diff options
| author | Richard Sandiford <rsandifo@linux.vnet.ibm.com> | 2013-08-05 11:23:46 +0000 |
|---|---|---|
| committer | Richard Sandiford <rsandifo@linux.vnet.ibm.com> | 2013-08-05 11:23:46 +0000 |
| commit | c212125d2797bb41cb83c7cf6d0b65a2f61e24de (patch) | |
| tree | c9673d8c1bccedb9bafe4b75921c5a663cb2e892 /llvm/test/ExecutionEngine/test-interp-vec-shift.ll | |
| parent | 483b9fbddb66e0db6388bae44b73cad6f704c77f (diff) | |
| download | bcm5719-llvm-c212125d2797bb41cb83c7cf6d0b65a2f61e24de.tar.gz bcm5719-llvm-c212125d2797bb41cb83c7cf6d0b65a2f61e24de.zip | |
[SystemZ] Use BRCT and BRCTG to eliminate add-&-compare sequences
This patch just uses a peephole test for "add; compare; branch" sequences
within a single block. The IR optimizers already convert loops to
decrement-and-branch-on-nonzero form in some cases, so even this
simplistic test triggers many times during a clang bootstrap and
projects/test-suite run. It looks like there are still cases where we
need to more strongly prefer branches on nonzero though. E.g. I saw a
case where a loop that started out with a check for 0 ended up with a
check for -1. I'll try to look at that sometime.
I ended up adding the Reference class because MachineInstr::readsRegister()
doesn't check for subregisters (by design, as far as I could tell).
llvm-svn: 187723
Diffstat (limited to 'llvm/test/ExecutionEngine/test-interp-vec-shift.ll')
0 files changed, 0 insertions, 0 deletions
