aboutsummaryrefslogtreecommitdiff
path: root/lib/Transforms
AgeCommit message (Collapse)Author
2009-01-07Implement the first half of PR3290: if there is a store of an Chris Lattner
integer to a (transitive) bitcast the alloca and if that integer has the full size of the alloca, then it clobbers the whole thing. Handle this by extracting pieces out of the stored integer and filing them away in the SROA'd elements. This triggers fairly frequently because the CFE uses integers to pass small structs by value and the inliner exposes these. For example, in kimwitu++, I see a bunch of these with i64 stores to "%struct.std::pair<std::_Rb_tree_const_iterator<kc::impl_abstract_phylum*>,bool>" In 176.gcc I see a few i32 stores to "%struct..0anon". In the testcase, this is a difference between compiling test1 to: _test1: subl $12, %esp movl 20(%esp), %eax movl %eax, 4(%esp) movl 16(%esp), %eax movl %eax, (%esp) movl (%esp), %eax addl 4(%esp), %eax addl $12, %esp ret vs: _test1: movl 8(%esp), %eax addl 4(%esp), %eax ret The second half of this will be to handle loads of the same form. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61853 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-07Factor a bunch of code out into a helper method.Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61852 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-07use continue to simplify code and reduce nesting, no functionalityChris Lattner
change. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61851 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-07Get TargetData once up front and cache as an ivar instead ofChris Lattner
requerying it all over the place. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61850 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-07Use the hasAllZeroIndices predicate to simplify some Chris Lattner
code, no functionality change. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61849 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-05Change m_ConstantInt and m_SelectCst to take their constant integersChris Lattner
as template arguments instead of as instance variables, exposing more optimization opportunities to the compiler earlier. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61776 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-05Teach the internalize pass to also internalizeDuncan Sands
global aliases. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61754 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-05Find loop back edges only after empty blocks are eliminated.Evan Cheng
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61752 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-05Not having an aliasee is a theoretical possibility.Duncan Sands
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61745 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-05Format more neatly.Duncan Sands
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61744 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-05Remove trailing spaces.Duncan Sands
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61743 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-05Delete unused global aliases with internal linkage.Duncan Sands
In fact this also deletes those with linkonce linkage, however this is currently dead because for the moment aliases aren't allowed to have this linkage type. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61742 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-05Tidy up #includes, deleting a bunch of unnecessary #includes.Dan Gohman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61715 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-05Move the libcall annotating part from doFinalization to doInitialization.Nick Lewycky
Finalization occurs after all the FunctionPasses in the group have run, which is clearly not what we want. This also means that we have to make sure that we apply the right param attributes when creating a new function. Also, add a missed optimization: strdup and strndup. NoCapture and NoAlias return! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61658 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-04Run a post-pass that marks known function declarations by name.Nick Lewycky
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61632 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-04Revert this transform. It was causing some dramatic slowdowns in a few ↵Bill Wendling
tests. See PR3266. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61623 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-03Any void readonly functions are provably dead, don't waste time adding Nick Lewycky
nocapture attributes to them. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61610 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-02Load tracking means that the value analyzed mayDuncan Sands
not have pointer type. In particular, it may be the condition argument for a select or a GEP index. While I was unable to construct a testcase for which some bits of the original pointer are captured due to one of these, it's very very close to being possible - so play safe and exclude these possibilities. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61580 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-02When calculating 'nocapture' argument attributes, allowDuncan Sands
the argument to be stored to an alloca by tracking uses of the alloca. This occurs 4 times (out of 7121, 0.05%) in MultiSource/Applications, so may not be worth it. On the other hand, it is easy to do and fairly cheap. The functions it helps are: W_addcom and W_addlit in spiff; process_args (argv) in d (make_dparser); ercPixConcealIMB in JM/ldecod. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61570 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-02Improve comments and reorganize a bit - no functionalityDuncan Sands
change. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61569 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-02Make adding nocapture a bit stronger. FreeInst is nocapture. Also, Nick Lewycky
functions that don't write can't leak a pointer except through the return value, so a void readonly function is implicitly nocapture. Test these, and add a test that verifies that f1 calling f2 with an otherwise dead pointer gets both of them marked nocapture. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61552 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-01Mention that this pass does escape analysis in theDuncan Sands
leading comments. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61548 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-01Fix comment.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61538 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-01Add transformation:Bill Wendling
xor (or (icmp, icmp), true) -> and(icmp, icmp) This is possible because of De Morgan's law. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61537 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-31Look through phi nodes and select instructions whenDuncan Sands
calculating nocapture attributes. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61535 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-31Don't analyze arguments already marked 'nocapture'.Duncan Sands
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61532 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-31Rename AddReadAttrs to FunctionAttrs, and teach it howDuncan Sands
to work out (in a very simplistic way) which function arguments (pointer arguments only) are only dereferenced and so do not escape. Mark such arguments 'nocapture'. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61525 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-29Experiments show that looking through phi nodesDuncan Sands
and select instructions doesn't buy anything here except extra complexity: the only difference in the entire testsuite was that a readonly function became readnone in MiBench/consumer-typeset. Add a comment about this. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61478 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-29Allow readnone functions to read (and write!) globalDuncan Sands
constants, since doing so is irrelevant for aliasing purposes. While this doesn't increase the total number of functions marked readonly or readnone in MultiSource/ Applications (3089), it does result in 12 functions being marked readnone rather than readonly. Before: readnone: 820 readonly: 2269 After: readnone: 832 readonly: 2257 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61469 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-23Revert 61362 and 61402 until SPEC breakage is fixed.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61403 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-23This fixes the bug in 175.vpr. It doesn't fix theDale Johannesen
other SPEC breakage. I'll be reverting all recent changes shortly, this checking is mostly so this change doesn't get lost. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61402 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-23Fix the time regression I introduced in 464.h264ref withDale Johannesen
my last patch to this file. The issue there was that all uses of an IV inside a loop are actually references to Base[IV*2], and there was one use outside that was the same but LSR didn't see the base or the scaling because it didn't recurse into uses outside the loop; thus, it used base+IV*scale mode inside the loop instead of pulling base out of the loop. This was extra bad because register pressure later forced both base and IV into memory. Doing that recursion, at least enough to figure out addressing modes, is a good idea in general; the change in AddUsersIfInteresting does this. However, there were side effects.... It is also possible for recursing outside the loop to introduce another IV where there was only 1 before (if the refs inside are not scaled and the ref outside is). I don't think this is a common case, but it's in the testsuite. It is right to be very aggressive about getting rid of such introduced IVs (CheckForIVReuse and the handling of nonzero RewriteFactor in StrengthReduceStridedIVUsers). In the testcase in question the new IV produced this way has both a nonconstant stride and a nonzero base, neither of which was handled before. And when inserting new code that feeds into a PHI, it's right to put such code at the original location rather than in the PHI's immediate predecessor(s) when the original location is outside the loop (a case that couldn't happen before) (RewriteInstructionToUseNewBase); better to avoid making multiple copies of it in this case. Also, the mechanism for keeping SCEV's corresponding to GEP's no longer works, as the GEP might change after its SCEV is remembered, invalidating the SCEV, and we might get a bad SCEV value when looking up the GEP again for a later loop. This also couldn't happen before, as we weren't recursing into GEP's outside the loop. I owe some testcases for this, want to get it in for nightly runs. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61362 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-23Don't forget to remove phi nodes from the value numbering table after we ↵Owen Anderson
collapse them. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61358 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-22Comment clean-ups. No functionality change.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61354 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-22Check that the instruction isn't in the value numbering scope.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61353 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-22Simplification: Negate the operator== method instead of implementing a full ↵Bill Wendling
operator!= method. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61352 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-22Add verification that deleted instruction isn't hiding in the PHI map.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61350 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-22Verify removed in a few more places.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61349 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-22Add verification functions to GVN which check to see that an instruction wasBill Wendling
truely deleted. These will be expanded with further checks of all of the data structures. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61347 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-21Turn strcmp into memcmp, such as strcmp(P, "x") --> memcmp(P, "x", 2).Nick Lewycky
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61297 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-20Remove redundant test for vector-nature. Scan the vector first to see whetherNick Lewycky
our optz'n will apply to it, then build the replacement vector only if needed. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61279 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-19- CodeGenPrepare does not split loop back edges but it only knows about back ↵Evan Cheng
edges of single block loops. It now does a DFS walk to find loop back edges. - Use SplitBlockPredecessors to factor out common predecessors of the critical edge destination. This is disabled for now due to some regressions. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61248 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-18Didn't mean to commit this.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61222 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-18Re-XFAIL this test until debug stuff settles down.Bill Wendling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61219 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-18Oops! Left out a line.Nick Lewycky
Simplifying the sdiv might allow further simplifications for our users. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61196 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-18Make all the vector elements positive in an srem of constant vector.Nick Lewycky
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61195 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-18Fix PR2929 by making bugpoint/code extract propagate the nothrowChris Lattner
bit from the original function to the cloned one. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61194 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-18Revert previous patch, appears to break bootstrap.Dale Johannesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61181 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-18Fix the time regression I introduced in 464.h264ref withDale Johannesen
my last patch to this file. The issue there was that all uses of an IV inside a loop are actually references to Base[IV*2], and there was one use outside that was the same but LSR didn't see the base or the scaling because it didn't recurse into uses outside the loop; thus, it used base+IV*scale mode inside the loop instead of pulling base out of the loop. This was extra bad because register pressure later forced both base and IV into memory. Doing that recursion, at least enough to figure out addressing modes, is a good idea in general; the change in AddUsersIfInteresting does this. However, there were side effects.... It is also possible for recursing outside the loop to introduce another IV where there was only 1 before (if the refs inside are not scaled and the ref outside is). I don't think this is a common case, but it's in the testsuite. It is right to be very aggressive about getting rid of such introduced IVs (CheckForIVReuse and the handling of nonzero RewriteFactor in StrengthReduceStridedIVUsers). In the testcase in question the new IV produced this way has both a nonconstant stride and a nonzero base, neither of which was handled before. (This patch does not handle all the cases where this can happen.) And when inserting new code that feeds into a PHI, it's right to put such code at the original location rather than in the PHI's immediate predecessor(s) when the original location is outside the loop (a case that couldn't happen before) (RewriteInstructionToUseNewBase); better to avoid making multiple copies of it in this case. Everything above is exercised in CodeGen/X86/lsr-negative-stride.ll (and ifcvt4 in ARM which is the same IR). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61178 91177308-0d34-0410-b5e6-96231b3b80d8
2008-12-18reapply this hunk from Bill's reversion in r61169, it is conservativeChris Lattner
and safe and orthogonal from turning off load pre. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61177 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-18-g5258 (git 2.32.0) at 2025-07-23 12:39:03 +0000