From: José Fonseca <jfonseca@vmware.com>
Date: Wed, 12 Oct 2011 19:42:05 +0000 (+0100)
Subject: scons: Use -fno-builtin-memcmp.
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=983fa4ad523535debf2e94cf6ac1fd4c5630c0d2;p=mesa.git

scons: Use -fno-builtin-memcmp.

ipers framerate on llmvpipe improves 60%.

Issue spotted by Adam Jackson <ajax at redhat.com>.

http://lists.freedesktop.org/archives/mesa-dev/2011-June/009077.html
---

diff --git a/scons/gallium.py b/scons/gallium.py
index 1c9c0ea32e4..c3350b3ca63 100755
--- a/scons/gallium.py
+++ b/scons/gallium.py
@@ -361,6 +361,9 @@ def generate(env):
             ccflags += ['-O0']
         else:
             ccflags += ['-O3']
+        # gcc's builtin memcmp is slower than glibc's
+        # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43052
+        ccflags += ['-fno-builtin-memcmp']
         # Work around aliasing bugs - developers should comment this out
         ccflags += ['-fno-strict-aliasing']
         ccflags += ['-g']