- May 01, 2010
-
-
Dale Johannesen authored
preventing the emission of the NOP on Darwin for a function with no actual code. From timberwolfmc with TEST=optllcdbg. llvm-svn: 102843
-
Anton Korobeynikov authored
when needed. This fixes PR7001 llvm-svn: 102838
-
Dan Gohman authored
handles argument lowering anyway, so there's no need for special casing here. llvm-svn: 102828
-
Dan Gohman authored
llvm-svn: 102826
-
Dan Gohman authored
try to put a kill flag on a DBG_INFO instruction. llvm-svn: 102820
-
Dale Johannesen authored
Seen in SingleSrc/Benchmarks/Misc/flops with TEST=optllcdbg. 7929951. llvm-svn: 102819
-
Dan Gohman authored
llvm-svn: 102817
-
Dan Gohman authored
modified. llvm-svn: 102816
-
Dan Gohman authored
code, and to eliminate the need for the SelectionDAGBuilder state to be live during CodeGenAndEmitDAG calls. Call SDB->clear() before CodeGenAndEmitDAG calls instead of before it, and move the CurDAG->clear() out of SelectionDAGBuilder, which doesn't own the DAG, and into CodeGenAndEmitDAG. llvm-svn: 102814
-
Dan Gohman authored
llvm-svn: 102810
-
Dan Gohman authored
changes before doing phi lowering for switches. llvm-svn: 102809
-
Bill Wendling authored
indexes could be of a different value type. Or not even using the same SDNode for the constant (weird, I know). Compare the actual values instead of the pointers. llvm-svn: 102791
-
- Apr 30, 2010
-
-
Dan Gohman authored
instruction selection is done; it's confusing to see parts of it printed, while other parts are omitted, along the way. llvm-svn: 102771
-
Jakob Stoklund Olesen authored
call that might throw. The landing pad assumes that all registers are in stack slots. We used to spill those dirty CSRs after the call, and the stack slots would be wrong when arriving at the landing pad. llvm-svn: 102770
-
Devang Patel authored
llvm-svn: 102743
-
Dan Gohman authored
llvm-svn: 102742
-
Jakob Stoklund Olesen authored
Rounding differences causes tests to fail on Linux. llvm-svn: 102729
-
Dan Gohman authored
comparisons sign-extended to a different bitwidth than the comparison operands. llvm-svn: 102721
-
Dan Gohman authored
llvm-svn: 102680
-
Dan Gohman authored
easy cases. llvm-svn: 102678
-
Jakob Stoklund Olesen authored
of different register classes. e.g. %reg1048:3<def> = EXTRACT_SUBREG %RAX<kill>, 3 Where %reg1048 is a GR32 register. This is not impossible to handle, but it is pretty hard and very rare. This should unbreak the dragonegg builder. llvm-svn: 102672
-
Dan Gohman authored
llvm-svn: 102666
-
Jakob Stoklund Olesen authored
llvm-svn: 102663
-
- Apr 29, 2010
-
-
Devang Patel authored
llvm-svn: 102661
-
Dale Johannesen authored
llvm-svn: 102657
-
Devang Patel authored
##DEBUG_VALUE: runOnMachineFunction:this <- RDI+0 ##DEBUG_VALUE: runOnMachineFunction:fn <- RSI+0 ##DEBUG_VALUE: DeadDefs <- undef ## SimpleRegisterCoalescing.cpp:2706 ##DEBUG_VALUE: getRegInfo:this <- [%rsp+$56]+$0 ##DEBUG_VALUE: getTarget:this <- [%rsp+$56]+$0 llvm-svn: 102655
-
Evan Cheng authored
llvm-svn: 102654
-
Devang Patel authored
llvm-svn: 102653
-
Evan Cheng authored
llvm-svn: 102606
-
Evan Cheng authored
llvm-svn: 102602
-
Evan Cheng authored
llvm-svn: 102590
-
Evan Cheng authored
llvm-svn: 102585
-
Dan Gohman authored
llvm-svn: 102584
-
Evan Cheng authored
llvm-svn: 102581
-
Evan Cheng authored
llvm-svn: 102573
-
Evan Cheng authored
- Also, update dbg_value is the value is being re-matted from a frame slot, e.g. fixed slots for arguments. llvm-svn: 102565
-
Devang Patel authored
llvm-svn: 102558
-
Evan Cheng authored
Replace r102368 with code that's less fragile. This creates DBG_VALUE instructions for function arguments early and insert them after instruction selection is done. llvm-svn: 102554
-
- Apr 28, 2010
-
-
Evan Cheng authored
Before: DBG_VALUE %RSI, 0, !-1; dbg:SimpleRegisterCoalescing.cpp:2707 Now: DBG_VALUE %RSI, 0, !"this"; dbg:SimpleRegisterCoalescing.cpp:2707 llvm-svn: 102518
-
Chris Lattner authored
alignment of globals to the preferred alignment, but only when there is no section specified on the global (by far the common case). llvm-svn: 102515
-