diff options
| author | Chris Lattner <sabre@nondot.org> | 2004-02-25 03:45:50 +0000 |
|---|---|---|
| committer | Chris Lattner <sabre@nondot.org> | 2004-02-25 03:45:50 +0000 |
| commit | d825d30f429c053a7cfb62b95f28f8152923829f (patch) | |
| tree | 07e1cac5ea5cd0a7f1461a9668bc361de5a94b8a /llvm/lib/Transforms | |
| parent | f85e33cd792e74e53163c0664172a0015f04b08e (diff) | |
| download | bcm5719-llvm-d825d30f429c053a7cfb62b95f28f8152923829f.tar.gz bcm5719-llvm-d825d30f429c053a7cfb62b95f28f8152923829f.zip | |
add an inefficient way of folding structure and constant array indexes together
into a single LEA instruction. This should improve the code generated for
things like X->A.B.C[12].D.
The bigger benefit is still coming though. Note that this uses an LEA instruction
instead of an add, giving the register allocator more freedom. We should probably
never generate ADDri32's.
llvm-svn: 11817
Diffstat (limited to 'llvm/lib/Transforms')
0 files changed, 0 insertions, 0 deletions
