projects / sbcl.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 2372ff8) | patch
Improved SIMD-PACK manipulation VOPs on x86-64
authorPaul Khuong <pvk@pvk.ca>
Wed, 22 May 2013 18:19:47 +0000 (14:19 -0400)
committerPaul Khuong <pvk@pvk.ca>
Wed, 22 May 2013 18:19:47 +0000 (14:19 -0400)
commit808d8109a9b4203ca690706c0b2b04ca759293df
tree9a18d2fa6c62f4626933b8e528a63f75cd0743fftree | snapshot
parent2372ff8da6e1099e8840b0815d75c414fff2f302commit | diff
Improved SIMD-PACK manipulation VOPs on x86-64

 * Tighten naive (and, in one case, wrong-looking) lifetime
   specifications to enable more register coalescing.

 * Exploit the coalescing by replacing explicit move instructions
   with MOVE, and specify the SC to get MOVAPS for float data.

 * Microoptimise away some PSRLDQ by 0.
src/compiler/x86-64/simd-pack.lisp diff | blob | history
Mirror of Steel Bank Common Lisp SBCL's official repository