in the process. (More information on tech-pkg.)
Bump PKGREVISION and BUILDLINK_DEPENDS of all packages using libtool and
installing .la files.
Bump PKGREVISION (only) of all packages depending directly on the above
via a buildlink3 include.
* make error message instaed of invalid output for -plain global option on
a "pam" program
* pnmtojpeg: fix bug with large exif files (negative malloc)
* ppntowinicon: fix bogus "out of memory" failre
* ppmtompeg: fix crash with input that is not a multiple of 16 pixels high
and wide
* pamtotga: fix crash due to freeing nonexistent color table
* pamtotga: fix bug: invalid TGA when you use -rgb with a non-color input
* pgmtopbm: fix calculations that divide by maxval+1; should divide by maxval
* pbm_readpbmrowpacked(): Fix bug: ORs bits into return value instead of
setting them
* Don't try to install pm_gamma (which does not exist in this level)
* pamflip: fix bug with random memory accesses due to signed/unsigned comparison
Changes 10.18.11:
* pnmtopng: fix bug with -transparent
Changes 10.18.10:
* Include pm_system.h, pm_gamma.h in install package
* Fix the ftell/ftello distinction for many platforms
* ppmtogif: Don't violate array bounds when GIF color map is larger
than internal color map.
* yuvsplittoppm: Fix -ccir601 option
* ppmtomap: Change #/bin/sh to #!/bin/sh
Changes 10.18.9:
* ppmtobmp: Fix bug: Generates invalid bits per pixel (e.g. 3) when input
image has small number of colors
* pbmtog3: fix buffer overrun when image > 1728 columns
* pnmtotiffcmyk: Correct abs to fabs so -theta works
* sldtoppm: Correct abs to fabs
* pgmramp: Correct abs to fabs so -rectangle and -ellipse work
* pgmramp: Don't crash with -rectangle and height or width = 1
* pnmtojpeg: interpret "length" field in exif data as unsigned 16 bit
instead of signed.
* ppmforge: Correct abs to fabs so the planet is more than 2 pixels wide
Changes 10.18.8:
* pnmrotate: fix bug: always crashes in memory free
* pnmrotate: remove debugging messages
Changes 10.18.7:
* ppmshadow: Fix crash in all invocations
* pnmquant: Fix bug wherein all options are rejected
Changes 10.18.6:
* Add default RPATHOPT = rpath setting to Makefile.config.in
* Add stdlib.h and pm_config.h to mallocvar.h
* lib/Makefile: $(LIBNETPBMSUFFIX) instead of .so
* fitstopnm: Don't produce maxval 0
Changes 10.18.5:
* pnmtopng: fix bug with array bounds violation with -alpha
Usual symptom: "internal inconsistency" message
Changes 10.18.4:
* anytopnm: Remove "function" so it works on FreeBSD
Changes 10.18.3:
* Fix stack corruption in pXm_check() on systems that have both 32 and 64 bit
file offsets
Changes 10.18.2:
* installnetpbm: don't crash on mkdir on old Perl
* bmptopnm: fix bug with BMPs that end prematurely
Changes 10.11.15:
* Bmptopnm goes into an infinite loop if you give it a BMP file that ends
prematurely. The bug is in the way the program interprets the return code
from fread() in readrow(). You can get the fix from the 10.18.1-10.18.2 diff.
* xwdtopnm: Doesn't work with at least one direct color 24 bits-per-pixel
xwd generated by Xfree86's Xwd.
* ppmquant does not choose the best set of colors. pnmquant does, though.
* giftopnm doesn't understand GIFs with the same color both transparent and
opaque.
pnmtopng -alpha fails with "internal inconsistency detected" message
Changes 10.11.14:
* pstopnm: fix bug in which Standard Input never works
* ppmtobmp: change 'char' to 'unsigned char' to work around a Solaris C library bug
* ppmtobmp: handle maxval properly for maxval != 256 truecolor BMP
Changes 10.11.13:
* Explicitly set SCRIPTS make variable to null so environment
variable of the same name doesn't screw things up.
* stamp_date: don't use LOGNAME as a variable (on AIX, you can't).
* configure: put space after -o in test compiles.
* ppmtompeg: adjust header files in parallel.c so it compiles on AIX
Changes 10.11.12:
* pnmpsnr: fix bug: Cr and Cb interchanged
* Fix crash in pnmstitch
Changes 10.11.10:
* Declare internal subroutines in pgmabel as 'static' so they don't conflict
with some C libraries
Changes 10.11.9:
* pstopnm: fix bug: encapsulated postscript confused with regular postscript,
so all pages get overlaid as one page
by moving the inclusion of buildlink3.mk files outside of the protected
region. This bug would be seen by users that have set PREFER_PKGSRC
or PREFER_NATIVE to non-default values.
BUILDLINK_PACKAGES should be ordered so that for any package in the
list, that package doesn't depend on any packages to the left of it
in the list. This ordering property is used to check for builtin
packages in the correct order. The problem was that including a
buildlink3.mk file for <pkg> correctly ensured that <pkg> was removed
from BUILDLINK_PACKAGES and appended to the end. However, since the
inclusion of any other buildlink3.mk files within that buildlink3.mk
was in a region that was protected against multiple inclusion, those
dependencies weren't also moved to the end of BUILDLINK_PACKAGES.
the normal case when BUILDLINK_DEPENDS.<pkg> isn't specified, it receives
a value only once due to the multiple inclusion protection in the
bulldlink3.mk files. In the case where a package includes several
buildlink3.mk files that each want a slightly different version of another
dependency, having BUILDLINK_DEPENDS.<pkg> be a list allows for the
strictest <pkg> dependency to be matched.
In some cases, gtar had problems, like:
/bin/tar: share/doc/netpbm: Cannot unlink: Directory not empty
/bin/tar: share/netpbm: Cannot unlink: Directory not empty
USE_GCC2 or USE_GCC3 where appropriate.
the functionality of the old gcc.buildlink2.mk has been rolled into
compiler.mk now, which is automatically used.
more changes to come later...
now and not NetBSD-*-arm32. Changes include one or more of:
- Change MACHINE_ARCH == arm32 to also match arm
- Where ONLY_FOR_PLATFORM includes NetBSD-*-arm32, add NetBSD-*-arm
- Where BROKEN or worked around for arm gcc bugs, set USE_GCC3
The last may shake out a few more broken packages the next bulk build.
- pbmtoepsi: fix bug: non six-digit lines between
%%BeginPreview: ... and %%EndImage .
- pnmtops: fix bug: 1 bit per sample output when maxval is 2 or 3.
Should be 4 bits per sample.
- pnmtops: fix bug: everything too dark when input maxval less
than Postscript maxval.
This fixes PR pkg/21271 by Dave Camp.
