| Commit message (Expand) | Author | Age | Files | Lines |
* | Fix up -fstack-protector on linux to use the segment | Eric Christopher | 2010-07-06 | 1 | -0/+25 |
* | Implement the "linker_private_weak" linkage type. This will be used for | Bill Wendling | 2010-07-01 | 1 | -1/+1 |
* | Eliminate the other half of the BRCOND optimization, and update | Dan Gohman | 2010-06-24 | 1 | -433/+0 |
* | Eliminate the first have of the optimization which eliminates BRCOND | Dan Gohman | 2010-06-24 | 1 | -21/+0 |
* | Remove the local register allocator. | Jakob Stoklund Olesen | 2010-06-15 | 5 | -5/+0 |
* | Fix some latency computation bugs: if the use is not a machine opcode do not ... | Evan Cheng | 2010-05-28 | 1 | -2/+2 |
* | Only use clairvoyance when defining a register, and then only if it has one use. | Jakob Stoklund Olesen | 2010-05-17 | 1 | -9/+5 |
* | Take allocation hints from copy instructions to/from physregs. | Jakob Stoklund Olesen | 2010-05-13 | 1 | -2/+2 |
* | Enable a bunch more -regalloc=fast tests | Jakob Stoklund Olesen | 2010-05-12 | 5 | -4/+14 |
* | Implement builtin_return_address(x) and builtin_frame_address(x) | Dale Johannesen | 2010-05-03 | 1 | -0/+24 |
* | Remove the -enable-sjlj-eh option, which doesn't do anything. | Duncan Sands | 2010-05-02 | 1 | -1/+1 |
* | on darwin empty functions need to codegen into something of non-zero length, | Chris Lattner | 2010-04-26 | 1 | -1/+1 |
* | Bill's change in r95336 broke empty aggregates embedded | Chris Lattner | 2010-04-20 | 1 | -1/+10 |
* | Start function numbering at 0. | Dan Gohman | 2010-04-17 | 1 | -2/+2 |
* | add newlines at the end of files. | Chris Lattner | 2010-04-07 | 1 | -1/+1 |
* | Split big test into multiple directories to cater to | Dale Johannesen | 2010-04-07 | 1 | -0/+33 |
* | After trivial coalescing, the MI being visited may have become a copy. Avoid ... | Evan Cheng | 2010-04-02 | 1 | -0/+70 |
* | add some nounwinds | Chris Lattner | 2010-03-28 | 1 | -12/+12 |
* | this takes an insane amount of time to run, disable it for now (PR6727) | Chris Lattner | 2010-03-28 | 1 | -1/+3 |
* | Turn calls to copysignl into an FCOPYSIGN node. Handle FCOPYSIGN nodes | Duncan Sands | 2010-03-14 | 1 | -0/+7 |
* | fix AsmPrinter::GetBlockAddressSymbol to always return a unique | Chris Lattner | 2010-03-14 | 1 | -4/+4 |
* | get MMI out of the label uniquing business, just go to MCContext | Chris Lattner | 2010-03-14 | 1 | -2/+2 |
* | Enable machine cse pass. | Evan Cheng | 2010-03-10 | 1 | -6/+6 |
* | The address of an indirect call must be in R12 on Darwin. | Dale Johannesen | 2010-03-09 | 1 | -0/+19 |
* | add some random nounwinds. | Chris Lattner | 2010-02-28 | 1 | -1/+1 |
* | Use the right floating point load/store instructions in PPCInstrInfo::foldMem... | Jakob Stoklund Olesen | 2010-02-26 | 1 | -0/+433 |
* | Change the scheduler from adding nodes in allnodes order | Chris Lattner | 2010-02-24 | 1 | -2/+2 |
* | When emitting an instruction which depends on both a post-incremented | Dan Gohman | 2010-02-22 | 1 | -0/+32 |
* | add some no-unwinds, other minor cleanups. | Chris Lattner | 2010-02-21 | 3 | -30/+30 |
* | add a triple so that this doesn't fail due to linux/ppc register printing | Chris Lattner | 2010-02-21 | 1 | -1/+1 |
* | filecheckize and add nouwinds. | Chris Lattner | 2010-02-21 | 1 | -9/+16 |
* | Make g5 target explicit; scheduling affects register choice. | Dale Johannesen | 2010-02-16 | 1 | -1/+1 |
* | Adjust register numbers in tests to compensate for the | Dale Johannesen | 2010-02-16 | 4 | -14/+14 |
* | When save/restoring CR at prolog/epilog, in a large | Dale Johannesen | 2010-02-12 | 1 | -0/+30 |
* | Fix alignment on ppc linux. This fixes the build of crtend.o | Rafael Espindola | 2010-02-06 | 1 | -9/+40 |
* | Make test more fucused eliminating extraneous bits. | Bill Wendling | 2010-02-05 | 1 | -44/+5 |
* | An empty global constant (one of size 0) may have a section immediately | Bill Wendling | 2010-02-05 | 1 | -0/+50 |
* | Reapply 95050 with a tweak to check the register class. | Dale Johannesen | 2010-02-03 | 1 | -2/+2 |
* | Test revert 95050; there's a good chance it's causing | Dale Johannesen | 2010-02-02 | 1 | -2/+2 |
* | Make local RA smarter about reusing input register of a copy | Dale Johannesen | 2010-02-02 | 1 | -2/+2 |
* | Give AsmPrinter the most common expected implementation of | Chris Lattner | 2010-01-28 | 1 | -1/+1 |
* | Attempt to unbreak test on Linux. Chris, please check. | Daniel Dunbar | 2010-01-25 | 1 | -3/+3 |
* | stop testing for invalid output. | Chris Lattner | 2010-01-23 | 1 | -1/+0 |
* | testcase for r94095 | Chris Lattner | 2010-01-21 | 1 | -0/+22 |
* | Now that we have everything nicely factored (e.g. asmprinter is not | Chris Lattner | 2010-01-19 | 1 | -1/+1 |
* | remove uses of deprecated functions, this generates slightly | Chris Lattner | 2010-01-13 | 1 | -4/+4 |
* | Delete useless trailing semicolons. | Dan Gohman | 2010-01-05 | 1 | -1/+1 |
* | Do better with physical reg operands (typically, from inline asm) | Dale Johannesen | 2009-12-16 | 1 | -2/+2 |
* | ProcessImplicitDefs should watch out for invalidated iterator and extra impli... | Evan Cheng | 2009-11-25 | 1 | -0/+56 |
* | Do not store R31 into the caller's link area on PPC. | Dale Johannesen | 2009-11-24 | 4 | -26/+26 |