comparison src/gccache-gtk.c @ 778:2923009caf47

[xemacs-hg @ 2002-03-16 10:38:59 by ben] cm.c, file-coding.c: fix warnings. .cvsignore: Those pesky *.tmp files. mule\arabic.el, mule\canna-leim.el, mule\china-util.el, mule\chinese.el, mule\cyril-util.el, mule\cyrillic.el, mule\devan-util.el, mule\devanagari.el, mule\english.el, mule\ethio-util.el, mule\ethiopic.el, mule\european.el, mule\greek.el, mule\hebrew.el, mule\indian.el, mule\japan-util.el, mule\japanese.el, mule\korea-util.el, mule\korean.el, mule\lao-util.el, mule\lao.el, mule\misc-lang.el, mule\mule-charset.el, mule\mule-cmds.el, mule\thai-util.el, mule\thai.el, mule\tibet-util.el, mule\tibetan.el, mule\viet-util.el, mule\vietnamese.el, unicode.el: Fix lots of warnings. Sync up some files to FSF 21.1. Copy over all charset definitions from FSF 21.1, convert them to our format, and stick them in the relevant files. Eventually we will actually be able to dump these files (though they may not quite work). autoload.el: Support defun*, defmacro*. mule/mule-composite.el, mule/mule-composite-stub.el: New file, stubs for nonexistent composition funs/vars. mule/viet-chars.el, dumped-lisp.el: Account for these changes. font.el, mouse.el, msw-font-menu.el, printer.el, startup.el: fix warnings.
author ben
date Sat, 16 Mar 2002 10:39:19 +0000
parents 0784d089fdc9
children a8d8f419b459
comparison
equal deleted inserted replaced
777:e65d9cf16707 778:2923009caf47
95 gc_cache_hash (const void *arg) 95 gc_cache_hash (const void *arg)
96 { 96 {
97 const struct gcv_and_mask *gcvm = (const struct gcv_and_mask *) arg; 97 const struct gcv_and_mask *gcvm = (const struct gcv_and_mask *) arg;
98 unsigned long *longs = (unsigned long *) &gcvm->gcv; 98 unsigned long *longs = (unsigned long *) &gcvm->gcv;
99 unsigned long hash = gcvm->mask; 99 unsigned long hash = gcvm->mask;
100 int i; 100 unsigned int i;
101 /* This could look at the mask and only use the used slots in the 101 /* This could look at the mask and only use the used slots in the
102 hash code. That would win in that we wouldn't have to initialize 102 hash code. That would win in that we wouldn't have to initialize
103 every slot of the gcv when calling gc_cache_lookup. But we need 103 every slot of the gcv when calling gc_cache_lookup. But we need
104 the hash function to be as fast as possible; some timings should 104 the hash function to be as fast as possible; some timings should
105 be done. */ 105 be done. */