[Crash-utility] [PATCH v2] diskdump: Add support for reading dumpfiles compressed by Zstandard

[Philipp Rudo]

Hi Kazu,

the patch follows the snappy and lzo example and all in all the
looks good to me. Two small comments below though. Nevertheless

Reviewed-by: Philipp Rudo <prudo at redhat.com>


On Fri, 17 Sep 2021 07:55:32 +0000
HAGIO KAZUHITO(萩尾　一仁) <k-hagio-ab at nec.com> wrote:

> -----Original Message-----
> > @@ -1002,6 +1006,11 @@ is_diskdump(char *file)
> >  	dd->flags |= SNAPPY_SUPPORTED;
> >  #endif
> > 
> > +#ifdef ZSTD
> > +	if ((dctx = ZSTD_createDCtx()) != NULL)
> > +		dd->flags |= ZSTD_SUPPORTED;  
> 
> I found that this part allocates dctx superfluously for split dumpfiles,
> so sending v2.
> 
> --
> From: Kazuhito Hagio <k-hagio at ab.jp.nec.com>
> Subject: [PATCH] diskdump: Add support for reading dumpfiles compressed by
>  Zstandard
> 
> Add support for reading dumpfiles compressed by Zstandard (zstd)
> using makedumpfile.
> 
> To build crash with zstd support, type "make zstd".
> 
> Signed-off-by: Kazuhito Hagio <k-hagio-ab at nec.com>
> ---
>  Makefile    |  4 ++++
>  README      |  4 ++--
>  configure.c | 20 +++++++++++++++++---
>  defs.h      |  4 ++++
>  diskdump.c  | 33 +++++++++++++++++++++++++++++++++
>  diskdump.h  |  1 +
>  help.c      |  4 ++--
>  7 files changed, 63 insertions(+), 7 deletions(-)

[...]

> --- a/configure.c
> +++ b/configure.c
> @@ -1746,6 +1746,7 @@ add_extra_lib(char *option)

You should probably add a 'For ZSTD:' section in the comment above this
function as well.

>  {
>  	int lzo, add_DLZO, add_llzo2; 
>  	int snappy, add_DSNAPPY, add_lsnappy;
> +	int zstd, add_DZSTD, add_lzstd;
>  	int valgrind, add_DVALGRIND;
>  	char *cflags, *ldflags;
>  	FILE *fp_cflags, *fp_ldflags;

[...]

> --- a/diskdump.c
> +++ b/diskdump.c
> @@ -96,6 +96,10 @@ static struct diskdump_data **dd_list = NULL;
>  static int num_dd = 0;
>  static int num_dumpfiles = 0;
>  
> +#ifdef ZSTD
> +static ZSTD_DCtx *dctx = NULL;
> +#endif
> +
>  int dumpfile_is_split(void)
>  {
>  	return KDUMP_SPLIT();
> @@ -1002,6 +1006,13 @@ is_diskdump(char *file)
>  	dd->flags |= SNAPPY_SUPPORTED;
>  #endif
>  
> +#ifdef ZSTD
> +	if (!dctx)
> +		dctx = ZSTD_createDCtx();
> +	if (dctx)
> +		dd->flags |= ZSTD_SUPPORTED;
> +#endif
> +

would it make sense to only set ZSTD_SUPPORTED here and move the
creation of dctx to cache_page where it is used? That at least would
save a couple of bytes when zstd is supported but the dump is not zstd
compressed.

Thanks
Philipp

>  	pc->read_vmcoreinfo = vmcoreinfo_read_string;
>  
>  	if ((pc->flags2 & GET_LOG) && KDUMP_CMPRS_VALID()) {
> @@ -1251,6 +1262,24 @@ cache_page(physaddr_t paddr)
>  			      ret);
>  			return READ_ERROR;
>  		}
> +#endif
> +	} else if (pd.flags & DUMP_DH_COMPRESSED_ZSTD) {
> +
> +		if (!(dd->flags & ZSTD_SUPPORTED)) {
> +			error(INFO, "%s: uncompess failed: no zstd compression support\n",
> +				DISKDUMP_VALID() ? "diskdump" : "compressed kdump");
> +			return READ_ERROR;
> +		}
> +#ifdef ZSTD
> +		retlen = ZSTD_decompressDCtx(dctx,
> +				dd->page_cache_hdr[i].pg_bufptr, block_size,
> +				dd->compressed_page, pd.size);
> +		if (ZSTD_isError(retlen)) {
> +			error(INFO, "%s: uncompress failed: %d (%s)\n",
> +				DISKDUMP_VALID() ? "diskdump" : "compressed kdump",
> +				retlen, ZSTD_getErrorName(retlen));
> +			return READ_ERROR;
> +		}
>  #endif
>  	} else
>  		memcpy(dd->page_cache_hdr[i].pg_bufptr,
> @@ -1806,6 +1835,8 @@ __diskdump_memory_dump(FILE *fp)
>  		fprintf(fp, "%sLZO_SUPPORTED", others++ ? "|" : "");
>  	if (dd->flags & SNAPPY_SUPPORTED)
>  		fprintf(fp, "%sSNAPPY_SUPPORTED", others++ ? "|" : "");
> +	if (dd->flags & ZSTD_SUPPORTED)
> +		fprintf(fp, "%sZSTD_SUPPORTED", others++ ? "|" : "");
>          fprintf(fp, ") %s\n", FLAT_FORMAT() ? "[FLAT]" : "");
>          fprintf(fp, "               dfd: %d\n", dd->dfd);
>          fprintf(fp, "               ofp: %lx\n", (ulong)dd->ofp);
> @@ -1872,6 +1903,8 @@ __diskdump_memory_dump(FILE *fp)
>  			fprintf(fp, "DUMP_DH_COMPRESSED_LZO");
>  		if (dh->status & DUMP_DH_COMPRESSED_SNAPPY)
>  			fprintf(fp, "DUMP_DH_COMPRESSED_SNAPPY");
> +		if (dh->status & DUMP_DH_COMPRESSED_ZSTD)
> +			fprintf(fp, "DUMP_DH_COMPRESSED_ZSTD");
>  		if (dh->status & DUMP_DH_COMPRESSED_INCOMPLETE)
>  			fprintf(fp, "DUMP_DH_COMPRESSED_INCOMPLETE");
>  		if (dh->status & DUMP_DH_EXCLUDED_VMEMMAP)
> diff --git a/diskdump.h b/diskdump.h
> index 28713407b841..c152c7b86616 100644
> --- a/diskdump.h
> +++ b/diskdump.h
> @@ -86,6 +86,7 @@ struct kdump_sub_header {
>  #define DUMP_DH_COMPRESSED_SNAPPY  0x4   /* page is compressed with snappy */
>  #define DUMP_DH_COMPRESSED_INCOMPLETE  0x8   /* dumpfile is incomplete */
>  #define DUMP_DH_EXCLUDED_VMEMMAP   0x10  /* unused vmemmap pages are excluded */
> +#define DUMP_DH_COMPRESSED_ZSTD    0x20  /* page is compressed with zstd */
>  
>  /* descriptor of each page for vmcore */
>  typedef struct page_desc {
> diff --git a/help.c b/help.c
> index 6c262a3ffcbb..f34838d59908 100644
> --- a/help.c
> +++ b/help.c
> @@ -9420,8 +9420,8 @@ README_ENTER_DIRECTORY,
>  "  Traditionally when vmcores are compressed via the makedumpfile(8) facility",
>  "  the libz compression library is used, and by default the crash utility",
>  "  only supports libz.  Recently makedumpfile has been enhanced to optionally",
> -"  use either the LZO or snappy compression libraries.  To build crash with",
> -"  either or both of those libraries, type \"make lzo\" or \"make snappy\".",
> +"  use the LZO, snappy or zstd compression libraries.  To build crash with any",
> +"  or all of those libraries, type \"make lzo\", \"make snappy\" or \"make zstd\".",
>  "",
>  "  crash supports valgrind Memcheck tool on the crash's custom memory allocator.",
>  "  To build crash with this feature enabled, type \"make valgrind\" and then run",