Code currently returns a fatal error if it tries to add more entries
than the array has room for, but it wasn't checking to make sure
the final null terminator entry would fit.
Reported by parfait 1.3:
Error: Buffer overrun
Buffer overflow (CWE 120): In array dereference of files[i] with index i
Array size is 256 elements (of 4 bytes each), index >= 0 and index <= 256
at line 250 of fc-glyphname/fc-glyphname.c in function 'main'.
Error: Buffer overrun
Buffer overflow (CWE 120): In array dereference of entries[i] with index i
Array size is 1024 elements (of 8 bytes each), index >= 0 and index <= 1024
at line 298 of fc-lang/fc-lang.c in function 'main'.
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Previously we were failing if CROSS_COMPILING and the generated headers
were not present. It works just fine now.
One caveat: the fix is not fully correct since config.h is being
included in the files built with CC_FOR_BUILD, but config.h has config
for the host system, not the build system. Should be fine though.
When ./configure'd in a directory other than the srcdir, we need to look
in ${top_builddir)/src for the generated header fcstdint.h
Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk>
If the generated header files are included in the distribution, a build
outside of the source directory will use them. For machine-specific files,
this generates the wrong result (fcarch.h). Leaving them out of the
distribution forces them to be built.
Glyph names (now used only for dingbats) were using many relocations,
causing startup latency plus per-process memory usage. Replace pointers with
table indices, shrinking table size and elimninating relocations.
Using a simple shell script that processes the public headers, two header
files are constructed that map public symbols to hidden internal aliases
avoiding the assocated PLT entry for referring to a public symbol.
A few mistakes in the FcPrivate/FcPublic annotations were also discovered
through this process
Instead of passing directory information around in separate variables,
collect it all in an FcCache structure. Numerous internal and tool
interfaces changed as a result of this.
Charsets are now pre-frozen before being serialized. This causes them to
share across multiple fonts in the same cache.
clean up .spec file; perhaps this will be useful to somebody...
Update to 2.2.93
Make 'scanopen' static so GCC doesn't whine about lacking prototype
Add WARN_CFLAGS to pass -W flags for GCC systems
Change various char types around to match across function calls. Fixed bug
in using available_sizes[i].height which is in pixels, not 64ths of a
pixel.
but there are still failures long into the process dealing with docs
(as always).
The big changes here are mostly to make $(srcdir) != "." work correctly,
fixing the docbook related sections and fc-lang were particularily
tricky. Docbook refuses to load system entities from anywhere other
than where the original .sgml file was located, so no luck looking in
"." for the configure-generated version.sgml and confdir.sgml files.
fc-lang needed help finding .orth files; added a -d option to set the
directory as the least evil of many options.
Now to go use a faster machine and try and wring out the last issues.
Alan Coopersmith
Behdad Esfahbod
Akira TAGOH
Jon TURNEY
Keith Packard
Keith Packard
Keith Packard
Patrick Lam
Manish Singh
Carl Worth