| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Fix test. | Eli Friedman | 2011-05-17 | 1 | -1/+3 |
| | | | | | llvm-svn: 131444 | ||||
| * | Add target triple so test doesn't fail on Windows machines. | Evan Cheng | 2011-05-17 | 1 | -1/+1 |
| | | | | | llvm-svn: 131439 | ||||
| * | Add x86 fast-isel for calls returning first-class aggregates. rdar://9435872. | Eli Friedman | 2011-05-17 | 1 | -1/+1 |
| | | | | | llvm-svn: 131438 | ||||
| * | Teach LiveInterval::isZeroLength about null SlotIndexes. | Jakob Stoklund Olesen | 2011-05-16 | 2 | -15/+4 |
| | | | | | | | | | | | | | When instructions are deleted, they leave tombstone SlotIndex entries. The isZeroLength method should ignore these null indexes. This causes RABasic to sometimes spill a callee-saved register in the abi-isel.ll test, so don't run that test with -regalloc=basic. Prioritizing register allocation according to spill weight can cause more registers to be used. llvm-svn: 131436 | ||||
| * | Remove dead code. Fix associated test to use FileCheck. | Eli Friedman | 2011-05-16 | 1 | -1/+14 |
| | | | | | llvm-svn: 131424 | ||||
| * | Make fast-isel work correctly s/uadd.with.overflow intrinsics. | Eli Friedman | 2011-05-16 | 1 | -0/+22 |
| | | | | | llvm-svn: 131420 | ||||
| * | Basic fast-isel of extractvalue. Not too helpful on its own, given the IR ↵ | Eli Friedman | 2011-05-16 | 1 | -0/+26 |
| | | | | | | | clang generates for cases like this, but it should become more useful soon. llvm-svn: 131417 | ||||
| * | sets bit 0 of the function address of thumb function in .symtab | Rafael Espindola | 2011-05-16 | 1 | -0/+20 |
| | | | | | | | | | | | ("T is 1 if the target symbol S has type STT_FUNC and the symbol addresses a Thumb instruction ;it is 0 otherwise." from "ELF for the ARM Architecture" 4.7.1.2) Patch by Koan-Sin Tan! llvm-svn: 131406 | ||||
| * | Don't do tail calls in a function that call setjmp. The stack might be | Rafael Espindola | 2011-05-16 | 1 | -0/+16 |
| | | | | | | | corrupted when setjmp returns again. llvm-svn: 131399 | ||||
| * | Disable test harder. | Benjamin Kramer | 2011-05-14 | 1 | -1/+2 |
| | | | | | llvm-svn: 131363 | ||||
| * | Disable this test while I revise it. rdar://problem/9267970 | Stuart Hastings | 2011-05-14 | 1 | -1/+1 |
| | | | | | llvm-svn: 131350 | ||||
| * | SimplifyCFG: Use ComputeMaskedBits to prune dead cases from switch instructions. | Benjamin Kramer | 2011-05-14 | 1 | -0/+38 |
| | | | | | llvm-svn: 131345 | ||||
| * | Avoid combining GEPs that might overflow at runtime. | Stuart Hastings | 2011-05-14 | 1 | -0/+19 |
| | | | | | | | | | rdar://problem/9267970 Patch by Julien Lerouge! llvm-svn: 131339 | ||||
| * | Don't produce a vmovntdq if we don't have AVX support. | Rafael Espindola | 2011-05-14 | 1 | -0/+12 |
| | | | | | llvm-svn: 131330 | ||||
| * | Move test. | Rafael Espindola | 2011-05-13 | 1 | -0/+0 |
| | | | | | llvm-svn: 131315 | ||||
| * | Move test. | Rafael Espindola | 2011-05-13 | 1 | -0/+0 |
| | | | | | llvm-svn: 131314 | ||||
| * | Move platform-dependent test to appropriate directory. | Galina Kistanova | 2011-05-13 | 2 | -0/+5 |
| | | | | | llvm-svn: 131302 | ||||
| * | Make codegen able to handle values of empty types. This is one way | Rafael Espindola | 2011-05-13 | 1 | -0/+81 |
| | | | | | | | to fix PR9900. I will keep it open until sable is able to comment on it. llvm-svn: 131294 | ||||
| * | Since I can't reproduce the failures from 131261, re-trying with a | Stuart Hastings | 2011-05-13 | 1 | -7/+10 |
| | | | | | | | simplified version. <rdar://problem/9298790> llvm-svn: 131274 | ||||
| * | Revert 131266 and 131261 due to buildbot complaints. | Stuart Hastings | 2011-05-13 | 1 | -10/+7 |
| | | | | | | | rdar://problem/9298790 llvm-svn: 131269 | ||||
| * | Tweak 131261 (thumb2-cbnz.ll) to generate the intended cbnz. | Stuart Hastings | 2011-05-13 | 1 | -8/+10 |
| | | | | | | | rdar://problem/9298790 llvm-svn: 131266 | ||||
| * | Non-fast-isel followup to 129634; correctly handle branches controlled | Stuart Hastings | 2011-05-12 | 1 | -3/+4 |
| | | | | | | | | | | | | | | by non-CMP expressions. The executable test case (129821) would test this as well, if we had an "-O0 -disable-arm-fast-isel" LLVM-GCC tester. Alas, the ARM assembly would be very difficult to check with FileCheck. The thumb2-cbnz.ll test is affected; it generates larger code (tst.w vs. cmp #0), but I believe the new version is correct. rdar://problem/9298790 llvm-svn: 131261 | ||||
| * | Correction. Use explicit target triple in the test. | Galina Kistanova | 2011-05-12 | 1 | -1/+1 |
| | | | | | llvm-svn: 131252 | ||||
| * | Re-enable branchfolding common code hoisting optimization. Fixed a liveness ↵ | Evan Cheng | 2011-05-12 | 1 | -0/+28 |
| | | | | | | | test bug and also taught it to update liveins. llvm-svn: 131241 | ||||
| * | Move this test to CodeGen/Thumb. rdar://problem/9416774 | Stuart Hastings | 2011-05-11 | 1 | -1/+1 |
| | | | | | llvm-svn: 131196 | ||||
| * | Identify end of prologue (and beginning of function body) using ↵ | Devang Patel | 2011-05-11 | 1 | -0/+55 |
| | | | | | | | DW_LNS_set_prologue_end line table opcode. llvm-svn: 131194 | ||||
| * | Reduced test case. rdar://problem/9416774 | Stuart Hastings | 2011-05-11 | 1 | -36/+36 |
| | | | | | llvm-svn: 131191 | ||||
| * | Fix encoding of Thumb BLX register instructions. Patch by Koan-Sin Tan. | Owen Anderson | 2011-05-11 | 1 | -0/+2 |
| | | | | | llvm-svn: 131189 | ||||
| * | And lo, I was given a testcase for 131152. rdar://problem/9416774 | Stuart Hastings | 2011-05-11 | 1 | -0/+60 |
| | | | | | llvm-svn: 131184 | ||||
| * | Fixes a bug in the DAGCombiner. LoadSDNodes have two values (data, chain). | Nadav Rotem | 2011-05-11 | 4 | -11/+28 |
| | | | | | | | | | If there is a store after the load node, then there is a chain, which means that there is another user. Thus, asking hasOneUser would fail. Instead we ask hasNUsesOfValue on the 'data' value. llvm-svn: 131183 | ||||
| * | Add custom lowering of X86 vector SRA/SRL/SHL when the shift amount is a ↵ | Nadav Rotem | 2011-05-11 | 1 | -0/+142 |
| | | | | | | | splat vector. llvm-svn: 131179 | ||||
| * | Revert 131172 as it is causing clang to miscompile itself. I will try | Rafael Espindola | 2011-05-11 | 1 | -28/+0 |
| | | | | | | | to provide a reduced testcase. llvm-svn: 131176 | ||||
| * | Add a late optimization to BranchFolding that hoist common instruction sequences | Evan Cheng | 2011-05-11 | 1 | -0/+28 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | at the start of basic blocks to their common predecessor. It's actually quite common (e.g. about 50 times in JM/lencod) and has shown to be a nice code size benefit. e.g. pushq %rax testl %edi, %edi jne LBB0_2 ## BB#1: xorb %al, %al popq %rdx ret LBB0_2: xorb %al, %al callq _foo popq %rdx ret => pushq %rax xorb %al, %al testl %edi, %edi je LBB0_2 ## BB#1: callq _foo LBB0_2: popq %rdx ret rdar://9145558 llvm-svn: 131172 | ||||
| * | Add triple. | Rafael Espindola | 2011-05-10 | 1 | -1/+1 |
| | | | | | llvm-svn: 131169 | ||||
| * | Produce a __debug_frame section on darwin ARM when appropriate. | Rafael Espindola | 2011-05-10 | 1 | -2/+2 |
| | | | | | llvm-svn: 131151 | ||||
| * | On MachO, unlike ELF, there should be no relocation to produce the CIE pointer. | Rafael Espindola | 2011-05-10 | 1 | -0/+38 |
| | | | | | llvm-svn: 131149 | ||||
| * | The EH symbols are only needed in eh_frame, not debug_frame. | Rafael Espindola | 2011-05-10 | 1 | -0/+18 |
| | | | | | llvm-svn: 131146 | ||||
| * | Use .cfi_sections to put the unwind info in .debug_frame when possible. With | Rafael Espindola | 2011-05-10 | 1 | -0/+18 |
| | | | | | | | | | | this clang will use .debug_frame in, for example, clang -g -c -m32 test.c This matches gcc's behaviour. It looks like .debug_frame is a bit bigger than .eh_frame, but has the big advantage of not being allocated. llvm-svn: 131140 | ||||
| * | In a debug_frame the cfi offset is to the start of the debug_frame section! | Rafael Espindola | 2011-05-10 | 1 | -6/+2 |
| | | | | | llvm-svn: 131129 | ||||
| * | PTX: add test cases for cvt, fneg, and selp | Justin Holewinski | 2011-05-10 | 3 | -0/+274 |
| | | | | | | | Patch by Dan Bailey llvm-svn: 131128 | ||||
| * | Add support for producing .deubg_frame sections. | Rafael Espindola | 2011-05-10 | 1 | -0/+42 |
| | | | | | llvm-svn: 131121 | ||||
| * | X86: Add a bunch of peeps for add and sub of SETB. | Benjamin Kramer | 2011-05-08 | 2 | -4/+85 |
| | | | | | | | | | | | | | | | | | "b + ((a < b) ? 1 : 0)" compiles into cmpl %esi, %edi adcl $0, %esi instead of cmpl %esi, %edi sbbl %eax, %eax andl $1, %eax addl %esi, %eax This saves a register, a false dependency on %eax (Intel's CPUs still don't ignore it) and it's shorter. llvm-svn: 131070 | ||||
| * | The comparision "max(x,y)==x" is equivalent to "x>=y". Since the max is | Duncan Sands | 2011-05-07 | 1 | -0/+36 |
| | | | | | | | | often expressed as "x >= y ? x : y", there is a good chance we can extract the existing "x >= y" from it and use that as a replacement for "max(x,y)==x". llvm-svn: 131049 | ||||
| * | Emit a proper error message when register allocators run out of registers. | Jakob Stoklund Olesen | 2011-05-06 | 1 | -0/+17 |
| | | | | | | | | This can't be just an assertion, users can always write impossible inline assembly. Such an assembly statement should be included in the error message. llvm-svn: 131024 | ||||
| * | It's valid to take the blockaddress of a different function, so remove this | Nick Lewycky | 2011-05-06 | 1 | -0/+30 |
| | | | | | | | | assert in the bitcode writer. No change needed because the ValueEnumerator holds a whole-module numbering anyhow. Fixes PR9857! llvm-svn: 131016 | ||||
| * | Move few target-dependant tests to appropriate directories. | Galina Kistanova | 2011-05-06 | 6 | -0/+15 |
| | | | | | llvm-svn: 131002 | ||||
| * | Pass -disable-cfi to llc. | Rafael Espindola | 2011-05-06 | 5 | -5/+5 |
| | | | | | llvm-svn: 130999 | ||||
| * | Pass -disable-cfi. | Rafael Espindola | 2011-05-06 | 1 | -1/+1 |
| | | | | | llvm-svn: 130995 | ||||
| * | PTX: add PTX 2.3 language target | Justin Holewinski | 2011-05-06 | 1 | -0/+1 |
| | | | | | | | Patch by Wei-Ren Chen llvm-svn: 130980 | ||||
| * | Fix PR9820: a read-only call differs from a load in that a load doesn't | Duncan Sands | 2011-05-06 | 1 | -0/+26 |
| | | | | | | | | return the pointer being dereferenced, it returns the pointee, but a call might return the pointer itself. llvm-svn: 130979 | ||||

