summaryrefslogtreecommitdiffstats
path: root/llvm
Commit message (Collapse)AuthorAgeFilesLines
...
* enhance isRemovable to refuse to delete volatile mem transfersChris Lattner2010-11-302-8/+35
| | | | | | | now that DSE hacks on them. This fixes a regression I introduced, by generalizing DSE to hack on transfers. llvm-svn: 120445
* The VLDMQ/VSTMQ instructions are reprented as true Pseudo-insts now (i.e.,Jim Grosbach2010-11-301-5/+0
| | | | | | | no extra encoding information), so we no longer need to special case them here. llvm-svn: 120444
* Tidy up.Jim Grosbach2010-11-301-2/+3
| | | | llvm-svn: 120443
* Pseudo-ize ARM MOVPCRXJim Grosbach2010-11-302-8/+19
| | | | llvm-svn: 120442
* Delete a few no longer needed references to pseudos.Jim Grosbach2010-11-301-4/+0
| | | | llvm-svn: 120441
* Change the basic block map in LoopInfo from a std::map to a DenseMap. This ↵Cameron Zwarich2010-11-301-4/+5
| | | | | | | | is a 16% speedup running loops on test-suite + SPEC2000. Reviewed by Eric Christopher. llvm-svn: 120440
* Provide encodings for a few more load/store variants.Owen Anderson2010-11-301-4/+16
| | | | llvm-svn: 120439
* Pseudo-ize BX_CALL and friends. Remove dead instruction format classes.Jim Grosbach2010-11-303-46/+66
| | | | | | rdar://8685712 llvm-svn: 120438
* Add tests for more forms of Thumb2 loads and stores.Owen Anderson2010-11-301-0/+32
| | | | llvm-svn: 120436
* add TLI support indicating that jumps are more expensive than logical operationsChris Lattner2010-11-303-2/+23
| | | | | | and use this to disable a specific optimization. Patch by Micah Villmow! llvm-svn: 120435
* ptx: add command-line options for gpu target and ptx versionChe-Liang Chiou2010-11-302-0/+24
| | | | llvm-svn: 120423
* Not all platforms use _<func>. Duh.Eric Christopher2010-11-301-2/+2
| | | | llvm-svn: 120418
* Fix some grammar in comments I noticed.Eric Christopher2010-11-301-5/+5
| | | | llvm-svn: 120416
* This defaults to GenericDomain.Eric Christopher2010-11-301-1/+1
| | | | llvm-svn: 120415
* PR5207: Make APInt::set(), APInt::clear() and APInt::flip() return void.Jay Foad2010-11-304-33/+39
| | | | llvm-svn: 120413
* Implement a PseudoI class and transfer the sse instructions over to useEric Christopher2010-11-302-12/+15
| | | | | | it. llvm-svn: 120412
* Fix insertion point in pcmp expander.Eric Christopher2010-11-301-9/+2
| | | | | | While I'm there, clean up too many \n even for me. llvm-svn: 120411
* Fix some cleanups from my last patch.Eric Christopher2010-11-302-5/+5
| | | | llvm-svn: 120410
* Add parsing for the Thumb t_addrmode_s4 addressing mode. This can almostBill Wendling2010-11-305-14/+92
| | | | | | | | | | certainly be made more generic. But it does allow us to parse something like: ldr r3, [r2, r4] correctly in Thumb mode. llvm-svn: 120408
* ptx: add ld instructionChe-Liang Chiou2010-11-303-9/+118
| | | | | | | support register and register-immediate addressing mode todo: immediate and register-register addressing mode llvm-svn: 120407
* Rewrite the main DSE loop to be written in terms of reasoningChris Lattner2010-11-303-77/+142
| | | | | | | | | | | | | | about pairs of AA::Location's instead of looking for MemDep's "Def" predicate. This is more powerful and general, handling memset/memcpy/store all uniformly, and implementing PR8701 and probably obsoleting parts of memcpyoptimizer. This also fixes an obscure bug with init.trampoline and i8 stores, but I'm not surprised it hasn't been hit yet. Enhancing init.trampoline to carry the size that it stores would allow DSE to be much more aggressive about optimizing them. llvm-svn: 120406
* death to extraneous \n's.Chris Lattner2010-11-301-2/+1
| | | | llvm-svn: 120405
* Rewrite mwait and monitor support and custom lower arguments.Eric Christopher2010-11-304-4/+101
| | | | | | Fixes PR8573. llvm-svn: 120404
* Add a puts optimization that converts puts() to putchar('\n').Anders Carlsson2010-11-302-3/+45
| | | | llvm-svn: 120398
* Fix a typo.Anders Carlsson2010-11-301-1/+1
| | | | llvm-svn: 120394
* Rename this test to FPuts.ll since it actually tests fputs.Anders Carlsson2010-11-301-0/+0
| | | | llvm-svn: 120393
* rename a function and reduce some indentation, no functionality change.Chris Lattner2010-11-301-19/+21
| | | | llvm-svn: 120391
* Make @llvm.invariant.start not be readonly, so that it has side-effects. ThisNick Lewycky2010-11-302-2/+2
| | | | | | | unbreaks test/Transforms/InstCombine/invariant.ll which was broken by r120382. This is a fix-forward to do what I think Chris intended. llvm-svn: 120388
* Stub out a new LiveDebugVariables pass.Jakob Stoklund Olesen2010-11-305-0/+110
| | | | | | | | | | | | | This analysis is going to run immediately after LiveIntervals. It will stay alive during register allocation and keep track of user variables mentioned in DBG_VALUE instructions. When the register allocator is moving values between registers and the stack, it is very hard to keep track of DBG_VALUE instructions. We usually get it wrong. This analysis maintains a data structure that makes it easy to update DBG_VALUE instructions. llvm-svn: 120385
* remove a use of llvm-disChris Lattner2010-11-301-2/+1
| | | | llvm-svn: 120383
* remove the pointless check of MemoryUseIntrinsic fromChris Lattner2010-11-302-27/+1
| | | | | | | is trivially dead, since these have side effects. This makes the (misnamed) MemoryUseIntrinsic class dead, so remove it. llvm-svn: 120382
* strength reduce this.Chris Lattner2010-11-301-1/+1
| | | | llvm-svn: 120381
* getLocationForDest should work for memset as well.Chris Lattner2010-11-302-2/+3
| | | | llvm-svn: 120380
* rename doesClobberMemory -> hasMemoryWrite to be more specific, andChris Lattner2010-11-301-11/+11
| | | | | | remove an actively-wrong comment. llvm-svn: 120378
* clean up handling of 'free', detangling it from everything else.Chris Lattner2010-11-302-22/+26
| | | | | | | It can be seriously improved, but at least now it isn't intertwined with the other logic. llvm-svn: 120377
* merge one more awayChris Lattner2010-11-302-9/+12
| | | | llvm-svn: 120375
* I already merged partial-overwrite.ll -> PartialStore.llChris Lattner2010-11-303-29/+20
| | | | | | Merge context-sensitive.ll -> simple.ll and upgrade it. llvm-svn: 120374
* clean up DSE tests, removing some poorly reduced and useless old test,Chris Lattner2010-11-3014-1301/+115
| | | | | | merging more into other larger .ll files, filecheckizing along the way. llvm-svn: 120373
* Minor cleanups. No functional change.Bill Wendling2010-11-301-24/+23
| | | | llvm-svn: 120372
* s/ARM::BRIND/ARM::BX/g to coincide with r120366.Bill Wendling2010-11-303-5/+5
| | | | llvm-svn: 120371
* enhance basicaa to return "Mod" for a memcpy call when theChris Lattner2010-11-302-2/+18
| | | | | | queried location doesn't overlap the source, and add a testcase. llvm-svn: 120370
* Add correct encoding for "bl __aeabi_read_tp". However, the asm matcher isn'tBill Wendling2010-11-301-3/+5
| | | | | | able to match this yet. llvm-svn: 120369
* Teach basicaa that memset's modref set is at worst "mod" and neverChris Lattner2010-11-303-7/+28
| | | | | | | | | | contains "ref". Enhance DSE to use a modref query instead of a store-specific hack to generalize the "ignore may-alias stores" optimization to handle memset and memcpy. llvm-svn: 120368
* Rename BX/BRIND/etc patterns to clarify which is actually the BX instructionJim Grosbach2010-11-302-9/+9
| | | | | | and which are pseudos. llvm-svn: 120366
* Add some encoding for the adr instruction. Labels still need to be finished.Bill Wendling2010-11-301-6/+16
| | | | llvm-svn: 120365
* Correct Thumb2 encodings for a much wider range of loads and stores.Owen Anderson2010-11-305-48/+111
| | | | llvm-svn: 120364
* my previous patch would cause us to start deleting some volatileChris Lattner2010-11-302-1/+21
| | | | | | stores, fix and add a testcase. llvm-svn: 120363
* Make a few more ARM pseudo instructions actually use the PseudoInst base class.Jim Grosbach2010-11-301-14/+13
| | | | llvm-svn: 120362
* Predicate encoding should be withing {}s. And general cleanup.Bill Wendling2010-11-302-8/+4
| | | | llvm-svn: 120361
* Predicate encoding should be withing {}s.Bill Wendling2010-11-301-2/+2
| | | | llvm-svn: 120360
OpenPOWER on IntegriCloud