Programming Languages Research Group: Git - oota-llvm.git/history

2012-05-31	Owen Anderson	Make this testcase independent of register allocation.	blob \| commitdiff \| raw
2012-05-30	Owen Anderson	Switch the canonical FMA term operand order to match...	blob \| commitdiff \| raw \| diff to current
2012-05-30	Owen Anderson	Teach DAGCombine to canonicalize the position of a...	blob \| commitdiff \| raw \| diff to current
2012-05-02	Owen Anderson	Teach DAGCombine the same multiply-by-1.0 folding trick...	blob \| commitdiff \| raw \| diff to current
2012-04-27	Lang Hames	Fix the order of the operands in the llvm.fma intrinsic...	blob \| commitdiff \| raw \| diff to current
2012-04-13	Evan Cheng	On Darwin targets, only use vfma etc. if the source...	blob \| commitdiff \| raw \| diff to current
2012-04-11	Evan Cheng	Add more fused mul+add/sub patterns. rdar://10139676	blob \| commitdiff \| raw \| diff to current
2012-04-11	Evan Cheng	Match (fneg (fma) to vfnma. rdar://10139676	blob \| commitdiff \| raw \| diff to current
2012-04-11	Evan Cheng	Merge fma.ll into fusedMAC.ll	blob \| commitdiff \| raw \| diff to current
2012-03-05	Sebastian Pop	updated patch for the ARM fused multiply add/sub	blob \| commitdiff \| raw \| diff to current
2012-01-22	Anton Korobeynikov	Add fused multiple+add instructions from VFPv4.	blob \| commitdiff \| raw \| diff to current

C/C++ LLVM-based compilers that forbids OOTA behaviors