| Name |
Last commit
|
Last update |
|---|---|---|
| crosstest | ||
| pydir | ||
| src | ||
| tests_lit | ||
| .gitignore | ||
| LICENSE.TXT | ||
| LOWERING.rst | ||
| Makefile | ||
| Makefile.standalone | ||
| OWNERS | ||
| README.rst | ||
| codereview.settings | ||
| llvm2iceinsts.py | ||
| szdiff.py |
For ebp, exclude as needed. For esp, don't mark it as an int register. Not sure exactly how to do a targeted test for this Om1 register allocator. The Om1 regalloc seems to start w/ a fresh whitelist after each instruction, so it may assign the same register (e.g., eax), as an earlier instruction. Without pre-colored registers, I'm not sure how to force it to allocate something other than the first few registers. I do have a test case that has a ton of pre-colored registers, (e.g., cmpxchg8b), but that is a different CL: https://codereview.chromium.org/362463002/ Encountered for: BUG= https://code.google.com/p/nativeclient/issues/detail?id=3882 R=stichnot@chromium.org Review URL: https://codereview.chromium.org/369573005
| Name |
Last commit
|
Last update |
|---|---|---|
| crosstest | Loading commit data... | |
| pydir | Loading commit data... | |
| src | Loading commit data... | |
| tests_lit | Loading commit data... | |
| .gitignore | Loading commit data... | |
| LICENSE.TXT | Loading commit data... | |
| LOWERING.rst | Loading commit data... | |
| Makefile | Loading commit data... | |
| Makefile.standalone | Loading commit data... | |
| OWNERS | Loading commit data... | |
| README.rst | Loading commit data... | |
| codereview.settings | Loading commit data... | |
| llvm2iceinsts.py | Loading commit data... | |
| szdiff.py | Loading commit data... |