linux-kernel - Re: [GIT PULL] Squashfs fixes for 2.6.29?

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Wed, 11 Mar 2009 21:46:50 +0100
From:	Stefan Lippers-Hollmann <s.L-H@....de>
To:	Phillip Lougher <phillip@...gher.demon.co.uk>
Cc:	Phillip Lougher <phil.lougher@...il.com>,
	Geert Uytterhoeven <Geert.Uytterhoeven@...ycom.com>,
	Linux-kernel@...r.kernel.org,
	Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [GIT PULL] Squashfs fixes for 2.6.29?

Hi

On Mittwoch, 11. März 2009, Phillip Lougher wrote:
> Phillip Lougher wrote:
[...]
> The following patch fixes the problems you've experienced with your test
> filesystems (thanks for making them available), and also deals with the corrupted
> filesystems issue.  It correctly works with all the corrupted filesystems sent
> earlier this year.
> 
> The problem arises due to an unexpected corner-case with zlib which the
> corrupted filesystems patch didn't address.  Very occasionally zlib exits
> needing a couple of extra bytes of input (1 and 2 bytes in the test filesystems),
> but with avail_out == 0 and no more output buffers available.  This situation
> was incorrectly flagged as an error by the corrupted filesystems patch. I
> unfortunately didn't anticipate this scenario because it seems contrary to
> expectation that zlib will be still reading input having produced all the
> expected output.  The fix is to not print an error if zlib wants more input
> bytes and there's more input bytes to be read.

I just got home and around to test it with 2.6.29-rc7-git4, no further 
issues with large filesystems (copying + live usage) and it's working 
reliably (after whitespace fixing), thank you a lot. 

> Geert I can put a signed off line from you on the patch if you like (as you
> sent the original fix).  I could add reported and tested lines from both of you?
> They seem to be being used more often and sound a good idea.

Feel free to add the tags as needed, thanks.
Reported-by: Stefan Lippers-Hollmann <s.L-H@....de>
Tested-by: Stefan Lippers-Hollmann <s.L-H@....de>

> Thanks
> 
> Phillip

Regards
	Stefan Lippers-Hollmann

-- 
> diff --git a/fs/squashfs/block.c b/fs/squashfs/block.c
> index 321728f..b85173f 100644
> --- a/fs/squashfs/block.c
> +++ b/fs/squashfs/block.c
> @@ -166,6 +166,22 @@ int squashfs_read_data(struct super_block *sb, void **buffer, u64 index,
> 
>   		bytes = length;
>   		do {
> +			if (msblk->stream.avail_out == 0) {
> +				if (page < pages) {
> +					msblk->stream.next_out = buffer[page++];
> +					msblk->stream.avail_out =
> +								PAGE_CACHE_SIZE;
> +				} else if (msblk->stream.avail_in > 0
> +								|| bytes == 0) {
> +					ERROR("zlib_inflate tried to "
> +						"decompress too much data, "
> +						"expected %d bytes.  Zlib "
> +						"data probably corrupt\n",
> +						srclength);
> +					goto release_mutex;
> +				}
> +			}
> +
>   			if (msblk->stream.avail_in == 0 && k < b) {
>   				avail = min(bytes, msblk->devblksize - offset);
>   				bytes -= avail;
> @@ -184,19 +200,6 @@ int squashfs_read_data(struct super_block *sb, void **buffer, u64 index,
>   				offset = 0;
>   			}
> 
> -			if (msblk->stream.avail_out == 0) {
> -				if (page == pages) {
> -					ERROR("zlib_inflate tried to "
> -						"decompress too much data, "
> -						"expected %d bytes.  Zlib "
> -						"data probably corrupt\n",
> -						srclength);
> -					goto release_mutex;
> -				}
> -				msblk->stream.next_out = buffer[page++];
> -				msblk->stream.avail_out = PAGE_CACHE_SIZE;
> -			}
> -
>   			if (!zlib_init) {
>   				zlib_err = zlib_inflateInit(&msblk->stream);
>   				if (zlib_err != Z_OK) {
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/