search for: reuiqrement

Displaying 3 results from an estimated 3 matches for "reuiqrement".

2008 Jan 17
1
[LLVMdev] Still Another LiveInterval Question
...ue number is if they participate in 2-address instructions where a copy had to be introduced to satisfy the read-write requirements of the instruction? And further, in that case, LiveIntervals for virtual registers should only have at most two value numbers? How about on targets without these reuiqrements? Should LiveIntervals for virtual registers only ever have one value number? -Dave
2013 Jan 14
0
[LLVMdev] [cfe-dev] RFC: Codifying (but not formalizing) the optimization levels in LLVM and Clang
...difficult to know statically whether vectorization will help or harm. One can do runtime code selection but that has its own costs and can be slower than O2 in some cases. I simply don't think it's a useful or achieveable guarantee. It's a good goal but I would be against making it a reuiqrement as it will severely limit what we do at O3 over O2. FWIW, the Cray compiler considers O3 to be "try to make it as fast as possible without unreasonably increasing compile time" but does not guarantee it absolutely will be faster than -O2. We have -Oaggress for "pull out all the sto...
2013 Jan 14
17
[LLVMdev] RFC: Codifying (but not formalizing) the optimization levels in LLVM and Clang
This has been an idea floating around in my head for a while and after several discussions with others it continues to hold up so I thought I would mail it out. Sorry for cross posting to both lists, but this is an issue that would significantly impact both LLVM and Clang. Essentially, LLVM provides canned optimization "levels" for frontends to re-use. This is nothing new. However, we