summaryrefslogtreecommitdiffstats
path: root/llvm
Commit message (Collapse)AuthorAgeFilesLines
* Update FIXME.Jim Grosbach2011-07-291-5/+2
| | | | llvm-svn: 136470
* Tweak comment.Jim Grosbach2011-07-291-1/+1
| | | | llvm-svn: 136468
* Fix a case where, when trying to track tied operands, we'd accidentally ↵Owen Anderson2011-07-291-3/+3
| | | | | | overwrite our mapping. llvm-svn: 136467
* Switch the CMake edis build over to add_llvm_library_dependenciesDouglas Gregor2011-07-291-3/+8
| | | | llvm-svn: 136463
* Avoid undefined behaviour if somehow NUM_GRAPHS equals 2^32 (orDuncan Sands2011-07-291-10/+6
| | | | | | | whatever the size of unsigned is), though this can't actually occur for any integer value of NUM_NODES. llvm-svn: 136460
* Teach Path::GetCurrentDirectory to use $PWD, to support users who like to doNick Lewycky2011-07-291-2/+5
| | | | | | | | | | | | screwy things by setting PWD != getcwd(). For example, some developers I know will use this to control the value in gcc's DW_AT_comp_dir value in debug output. With this patch, that trick will now work on clang too. The only other effect of this change is that the static analysis will now respect $PWD when reporting the directory of the files in its HTML output. I think that's fine. llvm-svn: 136459
* Fix a lot of typos, improve (but not necessarily fix) grammaros and reflow someNick Lewycky2011-07-292-24/+22
| | | | | | lines. No functionality change. llvm-svn: 136458
* Misc optimizer+codegen work for 'cmpxchg' and 'atomicrmw'. They appear to beEli Friedman2011-07-2911-41/+263
| | | | | | | | | working on x86 (at least for trivial testcases); other architectures will need more work so that they actually emit the appropriate instructions for orderings stricter than 'monotonic'. (As far as I can tell, the ARM, PPC, Mips, and Alpha backends need such changes.) llvm-svn: 136457
* Remove incEdgeWeight and decEdgeWeight. Set edge weight directly to avoidJakub Staszak2011-07-291-30/+35
| | | | | | rounding errors. llvm-svn: 136456
* Fix the MCDisassembler dependencies. These were just absurdly wrong.Chandler Carruth2011-07-291-34/+7
| | | | | | | | | | | | First off, only depend on the actual MC-ized disassemblers in the targets, not all of the libraries those in turn depend on. Second off, only depend on those MC-ized disassemblers for targets we're building. This should fix builds of fewer than all targets. llvm-svn: 136455
* Fix two tests that I crashed in the previous commits. The mask eltsBruno Cardoso Lopes2011-07-292-6/+18
| | | | | | on the second half must be reindexed. llvm-svn: 136454
* Match VPERMIL masks more strictly and update the target specific maskBruno Cardoso Lopes2011-07-292-7/+27
| | | | | | generation to always catch the weird cases. llvm-svn: 136453
* Add DecodeShuffle shuffle support for VPERMIPD variantesBruno Cardoso Lopes2011-07-294-29/+60
| | | | llvm-svn: 136452
* Add v8i32 and v4i64 vpermil patternsBruno Cardoso Lopes2011-07-292-0/+17
| | | | llvm-svn: 136451
* Fix a bug while generating target specific VPERMIL masks: skipBruno Cardoso Lopes2011-07-291-4/+12
| | | | | | undef mask elements. This fixes PR10529. llvm-svn: 136450
* Enable usage of SSE4 extracts and inserts in their 128-bit AVX forms.Bruno Cardoso Lopes2011-07-291-39/+29
| | | | | | Also tidy up code a bit. llvm-svn: 136449
* Cleanup PALIGNR handling and remove the old palign pattern fragment.Bruno Cardoso Lopes2011-07-294-43/+23
| | | | | | | Also make PALIGNR masks to don't match 256-bits, which isn't supported It's also a step to solve PR10489 llvm-svn: 136448
* Use the pointer type size.Bill Wendling2011-07-291-1/+1
| | | | | | With this, we can now compile a simple EH program. llvm-svn: 136446
* And now something that compiles...Bill Wendling2011-07-291-1/+1
| | | | llvm-svn: 136445
* Make sure to sext or trunc the result from the register.Bill Wendling2011-07-291-2/+2
| | | | llvm-svn: 136444
* Transfer implicit operands in NEONMoveFixPass.Jakob Stoklund Olesen2011-07-292-10/+41
| | | | | | | | | Later passes /are/ using this information when running the register scavenger. This fixes the second problem in PR10520. llvm-svn: 136440
* Add -verify-arm-pseudo-expand.Jakob Stoklund Olesen2011-07-292-1/+8
| | | | | | | | | | This hidden llc option runs the machine code verifier after expanding ARM pseudo-instructions, but before if-conversion. The machine code verifier is much better at pointing out liveness errors that can trip up the register scavenger. llvm-svn: 136439
* Make sure to correctly clear the exact/nuw/nsw flags off of shifts when they ↵Eli Friedman2011-07-292-21/+44
| | | | | | are combined together. <rdar://problem/9859829> llvm-svn: 136435
* Rewrite the CMake build to use explicit dependencies between libraries,Chandler Carruth2011-07-2987-150/+675
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | specified in the same file that the library itself is created. This is more idiomatic for CMake builds, and also allows us to correctly specify dependencies that are missed due to bugs in the GenLibDeps perl script, or change from compiler to compiler. On Linux, this returns CMake to a place where it can relably rebuild several targets of LLVM. I have tried not to change the dependencies from the ones in the current auto-generated file. The only places I've really diverged are in places where I was seeing link failures, and added a dependency. The goal of this patch is not to start changing the dependencies, merely to move them into the correct location, and an explicit form that we can control and change when necessary. This also removes a serialization point in the build because we don't have to scan all the libraries before we begin building various tools. We no longer have a step of the build that regenerates a file inside the source tree. A few other associated cleanups fall out of this. This isn't really finished yet though. After talking to dgregor he urged switching to a single CMake macro to construct libraries with both sources and dependencies in the arguments. Migrating from the two macros to that style will be a follow-up patch. Also, llvm-config is still generated with GenLibDeps.pl, which means it still has slightly buggy dependencies. The internal CMake 'llvm-config-like' macro uses the correct explicitly specified dependencies however. A future patch will switch llvm-config generation (when using CMake) to be based on these deps as well. This may well break Windows. I'm getting a machine set up now to dig into any failures there. If anyone can chime in with problems they see or ideas of how to solve them for Windows, much appreciated. llvm-svn: 136433
* Remove extra semicolon.Jakub Staszak2011-07-291-1/+1
| | | | llvm-svn: 136432
* Enhance the fixed-length decoder emitter to support tied operands.Owen Anderson2011-07-281-1/+13
| | | | llvm-svn: 136431
* Visit the landingpad instruction.Bill Wendling2011-07-281-1/+35
| | | | | | | | | This generates the correct SDNodes for the landingpad instruction. It makes an assumption that the result of the landingpad instruction has at least two values. And that the first value is a pointer to the exception object and the second value is the "selector." llvm-svn: 136430
* Add the AddLandingPadInfo function.Bill Wendling2011-07-282-0/+39
| | | | | | | AddLandingPadInfo takes a landingpad instruction and grabs all of the information from it that it needs for EH table generation. llvm-svn: 136429
* Change LBH_TAKEN_WEIGHT to 124 (from 128). Right now, sum ofJakub Staszak2011-07-281-4/+4
| | | | | | | LBH_TAKEN_WEIGHT + LBH_NONTAKEN_WEIGHT = 128 which in _most_ cases reduce number of rounding errors. llvm-svn: 136428
* PLD and PLI are not predicable in ARM mode.Jim Grosbach2011-07-281-0/+1
| | | | llvm-svn: 136427
* CBZ/CBNZ are Thumb2 only. No need for ARM mode tests for them.Jim Grosbach2011-07-281-4/+0
| | | | llvm-svn: 136408
* ARM assembly parsing and encoding for BLX (immediate).Jim Grosbach2011-07-283-3/+16
| | | | | | | | Add parsing support for BLX (immediate). Since the register operand version is predicated and the label operand version is not, we have to use some special handling to get the operand list right for matching. llvm-svn: 136406
* Enhance the fixed-length decoder emitter to support parsing scattered fields.Owen Anderson2011-07-282-80/+111
| | | | llvm-svn: 136405
* LangRef and basic memory-representation/reading/writing for 'cmpxchg' andEli Friedman2011-07-2818-43/+946
| | | | | | | | | | | | | | | | | | | | | 'atomicrmw' instructions, which allow representing all the current atomic rmw intrinsics. The allowed operands for these instructions are heavily restricted at the moment; we can probably loosen it a bit, but supporting general first-class types (where it makes sense) might get a bit complicated, given how SelectionDAG works. As an initial cut, these operations do not support specifying an alignment, but it would be possible to add if we think it's useful. Specifying an alignment lower than the natural alignment would be essentially impossible to support on anything other than x86, but specifying a greater alignment would be possible. I can't think of any useful optimizations which would use that information, but maybe someone else has ideas. Optimizer/codegen support coming soon. llvm-svn: 136404
* If run with -debug give more information about Cyclic Probability.Jakub Staszak2011-07-281-0/+2
| | | | llvm-svn: 136403
* Heuristics are in descending priority now. If we use one of them, skip the rest.Jakub Staszak2011-07-281-13/+23
| | | | llvm-svn: 136402
* Handle REG_SEQUENCE with implicitly defined operands.Jakob Stoklund Olesen2011-07-282-0/+27
| | | | | | | | | | | | | Code like that would only be produced by bugpoint, but we should still handle it correctly. When a register is defined by a REG_SEQUENCE of undefs, the register itself is undef. Previously, we would create a register with uses but no defs. Fixes part of PR10520. llvm-svn: 136401
* Remove obsolete FIXME reference in comment.Jim Grosbach2011-07-281-1/+1
| | | | llvm-svn: 136400
* ARM assembly parsing and encoding for BFC and BFI.Jim Grosbach2011-07-283-2/+130
| | | | | | | Add parsing support that handles converting the lsb+width source into the odd way we represent the instruction (an inverted bitfield mask). llvm-svn: 136399
* Add InEdges (edges from header to the loop) in Loop Branch Heuristics, soJakub Staszak2011-07-281-0/+17
| | | | | | | there is no frequency difference whether condition is in the header or in the latch. llvm-svn: 136398
* Use ArrayRef instead of requiring an std::vector.Bill Wendling2011-07-282-6/+8
| | | | llvm-svn: 136396
* The personality function should be a Function* and not just a Value*.Bill Wendling2011-07-287-16/+23
| | | | llvm-svn: 136392
* Fix Clang attribute reader tblgen output for a corresponding fix on the ↵Douglas Gregor2011-07-282-4/+7
| | | | | | Clang side llvm-svn: 136390
* Reverse order of RS_Split live ranges under -compact-regions.Jakob Stoklund Olesen2011-07-281-3/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | There are two conflicting strategies in play: - Under high register pressure, we want to assign large live ranges first. Smaller live ranges are easier to place afterwards. - Live range splitting is guided by interference, so splitting should be deferred until interference is as realistic as possible. With the recent changes to the live range stages, and with compact regions enabled, it is less traumatic to split a live range too early. If some of the split products were too big, they can often be split again. By reversing the RS_Split order, we get this queue order: 1. Normal live ranges, large to small. 2. RS_Split live ranges, large to small. The large-to-small order improves RAGreedy's puzzle solving skills under high register pressure. It may cause a bit more iterated splitting, but we handle that better now. With this change, -compact-regions is mostly an improvement on SPEC. llvm-svn: 136388
* Initial code to convert ResumeInsts into calls to _Unwind_Resume.Bill Wendling2011-07-281-1/+57
| | | | | | This should be the only code necessary for DWARF EH prepare. llvm-svn: 136387
* Fix stupid mistake from commit 136381.Jakub Staszak2011-07-281-1/+1
| | | | llvm-svn: 136384
* Speed up BlockFrequencyInfo a little bit.Jakub Staszak2011-07-281-1/+3
| | | | llvm-svn: 136381
* Add fixme.Jim Grosbach2011-07-281-1/+5
| | | | llvm-svn: 136375
* Use version 402 for the GCDA files when compiling for Apple.Bill Wendling2011-07-281-0/+4
| | | | llvm-svn: 136369
* Update comments.Owen Anderson2011-07-281-18/+10
| | | | llvm-svn: 136367
OpenPOWER on IntegriCloud