| Commit message (Collapse) | Author | Age | Files | Lines | ||
|---|---|---|---|---|---|---|
| ... | ||||||
| * | Fast ISel trivially coalesces away no-op casts, so check for this when | Dan Gohman | 2010-05-14 | 1 | -4/+16 | |
| | | | | | | | setting kill flags. llvm-svn: 103832 | |||||
| * | Don't bother spilling before a return | Jakob Stoklund Olesen | 2010-05-14 | 1 | -4/+13 | |
| | | | | | llvm-svn: 103831 | |||||
| * | RegAllocLocal can count copies too | Jakob Stoklund Olesen | 2010-05-14 | 1 | -1/+4 | |
| | | | | | llvm-svn: 103830 | |||||
| * | SystemZ really does mean "has calls" and not just "adjusts stack." Go ahead and | Bill Wendling | 2010-05-14 | 1 | -5/+5 | |
| | | | | | | | | | replace the check with the appropriate predicate. Modify the testcase to reflect the correct code. (It should be saving callee-saved registers on the stack allocated by the calling fuction.) llvm-svn: 103829 | |||||
| * | Track allocatable instead of reserved regs, and never take an unallocatable ↵ | Jakob Stoklund Olesen | 2010-05-14 | 1 | -8/+10 | |
| | | | | | | | hint. llvm-svn: 103828 | |||||
| * | Don't set kill flags for instructions which the scheduler has cloned. | Dan Gohman | 2010-05-14 | 2 | -20/+34 | |
| | | | | | llvm-svn: 103827 | |||||
| * | BR is a barrier. | Dan Gohman | 2010-05-14 | 1 | -1/+1 | |
| | | | | | llvm-svn: 103826 | |||||
| * | Several tail call tests apparently rely upon this being "adjusts stack" instead | Bill Wendling | 2010-05-14 | 1 | -1/+1 | |
| | | | | | | | | of "has calls". That's probably wrong, but it needs further investigation. Revert to the original behavior until this is settled. llvm-svn: 103824 | |||||
| * | Avoid scanning the long tail of physreg operands on calls | Jakob Stoklund Olesen | 2010-05-14 | 1 | -4/+9 | |
| | | | | | llvm-svn: 103823 | |||||
| * | Do not forget to mark prcessed arguments. | Devang Patel | 2010-05-14 | 1 | -0/+1 | |
| | | | | | llvm-svn: 103822 | |||||
| * | Count coalesced copies | Jakob Stoklund Olesen | 2010-05-14 | 1 | -0/+2 | |
| | | | | | llvm-svn: 103821 | |||||
| * | Allow virtreg redefines when verifying for RegAllocFast | Jakob Stoklund Olesen | 2010-05-14 | 1 | -1/+1 | |
| | | | | | llvm-svn: 103820 | |||||
| * | This should happen if there are no calls, not if it just doesn't adjust the | Bill Wendling | 2010-05-14 | 1 | -1/+1 | |
| | | | | | | | stack. llvm-svn: 103813 | |||||
| * | Revert r103804. The comment is correct. | Bill Wendling | 2010-05-14 | 2 | -2/+2 | |
| | | | | | llvm-svn: 103808 | |||||
| * | Remove trailing whitespace | Jim Grosbach | 2010-05-14 | 1 | -3/+3 | |
| | | | | | llvm-svn: 103807 | |||||
| * | 80 column and trailing whitespace cleanup | Jim Grosbach | 2010-05-14 | 1 | -24/+25 | |
| | | | | | llvm-svn: 103806 | |||||
| * | add cmd line option to leave dbgvalues in during post-RA sceduling. Useful | Jim Grosbach | 2010-05-14 | 1 | -5/+14 | |
| | | | | | | | while debugging what's mishandled about them in the post-RA pass. llvm-svn: 103805 | |||||
| * | Fix comment. | Bill Wendling | 2010-05-14 | 1 | -1/+1 | |
| | | | | | llvm-svn: 103804 | |||||
| * | Rename "HasCalls" in MachineFrameInfo to "AdjustsStack" to better describe what | Bill Wendling | 2010-05-14 | 16 | -54/+69 | |
| | | | | | | | | | | the variable actually tracks. N.B., several back-ends are using "HasCalls" as being synonymous for something that adjusts the stack. This isn't 100% correct and should be looked into. llvm-svn: 103802 | |||||
| * | Lowering of atomic instructions can result in operands being | Dan Gohman | 2010-05-14 | 1 | -1/+7 | |
| | | | | | | | | used more than once. If ISel had put a kill flag on one of them, it's not valid to transfer the kill flag to each new instance. llvm-svn: 103799 | |||||
| * | Add support to preserve type info for the variables that are removed by the ↵ | Devang Patel | 2010-05-14 | 3 | -16/+42 | |
| | | | | | | | optimizer. llvm-svn: 103798 | |||||
| * | When verifying two-address instructions, check the following: | Jakob Stoklund Olesen | 2010-05-14 | 2 | -12/+21 | |
| | | | | | | | | | | - Kill is implicit when use and def registers are identical. - Only virtual registers can differ. Add a -verify-fast-regalloc to run the verifier before the fast allocator. llvm-svn: 103797 | |||||
| * | Fix so "int3" is correctly accepted, added "into" and fixed "int" with an | Kevin Enderby | 2010-05-14 | 1 | -2/+4 | |
| | | | | | | | argument, like "int $4", to not get an Assertion error. llvm-svn: 103791 | |||||
| * | Model VLD*_UPD and VLD*odd_UPD pair with REG_SEQUENCE. | Evan Cheng | 2010-05-14 | 2 | -26/+88 | |
| | | | | | llvm-svn: 103790 | |||||
| * | MC/Mach-O/x86_64: Darwin's special "signed_N" relocation types should only be | Daniel Dunbar | 2010-05-14 | 1 | -23/+23 | |
| | | | | | | | used to replace a normal relocation, not a reference to a GOT entry. llvm-svn: 103789 | |||||
| * | Simplify the handling of physreg defs and uses in RegAllocFast. | Jakob Stoklund Olesen | 2010-05-14 | 1 | -109/+72 | |
| | | | | | | | | This adds extra security against using clobbered physregs, and it adds kill markers to physreg uses. llvm-svn: 103784 | |||||
| * | Set isTerminator on TRAP instructions. | Dan Gohman | 2010-05-14 | 2 | -2/+2 | |
| | | | | | llvm-svn: 103778 | |||||
| * | Don't use isBarrier for the PowerPC sync instruction. isBarrier is for | Dan Gohman | 2010-05-14 | 1 | -1/+0 | |
| | | | | | | | control barriers, not memory ordering barriers. llvm-svn: 103777 | |||||
| * | Add mayLoad and mayStore flags to instructions which missed them. | Dan Gohman | 2010-05-14 | 2 | -2/+15 | |
| | | | | | llvm-svn: 103776 | |||||
| * | XFAIL the test I added with vg_leak, apparently it is the first and only llc | Daniel Dunbar | 2010-05-14 | 1 | -0/+2 | |
| | | | | | | | | -filetype=obj test, and -filetype=obj leaks a few objects. Added a FIXME, we need to sort out the ownership model for the various MC objects. llvm-svn: 103769 | |||||
| * | Inline Asm: Ensure buffer is newline terminated to match how the text is ↵ | Daniel Dunbar | 2010-05-14 | 1 | -0/+14 | |
| | | | | | | | | | printed. - This is a hack, but I can't decide the best place to handle this. Chris? llvm-svn: 103765 | |||||
| * | Enable opportunistic coalescing | Jakob Stoklund Olesen | 2010-05-14 | 1 | -7/+18 | |
| | | | | | llvm-svn: 103764 | |||||
| * | Added a QQQQ register file to model 4-consecutive Q registers. | Evan Cheng | 2010-05-14 | 6 | -52/+189 | |
| | | | | | llvm-svn: 103760 | |||||
| * | Add AsmParser support for darwin tbss directive. | Eric Christopher | 2010-05-14 | 4 | -1/+84 | |
| | | | | | | | Nothing uses this yet. llvm-svn: 103757 | |||||
| * | Fix a couple of typos. | Eric Christopher | 2010-05-14 | 1 | -2/+2 | |
| | | | | | llvm-svn: 103756 | |||||
| * | MC: Switch to completely lazy layout. | Daniel Dunbar | 2010-05-14 | 1 | -19/+35 | |
| | | | | | | | - The eliminates the last major algorithmic problem with MC. llvm-svn: 103754 | |||||
| * | MC: Extend MCAsmLayout to explicitly track which fragments have been layed ↵ | Daniel Dunbar | 2010-05-14 | 1 | -10/+44 | |
| | | | | | | | out, and enforce several invariants to LayoutFragment to ensure we only do layout in a sensible order. llvm-svn: 103753 | |||||
| * | MC: Implicitly assign section addresses when the previous fragment is layed out. | Daniel Dunbar | 2010-05-14 | 1 | -1/+11 | |
| | | | | | llvm-svn: 103752 | |||||
| * | MC: Switch MCFragment to storing the layout order index, not its index in ↵ | Daniel Dunbar | 2010-05-14 | 1 | -13/+14 | |
| | | | | | | | the file. llvm-svn: 103751 | |||||
| * | MC: Change LayoutSection() to only do the section initializiation. | Daniel Dunbar | 2010-05-14 | 1 | -25/+13 | |
| | | | | | | | Also, elimminate MCAsmLayout::set*, which are no longer needed. llvm-svn: 103750 | |||||
| * | Fix comments. | Evan Cheng | 2010-05-14 | 1 | -2/+2 | |
| | | | | | llvm-svn: 103749 | |||||
| * | Trust kill flags from isel and later passes. | Jakob Stoklund Olesen | 2010-05-14 | 1 | -4/+0 | |
| | | | | | llvm-svn: 103748 | |||||
| * | Fix an embarrassing runtime regression for RegAllocFast. | Jakob Stoklund Olesen | 2010-05-14 | 1 | -9/+31 | |
| | | | | | | | | | | | | | This loop is quadratic in the capacity for a DenseMap: while(!map.empty()) map.erase(map.begin()); Instead we now do a normal begin() - end() iteration followed by map.clear(). That also has the nice sideeffect of shrinking the map capacity on demand. llvm-svn: 103747 | |||||
| * | Adding a v8i64 512-bit vector type. This will be used to model ARM NEON ↵ | Evan Cheng | 2010-05-13 | 1 | -0/+6 | |
| | | | | | | | intrinsics which translate into a pair of vld / vst instructions that can load / store 8 consecutive 64-bit (D) registers. llvm-svn: 103746 | |||||
| * | Implement a correct ui64->f32 conversion. The old | Dale Johannesen | 2010-05-13 | 1 | -0/+36 | |
| | | | | | | | one was subject to double rounding in extreme cases. llvm-svn: 103744 | |||||
| * | Clean up RegAllocFast debug output | Jakob Stoklund Olesen | 2010-05-13 | 1 | -15/+14 | |
| | | | | | llvm-svn: 103739 | |||||
| * | MC: Move Layout{Fragment,Section} into MCAsmLayout, and add LayoutFile(). | Daniel Dunbar | 2010-05-13 | 1 | -26/+27 | |
| | | | | | llvm-svn: 103738 | |||||
| * | Teach MachineLICM and MachineSink how to clear kill flags conservatively | Dan Gohman | 2010-05-13 | 3 | -1/+27 | |
| | | | | | | | when they move instructions. llvm-svn: 103737 | |||||
| * | Add comment about the pseudo registers QQ, each of which is a pair of Q ↵ | Evan Cheng | 2010-05-13 | 1 | -0/+5 | |
| | | | | | | | registers. llvm-svn: 103731 | |||||
| * | Fix pr7110: For non-Darwin targets UnspilledCS1GPRs may include high registers. | Bob Wilson | 2010-05-13 | 1 | -1/+3 | |
| | | | | | | | Do not use those for Thumb1 functions. llvm-svn: 103730 | |||||

