#if HAVE_cmpstrnsi usage in builtins.c
Mike Stump
mrs@apple.com
Thu Apr 6 20:02:00 GMT 2006
I encountered a bug in the compiler that caused a bootstrap failure
for me due to a member reference in the expansion of optimize in some
apple local code:
gcc/builtins.c:3679:5: error: token "." is not valid in preprocessor
expressions
so, I checked the sources, and these can be complex, so they should
do an #ifdef, not an #if, also this makes gcc consistent in how it
checks these two, as there are other places that use #ifdef for them
both.
I checked this in as obvious, let me know if you think you're
reasonable and a maintainer and disagree. :-)
2006-04-06 Mike Stump <mrs@apple.com>
* builtins.c (expand_builtin_longjmp):Use #ifdef instead of #if
for HAVE_nonlocal_goto to be consistent.
(expand_builtin_nonlocal_goto): Likewise.
(expand_builtin_strcmp): Use #ifdef instead of #if
for HAVE_cmpstrnsi to be consistent.
Doing diffs in .:
--- ./builtins.c.~1~ 2006-04-04 13:45:16.000000000 -0700
+++ ./builtins.c 2006-04-06 12:29:58.000000000 -0700
@@ -793,7 +793,7 @@ expand_builtin_longjmp (rtx buf_addr, rt
/* Pick up FP, label, and SP from the block and jump. This
code is
from expand_goto in stmt.c; see there for detailed
comments. */
-#if HAVE_nonlocal_goto
+#ifdef HAVE_nonlocal_goto
if (HAVE_nonlocal_goto)
/* We have to pass a value to the nonlocal_goto pattern that
will
get copied into the static_chain pointer, but it does not
matter
@@ -866,7 +866,7 @@ expand_builtin_nonlocal_goto (tree argli
current_function_has_nonlocal_goto = 1;
-#if HAVE_nonlocal_goto
+#ifdef HAVE_nonlocal_goto
/* ??? We no longer need to pass the static chain value, afaik. */
if (HAVE_nonlocal_goto)
emit_insn (gen_nonlocal_goto (const0_rtx, r_label, r_sp, r_fp));
@@ -3676,7 +3676,7 @@ expand_builtin_strcmp (tree exp, rtx tar
GEN_INT (MIN (arg1_align, arg2_align)));
}
#endif
-#if HAVE_cmpstrnsi
+#ifdef HAVE_cmpstrnsi
/* Try to determine at least one length and call cmpstrnsi. */
if (!insn && HAVE_cmpstrnsi)
{
--------------
More information about the Gcc-patches
mailing list