Commit Graph

98 Commits

Author SHA1 Message Date
Akira TAGOH c20ac78b01 Bug 57114 - regression on FcFontMatch with namelang
After 7587d1c99d applied, family,
style, and fullname is localized against current locale or lang
if any though, the string in other languages were dropped from
the pattern. this caused unexpected mismatch on the target="font"
rules.

This fix adds other strings at the end of the list.
2012-11-15 16:37:32 +09:00
Akira TAGOH 1b692d8ab9 Fix the wrong estimation for the memory usage information in fontconfig 2012-06-01 19:06:17 +09:00
Akira TAGOH 4a060729a1 fcpat: Increase the number of buckets in the shared string hash table
This is a reasonably conservative increase in the number of buckets in the hash
table to 251.  After FcInit(), there are 240 shared strings in use on my system
(from configuration files I assume).  The hash value is stored in each link in
the chains so comparison are actually not very expensive.  This change should
reduce the average length of chains by a factor of 8.  With the reference
counted strings, it should keep the average length of chains to about 2.  The
number of buckets is prime so as not to rely too much on the quality of the
hash function.

https://bugs.freedesktop.org/show_bug.cgi?id=17832#c5

Patch from Karl Tomlinson
2012-03-28 13:42:01 +09:00
Akira TAGOH d8dcff7b96 Bug 17832 - Memory leaks due to FcStrStaticName use for external patterns
Use the reference-counted strings instead of the static strings

Patch from Karl Tomlinson
2012-03-28 13:42:01 +09:00
Behdad Esfahbod 1c475d5c8c Bug 35587 - Add padding to make valgrind and glibc not hate each other 2011-03-28 16:33:12 -04:00
Behdad Esfahbod 5aaf466d38 Cleanup copyright notices to replace "Keith Packard" with "the author(s)" 2010-11-10 16:45:42 -05:00
Behdad Esfahbod 594dcef0f3 Remove all training whitespaces 2010-04-12 12:19:05 -04:00
Behdad Esfahbod 963820fcbf [int] Remove fc_value_* macros that did nothing other than renaming 2009-11-16 17:46:18 -05:00
Behdad Esfahbod 888f9427ae [int] Remove fc_storage_type() in favor of direct access to v->type 2009-11-16 17:46:18 -05:00
Behdad Esfahbod efe5eae26b Simplify FcValueSave() semantics 2009-06-24 13:52:11 -04:00
Behdad Esfahbod 3074a73b41 Replace 'KEITH PACKARD' with 'THE AUTHOR(S)' in license text in all files 2009-03-13 17:59:28 -04:00
Behdad Esfahbod 263f16ced2 Oops, fix FcPatternFilter 2009-02-13 16:54:06 -08:00
Behdad Esfahbod 799691c901 Don't leak FcValues string loaded through fcxml.c (#17661)
Patch from Caolan McNamara.
2009-02-13 16:54:00 -08:00
Behdad Esfahbod 74e16ceeea Fix docs re 'orig' argument of FcPatternBuild and family
Now call it 'p' or 'pattern', since it's modified in place.
There is no copying.
2009-02-13 16:53:57 -08:00
Behdad Esfahbod 41fc0fe68d Add FcPatternFilter() (#13016) 2009-02-13 16:53:56 -08:00
Keith Packard 13a14cbf56 Fix a few memory tracking mistakes.
The built-in memory tracking code in fontconfig relies on a lot of manual
function call tracking. A pain, but it helps debug leaks.
2008-05-04 01:26:40 -07:00
Keith Packard a0a1da22a4 Include fcftaliastail.h so that the freetype funcs are exported.
This header file needs to be included at the end of every file that
exports any freetype symbols.
2008-01-07 16:31:06 -08:00
Keith Packard dbd065ad31 Remove freetype requirement for build-time applications.
This avoids requiring the freetype development files when cross compiling
2008-01-02 08:47:14 -08:00
Stephan Kulow 9b74b78fe8 Make FcPatternDuplicate copy the binding instead of always using Strong.
I noticed that Qt always uses a different font than fc-match advertises.
Debugging the issue, I found that a call that looks pretty innocent is
changing all weak bindings to strong bindings and as such changes the
semantic of the match: FcPatternDuplicate.
2007-03-12 10:21:35 -07:00
Keith Packard 2b77216ee2 Avoid writing uninitialized structure pad bytes to cache files.
The union inside the FcValue structure contains pad bytes. Instead of
copying the whole structure to the cache block, copy only the initialized
fields to avoid writing whichever bytes serve as padding within the
structure.
2006-12-02 13:04:05 -08:00
Keith Packard 23816bf9ac Eliminate .so PLT entries for local symbols. (thanks to Arjan van de Ven)
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
2006-09-05 02:24:01 -07:00
Keith Packard 9e612141df Reference count cache objects.
Caches contain patterns and character sets which are reference counted and
visible to applications. Reference count the underlying cache object so that
it stays around until all reference objects are no longer in use.

This is less efficient than just leaving all caches around forever, but does
avoid eternal size increases in case applications ever bother to actually
look for changes in the font configuration.
2006-09-04 22:20:25 -07:00
Keith Packard bc5e487f2a Pass directory information around in FcCache structure. Freeze charsets.
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.
2006-09-01 01:15:14 -07:00
Keith Packard e9a564e2cd Serialized value lists were only including one value.
The next pointer in the serialized value list wasn't getting set, so they
were truncated at a single value.
2006-08-31 09:07:32 -07:00
Keith Packard 09f9f6f62a Rework Object name database to unify typechecking and object lookup.
Eliminate ancient list of object name databases and load names into single
hash table that includes type information. Typecheck all pattern values to
avoid mis-typed pattern elements.
2006-08-30 18:50:58 -07:00
Keith Packard c02886485b FcCharSetSerialize was using wrong offset for leaves. Make fc-cat work.
FcCharSetSerialize was computing the offset to the unserialized leaf,
which left it pointing at random data when the cache was reloaded.

fc-cat has been updated to work with the new cache structure.

Various debug messages extended to help diagnose serialization errors.
2006-08-30 13:51:03 -07:00
Keith Packard 7ce1967331 Rework cache files to use offsets for all data structures.
Replace all of the bank/id pairs with simple offsets, recode several
data structures to always use offsets inside the library to avoid
conditional paths. Exposed data structures use pointers to hold offsets,
setting the low bit to distinguish between offset and pointer.

Use offset-based data structures for lang charset encodings; eliminates
separate data structure format for that file.

Much testing will be needed; offsets are likely not detected everywhere in
the library yet.
2006-08-30 04:16:22 -07:00
Keith Packard 529291bef4 Eliminate pattern freezing 2006-04-27 07:54:07 +00:00
Keith Packard c1c3ba06d5 Make path names in cache files absolute (NB, cache format change) Stop
permitting cache files to be stored in font dirs. Bump cache magic.
    Don't include /fonts.cache-2 in cache hash construction.
reviewed by: Patrick Lam <plam@mit.edu>
2006-04-27 07:11:44 +00:00
Patrick Lam f045376c08 Include $(top_srcdir), $(top_srcdir)/src before anything else.
Shuffle order of includes for building out of srcdir on win32.
reviewed by: plam
2006-04-25 05:57:41 +00:00
Patrick Lam 2f02e38361 Fix memory leak (Coverity defect #2089).
Ignore script if subtable is missing (Coverity defect #2088).
Fix possible null pointer dereference (Coverity defect #784) and memory
    leak (Coverity defects #785, #786).
Don't copy FcCharSet if we're going to throw it away anyway. (Reported by
    Kenichi Handa).
reviewed by: plam
2006-04-12 14:36:36 +00:00
Patrick Lam 04f7d3e7fd Properly convert static charsets to dynamic charsets.
Fix memory leak in error case (Coverity defects #1820, #1821, #1822).
Fix memory leak (Coverity defect #1819).
prevent crash when invalid include line is parsed (Coverity defect #763).
Fix potential null pointer access (Coverity defect #1804).
Remove dead code (Coverity defect #1194).
Prevent potential null pointer access (Coverity defect #767), ensure error
    value is read (Coverity defect #1195).
reviewed by: plam
2006-04-11 14:20:59 +00:00
Patrick Lam 44415a079a Portability fixes for HP-UX (reported by Christoph Bauer). Replace
'__inline__' by AC_C_INLINE and 'inline'. Replace '__alignof__' by
    'fc_alignof'.
reviewed by: plam
2006-04-07 17:27:39 +00:00
Patrick Lam 392fa276dc Reduce amount of dirty rss by const'ing some data structures.
Don't fail if we can't create or remove $(pkgcachedir) i.e.
    /var/cache/fontconfig. (reported by Quanah Gibson-Mount).
reviewed by: plam
2006-04-06 04:52:21 +00:00
Patrick Lam 9226e04c69 Because we hacked FcPatternGet, we don't really need to expand the filename
again in FcPatternGetString.
2006-03-05 15:33:46 +00:00
Patrick Lam 618adbaf7b Ok, so some people (wine!) use FcPatternGet to fetch FC_FILE. Make that
work. Reported by Bernhard Rosenkraenzer.
2006-03-05 06:05:50 +00:00
Patrick Lam bb6b19938e Get rid of C++-style comments. 2006-03-03 06:35:53 +00:00
Patrick Lam efb11b36c4 Perf optimizations. Inline FcValueCanonicalize, reduce FcValueListPtrU
usage, remove redundant cast.
reviewed by: plam
2006-02-07 21:15:33 +00:00
Patrick Lam 575a37b797 Fix additional memory leaks reported by Ronny V. Vindenes: don't invoke
FcValueSave on hashed static strings in FcPatternAddWithBinding.
Add another st_dev check in FcDirCacheOpen.
2006-01-25 14:52:49 +00:00
Patrick Lam 986e35979e Fix for unaligned memory accesses.
reviewed by: plam
2006-01-19 19:20:30 +00:00
Patrick Lam c60ec7cc6d Add self to AUTHORS list.
Minor change to global cache file format to fix fc-cat bug reported by
    Frederic Crozat, and buglet with not globally caching directories with
    zero fonts cached.
2006-01-05 15:12:22 +00:00
Patrick Lam ebc157f9a1 Shut up GCC warnings on amd.
reviewed by: plam
2005-12-21 20:00:20 +00:00
Patrick Lam 204da5a8b8 Because we've changed FcPatternAddString to use FcStrStaticName and not
FcValueSave, explicitly handle the case of a null string.
2005-12-08 05:54:27 +00:00
Patrick Lam 9ede93f1dc Don't free strings that have been returned from FcStrStaticName. 2005-12-05 16:08:01 +00:00
Patrick Lam 51af050992 Don't make FcPatternFindFullFname available to fccfg, it's not really safe.
Instead go through FcPatternGetString (sorry, perf guys.) Also, use
    globs for dirs as well.
2005-11-29 06:09:18 +00:00
Patrick Lam ced3f0a0ab Fix problem dating back at least to 2.3.2 where globs weren't being applied
to patterns loaded from a cache.
Fix some obvious spelling mistakes.
2005-11-29 00:14:42 +00:00
Patrick Lam 81fe99fdd0 Rename fcpatterns, fcpatternelts, fcvaluelists to _fcPatterns,
_fcPatternElts, _fcValueLists for consistency.
2005-11-25 16:04:44 +00:00
Patrick Lam 61571f3f2e Pass around FcCache *s to the Unserialize functions for extra consistency
(and less overhead, for what that's worth).
2005-11-25 15:50:34 +00:00
Patrick Lam 9ab79bdfb7 Inline the *PtrU functions to gain perf. Remove unneeded params for the
FcCompare* functions.
reviewed by: plam
2005-11-25 03:00:51 +00:00
Patrick Lam 1c9fdccab9 Move FC_BANK_DYNAMIC, FC_BANK_FIRST to internal header.
Check for type validity during FcPatternAddWithBinding, don't verify type
    in FcFontMatch, don't call FcCanonicalize here (which always does a
    deep copy).
reviewed by: plam
2005-11-24 21:40:20 +00:00