mirror of
https://github.com/ibm-s390-linux/s390-tools.git
synced 2026-08-05 02:14:52 +00:00
Integrate zlib DFLTCC deflate compression to single volume dasd dumper using the existing s390 extended dump format. Compression takes place only if DFLTCC facility is available, otherwise dump is written uncompressed as before. First megabyte of memory is always written uncompressed and afterwards this area is used for zlib workspace and for the compression output buffer. The compression takes place in chunks of data of equal size (currently 1MB) and the offset of each compressed chunk is stored in the dump segment header. Since existing dump segment headers of 1 page size are used, we need to limit the maximum size of compressed dump segments. Chunk is written uncompressed in case of compression error or if deflate compression only makes it bigger. Thus every chunk of data is compressed separately and can be decompressed independently. The main reason for that is to enable zgetdump to make fast read seeks. Otherwise, zgetdump would need to decompress a big dump segment in the worst case to extract a single piece of data. Put compression related functions and structures to eckd2dump_zlib.c and eckd2dump_zlib.h Update zipl man page with the general information of zlib compression support. Signed-off-by: Mikhail Zaslonko <zaslonko@linux.ibm.com> Reviewed-by: Alexander Egorenkov <egorenar@linux.ibm.com> Signed-off-by: Jan Höppner <hoeppner@linux.ibm.com>
183 lines
6.0 KiB
C
183 lines
6.0 KiB
C
/*
|
|
* zipl - zSeries Initial Program Loader tool
|
|
*
|
|
* Common ECKD dump I/O functions
|
|
*
|
|
* Copyright IBM Corp. 2013, 2023
|
|
*
|
|
* s390-tools is free software; you can redistribute it and/or modify
|
|
* it under the terms of the MIT license. See LICENSE for details.
|
|
*/
|
|
|
|
#include "cio.h"
|
|
#include "eckd2dump.h"
|
|
#include "eckd2dump_zlib.h"
|
|
#include "boot/linux_layout.h"
|
|
#include "boot/s390.h"
|
|
#include "stage2dump.h"
|
|
#include "dump/s390_dump.h"
|
|
|
|
/*
|
|
* The first megabyte of memory is used for zlib workspace and
|
|
* the output buffer for compressed data.
|
|
*/
|
|
static void *zlib_out_buf; // zlib output buffer
|
|
static unsigned long zlib_buf_size; // zlib output buffer size
|
|
|
|
/* Zlib workspace memory offset (skipping first 64K for the dumper itself) */
|
|
#define ZLIB_WORKSPACE_OFFSET IMAGE_ENTRY
|
|
|
|
/*
|
|
* Zlib workspace initialization.
|
|
* Return -1 if zlib deflate workspace requires more than half of the
|
|
* reserved memory area. Otherwise return 0.
|
|
*/
|
|
int zlib_workarea_init(unsigned long addr, z_stream *strm)
|
|
{
|
|
strm->workspace = (void *)MAX(addr, ZLIB_WORKSPACE_OFFSET);
|
|
zlib_out_buf = (void *)ROUND_UP((unsigned long)strm->workspace +
|
|
zlib_deflate_workspacesize(MAX_WBITS, MAX_MEM_LEVEL),
|
|
PAGE_SIZE);
|
|
/* Make sure there is enough space left for zlib output buffer (at least half) */
|
|
if ((unsigned long)(zlib_out_buf - strm->workspace) > ZLIB_WORKSPACE_LIMIT / 2)
|
|
return -1;
|
|
/* Memory left for zlib output buffer (currently 692K) */
|
|
zlib_buf_size = addr + ZLIB_WORKSPACE_LIMIT - (unsigned long)zlib_out_buf;
|
|
return 0;
|
|
}
|
|
|
|
#define COMPRESSED 0
|
|
#define UNCOMPRESSED 1
|
|
#define COMPRESSION_ERROR -1
|
|
|
|
/*
|
|
* Compress chunk of data of given length at the given address using zlib
|
|
* deflate compression and write it starting from the given block number.
|
|
* Variable *blk is updated to the next free block number.
|
|
* Return 0 if memory chunk is successfully compressed and written.
|
|
* Return 1 if compression is ineffective and the entire chunk is written
|
|
* uncompressed.
|
|
* In case of deflate error, write the chunk uncompressed and return -1
|
|
*/
|
|
static int compress_write_next_chunk(unsigned long addr, unsigned long len,
|
|
unsigned long *blk, z_stream *strm)
|
|
{
|
|
unsigned long start_blk;
|
|
int rc;
|
|
|
|
/* Reset the stream for each new chunk */
|
|
zlib_deflateReset(strm);
|
|
strm->next_in = (void *)addr;
|
|
strm->avail_in = len;
|
|
strm->next_out = zlib_out_buf;
|
|
strm->avail_out = zlib_buf_size;
|
|
start_blk = *blk;
|
|
while (strm->avail_in != 0) {
|
|
rc = zlib_deflate(strm, Z_NO_FLUSH);
|
|
/*
|
|
* Compression error. Write this data chunk uncompressed and
|
|
* notify the caller.
|
|
*/
|
|
if (rc != Z_OK) {
|
|
*blk = write_addr_range(start_blk, addr, len);
|
|
return COMPRESSION_ERROR;
|
|
}
|
|
if (strm->avail_out == 0) {
|
|
/*
|
|
* If compressed output is larger than the input, write this chunk of
|
|
* data uncompressed and notify the caller with the return code.
|
|
* Do it only if no compressed output has been written yet.
|
|
*/
|
|
if (strm->total_out >= strm->total_in &&
|
|
*blk == start_blk) {
|
|
*blk = write_addr_range(start_blk, addr, len);
|
|
return UNCOMPRESSED;
|
|
}
|
|
*blk = write_addr_range(*blk, (unsigned long)zlib_out_buf,
|
|
zlib_buf_size);
|
|
strm->next_out = (void *)zlib_out_buf;
|
|
strm->avail_out = zlib_buf_size;
|
|
}
|
|
}
|
|
while (rc == Z_OK) {
|
|
strm->next_in = NULL;
|
|
strm->avail_in = 0;
|
|
rc = zlib_deflate(strm, Z_FINISH);
|
|
len = ROUND_UP(zlib_buf_size - strm->avail_out, PAGE_SIZE);
|
|
*blk = write_addr_range(*blk, (unsigned long)zlib_out_buf, len);
|
|
if (strm->avail_out == 0) {
|
|
strm->next_out = zlib_out_buf;
|
|
strm->avail_out = zlib_buf_size;
|
|
rc = Z_OK;
|
|
} else {
|
|
break;
|
|
}
|
|
}
|
|
return COMPRESSED;
|
|
}
|
|
|
|
/*
|
|
* Compress and write memory dump segment with the uncompressed header to DASD
|
|
* and return the next free block number.
|
|
* The compression takes place in chunks of data of equal size (currently 1MB)
|
|
* and the offset of each compressed chunk is stored in the dump segment
|
|
* header. Due to limited size of header, the maximum size of compressed dump
|
|
* segment is limited to DUMP_SEGM_ZLIB_MAXLEN.
|
|
* If compression of a memory chunk leads to the data expansion (due to
|
|
* incompressible input), the chunk of data is written uncompressed.
|
|
* Thus every chunk of data is compressed separately and can be
|
|
* decompressed independently. The main reason is to enable zgetdump to make
|
|
* fast read seeks. Otherwise, zgetdump would need to uncompress a big dump
|
|
* segment in the worst case to extract a single piece of data.
|
|
*/
|
|
unsigned long write_compressed_dump_segment(unsigned long blk,
|
|
struct df_s390_dump_segm_hdr *segm,
|
|
z_stream *strm)
|
|
{
|
|
unsigned long head_blk, start_blk, zero_page, len, offset = 0;
|
|
unsigned long chunk_size;
|
|
int rc;
|
|
|
|
head_blk = blk;
|
|
/* Skip one block for the header (written later on) */
|
|
blk += m2b(sizeof(struct df_s390_dump_segm_hdr));
|
|
/* Compress each data chunk of the dump segment separately */
|
|
chunk_size = dump_hdr->zlib_entry_size;
|
|
for (unsigned int i = 0; i <= segm->len / chunk_size; i++) {
|
|
/* Save starting block number */
|
|
start_blk = blk;
|
|
len = i < segm->len / chunk_size ?
|
|
chunk_size : segm->len % chunk_size;
|
|
if (len == 0)
|
|
break;
|
|
rc = compress_write_next_chunk(segm->start + i * chunk_size,
|
|
len, &blk, strm);
|
|
/*
|
|
* Store the offset to the compressed chunk of data written
|
|
* to disk in blocks.
|
|
*/
|
|
segm->entry_count++;
|
|
segm->entry_offset[i] = (uint32_t)offset;
|
|
/*
|
|
* Compression was ineffective or compression error ocurred,
|
|
* data chunk has benn written uncompressed.
|
|
*/
|
|
if (rc == UNCOMPRESSED || rc == COMPRESSION_ERROR)
|
|
segm->entry_offset[i] |= DUMP_SEGM_ENTRY_UNCOMPRESSED;
|
|
offset += blk - start_blk;
|
|
progress_print(segm->start + i * chunk_size);
|
|
}
|
|
/* Compression successful, store compressed size in the segment header */
|
|
segm->size_on_disk = (uint32_t)offset;
|
|
/*
|
|
* Write the dump segment header itself (1 page, uncompressed) to
|
|
* the predefined location.
|
|
*/
|
|
zero_page = get_zeroed_page();
|
|
writeblock(head_blk, (unsigned long)segm,
|
|
m2b(sizeof(struct df_s390_dump_segm_hdr)), zero_page);
|
|
free_page(zero_page);
|
|
|
|
return blk;
|
|
}
|