]> git.lizzy.rs Git - zlib.git/log
zlib.git
5 years agoDon't bother computing check value after successful inflateSync().
Mark Adler [Thu, 3 Jan 2019 02:10:40 +0000 (18:10 -0800)]
Don't bother computing check value after successful inflateSync().

inflateSync() is used to skip invalid deflate data, which means
that the check value that was being computed is no longer useful.
This commit turns off the check value computation, and furthermore
allows a successful return if the compressed data terminated in a
graceful manner. This commit also fixes a bug in the case that
inflateSync() is used before a header is ever processed. In that
case, there is no knowledge of a trailer, so the remainder is
treated as raw.

5 years agoUse atomic test and set, if available, for dynamic CRC tables.
Mark Adler [Wed, 26 Dec 2018 21:50:27 +0000 (13:50 -0800)]
Use atomic test and set, if available, for dynamic CRC tables.

5 years agoSpeed up software CRC-32 computation by a factor of 1.5 to 3.
Mark Adler [Tue, 11 Dec 2018 09:11:38 +0000 (01:11 -0800)]
Speed up software CRC-32 computation by a factor of 1.5 to 3.

Use the interleaved method of Kadatch and Jenkins in order to make
use of pipelined instructions through multiple ALUs in a single
core. This also speeds up and simplifies the combination of CRCs,
and updates the functions to pre-calculate and use an operator for
CRC combination.

5 years agoAdd crc32_combine_gen() and crc32_combine_op() for fast combines.
Mark Adler [Sun, 4 Nov 2018 18:31:46 +0000 (10:31 -0800)]
Add crc32_combine_gen() and crc32_combine_op() for fast combines.

When the same len2 is used repeatedly, it is faster to use
crc32_combine_gen() to generate an operator, that is then used to
combine CRCs with crc32_combine_op().

5 years agoAdd tables for crc32_combine(), to speed it up by a factor of 200.
Mark Adler [Sat, 3 Nov 2018 05:55:14 +0000 (22:55 -0700)]
Add tables for crc32_combine(), to speed it up by a factor of 200.

5 years agoFix the zran.c example to work on a multiple-member gzip file.
Mark Adler [Mon, 8 Oct 2018 04:26:43 +0000 (21:26 -0700)]
Fix the zran.c example to work on a multiple-member gzip file.

5 years agoAdd gznorm.c example, which normalizes gzip files.
Mark Adler [Sat, 6 Oct 2018 06:06:36 +0000 (23:06 -0700)]
Add gznorm.c example, which normalizes gzip files.

5 years agoShow all the codes for the maximum tables size in enough.c.
Mark Adler [Sat, 4 Aug 2018 21:34:01 +0000 (14:34 -0700)]
Show all the codes for the maximum tables size in enough.c.

5 years agoClarify that prefix codes are counted in enough.c.
Mark Adler [Sat, 4 Aug 2018 21:27:02 +0000 (14:27 -0700)]
Clarify that prefix codes are counted in enough.c.

There is no assurance that all prefix codes are reachable as
optimal Huffman codes for the numbers of symbols encountered in
a deflate block. This code considers all possible prefix codes,
which might be a larger set than all possible Huffman codes,
depending on the constraints.

5 years agoUse inline function instead of macro for index in enough.c.
Mark Adler [Thu, 2 Aug 2018 01:24:43 +0000 (18:24 -0700)]
Use inline function instead of macro for index in enough.c.

5 years agoClean up code style in enough.c, update version.
Mark Adler [Wed, 1 Aug 2018 08:49:45 +0000 (01:49 -0700)]
Clean up code style in enough.c, update version.

5 years agoUse a macro for the printf format of big_t in enough.c.
Mark Adler [Wed, 1 Aug 2018 08:39:47 +0000 (01:39 -0700)]
Use a macro for the printf format of big_t in enough.c.

5 years agoUse a structure to make globals in enough.c evident.
Mark Adler [Wed, 1 Aug 2018 08:37:03 +0000 (01:37 -0700)]
Use a structure to make globals in enough.c evident.

6 years agoAssure that the number of bits for deflatePrime() is valid.
Mark Adler [Wed, 18 Apr 2018 05:44:41 +0000 (22:44 -0700)]
Assure that the number of bits for deflatePrime() is valid.

6 years agoFix a bug that can crash deflate on some input when using Z_FIXED.
Mark Adler [Wed, 18 Apr 2018 05:09:22 +0000 (22:09 -0700)]
Fix a bug that can crash deflate on some input when using Z_FIXED.

This bug was reported by Danilo Ramos of Eideticom, Inc. It has
lain in wait 13 years before being found! The bug was introduced
in zlib 1.2.2.2, with the addition of the Z_FIXED option. That
option forces the use of fixed Huffman codes. For rare inputs with
a large number of distant matches, the pending buffer into which
the compressed data is written can overwrite the distance symbol
table which it overlays. That results in corrupted output due to
invalid distances, and can result in out-of-bound accesses,
crashing the application.

The fix here combines the distance buffer and literal/length
buffers into a single symbol buffer. Now three bytes of pending
buffer space are opened up for each literal or length/distance
pair consumed, instead of the previous two bytes. This assures
that the pending buffer cannot overwrite the symbol table, since
the maximum fixed code compressed length/distance is 31 bits, and
since there are four bytes of pending space for every three bytes
of symbol space.

6 years agoCorrect the initialization requirements for deflateInit2().
Mark Adler [Wed, 31 Jan 2018 21:24:12 +0000 (13:24 -0800)]
Correct the initialization requirements for deflateInit2().

6 years agoEmphasize the need to continue decompressing gzip members.
Mark Adler [Tue, 9 Jan 2018 02:16:05 +0000 (18:16 -0800)]
Emphasize the need to continue decompressing gzip members.

6 years agoAdd legal disclaimer to README.
Mark Adler [Fri, 13 Oct 2017 04:57:51 +0000 (21:57 -0700)]
Add legal disclaimer to README.

6 years agoFix deflateEnd() to not report an error at start of raw deflate.
Mark Adler [Fri, 13 Oct 2017 04:07:22 +0000 (21:07 -0700)]
Fix deflateEnd() to not report an error at start of raw deflate.

6 years agoRemove old assembler code in which bugs have manifested.
Mark Adler [Fri, 13 Oct 2017 03:08:53 +0000 (20:08 -0700)]
Remove old assembler code in which bugs have manifested.

In addition, there is not sufficient gain from the inflate
assembler code to warrant its inclusion.

6 years agoMake the names in functions declarations identical to definitions.
Mark Adler [Fri, 13 Oct 2017 03:03:51 +0000 (20:03 -0700)]
Make the names in functions declarations identical to definitions.

6 years agoAvoid an undefined behavior of memcpy() in _tr_stored_block().
Mark Adler [Fri, 13 Oct 2017 02:44:01 +0000 (19:44 -0700)]
Avoid an undefined behavior of memcpy() in _tr_stored_block().

Allegedly the behavior of memcpy() is undefined if the source
pointer is NULL, even if the number of bytes to copy is zero.

6 years agoAvoid undefined behaviors of memcpy() in gz*printf().
Mark Adler [Fri, 13 Oct 2017 02:34:51 +0000 (19:34 -0700)]
Avoid undefined behaviors of memcpy() in gz*printf().

6 years agoAvoid an undefined behavior of memcpy() in gzappend().
Mark Adler [Fri, 13 Oct 2017 02:27:59 +0000 (19:27 -0700)]
Avoid an undefined behavior of memcpy() in gzappend().

6 years agoAvoid the use of ptrdiff_t.
Mark Adler [Sat, 3 Jun 2017 16:49:39 +0000 (09:49 -0700)]
Avoid the use of ptrdiff_t.

This isn't the right type anyway to assure that it contains a
pointer. That type would be intptr_t or uintptr_t. However the C99
standard says that those types are optional, so their use would not
be portable. This commit simply uses size_t or whatever configure
decided to use for size_t. That would be the same length as
ptrdiff_t, and so will work just as well. The code checks to see if
the length of the type used is the same as the length of a void
pointer, so there is already protection against the use of the
wrong type. The use of size_t (or ptrdiff_t) will almost always
work, as all modern architectures have an array size that is the
same as the pointer size. Only old segmented architectures would
have to fall back to the slower CRC-32 calculation, where the
amount of memory that can be accessed is larger than the maximum
array size.

7 years agoHandle case where inflateSync used when header never processed.
Mark Adler [Sun, 16 Apr 2017 15:35:33 +0000 (08:35 -0700)]
Handle case where inflateSync used when header never processed.

If zlib and/or gzip header processing was requested, but a header
was never provided and inflateSync was used successfully, then the
inflate state would be inconsistent, trying to compute a check
value but with no flags set. This commit sets the inflate mode to
raw in this case, since there is no other assumption that can be
made if a header was requested but never seen.

7 years agoDon't compute check value for raw inflate if asked to validate.
Mark Adler [Thu, 30 Mar 2017 21:48:43 +0000 (14:48 -0700)]
Don't compute check value for raw inflate if asked to validate.

7 years agoAdd address checking in clang to -w option of configure.
Mark Adler [Sun, 19 Feb 2017 07:07:01 +0000 (23:07 -0800)]
Add address checking in clang to -w option of configure.

7 years agoReturn an error if the gzputs string length can't fit in an int.
Mark Adler [Sun, 12 Feb 2017 07:54:17 +0000 (23:54 -0800)]
Return an error if the gzputs string length can't fit in an int.

7 years agoSmall speedup to inflate [psumbera].
Mark Adler [Sun, 12 Feb 2017 07:21:41 +0000 (23:21 -0800)]
Small speedup to inflate [psumbera].

Seeing a few percent speedup by using a pointer instead of an
assigned structure. This seems to help the compiler to optimize
better.

7 years agoUpdate use of errno for newer Windows CE versions.
Mark Adler [Sun, 12 Feb 2017 06:48:31 +0000 (22:48 -0800)]
Update use of errno for newer Windows CE versions.

7 years agoAvoid some conversion warnings in gzread.c and gzwrite.c.
Mark Adler [Sun, 12 Feb 2017 06:45:27 +0000 (22:45 -0800)]
Avoid some conversion warnings in gzread.c and gzwrite.c.

7 years agoHave Makefile return non-zero error code on test failure.
Mark Adler [Sun, 12 Feb 2017 06:38:09 +0000 (22:38 -0800)]
Have Makefile return non-zero error code on test failure.

7 years agoAvoid a conversion error in gzseek when off_t type too small.
Mark Adler [Sun, 5 Feb 2017 07:58:37 +0000 (23:58 -0800)]
Avoid a conversion error in gzseek when off_t type too small.

This is a problem in the odd case that the second argument of
LSEEK is a larger type than off_t. Apparently MinGW defines off_t
to be 32 bits, but _lseeki64 has a 64-bit second argument.

Also undo a previous commit to permit MinGW to use _lseeki64.

7 years agoFix CLEAR_HASH macro to be usable as a single statement.
Mark Adler [Mon, 23 Jan 2017 07:38:52 +0000 (23:38 -0800)]
Fix CLEAR_HASH macro to be usable as a single statement.

As it is used in deflateParams().

7 years agoFix bug when window full in deflate_stored().
Mark Adler [Sat, 21 Jan 2017 20:13:25 +0000 (12:13 -0800)]
Fix bug when window full in deflate_stored().

7 years agoLimit hash table inserts after switch from stored deflate.
Mark Adler [Sat, 21 Jan 2017 09:50:26 +0000 (01:50 -0800)]
Limit hash table inserts after switch from stored deflate.

This limits hash table inserts to the available data in the window
and to the sliding window size in deflate_stored(). The hash table
inserts are deferred until deflateParams() switches to a non-zero
compression level.

7 years agoPermit a deflateParams() parameter change as soon as possible.
Mark Adler [Mon, 16 Jan 2017 17:49:35 +0000 (09:49 -0800)]
Permit a deflateParams() parameter change as soon as possible.

This commit allows a parameter change even if the input data has
not all been compressed and copied to the application output
buffer, so long as all of the input data has been compressed to
the internal pending output buffer. This also allows an immediate
deflateParams change so long as there have been no deflate calls
since initialization or reset.

7 years agoCygwin does not have _wopen(), so do not create gzopen_w() there.
Mark Adler [Mon, 16 Jan 2017 17:38:36 +0000 (09:38 -0800)]
Cygwin does not have _wopen(), so do not create gzopen_w() there.

7 years agoChange version number to 1.2.11.1.
Mark Adler [Mon, 16 Jan 2017 06:46:03 +0000 (22:46 -0800)]
Change version number to 1.2.11.1.

7 years agozlib 1.2.11 v1.2.11
Mark Adler [Sun, 15 Jan 2017 17:18:46 +0000 (09:18 -0800)]
zlib 1.2.11

7 years agoPermit immediate deflateParams changes before any deflate input.
Mark Adler [Sun, 15 Jan 2017 16:22:16 +0000 (08:22 -0800)]
Permit immediate deflateParams changes before any deflate input.

This permits deflateParams to change the strategy and level right
after deflateInit, without having to wait until a header has been
written. The parameters can be changed immediately up until the
first deflate call that consumes any input data.

7 years agoUpdate high water mark in deflate_stored.
Mark Adler [Sun, 15 Jan 2017 16:15:55 +0000 (08:15 -0800)]
Update high water mark in deflate_stored.

This avoids unnecessary filling of bytes in the sliding window
buffer when switching from level zero to a non-zero level. This
also provides a consistent indication of deflate having taken
input for a later commit ...

7 years agoUpdate vestigial comment from very old Info-ZIP deflate.
Mark Adler [Sat, 14 Jan 2017 07:10:03 +0000 (23:10 -0800)]
Update vestigial comment from very old Info-ZIP deflate.

7 years agoFix deflate stored bug when pulling last block from window.
Mark Adler [Fri, 13 Jan 2017 05:51:20 +0000 (21:51 -0800)]
Fix deflate stored bug when pulling last block from window.

And some cosmetic cleanups.

7 years agoUpdate location of Visual Studio project files.
Mark Adler [Thu, 5 Jan 2017 07:18:55 +0000 (23:18 -0800)]
Update location of Visual Studio project files.

7 years agoDelete user-specific Visual Studio project files.
Mark Adler [Thu, 5 Jan 2017 06:56:55 +0000 (22:56 -0800)]
Delete user-specific Visual Studio project files.

7 years agoChange version number to 1.2.10.1.
Mark Adler [Sun, 15 Jan 2017 16:54:04 +0000 (08:54 -0800)]
Change version number to 1.2.10.1.

7 years agozlib 1.2.10 v1.2.10
Mark Adler [Tue, 3 Jan 2017 01:53:59 +0000 (17:53 -0800)]
zlib 1.2.10

7 years agoFix compilation with --solo and --debug combined.
Mark Adler [Tue, 3 Jan 2017 02:18:31 +0000 (18:18 -0800)]
Fix compilation with --solo and --debug combined.

However this ends up not really being solo, since it has to
include external libraries.

7 years agoAdd warnings when compiling with assembler code.
Mark Adler [Tue, 3 Jan 2017 01:25:27 +0000 (17:25 -0800)]
Add warnings when compiling with assembler code.

There have been many reports of bugs in the assembler codes
intended to speed up deflate and inflate. They are third-party
contributions in contrib, and so are not supported by the zlib
maintainers.

7 years agoRemove files to be installed before copying them in Makefile.in.
Mark Adler [Tue, 3 Jan 2017 01:04:38 +0000 (17:04 -0800)]
Remove files to be installed before copying them in Makefile.in.

7 years agoFix bug in gzwrite.c that produced corrupt gzip files.
Mark Adler [Tue, 3 Jan 2017 00:17:43 +0000 (16:17 -0800)]
Fix bug in gzwrite.c that produced corrupt gzip files.

7 years agoFix bug in deflate_stored() for zero-length input.
Mark Adler [Mon, 2 Jan 2017 23:17:59 +0000 (15:17 -0800)]
Fix bug in deflate_stored() for zero-length input.

7 years agoMinor edits and clarifications of comments.
Mark Adler [Mon, 2 Jan 2017 04:02:24 +0000 (20:02 -0800)]
Minor edits and clarifications of comments.

7 years agoAvoid warnings on snprintf() return value.
Mark Adler [Sun, 1 Jan 2017 20:23:04 +0000 (12:23 -0800)]
Avoid warnings on snprintf() return value.

7 years agoChange version number to zlib 1.2.9.1.
Mark Adler [Mon, 2 Jan 2017 06:21:21 +0000 (22:21 -0800)]
Change version number to zlib 1.2.9.1.

7 years agoFix some stray 1.2.8.1 version numbers.
Mark Adler [Mon, 2 Jan 2017 06:12:12 +0000 (22:12 -0800)]
Fix some stray 1.2.8.1 version numbers.

7 years agozlib 1.2.9 v1.2.9
Mark Adler [Sat, 31 Dec 2016 19:41:31 +0000 (11:41 -0800)]
zlib 1.2.9

7 years agoUpdate Visual Studio project files (AraHaan).
Mark Adler [Sun, 1 Jan 2017 02:51:34 +0000 (18:51 -0800)]
Update Visual Studio project files (AraHaan).

7 years agoAdd crc32_z() and adler32_z() functions with size_t lengths.
Mark Adler [Sun, 1 Jan 2017 00:57:26 +0000 (16:57 -0800)]
Add crc32_z() and adler32_z() functions with size_t lengths.

7 years agoMake z_size_t unsigned long for non-standard C.
Mark Adler [Sun, 1 Jan 2017 00:41:36 +0000 (16:41 -0800)]
Make z_size_t unsigned long for non-standard C.

Also declare z_size_t when compiling solo.

7 years agoAvoid the need for ssize_t.
Mark Adler [Sat, 31 Dec 2016 18:03:09 +0000 (10:03 -0800)]
Avoid the need for ssize_t.

Limit read() and write() requests to sizes that fit in an int.
This allows storing the return value in an int, and avoiding the
need to use or construct an ssize_t type. This is required for
Microsoft C, whose _read and _write functions take an unsigned
request and return an int.

7 years agoUse a uniform approach for the largest value of an unsigned type.
Mark Adler [Sat, 31 Dec 2016 16:49:17 +0000 (08:49 -0800)]
Use a uniform approach for the largest value of an unsigned type.

7 years agoUse intptr_t for z_ssize_t on MSVC.
Mark Adler [Sat, 31 Dec 2016 07:40:43 +0000 (23:40 -0800)]
Use intptr_t for z_ssize_t on MSVC.

7 years agoAvoid some random compiler warnings on various platforms.
Mark Adler [Sat, 31 Dec 2016 06:05:05 +0000 (22:05 -0800)]
Avoid some random compiler warnings on various platforms.

7 years agoAllow minigzip to compile when testing with ./configure --solo.
Mark Adler [Sat, 31 Dec 2016 06:04:22 +0000 (22:04 -0800)]
Allow minigzip to compile when testing with ./configure --solo.

7 years agoReplace as400 with os400 for OS/400 support (Monnerat).
Mark Adler [Sat, 31 Dec 2016 03:48:07 +0000 (19:48 -0800)]
Replace as400 with os400 for OS/400 support (Monnerat).

7 years agoDetect clang in cc version.
Mark Adler [Sat, 31 Dec 2016 02:58:46 +0000 (18:58 -0800)]
Detect clang in cc version.

7 years agoFix init macros to use z_ prefix when requested.
Mark Adler [Sat, 31 Dec 2016 02:18:48 +0000 (18:18 -0800)]
Fix init macros to use z_ prefix when requested.

7 years agoFix character encoding and link in contrib README.
Mark Adler [Sat, 31 Dec 2016 01:20:44 +0000 (17:20 -0800)]
Fix character encoding and link in contrib README.

7 years agoUse snprintf() for later versions of Microsoft C.
Mark Adler [Sat, 31 Dec 2016 01:13:31 +0000 (17:13 -0800)]
Use snprintf() for later versions of Microsoft C.

7 years agoAdd deflateGetDictionary() function.
Mark Adler [Sat, 31 Dec 2016 00:29:56 +0000 (16:29 -0800)]
Add deflateGetDictionary() function.

Per request, but its utility is likely to be very limited. See the
comments in zlib.h.

7 years agoNo need to check for NULL argument to free().
Mark Adler [Fri, 30 Dec 2016 22:30:52 +0000 (14:30 -0800)]
No need to check for NULL argument to free().

7 years agoAdd gzfwrite(), duplicating the interface of fwrite().
Mark Adler [Mon, 5 Dec 2016 02:29:43 +0000 (18:29 -0800)]
Add gzfwrite(), duplicating the interface of fwrite().

7 years agoAdd gzfread(), duplicating the interface of fread().
Mark Adler [Mon, 5 Dec 2016 01:09:33 +0000 (17:09 -0800)]
Add gzfread(), duplicating the interface of fread().

7 years agoFix compile option for when z_size_t needs to be a long long.
Mark Adler [Mon, 5 Dec 2016 00:50:49 +0000 (16:50 -0800)]
Fix compile option for when z_size_t needs to be a long long.

7 years agoCreate z_size_t and z_ssize_t types.
Mark Adler [Sat, 3 Dec 2016 18:27:14 +0000 (10:27 -0800)]
Create z_size_t and z_ssize_t types.

Normally these are set to size_t and ssize_t. But if they do not
exist, then they are set to the smallest integer type that can
contain a pointer. size_t is unsigned and ssize_t is signed.

7 years agoDon't need to emit an empty fixed block when changing parameters.
Mark Adler [Sat, 3 Dec 2016 16:29:57 +0000 (08:29 -0800)]
Don't need to emit an empty fixed block when changing parameters.

gzsetparams() was using Z_PARTIAL_FLUSH when it could use Z_BLOCK
instead. This commit uses Z_BLOCK, which avoids emitting an
unnecessary ten bits into the stream.

7 years agoClean up gz* function return values.
Mark Adler [Sat, 3 Dec 2016 16:18:56 +0000 (08:18 -0800)]
Clean up gz* function return values.

In some cases the return values did not match the documentation,
or the documentation did not document all of the return values.
gzprintf() now consistently returns negative values on error,
which matches the behavior of the stdio fprintf() function.

7 years agoSpeed up deflation for level 0 (storing).
Mark Adler [Sat, 5 Nov 2016 15:43:29 +0000 (08:43 -0700)]
Speed up deflation for level 0 (storing).

The previous code slid the window and the hash table and copied
every input byte three times in order to just write the data as
stored blocks with no compression. This commit minimizes sliding
and copying, especially for large input and output buffers.

Level 0 compression is now more than 20 times faster than before
the commit.

Most of the speedup is due to deferring hash table slides until
deflateParams() is called to change the compression level away
from 0. More speedup is due to copying directly from next_in to
next_out when the amounts of available input data and output space
permit it, avoiding the intermediate pending buffer. Additionally,
only the last 32K of the used input data is copied back to the
sliding window when large input buffers are provided.

7 years agoAssure that deflateParams() will not switch functions mid-block.
Mark Adler [Wed, 23 Nov 2016 07:29:19 +0000 (23:29 -0800)]
Assure that deflateParams() will not switch functions mid-block.

This alters the specification in zlib.h, so that deflateParams()
will not change any parameters if there is not enough output space
in the event that a block is emitted in order to allow switching
the compression function.

7 years agoExplicitly ignore a return value in gzwrite.c.
Mark Adler [Tue, 22 Nov 2016 20:02:29 +0000 (12:02 -0800)]
Explicitly ignore a return value in gzwrite.c.

7 years agoIncrease verbosity required to warn about bit length overflow.
Mark Adler [Sun, 20 Nov 2016 19:36:15 +0000 (11:36 -0800)]
Increase verbosity required to warn about bit length overflow.

When debugging the Huffman coding would warn about resulting codes
greater than 15 bits in length. This is handled properly, and is
not uncommon. This increases the verbosity of the warning by one,
so that it is not displayed by default.

7 years agoAdd uncompress2() function, which returns the input size used.
Mark Adler [Wed, 16 Nov 2016 03:45:01 +0000 (20:45 -0700)]
Add uncompress2() function, which returns the input size used.

7 years agoMinor edits to the documentation in source file contents.
Mark Adler [Mon, 14 Nov 2016 17:19:25 +0000 (10:19 -0700)]
Minor edits to the documentation in source file contents.

7 years agoFix bugs in creating a very large gzip header.
Mark Adler [Sun, 6 Nov 2016 05:55:34 +0000 (22:55 -0700)]
Fix bugs in creating a very large gzip header.

7 years agoAdd --debug (-d) option to ./configure to define ZLIB_DEBUG.
Mark Adler [Sun, 30 Oct 2016 16:33:25 +0000 (09:33 -0700)]
Add --debug (-d) option to ./configure to define ZLIB_DEBUG.

7 years agoUse memcpy for stored blocks.
Mark Adler [Sun, 30 Oct 2016 16:25:32 +0000 (09:25 -0700)]
Use memcpy for stored blocks.

This speeds up level 0 by about a factor of three, as compared to
the previous byte-at-a-time loop. We can do much better though. A
later commit avoids this copy for level 0 with large buffers,
instead copying directly from the input to the output. This commit
still speeds up storing incompressible data found when compressing
normally.

7 years agoFix some typos.
Mark Adler [Sun, 30 Oct 2016 15:36:13 +0000 (08:36 -0700)]
Fix some typos.

7 years agoFix bug when level 0 used with Z_HUFFMAN or Z_RLE.
Mark Adler [Fri, 28 Oct 2016 05:50:43 +0000 (22:50 -0700)]
Fix bug when level 0 used with Z_HUFFMAN or Z_RLE.

Compression level 0 requests no compression, using only stored
blocks. When Z_HUFFMAN or Z_RLE was used with level 0 (granted,
an odd choice, but permitted), the resulting blocks were mostly
fixed or dynamic. The reason is that deflate_stored() was not
being called in that case. The compressed data was valid, but it
was not what the application requested. This commit assures that
only stored blocks are emitted for compression level 0, regardless
of the strategy selected.

7 years agoClean up and comment the use of local for static.
Mark Adler [Wed, 26 Oct 2016 17:25:10 +0000 (10:25 -0700)]
Clean up and comment the use of local for static.

7 years agoMake a noble effort at setting OS_CODE correctly.
Mark Adler [Wed, 26 Oct 2016 03:45:41 +0000 (20:45 -0700)]
Make a noble effort at setting OS_CODE correctly.

This updates the OS_CODE determination at compile time to match as
closely as possible the operating system mappings documented in
the PKWare APPNOTE.TXT version 6.3.4, section 4.4.2.2. That byte
in the gzip header is used by nobody for anything, as far as I can
tell. However we might as well try to set it appropriately.

7 years agoDo a more thorough check of the state for every stream call.
Mark Adler [Tue, 25 Oct 2016 03:11:41 +0000 (20:11 -0700)]
Do a more thorough check of the state for every stream call.

This verifies that the state has been initialized, that it is the
expected type of state, deflate or inflate, and that at least the
first several bytes of the internal state have not been clobbered.

7 years agoDocument the rejection of 256-byte window requests in zlib.h.
Mark Adler [Mon, 24 Oct 2016 23:00:51 +0000 (16:00 -0700)]
Document the rejection of 256-byte window requests in zlib.h.

7 years agoReject a window size of 256 bytes if not using the zlib wrapper.
Mark Adler [Mon, 24 Oct 2016 22:52:19 +0000 (15:52 -0700)]
Reject a window size of 256 bytes if not using the zlib wrapper.

There is a bug in deflate for windowBits == 8 (256-byte window).
As a result, zlib silently changes a request for 8 to a request
for 9 (512-byte window), and sets the zlib header accordingly so
that the decompressor knows to use a 512-byte window. However if
deflateInit2() is used for raw deflate or gzip streams, then there
is no indication that the request was not honored, and the
application might assume that it can use a 256-byte window when
decompressing. This commit returns an error if the user requests
a 256-byte window when using raw deflate or gzip encoding.

7 years agoAvoid obfuscating use of default case in inftrees.c.
Mark Adler [Fri, 14 Oct 2016 20:30:18 +0000 (13:30 -0700)]
Avoid obfuscating use of default case in inftrees.c.

7 years agoMove macro definition in deflate.c to where it is used.
Mark Adler [Fri, 14 Oct 2016 20:18:58 +0000 (13:18 -0700)]
Move macro definition in deflate.c to where it is used.

This avoid defining a macro that is never used when not debugging.

7 years agoAvoid recursive gzgetc() macro call.
Mark Adler [Fri, 14 Oct 2016 20:16:07 +0000 (13:16 -0700)]
Avoid recursive gzgetc() macro call.

Recursive macro calls are normally caught by the preprocessor and
avoided. This commit avoids the possibility of a problem entirely.

7 years agoMake globals in examples local to compilation unit.
Mark Adler [Fri, 14 Oct 2016 20:10:54 +0000 (13:10 -0700)]
Make globals in examples local to compilation unit.