summaryrefslogtreecommitdiff
path: root/docs/gcc-loops.png
diff options
context:
space:
mode:
authorNadav Rotem <nrotem@apple.com>2012-12-27 08:15:45 +0000
committerNadav Rotem <nrotem@apple.com>2012-12-27 08:15:45 +0000
commitd6fb53adb19ccfbfb1eedec11c899aaa8401d036 (patch)
tree5a44210f5aa0c0130845b720b747cabe76301e34 /docs/gcc-loops.png
parent3c22a444001ac3f7a89c9888d60aec69352d0e58 (diff)
downloadllvm-d6fb53adb19ccfbfb1eedec11c899aaa8401d036.tar.gz
llvm-d6fb53adb19ccfbfb1eedec11c899aaa8401d036.tar.bz2
llvm-d6fb53adb19ccfbfb1eedec11c899aaa8401d036.tar.xz
On AVX/AVX2 the type v8i1 is legalized to v8i16, which is an XMM sized
register. In most cases we actually compare or select YMM-sized registers and mixing the two types creates horrible code. This commit optimizes some of the transition sequences. PR14657. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@171148 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'docs/gcc-loops.png')
0 files changed, 0 insertions, 0 deletions