diff options
| author | Daniel Stenberg <daniel@haxx.se> | 2005-08-19 15:07:50 +0000 |
|---|---|---|
| committer | Daniel Stenberg <daniel@haxx.se> | 2005-08-19 15:07:50 +0000 |
| commit | c7b7bf32f8ab6020047db8b4dbb33e58bae2b145 (patch) | |
| tree | 2a300539a29a00a0b2b652979c752ec721268592 | |
| parent | 710ee3b0e0858a3ee8283fd1de1bc35f24c2bb5b (diff) | |
removed the unreachable code warning from gcc debug builds, even the most
recent gcc versions give far too many false positives for this to be valuable
| -rw-r--r-- | acinclude.m4 | 7 |
1 files changed, 5 insertions, 2 deletions
diff --git a/acinclude.m4 b/acinclude.m4 index ea62ce3bc..5563414b5 100644 --- a/acinclude.m4 +++ b/acinclude.m4 @@ -758,7 +758,10 @@ AC_DEFUN([CURL_CC_DEBUG_OPTS], fi dnl -Wunreachable-code seems totally unreliable on my gcc 3.3.2 on - dnl on i686-Linux as it gives us heaps with false positives + dnl on i686-Linux as it gives us heaps with false positives. + dnl Also, on gcc 4.0.X it is totally unbearable and complains all + dnl over making it unusable for generic purposes. Let's not use it. + if test "$gccnum" -ge "303"; then dnl gcc 3.3 and later WARN="$WARN -Wendif-labels -Wstrict-prototypes" @@ -766,7 +769,7 @@ AC_DEFUN([CURL_CC_DEBUG_OPTS], if test "$gccnum" -ge "304"; then # try these on gcc 3.4 - WARN="$WARN -Wunreachable-code -Wdeclaration-after-statement" + WARN="$WARN -Wdeclaration-after-statement" fi for flag in $CPPFLAGS; do |