Move immintrin/arm_neon includes to where they are used.
On my Mac (so, immintrin), this improves compile time, both wall and cpu,
by about 16%. To test I ran this on an SSD with files hot in their caches:
$ env CC=/usr/bin/clang CXX=/usr/bin/clang++ ./gyp_skia && \
ninja -C out/Release -t clean && \
time ninja -C out/Release
Before: 159 wall / 3367 cpu
159 wall / 3368 cpu
After: 137 wall / 2860 cpu
136 wall / 2863 cpu
I also tried further refining immintrin down to emmintrin / tmmintrin / smmintrin etc.
That made no signficant difference, so I've kept immintrin for its simplicity.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2045633002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
TBR=reed@google.com
No public API changes.
Committed: https://skia.googlesource.com/skia/+/12dfaaa53c23f3d03050bde8f64136ac1f44164a
Review-Url: https://codereview.chromium.org/2045633002
diff --git a/include/core/SkTypes.h b/include/core/SkTypes.h
index a47225d..4e95c69 100644
--- a/include/core/SkTypes.h
+++ b/include/core/SkTypes.h
@@ -14,12 +14,6 @@
#include "SkPostConfig.h"
#include <stddef.h>
#include <stdint.h>
-
-#if defined(SK_ARM_HAS_NEON)
- #include <arm_neon.h>
-#elif SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_SSE2
- #include <immintrin.h>
-#endif
// IWYU pragma: end_exports
#include <string.h>
diff --git a/include/private/SkFloatingPoint.h b/include/private/SkFloatingPoint.h
index 6ed6144..a7aa50c 100644
--- a/include/private/SkFloatingPoint.h
+++ b/include/private/SkFloatingPoint.h
@@ -15,6 +15,12 @@
#include <math.h>
#include <float.h>
+#if SK_CPU_SSE_LEVEL >= SK_CPU_SSE_LEVEL_SSE1
+ #include <xmmintrin.h>
+#elif defined(SK_ARM_HAS_NEON)
+ #include <arm_neon.h>
+#endif
+
// For _POSIX_VERSION
#if defined(__unix__) || (defined(__APPLE__) && defined(__MACH__))
#include <unistd.h>