index
:
bcm5719-llvm
meklort-10.0.0
meklort-10.0.1
ortega-7.0.1
Project Ortega BCM5719 LLVM
Raptor Computing Systems
summary
refs
log
tree
commit
diff
stats
log msg
author
committer
range
path:
root
/
llvm
/
lib
/
Target
/
R600
/
SIInstrInfo.cpp
Commit message (
Expand
)
Author
Age
Files
Lines
*
R600/SI: Fix verifier error when producing v_madmk_f32
Matt Arsenault
2015-04-24
1
-0
/
+3
*
R600/SI: Special case v_mov_b32 as really rematerializable
Matt Arsenault
2015-04-23
1
-0
/
+14
*
R600/SI: Improve BFM support
Marek Olsak
2015-03-24
1
-0
/
+1
*
R600/SI: Use V_FRACT_F64 for faster 64-bit floor on SI
Marek Olsak
2015-03-24
1
-0
/
+20
*
R600/SI: Merge tables for commuting
Matt Arsenault
2015-03-23
1
-20
/
+0
*
R600/SI: Allow commuting compares
Matt Arsenault
2015-03-23
1
-2
/
+24
*
R600/SI: Fix crash in SIInstrInfo::areLoadsFromSameBasePtr()
Tom Stellard
2015-03-23
1
-2
/
+10
*
Remove the need to cache the subtarget in the R600 TargetRegisterInfo
Eric Christopher
2015-03-11
1
-29
/
+6
*
R600/SI: Re-order MUBUF operands to match asm strings.
Tom Stellard
2015-03-10
1
-4
/
+3
*
R600/SI: Move kill flag to second instruction when splitting SMRD
Tom Stellard
2015-03-10
1
-5
/
+12
*
R600/SI: Add an intrinsic for S_FLBIT_I32 / V_FFBH_I32
Marek Olsak
2015-03-04
1
-0
/
+1
*
Make some non-constant static variables non-static or fully const.
Benjamin Kramer
2015-03-01
1
-4
/
+1
*
R600/SI: Add slc, glc, and tfe to non-atomic _ADDR64 instructions
Tom Stellard
2015-02-27
1
-1
/
+7
*
R600/SI: Use v_madmk_f32
Matt Arsenault
2015-02-21
1
-4
/
+51
*
R600/SI: Try to use v_madak_f32
Matt Arsenault
2015-02-21
1
-0
/
+78
*
R600/SI: Simplify verification of AMDGPU::OPERAND_REG_INLINE_C
Marek Olsak
2015-02-18
1
-8
/
+6
*
R600/SI: Remove explicit VOP operand checking
Marek Olsak
2015-02-18
1
-28
/
+0
*
R600/SI: Fix brace identation
Matt Arsenault
2015-02-18
1
-1
/
+1
*
R600/SI: Fix copies from SGPR to VCC
Matt Arsenault
2015-02-14
1
-5
/
+10
*
R600/SI: Add hack to copy from a VGPR to VCC
Matt Arsenault
2015-02-14
1
-0
/
+10
*
R600/SI: Allow f64 inline immediates in i64 operands
Matt Arsenault
2015-02-13
1
-21
/
+41
*
R600/SI: Remove unnecessary check for fpimm
Matt Arsenault
2015-02-13
1
-1
/
+1
*
R600/SI: Add soffset operand to mubuf addr64 instruction
Tom Stellard
2015-02-11
1
-3
/
+2
*
R600/SI: Fix B64 VALU shifts on VI
Marek Olsak
2015-02-03
1
-0
/
+18
*
R600/SI: Don't generate non-existent LSHL, LSHR, ASHR B32 variants on VI
Marek Olsak
2015-02-03
1
-2
/
+6
*
Reuse a bunch of cached subtargets and remove getSubtarget calls
Eric Christopher
2015-01-30
1
-3
/
+2
*
R600/SI: Add subtarget feature to enable VGPR spilling for all shader types
Tom Stellard
2015-01-20
1
-11
/
+3
*
R600/SI: Use external symbols for scratch buffer
Tom Stellard
2015-01-20
1
-2
/
+2
*
R600/SI: Update SIInstrInfo:verifyInstruction() after r225662
Tom Stellard
2015-01-20
1
-6
/
+12
*
R600/SI: Use 64-bit encoding by default for opcodes that are VOP3-only on VI
Marek Olsak
2015-01-15
1
-3
/
+3
*
R600/SI: Don't shrink instructions whose e32 encoding doesn't exist
Marek Olsak
2015-01-15
1
-1
/
+5
*
R600/SI: Use IMPLICIT_DEF and KILL when failing to spill VGPRs
Tom Stellard
2015-01-14
1
-3
/
+2
*
R600/SI: Spill VGPRs to scratch space for compute shaders
Tom Stellard
2015-01-14
1
-9
/
+17
*
R600/SI: Add pattern for bitcasting fp immediates to integers
Tom Stellard
2015-01-13
1
-18
/
+14
*
R600/SI: Use RegisterOperands to specify which operands can accept immediates
Tom Stellard
2015-01-12
1
-2
/
+2
*
R600/SI: Commute instructions to enable more folding opportunities
Tom Stellard
2015-01-07
1
-1
/
+4
*
R600/SI: Remove VReg_32 register class
Tom Stellard
2015-01-07
1
-11
/
+11
*
R600/SI: Add a V_MOV_B64 pseudo instruction
Tom Stellard
2015-01-07
1
-0
/
+31
*
R600/SI: Teach SIFoldOperands to split 64-bit constants when folding
Tom Stellard
2015-01-07
1
-0
/
+10
*
R600/SI: isLegalOperand() shouldn't check constant bus for SALU instructions
Tom Stellard
2014-12-19
1
-1
/
+1
*
R600/SI: Fix f64 inline immediates
Matt Arsenault
2014-12-17
1
-12
/
+26
*
R600/SI: Handle physical registers in getOpRegClass
Matt Arsenault
2014-12-11
1
-2
/
+7
*
R600/SI: Don't verify constant bus usage of flag ops
Matt Arsenault
2014-12-11
1
-2
/
+10
*
R600/SI: Set 20-bit immediate byte offset for SMRD on VI
Marek Olsak
2014-12-07
1
-13
/
+23
*
R600/SI: Update instruction conversions for VI
Marek Olsak
2014-12-07
1
-0
/
+27
*
R600/SI: Set the ATC bit on all resource descriptors for the HSA runtime
Tom Stellard
2014-12-02
1
-4
/
+14
*
R600/SI: Various instruction format bit test cleanups
Matt Arsenault
2014-12-01
1
-52
/
+0
*
R600/SI: Add an s_mov_b32 to patterns which use the M0RegClass
Tom Stellard
2014-11-21
1
-20
/
+0
*
R600/SI: Make SIInstrInfo::isOperandLegal() more strict
Tom Stellard
2014-11-19
1
-1
/
+10
*
R600/SI: Implement areMemAccessesTriviallyDisjoint
Matt Arsenault
2014-11-19
1
-0
/
+83
[next]