aboutsummaryrefslogtreecommitdiffstats
path: root/lib/Transforms
Commit message (Collapse)AuthorAgeFilesLines
* Add a comment to the commit r165187.Jakub Staszak2012-10-041-1/+3
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165238 91177308-0d34-0410-b5e6-96231b3b80d8
* SimplifyCFG: Enhance the "remove CFG edge that leads to null pointer ↵Benjamin Kramer2012-10-041-2/+3
| | | | | | | | | | dereference" optimization to also handle instructions with multiple uses. We conservatively only check the first use to avoid walking long use chains. This catches the common case of having both a load and a store to a pointer supplied by a PHI node. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165232 91177308-0d34-0410-b5e6-96231b3b80d8
* In my recent change to avoid use of underaligned memory I didn't notice thatDuncan Sands2012-10-041-7/+7
| | | | | | | | | cpyDest can be mutated in some cases, which would then cause a crash later if indeed the memory was underaligned. This brought down several buildbots, so I guess the underaligned case is much more common than I thought! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165228 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix PR13969, a mini-phase-ordering issue with the new SROA pass.Chandler Carruth2012-10-041-23/+50
| | | | | | | | | | | | | | | | | | | | | Currently, we re-visit allocas when something changes about the way they might be *split* to allow better scalarization to take place. However, we weren't handling the case when the *promotion* is what would change the behavior of SROA. When an address derived from an alloca is stored into another alloca, we consider the first to have escaped. If the second is ever promoted to an SSA value, we will suddenly be able to run the SROA pass on the first alloca. This patch adds explicit support for this form if iteration. When we detect a store of a pointer derived from an alloca, we flag the underlying alloca for reprocessing after promotion. The logic works hard to only do this when there is definitely going to be promotion and it might remove impediments to the analysis of the alloca. Thanks to Nick for the great test case and Benjamin for some sanity check review. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165223 91177308-0d34-0410-b5e6-96231b3b80d8
* The memcpy optimizer was happily doing call slot forwarding when the new memoryDuncan Sands2012-10-041-4/+30
| | | | | | | | | | | was less aligned than the old. In the testcase this results in an overaligned memset: the memset alignment was correct for the original memory but is too much for the new memory. Fix this by either increasing the alignment of the new memory or bailing out if that isn't possible. Should fix the gcc-4.7 self-host buildbot failure. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165220 91177308-0d34-0410-b5e6-96231b3b80d8
* Teach the integer-promotion rewrite strategy to be endianness aware.Chandler Carruth2012-10-041-9/+29
| | | | | | | | | | | | | | | | | | | | | | | Sorry for this being broken so long. =/ As part of this, switch all of the existing tests to be Little Endian, which is the behavior I was asserting in them anyways! Add in a new big-endian test that checks the interesting behavior there. Another part of this is to tighten the rules abotu when we perform the full-integer promotion. This logic now rejects cases where there fully promoted integer is a non-multiple-of-8 bitwidth or cases where the loads or stores touch bits which are in the allocated space of the alloca but are not loaded or stored when accessing the integer. Sadly, these aren't really observable today as the rest of the pass will already ensure the invariants hold. However, the latter situation is likely to become a potential concern in the future. Thanks to Benjamin and Duncan for early review of this patch. I'm still looking into whether there are further endianness issues, please let me know if anyone sees BE failures persisting past this. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165219 91177308-0d34-0410-b5e6-96231b3b80d8
* Use method to query for attributes.Bill Wendling2012-10-042-2/+2
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165209 91177308-0d34-0410-b5e6-96231b3b80d8
* Add method to query for 'NoAlias' attribute on call/invoke instructions.Bill Wendling2012-10-041-1/+1
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165208 91177308-0d34-0410-b5e6-96231b3b80d8
* Use method to query for attributes.Bill Wendling2012-10-041-1/+1
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165207 91177308-0d34-0410-b5e6-96231b3b80d8
* Query for attributes via the correct method call.Bill Wendling2012-10-041-2/+2
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165206 91177308-0d34-0410-b5e6-96231b3b80d8
* [tsan] add 3 internal flags for fine-grain control of what is instrumented ↵Kostya Serebryany2012-10-041-7/+18
| | | | | | and what is not. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165204 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix PR13967.Jakub Staszak2012-10-031-1/+4
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165187 91177308-0d34-0410-b5e6-96231b3b80d8
* This Patch corrects a problem whereby the optimization to use a faster dividePreston Gurd2012-10-031-5/+15
| | | | | | | | | | | | | | | | instruction (for Intel Atom) was not being done by Clang, because the type context used by Clang is not the default context. It fixes the problem by getting the global context types for each div/rem instruction in order to compare them against the types in the BypassTypeMap. Tests for this will be done as a separate patch to Clang. Patch by Tyler Nowicki. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165126 91177308-0d34-0410-b5e6-96231b3b80d8
* tsan: prepare for migration to new memory_order enum values (ABI compatible)Dmitry Vyukov2012-10-031-1/+2
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165107 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix an issue where we failed to adjust the alignment constraint onChandler Carruth2012-10-031-0/+6
| | | | | | | | a memcpy to reflect that '0' has a different meaning when applied to a load or store. Now we correctly use underaligned loads and stores for the test case added. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165101 91177308-0d34-0410-b5e6-96231b3b80d8
* Try to use a better set of abstractions for computing the alignmentChandler Carruth2012-10-031-16/+28
| | | | | | | | | | | | | | | | | | | | necessary during rewriting. As part of this, fix a real think-o here where we might have left off an alignment specification when the address is in fact underaligned. I haven't come up with any way to trigger this, as there is always some other factor that reduces the alignment, but it certainly might have been an observable bug in some way I can't think of. This also slightly changes the strategy for placing explicit alignments on loads and stores to only do so when the alignment does not match that required by the ABI. This causes a few redundant alignments to go away from test cases. I've also added a couple of tests that really push on the alignment that we end up with on loads and stores. More to come here as I try to fix an underlying bug I have conjectured and produced test cases for, although it's not clear if this bug is the one currently hitting dragonegg's gcc47 bootstrap. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165100 91177308-0d34-0410-b5e6-96231b3b80d8
* Switch the SetVector::remove_if implementation to use partition whichChandler Carruth2012-10-031-1/+3
| | | | | | | | | | | | | | preserves the values of the relocated entries, unlikely remove_if. This allows walking them and erasing them. Also flesh out the predicate we are using for this to support the various constraints actually imposed on a UnaryPredicate -- without this we can't compose it with std::not1. Thanks to Sean Silva for the review here and noticing the issue with std::remove_if. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165073 91177308-0d34-0410-b5e6-96231b3b80d8
* Teach the new SROA to handle cases where an alloca that has already beenChandler Carruth2012-10-021-0/+4
| | | | | | | | | | | | | | | | scheduled for processing on the worklist eventually gets deleted while we are processing another alloca, fixing the original test case in PR13990. To facilitate this, add a remove_if helper to the SetVector abstraction. It's not easy to use the standard abstractions for this because of the specifics of SetVectors types and implementation. Finally, a nice small test case is included. Thanks to Benjamin for the fantastic reduced test case here! All I had to do was delete some empty basic blocks! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165065 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix another crasher in SROA, reported by Joel.Chandler Carruth2012-10-021-9/+32
| | | | | | | | | | | | | | | | | | We require that the indices into the use lists are stable in order to build fast lookup tables to locate a particular partition use from an operand of a PHI or select. This is (obviously in hind sight) incompatible with erasing elements from the array. Really, we don't want to erase anyways. It is expensive, and a rare operation. Instead, simply weaken the contract of the PartitionUse structure to allow null Use pointers to represent dead uses. Now we can clear out the pointer to mark things as dead, and all it requires is adding some 'continue' checks to the various loops. I'm still reducing a test case for this, as the test case I have is huge. I think this one I can get a nice test case for though, as it was much more deterministic. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165032 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix a silly coding error on my part. The whole point of the speculatorChandler Carruth2012-10-021-16/+28
| | | | | | | | | | | | | | | being separate was that it can grow the use list. As a consequence, we can't use the iterator-pair interface, we need an index based interface. Expose such an interface from the AllocaPartitioning, and use it in the speculator. This should at least fix a use-after-free bug found by Duncan, and may fix some of the other crashers. I don't have a nice deterministic test case yet, but if I get a good one, I'll add it. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@165027 91177308-0d34-0410-b5e6-96231b3b80d8
* Turn the new SROA pass back on. Let's see if it sticks this time. =]Chandler Carruth2012-10-021-1/+1
| | | | | | Again, let me know if anything breaks due to this! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164986 91177308-0d34-0410-b5e6-96231b3b80d8
* Make this plural. Spotted by Duncan in review (and a very old typo, thisChandler Carruth2012-10-011-1/+1
| | | | | | is the second time I've moved this comment around...) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164939 91177308-0d34-0410-b5e6-96231b3b80d8
* Prune some unnecessary includes.Chandler Carruth2012-10-011-4/+0
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164938 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix several issues with alignment. We weren't always accounting for typeChandler Carruth2012-10-011-31/+32
| | | | | | | | | | alignment requirements of the new alloca. As one consequence which was reported as a bug by Duncan, we overaligned memcpy calls to ranges of allocas after they were rewritten to types with lower alignment requirements. Other consquences are possible, but I don't have any test cases for them. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164937 91177308-0d34-0410-b5e6-96231b3b80d8
* SimplifyCFG: Don't crash when forming a switch bitmap with an undef default ↵Benjamin Kramer2012-10-011-2/+5
| | | | | | | | value. Fixes PR13985. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164934 91177308-0d34-0410-b5e6-96231b3b80d8
* Factor the PHI and select speculation into a separate rewriter. ThisChandler Carruth2012-10-011-263/+294
| | | | | | | | | | | | | | | could probably be factored still further to hoist this logic into a generic helper, but currently I don't have particularly clean ideas about how to handle that. This at least allows us to drop custom load rewriting from the speculation logic, which in turn allows the existing load rewriting logic to fire. In theory, this could enable vector promotion or other tricks after speculation occurs, but I've not dug into such issues. This is primarily just cleaning up the factoring of the code and the resulting logic. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164933 91177308-0d34-0410-b5e6-96231b3b80d8
* Refactor the PartitionUse structure to actually use the Use* instead ofChandler Carruth2012-10-011-85/+79
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | a pair of instructions, one for the used pointer and the second for the user. This simplifies the representation and also makes it more dense. This was noticed because of the miscompile in PR13926. In that case, we were running up against a fundamental "bad idea" in the speculation of PHI and select instructions: the speculation and rewriting are interleaved, which requires phi speculation to also perform load rewriting! This is bad, and causes us to miss opportunities to do (for example) vector rewriting only exposed after PHI speculation, etc etc. It also, in the old system, required us to insert *new* load uses into the current partition's use list, which would then be ignored during rewriting because we had already extracted an end iterator for the use list. The appending behavior (and much of the other oddities) stem from the strange de-duplication strategy in the PartitionUse builder. Amusingly, all this went without notice for so long because it could only be triggered by having *different* GEPs into the same partition of the same alloca, where both different GEPs were operands of a single PHI, and where the GEP which was not encountered first also had multiple uses within that same PHI node... Hence the insane steps required to reproduce. So, step one in fixing this fundamental bad idea is to make the PartitionUse actually contain a Use*, and to make the builder do proper deduplication instead of funky de-duplication. This is enough to remove the appending behavior, and fix the miscompile in PR13926, but there is more work to be done here. Subsequent commits will lift the speculation into its own visitor. It'll be a useful step toward potentially extracting all of the speculation logic into a generic utility transform. The existing PHI test case for repeated operands has been made more extreme to catch even these issues. This test case, run through the old pass, will exactly reproduce the miscompile from PR13926. ;] We were so close here! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164925 91177308-0d34-0410-b5e6-96231b3b80d8
* SimplifyCFG: Enumerating all predecessors of a BB can be expensive ↵Benjamin Kramer2012-09-301-3/+7
| | | | | | | | (switches), avoid it if possible. No functionality change. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164923 91177308-0d34-0410-b5e6-96231b3b80d8
* ArgumentPromotion: Remove ancient workaround for a bug in the C backend.Benjamin Kramer2012-09-301-19/+1
| | | | | | Fun fact: The CBE learned how to deal with this situation before it was removed. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164918 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix a somewhat surprising miscompile where code relying on an ABIChandler Carruth2012-09-291-3/+13
| | | | | | | | | | | | | | | alignment could lose it due to the alloca type moving down to a much smaller alignment guarantee. Now SROA will actively compute a proper alignment, factoring the target data, any explicit alignment, and the offset within the struct. This will in some cases lower the alignment requirements, but when we lower them below those of the type, we drop the alignment entirely to give freedom to the code generator to align it however is convenient. Thanks to Duncan for the lovely test case that pinned this down. =] git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164891 91177308-0d34-0410-b5e6-96231b3b80d8
* Do not delete BBs if their addresses are taken. rdar://12396696Evan Cheng2012-09-281-2/+3
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164866 91177308-0d34-0410-b5e6-96231b3b80d8
* GlobalDCE should be run at -O2 / -Os to eliminate unused dtor, etc. ↵Evan Cheng2012-09-281-4/+3
| | | | | | rdar://9142819 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164850 91177308-0d34-0410-b5e6-96231b3b80d8
* CorrelatedPropagation: BasicBlock::removePredecessor can simplify PHI nodes. ↵Benjamin Kramer2012-09-281-0/+5
| | | | | | | | If the it's the condition of a SwitchInst, reload it. Fixes PR13972. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164818 91177308-0d34-0410-b5e6-96231b3b80d8
* GlobalOpt: non-constexpr bitcasts or GEPs can occur even if the global value ↵Benjamin Kramer2012-09-281-1/+3
| | | | | | | | is only stored once. Fixes PR13968. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164815 91177308-0d34-0410-b5e6-96231b3b80d8
* Surprisingly, we missed a trivial case here. Fix that!Nick Lewycky2012-09-281-0/+4
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164814 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix a integer overflow in SimplifyCFG's look up table formation logic.Benjamin Kramer2012-09-271-0/+4
| | | | | | | | If the width is very large it gets truncated from uint64_t to uint32_t when passed to TD->fitsInLegalInteger. The truncated value can fit in a register. This manifested in massive memory usage or crashes (PR13946). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164784 91177308-0d34-0410-b5e6-96231b3b80d8
* Revert 'Fix a typo 'iff' => 'if''. iff is an abreviation of if and only if. ↵Sylvestre Ledru2012-09-2715-46/+46
| | | | | | See: http://en.wikipedia.org/wiki/If_and_only_if Commit 164767 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164768 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix a typo 'iff' => 'if'Sylvestre Ledru2012-09-2715-46/+46
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164767 91177308-0d34-0410-b5e6-96231b3b80d8
* Prefer shuffles to selects. Backends love shuffles!Nick Lewycky2012-09-271-1/+19
| | | | git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164763 91177308-0d34-0410-b5e6-96231b3b80d8
* Disable the new SROA pass to get the tree back in working order. We don't yetNick Lewycky2012-09-261-1/+1
| | | | | | | have testcases for the current problems. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164731 91177308-0d34-0410-b5e6-96231b3b80d8
* Remove the `hasFnAttr' method from Function.Bill Wendling2012-09-266-12/+12
| | | | | | | | The hasFnAttr method has been replaced by querying the Attributes explicitly. No intended functionality change. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164725 91177308-0d34-0410-b5e6-96231b3b80d8
* Address Duncan's comments on r164684:Hans Wennborg2012-09-261-9/+5
| | | | | | | | - Put statistics in alphabetical order - Don't use getZextValue when building TableInt, just use APInts - Introduce Create{Z,S}ExtOrTrunc in IRBuilder. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164696 91177308-0d34-0410-b5e6-96231b3b80d8
* Address Duncan's comments on r164682:Hans Wennborg2012-09-261-6/+4
| | | | | | | - Finish assert messages with exclamation mark - Move overflow checking into ShouldBuildLookupTable. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164692 91177308-0d34-0410-b5e6-96231b3b80d8
* Analogous fix to memset and memcpy rewriting. Don't have a test caseChandler Carruth2012-09-261-0/+16
| | | | | | | contrived for these yet, as I spotted them by inspection and the test cases are a bit more tricky to phrase. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164691 91177308-0d34-0410-b5e6-96231b3b80d8
* When rewriting the pointer operand to a load or store which hasChandler Carruth2012-09-261-0/+6
| | | | | | | alignment guarantees attached, re-compute the alignment so that we consider offsets which impact alignment. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164690 91177308-0d34-0410-b5e6-96231b3b80d8
* Teach all of the loads, stores, memsets and memcpys created by theChandler Carruth2012-09-261-32/+49
| | | | | | | | | | | | | rewriter in SROA to carry a proper alignment. This involves interrogating various sources of alignment, etc. This is a more complete and principled fix to PR13920 as well as related bugs pointed out by Eli in review and by inspection in the area. Also by inspection fix the integer and vector promotion paths to create aligned loads and stores. I still need to work up test cases for these... Sorry for the delay, they were found purely by inspection. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164689 91177308-0d34-0410-b5e6-96231b3b80d8
* SimplifyCFG: Make the switch-to-lookup table transformation store theHans Wennborg2012-09-261-12/+89
| | | | | | | | | | | | | | | | tables in bitmaps when they fit in a target-legal register. This saves some space, and it also allows for building tables that would otherwise be deemed too sparse. One interesting case that this hits is example 7 from http://blog.regehr.org/archives/320. We currently generate good code for this when lowering the switch to the selection DAG: we build a bitmask to decide whether to jump to one block or the other. My patch will result in the same bitmask, but it removes the need for the jump, as the return value can just be retrieved from the mask. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164684 91177308-0d34-0410-b5e6-96231b3b80d8
* SimplifyCFG: Refactor the switch-to-lookup table transformation byHans Wennborg2012-09-261-72/+115
| | | | | | breaking out the building of lookup tables into a separate class. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164682 91177308-0d34-0410-b5e6-96231b3b80d8
* Revert the business end of r164636 and try again. I'll come in again. ;]Chandler Carruth2012-09-261-8/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This should really, really fix PR13916. For real this time. The underlying bug is... a bit more subtle than I had imagined. The setup is a code pattern that leads to an @llvm.memcpy call with two equal pointers to an alloca in the source and dest. Now, not any pattern will do. The alloca needs to be formed just so, and both pointers should be wrapped in different bitcasts etc. When this precise pattern hits, a funny sequence of events transpires. First, we correctly detect the potential for overlap, and correctly optimize the memcpy. The first time. However, we do simplify the set of users of the alloca, and that causes us to run the alloca back through the SROA pass in case there are knock-on simplifications. At this point, a curious thing has happened. If we happen to have an i8 alloca, we have direct i8 pointer values. So we don't bother creating a cast, we rewrite the arguments to the memcpy to dircetly refer to the alloca. Now, in an unrelated area of the pass, we have clever logic which ensures that when visiting each User of a particular pointer derived from an alloca, we only visit that User once, and directly inspect all of its operands which refer to that particular pointer value. However, the mechanism used to detect memcpy's with the potential to overlap relied upon getting visited once per *Use*, not once per *User*. This is always true *unless* the same exact value is both source and dest. It turns out that almost nothing actually produces that pattern though. We can hand craft test cases that more directly test this behavior of course, and those are included. Also, note that there is a significant missed optimization here -- we prove in many cases that there is a non-volatile memcpy call with identical source and dest addresses. We shouldn't prevent splitting the alloca in that case, and in fact we should just remove such memcpy calls eagerly. I'll address that in a subsequent commit. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164669 91177308-0d34-0410-b5e6-96231b3b80d8
* Rename virtual table anchors from Anchor() to anchor() for consistency with ↵Craig Topper2012-09-261-1/+1
| | | | | | the rest of the tree. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@164666 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.1 at 2025-01-15 03:31:46 +0000