| commit | d6d82aa8894212d0b665de478ebd277d56b98d2c | [log] [tgz] |
|---|---|---|
| author | Chris Lattner <sabre@nondot.org> | Tue Apr 18 03:57:35 2006 +0000 |
| committer | Chris Lattner <sabre@nondot.org> | Tue Apr 18 03:57:35 2006 +0000 |
| tree | b0876547c1859b0a03878e2b9adc272fe1f0264d | |
| parent | 48786e488712642a46cfa9700387152d82a0e719 [diff] |
Implement v16i8 multiply with this code:
vmuloub v5, v3, v2
vmuleub v2, v3, v2
vperm v2, v2, v5, v4
This implements CodeGen/PowerPC/vec_mul.ll. With this, v16i8 multiplies are
6.79x faster than before.
Overall, UnitTests/Vector/multiplies.c is now 2.45x faster with LLVM than with
GCC.
Remove the 'integer multiplies' todo from the README file.
llvm-svn: 27792