diff options
author | Stan Shebs <shebs@codesourcery.com> | 1994-04-15 21:55:49 +0000 |
---|---|---|
committer | Stan Shebs <shebs@codesourcery.com> | 1994-04-15 21:55:49 +0000 |
commit | fad466eb6de0629edc89081998337fd13566e989 (patch) | |
tree | 0190472f3bd8a3ce36a30ba310c4aa2acfb5cfc0 /gdb/defs.h | |
parent | 97ea252ef8065df3fcc039801fcc370aa3e3c00f (diff) | |
download | gdb-fad466eb6de0629edc89081998337fd13566e989.zip gdb-fad466eb6de0629edc89081998337fd13566e989.tar.gz gdb-fad466eb6de0629edc89081998337fd13566e989.tar.bz2 |
Fri Apr 15 11:53:46 1994 Stan Shebs (shebs@andros.cygnus.com)
* source.c (DIRNAME_SEPARATOR): New macro, replaces all references
to : in search path processing.
* defs.h (qsort): Rename argument in prototype.
* symtab.h (SAYMBOL_VALUE): Rename value field, avoids bugs in
some compilers.
* breakpoint.c, exec.c, mdebugread.c, mipsread.c, xcoffexec.c
(false): Eliminate usages.
Diffstat (limited to 'gdb/defs.h')
-rw-r--r-- | gdb/defs.h | 28 |
1 files changed, 14 insertions, 14 deletions
@@ -520,24 +520,24 @@ enum val_prettyprint #else /* No BFD64 */ -/* Default to support for "long long" if the host compiler being used is gcc. - Config files must define CC_HAS_LONG_LONG to use other host compilers - that are capable of supporting "long long", and to cause gdb to use that - support. Not defining CC_HAS_LONG_LONG will suppress use of "long long" - regardless of what compiler is used. - - FIXME: For now, automatic selection of "long long" as the default when - gcc is used is disabled, pending further testing. Concerns include the - impact on gdb performance and the universality of bugfree long long - support on platforms that do have gcc. Compiling with FORCE_LONG_LONG - will select "long long" use for testing purposes. -fnf */ +/* If all compilers for this host support "long long" and we want to + use it for LONGEST (the performance hit is about 10% on a testsuite + run based on one DECstation test), then the xm.h file can define + CC_HAS_LONG_LONG. + + Using GCC 1.39 on BSDI with long long causes about 700 new + testsuite failures. Using long long for LONGEST on the DECstation + causes 3 new FAILs in the testsuite and many heuristic fencepost + warnings. These are not investigated, but a first guess would be + that the BSDI problems are GCC bugs in long long support and the + latter are GDB bugs. */ #ifndef CC_HAS_LONG_LONG -# if defined (__GNUC__) /*&& defined (FORCE_LONG_LONG)*/ /* See FIXME above */ +# if defined (__GNUC__) && defined (FORCE_LONG_LONG) # define CC_HAS_LONG_LONG 1 # endif #endif - + /* LONGEST should not be a typedef, because "unsigned LONGEST" needs to work. CC_HAS_LONG_LONG is defined if the host compiler supports "long long" variables and we wish to make use of that support. */ @@ -739,7 +739,7 @@ free PARAMS ((void *)); /* 4.10.3.2 */ extern void qsort PARAMS ((void *base, size_t nmemb, /* 4.10.5.2 */ size_t size, - int (*comp)(const void *, const void *))); + int (*compar)(const void *, const void *))); #ifndef MEM_FNS_DECLARED /* Some non-ANSI use void *, not char *. */ extern PTR |