Commits

Branch selector

 bee78fe

User selector

All users

Datepicker

All time

Commit History

Commits on Apr 11, 2012

• Clean up ARM fused multiply + add/sub support some more: rename some isel

  Show description for bee78fe
  

  Evan Cheng

  committed
  bee78fe

• Fix an overly indented line. Remove an 'else' after an 'if' that returns.

  Show description for 7eb46d8
  topperc

  committed
  7eb46d8

• Inline implVisitAluOverflow by introducing a nested switch to convert the intrinsic to an nodetype.

  Show description for c42e640
  topperc

  committed
  c42e640

• Tablegen'd regpressure: emit the weighted pressure limit.

  Show description for 680363b
  atrick

  committed
  680363b

• Table-generated register pressure fixes.

  Show description for aa744e2
  atrick

  committed
  aa744e2

• Optimize code a bit by calling push_back only once in some loops. Reduces compiled code size a bit.

  Show description for 23de31b
  topperc

  committed
  23de31b

• Match (fneg (fma) to vfnma. rdar://10139676

  Show description for 92c9045
  

  Evan Cheng

  committed
  92c9045

• Add retw and lretw instructions. Also, fix Intel syntax parsing for all

  Show description for 0d82fe7
  

  Charles Davis

  committed
  0d82fe7

• Merge fma.ll into fusedMAC.ll

  Show description for a0908d0
  

  Evan Cheng

  committed
  a0908d0

• Fix ARM disassembly of VLD instructions with writebacks.  And add test a case

  Show description for a69da35
  enderby

  committed
  a69da35

• ARM add missing Thumb1 two-operand aliases for shift-by-immediate.

  Show description for a5378eb
  

  Jim Grosbach

  committed
  a5378eb

• Fix a number of problems with ARM fused multiply add/subtract instructions.

  Show description for 82509e5
  

  Evan Cheng

  committed
  82509e5

• Tweak MachineLICM heuristics for cheap instructions.

  Show description for 71fbed4
  stoklund

  committed
  71fbed4

• Only check for PHI uses inside the current loop.

  Show description for 8b560b8
  stoklund

  committed
  8b560b8

• Fix test to be register assignment invariant.

  Show description for 89cdaf4
  stoklund

  committed
  89cdaf4

Commits on Apr 10, 2012

• TableGen/reginfo potential bug: typo from previous checkin.

  Show description for 4b74558
  atrick

  committed
  4b74558

• Move the constant-folding support for FP_ROUND in SelectionDAG from the one-operand version of getNode() to the two-operand version, since it became a two-operand node at sound point.

  Show description for 06886aa
  resistor

  committed
  06886aa

• llvm-stress: stop abusing ConstantFP::get()

  Show description for 83f17f2
  nobled

  committed
  83f17f2

• llvm-stress: don't make vectors of x86_mmx type

  Show description for 701de8f
  nobled

  committed
  701de8f

• [tsan] two more compile-time optimizations:

  Show description for cff60c1
  kcc

  committed
  cff60c1

• Handle llvm.fma.* intrinsics. rdar://10914096

  Show description for 3aef2ff
  

  Evan Cheng

  committed
  3aef2ff

• Add a comment noting that the fdiv -> fmul conversion won't generate

  Show description for 507bb7a
  CunningBaldrick

  committed
  507bb7a

• The MDString class stored a StringRef to the string which was already in a

  Show description for 3ecb447
  isanbard

  committed
  3ecb447

• Whitespace.

  Show description for 7f1f145
  

  Chad Rosier

  committed
  7f1f145

• Revert r154396, which looks to be the real culprit behind the bot failures.

  Show description for 9142ed5
  

  Chad Rosier

  committed
  9142ed5

• Temporarily revert this patch to see if it brings the buildbots back.

  Show description for a139051
  echristo

  committed
  a139051

• [tsan] compile-time instrumentation: do not instrument a read if

  Show description for 2076af0
  kcc

  committed
  2076af0

• To ensure that we have more accurate line information for a block

  Show description for 18112d8
  echristo

  committed
  18112d8

• Revert r154397, which was causing make check failures on the buildbots.

  Show description for b79855e
  resistor

  committed
  b79855e

• ARM fix cc_out operand handling for t2SUBrr instructions.

  Show description for a23ecc2
  

  Jim Grosbach

  committed
  a23ecc2

• Remove unused variable.

  Show description for 2330636
  dwblaikie

  committed
  2330636

• Fix a dagcombine optimization which assumes that the vsetcc result type is always

  Show description for 77dde08
  nadavrot

  committed
  77dde08

• Modify the code that lowers shuffles to blends from using blendvXX to vblendXX.

  Show description for 50e64cf
  nadavrot

  committed
  50e64cf

• Make a somewhat subtle change in the logic of block placement. Sometimes

  Show description for 45fb79b
  chandlerc

  committed
  45fb79b

• Transform div to mul with reciprocal only when fp imm is legal.

  Show description for 999821c
  asl

  committed
  999821c

Pagination

Previous

Next