Matt Pharr ee7e367981 Do global dead code elimination early in optimization. ...

This gives a 15-20% speedup in compilation time for simple
programs (but only ~2% for the big 21k monster program).

2012-05-05 15:47:19 -07:00

..

builtins.c

Mingw does not have sysconf, use the msc way of finding processors.

2012-01-09 09:45:40 +01:00

dispatch.ll

Update dynamic target dispatch code to support AVX2.

2012-01-12 08:37:18 -08:00

target-avx1-x2.ll

Use native float/half conversion instructions with the AVX2 target.

2012-01-24 15:33:38 -08:00

target-avx1.ll

Use native float/half conversion instructions with the AVX2 target.

2012-01-24 15:33:38 -08:00

target-avx2-x2.ll

Use native float/half conversion instructions with the AVX2 target.

2012-01-24 15:33:38 -08:00

target-avx2.ll

Use native float/half conversion instructions with the AVX2 target.

2012-01-24 15:33:38 -08:00

target-avx-common.ll

Have aos/soa and broadcast/shuffle/rotate functions provided by the target.

2012-01-04 12:59:03 -08:00

target-avx-x2.ll

Add separate set of builtins for AVX2.

2012-01-13 14:40:01 -08:00

target-avx.ll

Add separate set of builtins for AVX2.

2012-01-13 14:40:01 -08:00

target-generic-1.ll

Scalar target builtins bugfixes.

2012-01-29 13:47:44 -08:00

target-generic-4.ll

Add "generic" 4, 8, and 16-wide targets.

2011-12-19 13:46:50 -08:00

target-generic-8.ll

Add "generic" 4, 8, and 16-wide targets.

2011-12-19 13:46:50 -08:00

target-generic-16.ll

Add "generic" 4, 8, and 16-wide targets.

2011-12-19 13:46:50 -08:00

target-generic-32.ll

Add generic-32 target.

2012-05-03 11:11:06 -07:00

target-generic-common.ll

Fix generic targets: half conversion functions weren't declared.

2012-01-27 14:44:43 -08:00

target-sse2-common.ll

Have aos/soa and broadcast/shuffle/rotate functions provided by the target.

2012-01-04 12:59:03 -08:00

target-sse2-x2.ll

Use native float/half conversion instructions with the AVX2 target.

2012-01-24 15:33:38 -08:00

target-sse2.ll

Use native float/half conversion instructions with the AVX2 target.

2012-01-24 15:33:38 -08:00

target-sse4-common.ll

Have aos/soa and broadcast/shuffle/rotate functions provided by the target.

2012-01-04 12:59:03 -08:00

target-sse4-x2.ll

Use native float/half conversion instructions with the AVX2 target.

2012-01-24 15:33:38 -08:00

target-sse4.ll

Use native float/half conversion instructions with the AVX2 target.

2012-01-24 15:33:38 -08:00

util.m4

Do global dead code elimination early in optimization.

2012-05-05 15:47:19 -07:00