Commit Graph

30 Commits

Author SHA1 Message Date
Xin LI
dec32474f0 r281540 was upstreamed as NetBSD r1.108 of gzip.c, note it as merged. 2015-04-16 22:30:57 +00:00
Xin LI
c1c4677aec When reading in the original file name from gzip header, we read
in PATH_MAX + 1 bytes from the file.  In r281500, strrchr() is
used to strip possible path portion of the file name to mitigate
a possible attack.  Unfortunately, strrchr() expects a buffer
that is NUL-terminated, and since we are processing potentially
untrusted data, we can not assert that be always true.

Solve this by reading in one less byte (now PATH_MAX) and
explicitly terminate the buffer after the read size with NUL.

Reported by:	Coverity
CID:		1264915
X-MFC-with:	281500
MFC after:	13 days
2015-04-15 00:07:21 +00:00
Xin LI
0eeac58930 Sync with NetBSD:
- Mention xz(1) in gzip(1).
 - Strip away path from header name when decompressing.

MFC after:	2 weeks
2015-04-13 19:46:30 +00:00
Jilles Tjoelker
d6b3ef634c compress,gzip,xz: Preserve timestamps with nanosecond precision. 2015-02-17 13:12:54 +00:00
Xin LI
060ea80ecb Sync with NetBSD.
MFC after:	2 weeks
2014-10-23 01:22:29 +00:00
Xin LI
140c037a0c Incorporate recent changes from NetBSD. Most notable change is the addition
of support of decompressing xz files.

Obtained from:	NetBSD
2011-10-10 06:37:32 +00:00
Xin LI
5106ce89d0 Fix a regression introduced with previous changeset: if output is stdout,
do not check for symbolic link.
2011-05-25 18:04:11 +00:00
Xin LI
fc26891004 Match symbolic link handling behavior with GNU gzip, bzip2 and xz:
When we are operating on a symbolic link pointing to an existing
file, bail out by default, but go ahead if -f is specified.

Submitted by:	arundel
MFC after:	2 weeks
2011-05-23 09:40:21 +00:00
Xin LI
be07528c53 Diff reduction against NetBSD. The most notable change is to zdiff(1) to
handle more file formats including bzip2 and xz.

MFC after:	2 weeks
2011-05-23 09:02:44 +00:00
Benedict Reuschling
c4fadc2b78 Correct some typos in comments, no functional changes. 2010-10-16 15:24:04 +00:00
Xin LI
fb2924e4ac In the past gunzip(1) write()'s after each inflate return. This is
not optimal from a performance standpoint since the write buffer is
not necessarily be filled up when the inflate rountine reached the
end of input buffer and it's not the end of file.

This problem gets uncovered by trying to pipe gunzip -c output to
a GEOM device directly, which enforces the writes be multiple of
sector size.

Sponsored by:	iXsystems, Inc.
Reported by:	jpaetzel
MFC after:	2 weeks
2010-09-23 01:24:33 +00:00
Xin LI
08b300a1b7 Several style fixes as prompted by bde@.
While I'm there, loosen the st_nlink check and fix grammar for 1 extra
links.
2010-06-10 20:59:28 +00:00
Xin LI
1371ad1594 Follow up revision 208888 with purely ident changes.
MFC after:	1 week
2010-06-07 10:10:44 +00:00
Xin LI
09e979e915 Correct a bug in gzip(1): make sure that initialize isb with fstat() on
input file before using it.

PR:		bin/147275
Submitted by:	thomas
MFC after:	1 week
2010-06-07 10:09:40 +00:00
Xin LI
637bfa488b Use _exit(2) system call directly instead of using exit(3) in signal
handler, as the latter is not guaranteed to be signal safe, and we
do not really care about flushing the stream during SIGINT.

Suggested by:	Maxim Konovalov <maxim.konovalov gmail com>
MFC after:	13 days
2010-04-27 17:57:04 +00:00
Xin LI
18333f544e Add a signal handler for SIGINT which removes output file when necessary.
While I'm there, move unlink_input() slightly down to after closing the
output file, in uncompression path.

MFC after:	2 weeks
2010-04-26 20:05:48 +00:00
Xin LI
feeb03b2c8 Diff reduction against NetBSD and add myself to AUTHORS section of the
manual page as I wrote the unpack functionality.  No actual executable
code change verified with md5(1).
2010-04-07 22:54:53 +00:00
Ed Schouten
99742a231f Change all our own code to use st_*tim instead of st_*timespec.
Also remove some local patches to diff(1) which are now unneeded.
2010-03-28 13:16:08 +00:00
Xin LI
deb6327793 Correct a stack underflow in gzip:
- Limit suffix to be no more than 30 bytes long.  This matches GNU
   behavior.
 - Correct usage of memcpy().

Note that this commit only corrects the stack underflow issue, we
still need some other fixes to cover other edges. [1]

Reported by:	Ron Jude <ronj wytheville org>
Discussed with:	Matthew Green (original NetBSD gzip author),
		Eygene Ryabinkin <rea-fbsd codelabs ru> [1]
Approved by:	re (kib)
2009-07-31 08:37:27 +00:00
Xin LI
b47ed83de8 Sync with NetBSD:
- gzip -n does not store timestamp; [1]
 - Reduce diff against NetBSD by moving some casts in our local
   versions.

PR:		bin/134955
Obtained from:	NetBSD
MFC after:	1 month
2009-06-24 21:51:02 +00:00
Xin LI
4ca8a62bb3 Add support for uncompressing pack(1)'ed files. Pack(1) is a program found
in some commercial Unix systems, which utilizes Huffman minimum redundancy
code tree to compress files.  This implementation supports the "new" pack
format only, just like GNU gzip did.

Thanks for oliver@'s archive set which I can test against, and Mingyan Guo
for providing helpful review of my code.

PR:		bin/109567
MFC after:	1 month
2009-06-21 09:39:43 +00:00
Xin LI
80b807fef2 Two fixes for SMALL case when compiling with WARNS=6:
- Reduce scope where return value can be referenced.
 - Add a dummy access to timestamp to silence warning.

Submitted by:	Mingyan Guo <guomingyan gmail com>
2009-06-19 19:28:21 +00:00
Colin Percival
80d61a6a48 Mark functions as __dead2 in order to help the LLVM static checker
understand which code paths aren't possible.

Reported by:	edwin
2008-08-04 07:36:53 +00:00
Xin LI
bf15662677 Sync with NetBSD's license changes. 2008-07-23 00:03:36 +00:00
Robert Watson
3c23548899 Teach gunzip that .tbz and .tbz2 uncompress to .tar, in a manner similar to
its existing understanding that .tgz uncompresses to .tar.

MFC after:	3 days
PR:		121519
Submitted by:	Ighighi <ighighi at gmail.com>
2008-03-09 13:49:09 +00:00
Robert Watson
c7b65254c4 When gzipping to a target file, only set the source file flags on the
target file after the timestamp has been set; otherwise setting the
timestamp will fail if the flags don't permit it (i.e., uchg).

MFC after:	1 week
PR:		120208
Submitted by:	Ighighi <ighighi at gmail.com>
2008-03-09 11:16:42 +00:00
Xin LI
78b8781ab1 - Simulate GNU gzip(1) behavior where full filename is
being output in verbose mode when doing recursive[1].
 - Use better representation of S:

PR:		bin/114470
Submitted by:	Ighighi <ighighi gmail com> [1]
Approved by:	re (hrs)
2007-07-12 01:17:14 +00:00
Xin LI
2a8e7ac9c8 Add a '-k' flag which is similar with its bzip2(1) counterpart,
meaning that the user wants the input file to be left intact.

Feature request: Ighighi <ighighi gmail.com>
PR:		 bin/103006
MFC after:	 1 month
2007-05-28 08:20:46 +00:00
Xin LI
161c3dc4b0 Fix typo.
Obtained from:	DragonFly
2007-01-31 07:13:25 +00:00
Xin LI
9a9ea25f4a Replace the GNU gzip with a slightly modified NetBSD gzip. The
NetBSD version is a feature-to-feature re-implementation of GNU
gzip using the freely-redistributable zlib and this version is
expected to be mostly bug-to-bug compatible with the GNU
implementation.

 - Because this is a piece of mature code and we want to make
   changes so it is added directly rather than importing to
   src/contrib.
 - Connect newly added code to src/usr.bin/ and rescue/rescue
   build.
 - Disconnect the GNU gzip code from build for now, they will
   be eventually removed completely.
 - Provide two new src.conf(5) knobs, WITHOUT_BZIP2_SUPPORT and
   WITHOUT_BZIP2.

Tested by:	kris (full exp-7 pointyhat build)
Approved by:	core (importing a 4-clause BSD licensed file)
Approved by:	re (adding new utility during -HEAD code slush)
2007-01-26 10:19:08 +00:00