Apple removed support for gcov in the default gcc compiler chain,
when they moved to llvm. This can be circumvented in XCode 4.2 by
using the gcc chain with gcc-4.2. This patch allows setting
GCC_CLASSIC to the name of a real gcc executable (e.g. "gcc-4.2")
to allow coverage testing.
crc32.c was #including limits.h in order to find a four-byte integer
type. It was doing this even if Z_SOLO were defined, violating the
intent of Z_SOLO, which is to include no library headers and require
no library functions. Now crc32.c obeys the intent of Z_SOLO, but
with the downside that crc32() will be slower than when not compiled
with Z_SOLO. This can be remedied manually by typedefing u4 to a
known four-byte unsigned integer type, and #defining BYFOUR in
crc32.c.
gzflags() was put in gzwrite.c in order to be compiled exactly the
same as gzprintf(), so that it was guaranteed to return the correct
information. However that causes a static linkage to zlib to bring
in many routines that are often not used. All that is required to
duplicate the compilation environment of gzprintf() is to include
gzguts.h. So that is now done in zutil.c to assure that the correct
flags are returned.
When successful, gzputc would return the second argument. If the
second argument were -1, gzputc would return -1 instead of the
character written, which was 255. However the -1 would not be
distinguishable from an error. Now gzputc returns 255 in that
case.
This makes build-testing and installing the minizip/miniunzip programs
as simple as "autoreconf -if && ./configure --enable-demos && make &&
make install". Without --enable-demos, the makefile will only build
and install the library, as before. Helped by Mike Frysinger.
minizip/miniunzip were not intended to be general-purpose installed
utilities, but they can be useful from time to time as a lightweight
substitute for zip/unzip. You can also use them to quickly test that
the library installation procedure worked.
Instead of using relative paths directly, use paths relative to
top_srcdir and top_builddir to refer to source files and built files,
respectively.
Note that the toplevel zlib configure script still does not have any
special support for out-of-tree builds. But now you can do
(cd contrib/minizip && autoreconf -fis)
mkdir -p BUILD/test
cp *.c *.h *.in zlib.map configure zlib.3 BUILD
cp test/*.c BUILD/test
(cd BUILD && ./configure --shared)
(cd BUILD && make)
mkdir -p BUILD/contrib/minizip
cd BUILD/contrib/minizip
../../../contrib/minizip/configure
make
While at it, move the include path and library path settings to
CPPFLAGS and LDFLAGS respectively instead of setting both in CFLAGS.
Thanks to Mike Frysinger for advice.
Trying to build the minizip utility from contrib/minizip after an
autoreconf -f:
libtool: link: gcc -g -O2 -o minizip minizip.o
minizip.o: In function `getFileCrc':
/tmp/zlib/contrib/minizip/minizip.c:211: undefined reference to `crc32'
minizip.o: In function `main':
/tmp/zlib/contrib/minizip/minizip.c:378: undefined reference to `zipOpen64'
/tmp/zlib/contrib/minizip/minizip.c:451: undefined reference to `zipOpenNewFileInZip3_64'
/tmp/zlib/contrib/minizip/minizip.c:502: undefined reference to `zipCloseFileInZip'
/tmp/zlib/contrib/minizip/minizip.c:509: undefined reference to `zipClose'
/tmp/zlib/contrib/minizip/minizip.c:485: undefined reference to `zipWriteInFileInZip'
collect2: error: ld returned 1 exit status
The cause: contrib/minizip/Makefile.am does not specify that minizip
needs to be linked to libminizip. With some linkers (e.g., GNU
binutils without --copy-dt-needed-entries), an indirect dependency
cannot be used to resolve symbols, so link to libz for crc32(), too.
Trying to build miniunzip utility from contrib/minizip after an
autoreconf -f produces
[...]
In file included from minizip.c:61:0:
zip.h:50:18: fatal error: zlib.h: No such file or directory
unless zlib is already installed. Use AM_CFLAGS to set the include
path and library path to point to the just-build copy of zlib to
fix this. (This was already done for libminizip but not the binaries
that use it before this patch.)
Previously ./configure would use any output on stderr as an indication
that the compilation failed. However if some compiler wrapper uses
stderr for some other purpose, e.g. distcc for nodes going down, then
./configure would not properly configure the build. This problem was
noted by Mike Frysinger. For backwards compatibility, ./configure
will revert to the old way, i.e. checking for anything on stderr, if
when it deliberately runs the compiler with an error, a zero exit
status is returned.
This allows deflate to generate the same output when continuing after
a Z_SYNC_FLUSH vs. using deflateSetDictionary() after a Z_FULL_FLUSH
or a deflateReset(). It also slightly improves compression when
flushing by providing two more strings to possibly match at the start
of the new block.
Previously, the bit buffer would hold 1 to 16 bits after "all" of the
output is provided after a Z_BLOCK deflate() call. Now at most seven
bits remain in the output buffer after Z_BLOCK. flush_pending() now
flushes the bit buffer before copying out the byte buffer, in order
for it to really flush as much as possible.
Z_PARTIAL_FLUSH would sometimes emit two empty static blocks instead
of one in order to provide enough lookahead for inflate to be able
to decode what was last compressed. inflate no longer needs that
much lookahead, so this removes the possibility of emitting the
second empty static block. Z_PARTIAL_FLUSH will now emit only one
empty static block.
Previously when doing an empty flush, a extra static or stored block
could be emitted before the requested empty static or stored block.
This patch prevents the emission of empty blocks by the deflate_*
functions.
The incorporation of the Z_BLOCK flush did not update the rejection
of lower ranked flushes immediately after higher ranked flushes with
no more input data. This prevented an empty Z_SYNC_FLUSH right after
a Z_BLOCK flush, which would be desired to bring the deflate stream
to a byte boundary conditionally on whether or not it was already at
a byte boundary. This patch re-ranks Z_BLOCK above Z_NO_FLUSH but
below Z_PARTIAL_FLUSH, allowing stronger empty flushes to follow a
Z_BLOCK flush.
Moves new function definitions in order to preserve shared library
compatibility with previous versions. Also increases buffer size in
RPG binding and updates the documentation.
Mark Adler
Jonathan Nieder