vulkancts/README: Recommend -msse2 -mfpmath=sse for 32-bit x86 builds

Otherwise, GCC may produce code that doesn't follow the IEEE floating
point rules.  This can result in spurious test failures due to precision
issues with the CTS.  These issues won't be noticed with 64-bit builds
because GCC enables SSE floating-point math by default on x86_64.

Change-Id: Id0a60464d7607ec70b62e62c73558b8750b74849

diff --git a/external/vulkancts/README.md b/external/vulkancts/README.md
index b8dbd95..3456c5a 100644 (file)
--- a/external/vulkancts/README.md
+++ b/external/vulkancts/README.md
@@ -74,6 +74,9 @@ using cmake.
 
 Release build can be done by using -DCMAKE_BUILD_TYPE=Release
 
+If building for 32-bit x86 with GCC, you probably also want to add `-msse2
+-mfpmath=sse` to ensure that you get correct IEEE floating-point behavior.
+
 ### Linux 64-bit Debug
 
        cmake <path to vulkancts> -DCMAKE_BUILD_TYPE=Debug -DCMAKE_C_FLAGS=-m64 -DCMAKE_CXX_FLAGS=-m64