdered(x, 0, n, bits.Len(uint(n)))
+}
+
+// SortFunc sorts the slice x in ascending order as determined by the cmp
+// function. This sort is not guaranteed to be stable.
+// cmp(a, b) should return a negative number when a < b, a positive number when
+// a > b and zero when a == b.
+//
+// SortFunc requires that cmp is a strict weak ordering.
+// See https://en.wikipedia.org/wiki/Weak_ordering#Strict_weak_orderings.
+func SortFunc[S ~[]E, E any](x S, cmp func(a, b E) int) {
+	n := len(x)
+	pdqsortCmpFunc(x, 0, n, bits.Len(uint(n)), cmp)
+}
+
+// SortStableFunc sorts the slice x while keeping the original order of equal
+// elements, using cmp to compare elements in the same way as [SortFunc].
+func SortStableFunc[S ~[]E, E any](x S, cmp func(a, b E) int) {
+	stableCmpFunc(x, len(x), cmp)
+}
+
+// IsSorted reports whether x is sorted in ascending order.
+func IsSorted[S ~[]E, E constraints.Ordered](x S) bool {
+	for i := len(x) - 1; i > 0; i-- {
+		if cmpLess(x[i], x[i-1]) {
+			return false
+		}
+	}
+	return true
+}
+
+// IsSortedFunc reports whether x is sorted in ascending order, with cmp as the
+// comparison function as defined by [SortFunc].
+func IsSortedFunc[S ~[]E, E any](x S, cmp func(a, b E) int) bool {
+	for i := len(x) - 1; i > 0; i-- {
+		if cmp(x[i], x[i-1]) < 0 {
+			return false
+		}
+	}
+	return true
+}
+
+// Min returns the minimal value in x. It panics if x is empty.
+// For floating-point numbers, Min propagates NaNs (any NaN value in x
+// forces the output to be NaN).
+func Min[S ~[]E, E constraints.Ordered](x S) E {
+	if len(x) < 1 {
+		panic("slices.Min: empty list")
+	}
+	m := x[0]
+	for i := 1; i < len(x); i++ {
+		m = min(m, x[i])
+	}
+	return m
+}
+
+// MinFunc returns the minimal value in x, using cmp to compare elements.
+// It panics if x is empty. If there is more than one minimal element
+// according to the cmp function, MinFunc returns the first one.
+func MinFunc[S ~[]E, E any](x S, cmp func(a, b E) int) E {
+	if len(x) < 1 {
+		panic("slices.MinFunc: empty list")
+	}
+	m := x[0]
+	for i := 1; i < len(x); i++ {
+		if cmp(x[i], m) < 0 {
+			m = x[i]
+		}
+	}
+	return m
+}
+
+// Max returns the maximal value in x. It panics if x is empty.
+// For floating-point E, Max propagates NaNs (any NaN value in x
+// forces the output to be NaN).
+func Max[S ~[]E, E constraints.Ordered](x S) E {
+	if len(x) < 1 {
+		panic("slices.Max: empty list")
+	}
+	m := x[0]
+	for i := 1; i < len(x); i++ {
+		m = max(m, x[i])
+	}
+	return m
+}
+
+// MaxFunc returns the maximal value in x, using cmp to compare elements.
+// It panics if x is empty. If there is more than one maximal element
+// according to the cmp function, MaxFunc returns the first one.
+func MaxFunc[S ~[]E, E any](x S, cmp func(a, b E) int) E {
+	if len(x) < 1 {
+		panic("slices.MaxFunc: empty list")
+	}
+	m := x[0]
+	for i := 1; i < len(x); i++ {
+		if cmp(x[i], m) > 0 {
+			m = x[i]
+		}
+	}
+	return m
+}
+
+// BinarySearch searches for target in a sorted slice and returns the position
+// where target is found, or the position where target would appear in the
+// sort order; it also returns a bool saying whether the target is really found
+// in the slice. The slice must be sorted in increasing order.
+func BinarySearch[S ~[]E, E constraints.Ordered](x S, target E) (int, bool) {
+	// Inlining is faster than calling BinarySearchFunc with a lambda.
+	n := len(x)
+	// Define x[-1] < target and x[n] >= target.
+	// Invariant: x[i-1] < target, x[j] >= target.
+	i, j := 0, n
+	for i < j {
+		h := int(uint(i+j) >> 1) // avoid overflow when computing h
+		// i ≤ h < j
+		if cmpLess(x[h], target) {
+			i = h + 1 // preserves x[i-1] < target
+		} else {
+			j = h // preserves x[j] >= target
+		}
+	}
+	// i == j, x[i-1] < target, and x[j] (= x[i]) >= target  =>  answer is i.
+	return i, i < n && (x[i] == target || (isNaN(x[i]) && isNaN(target)))
+}
+
+// BinarySearchFunc works like [BinarySearch], but uses a custom comparison
+// function. The slice must be sorted in increasing order, where "increasing"
+// is defined by cmp. cmp should return 0 if the slice element matches
+// the target, a negative number if the slice element precedes the target,
+// or a positive number if the slice element follows the target.
+// cmp must implement the same ordering as the slice, such that if
+// cmp(a, t) < 0 and cmp(b, t) >= 0, then a must precede b in the slice.
+func BinarySearchFunc[S ~[]E, E, T any](x S, target T, cmp func(E, T) int) (int, bool) {
+	n := len(x)
+	// Define cmp(x[-1], target) < 0 and cmp(x[n], target) >= 0 .
+	// Invariant: cmp(x[i - 1], target) < 0, cmp(x[j], target) >= 0.
+	i, j := 0, n
+	for i < j {
+		h := int(uint(i+j) >> 1) // avoid overflow when computing h
+		// i ≤ h < j
+		if cmp(x[h], target) < 0 {
+			i = h + 1 // preserves cmp(x[i - 1], target) < 0
+		} else {
+			j = h // preserves cmp(x[j], target) >= 0
+		}
+	}
+	// i == j, cmp(x[i-1], target) < 0, and cmp(x[j], target) (= cmp(x[i], target)) >= 0  =>  answer is i.
+	return i, i < n && cmp(x[i], target) == 0
+}
+
+type sortedHint int // hint for pdqsort when choosing the pivot
+
+const (
+	unknownHint sortedHint = iota
+	increasingHint
+	decreasingHint
+)
+
+// xorshift paper: https://www.jstatsoft.org/article/view/v008i14/xorshift.pdf
+type xorshift uint64
+
+func (r *xorshift) Next() uint64 {
+	*r ^= *r << 13
+	*r ^= *r >> 17
+	*r ^= *r << 5
+	return uint64(*r)
+}
+
+func nextPowerOfTwo(length int) uint {
+	return 1 << bits.Len(uint(length))
+}
+
+// isNaN reports whether x is a NaN without requiring the math package.
+// This will always return false if T is not floating-point.
+func isNaN[T constraints.Ordered](x T) bool {
+	return x != x
+}

vendor/golang.org/x/exp/slices/zsortanyfunc.go

@@ -0,0 +1,479 @@
+// Code generated by gen_sort_variants.go; DO NOT EDIT.
+
+// Copyright 2022 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package slices
+
+// insertionSortCmpFunc sorts data[a:b] using insertion sort.
+func insertionSortCmpFunc[E any](data []E, a, b int, cmp func(a, b E) int) {
+	for i := a + 1; i < b; i++ {
+		for j := i; j > a && (cmp(data[j], data[j-1]) < 0); j-- {
+			data[j], data[j-1] = data[j-1], data[j]
+		}
+	}
+}
+
+// siftDownCmpFunc implements the heap property on data[lo:hi].
+// first is an offset into the array where the root of the heap lies.
+func siftDownCmpFunc[E any](data []E, lo, hi, first int, cmp func(a, b E) int) {
+	root := lo
+	for {
+		child := 2*root + 1
+		if child >= hi {
+			break
+		}
+		if child+1 < hi && (cmp(data[first+child], data[first+child+1]) < 0) {
+			child++
+		}
+		if !(cmp(data[first+root], data[first+child]) < 0) {
+			return
+		}
+		data[first+root], data[first+child] = data[first+child], data[first+root]
+		root = child
+	}
+}
+
+func heapSortCmpFunc[E any](data []E, a, b int, cmp func(a, b E) int) {
+	first := a
+	lo := 0
+	hi := b - a
+
+	// Build heap with greatest element at top.
+	for i := (hi - 1) / 2; i >= 0; i-- {
+		siftDownCmpFunc(data, i, hi, first, cmp)
+	}
+
+	// Pop elements, largest first, into end of data.
+	for i := hi - 1; i >= 0; i-- {
+		data[first], data[first+i] = data[first+i], data[first]
+		siftDownCmpFunc(data, lo, i, first, cmp)
+	}
+}
+
+// pdqsortCmpFunc sorts data[a:b].
+// The algorithm based on pattern-defeating quicksort(pdqsort), but without the optimizations from BlockQuicksort.
+// pdqsort paper: https://arxiv.org/pdf/2106.05123.pdf
+// C++ implementation: https://github.com/orlp/pdqsort
+// Rust implementation: https://docs.rs/pdqsort/latest/pdqsort/
+// limit is the number of allowed bad (very unbalanced) pivots before falling back to heapsort.
+func pdqsortCmpFunc[E any](data []E, a, b, limit int, cmp func(a, b E) int) {
+	const maxInsertion = 12
+
+	var (
+		wasBalanced    = true // whether the last partitioning was reasonably balanced
+		wasPartitioned = true // whether the slice was already partitioned
+	)
+
+	for {
+		length := b - a
+
+		if length <= maxInsertion {
+			insertionSortCmpFunc(data, a, b, cmp)
+			return
+		}
+
+		// Fall back to heapsort if too many bad choices were made.
+		if limit == 0 {
+			heapSortCmpFunc(data, a, b, cmp)
+			return
+		}
+
+		// If the last partitioning was imbalanced, we need to breaking patterns.
+		if !wasBalanced {
+			breakPatternsCmpFunc(data, a, b, cmp)
+			limit--
+		}
+
+		pivot, hint := choosePivotCmpFunc(data, a, b, cmp)
+		if hint == decreasingHint {
+			reverseRangeCmpFunc(data, a, b, cmp)
+			// The chosen pivot was pivot-a elements after the start of the array.
+			// After reversing it is pivot-a elements before the end of the array.
+			// The idea came from Rust's implementation.
+			pivot = (b - 1) - (pivot - a)
+			hint = increasingHint
+		}
+
+		// The slice is likely already sorted.
+		if wasBalanced && wasPartitioned && hint == increasingHint {
+			if partialInsertionSortCmpFunc(data, a, b, cmp) {
+				return
+			}
+		}
+
+		// Probably the slice contains many duplicate elements, partition the slice into
+		// elements equal to and elements greater than the pivot.
+		if a > 0 && !(cmp(data[a-1], data[pivot]) < 0) {
+			mid := partitionEqualCmpFunc(data, a, b, pivot, cmp)
+			a = mid
+			continue
+		}
+
+		mid, alreadyPartitioned := partitionCmpFunc(data, a, b, pivot, cmp)
+		wasPartitioned = alreadyPartitioned
+
+		leftLen, rightLen := mid-a, b-mid
+		balanceThreshold := length / 8
+		if leftLen < rightLen {
+			wasBalanced = leftLen >= balanceThreshold
+			pdqsortCmpFunc(data, a, mid, limit, cmp)
+			a = mid + 1
+		} else {
+			wasBalanced = rightLen >= balanceThreshold
+			pdqsortCmpFunc(data, mid+1, b, limit, cmp)
+			b = mid
+		}
+	}
+}
+
+// partitionCmpFunc does one quicksort partition.
+// Let p = data[pivot]
+// Moves elements in data[a:b] around, so that data[i]<p and data[j]>=p for i<newpivot and j>newpivot.
+// On return, data[newpivot] = p
+func partitionCmpFunc[E any](data []E, a, b, pivot int, cmp func(a, b E) int) (newpivot int, alreadyPartitioned bool) {
+	data[a], data[pivot] = data[pivot], data[a]
+	i, j := a+1, b-1 // i and j are inclusive of the elements remaining to be partitioned
+
+	for i <= j && (cmp(data[i], data[a]) < 0) {
+		i++
+	}
+	for i <= j && !(cmp(data[j], data[a]) < 0) {
+		j--
+	}
+	if i > j {
+		data[j], data[a] = data[a], data[j]
+		return j, true
+	}
+	data[i], data[j] = data[j], data[i]
+	i++
+	j--
+
+	for {
+		for i <= j && (cmp(data[i], data[a]) < 0) {
+			i++
+		}
+		for i <= j && !(cmp(data[j], data[a]) < 0) {
+			j--
+		}
+		if i > j {
+			break
+		}
+		data[i], data[j] = data[j], data[i]
+		i++
+		j--
+	}
+	data[j], data[a] = data[a], data[j]
+	return j, false
+}
+
+// partitionEqualCmpFunc partitions data[a:b] into elements equal to data[pivot] followed by elements greater than data[pivot].
+// It assumed that data[a:b] does not contain elements smaller than the data[pivot].
+func partitionEqualCmpFunc[E any](data []E, a, b, pivot int, cmp func(a, b E) int) (newpivot int) {
+	data[a], data[pivot] = data[pivot], data[a]
+	i, j := a+1, b-1 // i and j are inclusive of the elements remaining to be partitioned
+
+	for {
+		for i <= j && !(cmp(data[a], data[i]) < 0) {
+			i++
+		}
+		for i <= j && (cmp(data[a], data[j]) < 0) {
+			j--
+		}
+		if i > j {
+			break
+		}
+		data[i], data[j] = data[j], data[i]
+		i++
+		j--
+	}
+	return i
+}
+
+// partialInsertionSortCmpFunc partially sorts a slice, returns true if the slice is sorted at the end.
+func partialInsertionSortCmpFunc[E any](data []E, a, b int, cmp func(a, b E) int) bool {
+	const (
+		maxSteps         = 5  // maximum number of adjacent out-of-order pairs that will get shifted
+		shortestShifting = 50 // don't shift any elements on short arrays
+	)
+	i := a + 1
+	for j := 0; j < maxSteps; j++ {
+		for i < b && !(cmp(data[i], data[i-1]) < 0) {
+			i++
+		}
+
+		if i == b {
+			return true
+		}
+
+		if b-a < shortestShifting {
+			return false
+		}
+
+		data[i], data[i-1] = data[i-1], data[i]
+
+		// Shift the smaller one to the left.
+		if i-a >= 2 {
+			for j := i - 1; j >= 1; j-- {
+				if !(cmp(data[j], data[j-1]) < 0) {
+					break
+				}
+				data[j], data[j-1] = data[j-1], data[j]
+			}
+		}
+		// Shift the greater one to the right.
+		if b-i >= 2 {
+			for j := i + 1; j < b; j++ {
+				if !(cmp(data[j], data[j-1]) < 0) {
+					break
+				}
+				data[j], data[j-1] = data[j-1], data[j]
+			}
+		}
+	}
+	return false
+}
+
+// breakPatternsCmpFunc scatters some elements around in an attempt to break some patterns
+// that might cause imbalanced partitions in quicksort.
+func breakPatternsCmpFunc[E any](data []E, a, b int, cmp func(a, b E) int) {
+	length := b - a
+	if length >= 8 {
+		random := xorshift(length)
+		modulus := nextPowerOfTwo(length)
+
+		for idx := a + (length/4)*2 - 1; idx <= a+(length/4)*2+1; idx++ {
+			other := int(uint(random.Next()) & (modulus - 1))
+			if other >= length {
+				other -= length
+			}
+			data[idx], data[a+other] = data[a+other], data[idx]
+		}
+	}
+}
+
+// choosePivotCmpFunc chooses a pivot in data[a:b].
+//
+// [0,8): chooses a static pivot.
+// [8,shortestNinther): uses the simple median-of-three method.
+// [shortestNinther,∞): uses the Tukey ninther method.
+func choosePivotCmpFunc[E any](data []E, a, b int, cmp func(a, b E) int) (pivot int, hint sortedHint) {
+	const (
+		shortestNinther = 50
+		maxSwaps        = 4 * 3
+	)
+
+	l := b - a
+
+	var (
+		swaps int
+		i     = a + l/4*1
+		j     = a + l/4*2
+		k     = a + l/4*3
+	)
+
+	if l >= 8 {
+		if l >= shortestNinther {
+			// Tukey ninther method, the idea came from Rust's implementation.
+			i = medianAdjacentCmpFunc(data, i, &swaps, cmp)
+			j = medianAdjacentCmpFunc(data, j, &swaps, cmp)
+			k = medianAdjacentCmpFunc(data, k, &swaps, cmp)
+		}
+		// Find the median among i, j, k and stores it into j.
+		j = medianCmpFunc(data, i, j, k, &swaps, cmp)
+	}
+
+	switch swaps {
+	case 0:
+		return j, increasingHint
+	case maxSwaps:
+		return j, decreasingHint
+	default:
+		return j, unknownHint
+	}
+}
+
+// order2CmpFunc returns x,y where data[x] <= data[y], where x,y=a,b or x,y=b,a.
+func order2CmpFunc[E any](data []E, a, b int, swaps *int, cmp func(a, b E) int) (int, int) {
+	if cmp(data[b], data[a]) < 0 {
+		*swaps++
+		return b, a
+	}
+	return a, b
+}
+
+// medianCmpFunc returns x where data[x] is the median of data[a],data[b],data[c], where x is a, b, or c.
+func medianCmpFunc[E any](data []E, a, b, c int, swaps *int, cmp func(a, b E) int) int {
+	a, b = order2CmpFunc(data, a, b, swaps, cmp)
+	b, c = order2CmpFunc(data, b, c, swaps, cmp)
+	a, b = order2CmpFunc(data, a, b, swaps, cmp)
+	return b
+}
+
+// medianAdjacentCmpFunc finds the median of data[a - 1], data[a], data[a + 1] and stores the index into a.
+func medianAdjacentCmpFunc[E any](data []E, a int, swaps *int, cmp func(a, b E) int) int {
+	return medianCmpFunc(data, a-1, a, a+1, swaps, cmp)
+}
+
+func reverseRangeCmpFunc[E any](data []E, a, b int, cmp func(a, b E) int) {
+	i := a
+	j := b - 1
+	for i < j {
+		data[i], data[j] = data[j], data[i]
+		i++
+		j--
+	}
+}
+
+func swapRangeCmpFunc[E any](data []E, a, b, n int, cmp func(a, b E) int) {
+	for i := 0; i < n; i++ {
+		data[a+i], data[b+i] = data[b+i], data[a+i]
+	}
+}
+
+func stableCmpFunc[E any](data []E, n int, cmp func(a, b E) int) {
+	blockSize := 20 // must be > 0
+	a, b := 0, blockSize
+	for b <= n {
+		insertionSortCmpFunc(data, a, b, cmp)
+		a = b
+		b += blockSize
+	}
+	insertionSortCmpFunc(data, a, n, cmp)
+
+	for blockSize < n {
+		a, b = 0, 2*blockSize
+		for b <= n {
+			symMergeCmpFunc(data, a, a+blockSize, b, cmp)
+			a = b
+			b += 2 * blockSize
+		}
+		if m := a + blockSize; m < n {
+			symMergeCmpFunc(data, a, m, n, cmp)
+		}
+		blockSize *= 2
+	}
+}
+
+// symMergeCmpFunc merges the two sorted subsequences data[a:m] and data[m:b] using
+// the SymMerge algorithm from Pok-Son Kim and Arne Kutzner, "Stable Minimum
+// Storage Merging by Symmetric Comparisons", in Susanne Albers and Tomasz
+// Radzik, editors, Algorithms - ESA 2004, volume 3221 of Lecture Notes in
+// Computer Science, pages 714-723. Springer, 2004.
+//
+// Let M = m-a and N = b-n. Wolog M < N.
+// The recursion depth is bound by ceil(log(N+M)).
+// The algorithm needs O(M*log(N/M + 1)) calls to data.Less.
+// The algorithm needs O((M+N)*log(M)) calls to data.Swap.
+//
+// The paper gives O((M+N)*log(M)) as the number of assignments assuming a
+// rotation algorithm which uses O(M+N+gcd(M+N)) assignments. The argumentation
+// in the paper carries through for Swap operations, especially as the block
+// swapping rotate uses only O(M+N) Swaps.
+//
+// symMerge assumes non-degenerate arguments: a < m && m < b.
+// Having the caller check this condition eliminates many leaf recursion calls,
+// which improves performance.
+func symMergeCmpFunc[E any](data []E, a, m, b int, cmp func(a, b E) int) {
+	// Avoid unnecessary recursions of symMerge
+	// by direct insertion of data[a] into data[m:b]
+	// if data[a:m] only contains one element.
+	if m-a == 1 {
+		// Use binary search to find the lowest index i
+		// such that data[i] >= data[a] for m <= i < b.
+		// Exit the search loop with i == b in case no such index exists.
+		i := m
+		j := b
+		for i < j {
+			h := int(uint(i+j) >> 1)
+			if cmp(data[h], data[a]) < 0 {
+				i = h + 1
+			} else {
+				j = h
+			}
+		}
+		// Swap values until data[a] reaches the position before i.
+		for k := a; k < i-1; k++ {
+			data[k], data[k+1] = data[k+1], data[k]
+		}
+		return
+	}
+
+	// Avoid unnecessary recursions of symMerge
+	// by direct insertion of data[m] into data[a:m]
+	// if data[m:b] only contains one element.
+	if b-m == 1 {
+		// Use binary search to find the lowest index i
+		// such that data[i] > data[m] for a <= i < m.
+		// Exit the search loop with i == m in case no such index exists.
+		i := a
+		j := m
+		for i < j {
+			h := int(uint(i+j) >> 1)
+			if !(cmp(data[m], data[h]) < 0) {
+				i = h + 1
+			} else {
+				j = h
+			}
+		}
+		// Swap values until data[m] reaches the position i.
+		for k := m; k > i; k-- {
+			data[k], data[k-1] = data[k-1], data[k]
+		}
+		return
+	}
+
+	mid := int(uint(a+b) >> 1)
+	n := mid + m
+	var start, r int
+	if m > mid {
+		start = n - b
+		r = mid
+	} else {
+		start = a
+		r = m
+	}
+	p := n - 1
+
+	for start < r {
+		c := int(uint(start+r) >> 1)
+		if !(cmp(data[p-c], data[c]) < 0) {
+			start = c + 1
+		} else {
+			r = c
+		}
+	}
+
+	end := n - start
+	if start < m && m < end {
+		rotateCmpFunc(data, start, m, end, cmp)
+	}
+	if a < start && start < mid {
+		symMergeCmpFunc(data, a, start, mid, cmp)
+	}
+	if mid < end && end < b {
+		symMergeCmpFunc(data, mid, end, b, cmp)
+	}
+}
+
+// rotateCmpFunc rotates two consecutive blocks u = data[a:m] and v = data[m:b] in data:
+// Data of the form 'x u v y' is changed to 'x v u y'.
+// rotate performs at most b-a many calls to data.Swap,
+// and it assumes non-degenerate arguments: a < m && m < b.
+func rotateCmpFunc[E any](data []E, a, m, b int, cmp func(a, b E) int) {
+	i := m - a
+	j := b - m
+
+	for i != j {
+		if i > j {
+			swapRangeCmpFunc(data, m-i, m, j, cmp)
+			i -= j
+		} else {
+			swapRangeCmpFunc(data, m-i, m+j-i, i, cmp)
+			j -= i
+		}
+	}
+	// i == j
+	swapRangeCmpFunc(data, m-i, m, i, cmp)
+}

vendor/golang.org/x/exp/slices/zsortordered.go

@@ -0,0 +1,481 @@
+// Code generated by gen_sort_variants.go; DO NOT EDIT.
+
+// Copyright 2022 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package slices
+
+import "golang.org/x/exp/constraints"
+
+// insertionSortOrdered sorts data[a:b] using insertion sort.
+func insertionSortOrdered[E constraints.Ordered](data []E, a, b int) {
+	for i := a + 1; i < b; i++ {
+		for j := i; j > a && cmpLess(data[j], data[j-1]); j-- {
+			data[j], data[j-1] = data[j-1], data[j]
+		}
+	}
+}
+
+// siftDownOrdered implements the heap property on data[lo:hi].
+// first is an offset into the array where the root of the heap lies.
+func siftDownOrdered[E constraints.Ordered](data []E, lo, hi, first int) {
+	root := lo
+	for {
+		child := 2*root + 1
+		if child >= hi {
+			break
+		}
+		if child+1 < hi && cmpLess(data[first+child], data[first+child+1]) {
+			child++
+		}
+		if !cmpLess(data[first+root], data[first+child]) {
+			return
+		}
+		data[first+root], data[first+child] = data[first+child], data[first+root]
+		root = child
+	}
+}
+
+func heapSortOrdered[E constraints.Ordered](data []E, a, b int) {
+	first := a
+	lo := 0
+	hi := b - a
+
+	// Build heap with greatest element at top.
+	for i := (hi - 1) / 2; i >= 0; i-- {
+		siftDownOrdered(data, i, hi, first)
+	}
+
+	// Pop elements, largest first, into end of data.
+	for i := hi - 1; i >= 0; i-- {
+		data[first], data[first+i] = data[first+i], data[first]
+		siftDownOrdered(data, lo, i, first)
+	}
+}
+
+// pdqsortOrdered sorts data[a:b].
+// The algorithm based on pattern-defeating quicksort(pdqsort), but without the optimizations from BlockQuicksort.
+// pdqsort paper: https://arxiv.org/pdf/2106.05123.pdf
+// C++ implementation: https://github.com/orlp/pdqsort
+// Rust implementation: https://docs.rs/pdqsort/latest/pdqsort/
+// limit is the number of allowed bad (very unbalanced) pivots before falling back to heapsort.
+func pdqsortOrdered[E constraints.Ordered](data []E, a, b, limit int) {
+	const maxInsertion = 12
+
+	var (
+		wasBalanced    = true // whether the last partitioning was reasonably balanced
+		wasPartitioned = true // whether the slice was already partitioned
+	)
+
+	for {
+		length := b - a
+
+		if length <= maxInsertion {
+			insertionSortOrdered(data, a, b)
+			return
+		}
+
+		// Fall back to heapsort if too many bad choices were made.
+		if limit == 0 {
+			heapSortOrdered(data, a, b)
+			return
+		}
+
+		// If the last partitioning was imbalanced, we need to breaking patterns.
+		if !wasBalanced {
+			breakPatternsOrdered(data, a, b)
+			limit--
+		}
+
+		pivot, hint := choosePivotOrdered(data, a, b)
+		if hint == decreasingHint {
+			reverseRangeOrdered(data, a, b)
+			// The chosen pivot was pivot-a elements after the start of the array.
+			// After reversing it is pivot-a elements before the end of the array.
+			// The idea came from Rust's implementation.
+			pivot = (b - 1) - (pivot - a)
+			hint = increasingHint
+		}
+
+		// The slice is likely already sorted.
+		if wasBalanced && wasPartitioned && hint == increasingHint {
+			if partialInsertionSortOrdered(data, a, b) {
+				return
+			}
+		}
+
+		// Probably the slice contains many duplicate elements, partition the slice into
+		// elements equal to and elements greater than the pivot.
+		if a > 0 && !cmpLess(data[a-1], data[pivot]) {
+			mid := partitionEqualOrdered(data, a, b, pivot)
+			a = mid
+			continue
+		}
+
+		mid, alreadyPartitioned := partitionOrdered(data, a, b, pivot)
+		wasPartitioned = alreadyPartitioned
+
+		leftLen, rightLen := mid-a, b-mid
+		balanceThreshold := length / 8
+		if leftLen < rightLen {
+			wasBalanced = leftLen >= balanceThreshold
+			pdqsortOrdered(data, a, mid, limit)
+			a = mid + 1
+		} else {
+			wasBalanced = rightLen >= balanceThreshold
+			pdqsortOrdered(data, mid+1, b, limit)
+			b = mid
+		}
+	}
+}
+
+// partitionOrdered does one quicksort partition.
+// Let p = data[pivot]
+// Moves elements in data[a:b] around, so that data[i]<p and data[j]>=p for i<newpivot and j>newpivot.
+// On return, data[newpivot] = p
+func partitionOrdered[E constraints.Ordered](data []E, a, b, pivot int) (newpivot int, alreadyPartitioned bool) {
+	data[a], data[pivot] = data[pivot], data[a]
+	i, j := a+1, b-1 // i and j are inclusive of the elements remaining to be partitioned
+
+	for i <= j && cmpLess(data[i], data[a]) {
+		i++
+	}
+	for i <= j && !cmpLess(data[j], data[a]) {
+		j--
+	}
+	if i > j {
+		data[j], data[a] = data[a], data[j]
+		return j, true
+	}
+	data[i], data[j] = data[j], data[i]
+	i++
+	j--
+
+	for {
+		for i <= j && cmpLess(data[i], data[a]) {
+			i++
+		}
+		for i <= j && !cmpLess(data[j], data[a]) {
+			j--
+		}
+		if i > j {
+			break
+		}
+		data[i], data[j] = data[j], data[i]
+		i++
+		j--
+	}
+	data[j], data[a] = data[a], data[j]
+	return j, false
+}
+
+// partitionEqualOrdered partitions data[a:b] into elements equal to data[pivot] followed by elements greater than data[pivot].
+// It assumed that data[a:b] does not contain elements smaller than the data[pivot].
+func partitionEqualOrdered[E constraints.Ordered](data []E, a, b, pivot int) (newpivot int) {
+	data[a], data[pivot] = data[pivot], data[a]
+	i, j := a+1, b-1 // i and j are inclusive of the elements remaining to be partitioned
+
+	for {
+		for i <= j && !cmpLess(data[a], data[i]) {
+			i++
+		}
+		for i <= j && cmpLess(data[a], data[j]) {
+			j--
+		}
+		if i > j {
+			break
+		}
+		data[i], data[j] = data[j], data[i]
+		i++
+		j--
+	}
+	return i
+}
+
+// partialInsertionSortOrdered partially sorts a slice, returns true if the slice is sorted at the end.
+func partialInsertionSortOrdered[E constraints.Ordered](data []E, a, b int) bool {
+	const (
+		maxSteps         = 5  // maximum number of adjacent out-of-order pairs that will get shifted
+		shortestShifting = 50 // don't shift any elements on short arrays
+	)
+	i := a + 1
+	for j := 0; j < maxSteps; j++ {
+		for i < b && !cmpLess(data[i], data[i-1]) {
+			i++
+		}
+
+		if i == b {
+			return true
+		}
+
+		if b-a < shortestShifting {
+			return false
+		}
+
+		data[i], data[i-1] = data[i-1], data[i]
+
+		// Shift the smaller one to the left.
+		if i-a >= 2 {
+			for j := i - 1; j >= 1; j-- {
+				if !cmpLess(data[j], data[j-1]) {
+					break
+				}
+				data[j], data[j-1] = data[j-1], data[j]
+			}
+		}
+		// Shift the greater one to the right.
+		if b-i >= 2 {
+			for j := i + 1; j < b; j++ {
+				if !cmpLess(data[j], data[j-1]) {
+					break
+				}
+				data[j], data[j-1] = data[j-1], data[j]
+			}
+		}
+	}
+	return false
+}
+
+// breakPatternsOrdered scatters some elements around in an attempt to break some patterns
+// that might cause imbalanced partitions in quicksort.
+func breakPatternsOrdered[E constraints.Ordered](data []E, a, b int) {
+	length := b - a
+	if length >= 8 {
+		random := xorshift(length)
+		modulus := nextPowerOfTwo(length)
+
+		for idx := a + (length/4)*2 - 1; idx <= a+(length/4)*2+1; idx++ {
+			other := int(uint(random.Next()) & (modulus - 1))
+			if other >= length {
+				other -= length
+			}
+			data[idx], data[a+other] = data[a+other], data[idx]
+		}
+	}
+}
+
+// choosePivotOrdered chooses a pivot in data[a:b].
+//
+// [0,8): chooses a static pivot.
+// [8,shortestNinther): uses the simple median-of-three method.
+// [shortestNinther,∞): uses the Tukey ninther method.
+func choosePivotOrdered[E constraints.Ordered](data []E, a, b int) (pivot int, hint sortedHint) {
+	const (
+		shortestNinther = 50
+		maxSwaps        = 4 * 3
+	)
+
+	l := b - a
+
+	var (
+		swaps int
+		i     = a + l/4*1
+		j     = a + l/4*2
+		k     = a + l/4*3
+	)
+
+	if l >= 8 {
+		if l >= shortestNinther {
+			// Tukey ninther method, the idea came from Rust's implementation.
+			i = medianAdjacentOrdered(data, i, &swaps)
+			j = medianAdjacentOrdered(data, j, &swaps)
+			k = medianAdjacentOrdered(data, k, &swaps)
+		}
+		// Find the median among i, j, k and stores it into j.
+		j = medianOrdered(data, i, j, k, &swaps)
+	}
+
+	switch swaps {
+	case 0:
+		return j, increasingHint
+	case maxSwaps:
+		return j, decreasingHint
+	default:
+		return j, unknownHint
+	}
+}
+
+// order2Ordered returns x,y where data[x] <= data[y], where x,y=a,b or x,y=b,a.
+func order2Ordered[E constraints.Ordered](data []E, a, b int, swaps *int) (int, int) {
+	if cmpLess(data[b], data[a]) {
+		*swaps++
+		return b, a
+	}
+	return a, b
+}
+
+// medianOrdered returns x where data[x] is the median of data[a],data[b],data[c], where x is a, b, or c.
+func medianOrdered[E constraints.Ordered](data []E, a, b, c int, swaps *int) int {
+	a, b = order2Ordered(data, a, b, swaps)
+	b, c = order2Ordered(data, b, c, swaps)
+	a, b = order2Ordered(data, a, b, swaps)
+	return b
+}
+
+// medianAdjacentOrdered finds the median of data[a - 1], data[a], data[a + 1] and stores the index into a.
+func medianAdjacentOrdered[E constraints.Ordered](data []E, a int, swaps *int) int {
+	return medianOrdered(data, a-1, a, a+1, swaps)
+}
+
+func reverseRangeOrdered[E constraints.Ordered](data []E, a, b int) {
+	i := a
+	j := b - 1
+	for i < j {
+		data[i], data[j] = data[j], data[i]
+		i++
+		j--
+	}
+}
+
+func swapRangeOrdered[E constraints.Ordered](data []E, a, b, n int) {
+	for i := 0; i < n; i++ {
+		data[a+i], data[b+i] = data[b+i], data[a+i]
+	}
+}
+
+func stableOrdered[E constraints.Ordered](data []E, n int) {
+	blockSize := 20 // must be > 0
+	a, b := 0, blockSize
+	for b <= n {
+		insertionSortOrdered(data, a, b)
+		a = b
+		b += blockSize
+	}
+	insertionSortOrdered(data, a, n)
+
+	for blockSize < n {
+		a, b = 0, 2*blockSize
+		for b <= n {
+			symMergeOrdered(data, a, a+blockSize, b)
+			a = b
+			b += 2 * blockSize
+		}
+		if m := a + blockSize; m < n {
+			symMergeOrdered(data, a, m, n)
+		}
+		blockSize *= 2
+	}
+}
+
+// symMergeOrdered merges the two sorted subsequences data[a:m] and data[m:b] using
+// the SymMerge algorithm from Pok-Son Kim and Arne Kutzner, "Stable Minimum
+// Storage Merging by Symmetric Comparisons", in Susanne Albers and Tomasz
+// Radzik, editors, Algorithms - ESA 2004, volume 3221 of Lecture Notes in
+// Computer Science, pages 714-723. Springer, 2004.
+//
+// Let M = m-a and N = b-n. Wolog M < N.
+// The recursion depth is bound by ceil(log(N+M)).
+// The algorithm needs O(M*log(N/M + 1)) calls to data.Less.
+// The algorithm needs O((M+N)*log(M)) calls to data.Swap.
+//
+// The paper gives O((M+N)*log(M)) as the number of assignments assuming a
+// rotation algorithm which uses O(M+N+gcd(M+N)) assignments. The argumentation
+// in the paper carries through for Swap operations, especially as the block
+// swapping rotate uses only O(M+N) Swaps.
+//
+// symMerge assumes non-degenerate arguments: a < m && m < b.
+// Having the caller check this condition eliminates many leaf recursion calls,
+// which improves performance.
+func symMergeOrdered[E constraints.Ordered](data []E, a, m, b int) {
+	// Avoid unnecessary recursions of symMerge
+	// by direct insertion of data[a] into data[m:b]
+	// if data[a:m] only contains one element.
+	if m-a == 1 {
+		// Use binary search to find the lowest index i
+		// such that data[i] >= data[a] for m <= i < b.
+		// Exit the search loop with i == b in case no such index exists.
+		i := m
+		j := b
+		for i < j {
+			h := int(uint(i+j) >> 1)
+			if cmpLess(data[h], data[a]) {
+				i = h + 1
+			} else {
+				j = h
+			}
+		}
+		// Swap values until data[a] reaches the position before i.
+		for k := a; k < i-1; k++ {
+			data[k], data[k+1] = data[k+1], data[k]
+		}
+		return
+	}
+
+	// Avoid unnecessary recursions of symMerge
+	// by direct insertion of data[m] into data[a:m]
+	// if data[m:b] only contains one element.
+	if b-m == 1 {
+		// Use binary search to find the lowest index i
+		// such that data[i] > data[m] for a <= i < m.
+		// Exit the search loop with i == m in case no such index exists.
+		i := a
+		j := m
+		for i < j {
+			h := int(uint(i+j) >> 1)
+			if !cmpLess(data[m], data[h]) {
+				i = h + 1
+			} else {
+				j = h
+			}
+		}
+		// Swap values until data[m] reaches the position i.
+		for k := m; k > i; k-- {
+			data[k], data[k-1] = data[k-1], data[k]
+		}
+		return
+	}
+
+	mid := int(uint(a+b) >> 1)
+	n := mid + m
+	var start, r int
+	if m > mid {
+		start = n - b
+		r = mid
+	} else {
+		start = a
+		r = m
+	}
+	p := n - 1
+
+	for start < r {
+		c := int(uint(start+r) >> 1)
+		if !cmpLess(data[p-c], data[c]) {
+			start = c + 1
+		} else {
+			r = c
+		}
+	}
+
+	end := n - start
+	if start < m && m < end {
+		rotateOrdered(data, start, m, end)
+	}
+	if a < start && start < mid {
+		symMergeOrdered(data, a, start, mid)
+	}
+	if mid < end && end < b {
+		symMergeOrdered(data, mid, end, b)
+	}
+}
+
+// rotateOrdered rotates two consecutive blocks u = data[a:m] and v = data[m:b] in data:
+// Data of the form 'x u v y' is changed to 'x v u y'.
+// rotate performs at most b-a many calls to data.Swap,
+// and it assumes non-degenerate arguments: a < m && m < b.
+func rotateOrdered[E constraints.Ordered](data []E, a, m, b int) {
+	i := m - a
+	j := b - m
+
+	for i != j {
+		if i > j {
+			swapRangeOrdered(data, m-i, m, j)
+			i -= j
+		} else {
+			swapRangeOrdered(data, m-i, m+j-i, i)
+			j -= i
+		}
+	}
+	// i == j
+	swapRangeOrdered(data, m-i, m, i)
+}

vendor/golang.org/x/tools/go/internal/packagesdriver/sizes.go

@@ -13,16 +13,17 @@ import (
 	"golang.org/x/tools/internal/gocommand"
 )
 
-var debug = false
-
 func GetSizesForArgsGolist(ctx context.Context, inv gocommand.Invocation, gocmdRunner *gocommand.Runner) (string, string, error) {
 	inv.Verb = "list"
 	inv.Args = []string{"-f", "{{context.GOARCH}} {{context.Compiler}}", "--", "unsafe"}
 	stdout, stderr, friendlyErr, rawErr := gocmdRunner.RunRaw(ctx, inv)
 	var goarch, compiler string
 	if rawErr != nil {
-		if rawErrMsg := rawErr.Error(); strings.Contains(rawErrMsg, "cannot find main module") || strings.Contains(rawErrMsg, "go.mod file not found") {
+		rawErrMsg := rawErr.Error()
-			// User's running outside of a module. All bets are off. Get GOARCH and guess compiler is gc.
+		if strings.Contains(rawErrMsg, "cannot find main module") ||
+			strings.Contains(rawErrMsg, "go.mod file not found") {
+			// User's running outside of a module.
+			// All bets are off. Get GOARCH and guess compiler is gc.
 			// TODO(matloob): Is this a problem in practice?
 			inv.Verb = "env"
 			inv.Args = []string{"GOARCH"}
@@ -32,8 +33,12 @@ func GetSizesForArgsGolist(ctx context.Context, inv gocommand.Invocation, gocmdR
 			}
 			goarch = strings.TrimSpace(envout.String())
 			compiler = "gc"
-		} else {
+		} else if friendlyErr != nil {
 			return "", "", friendlyErr
+		} else {
+			// This should be unreachable, but be defensive
+			// in case RunRaw's error results are inconsistent.
+			return "", "", rawErr
 		}
 	} else {
 		fields := strings.Fields(stdout.String())