aboutsummaryrefslogtreecommitdiff
path: root/lib/Transforms
AgeCommit message (Collapse)Author
2003-10-07Fix bug in previous checkinChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8922 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-06Minor speedups for the instcombine passChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8894 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-06Speed up the predicate used to decide when to inline by caching the sizeChris Lattner
of callees between executions. On eon, in release mode, this changes the inliner from taking 11.5712s to taking 2.2066s. In debug mode, it went from taking 14.4148s to taking 7.0745s. In release mode, this is a 24.7% speedup of gccas, in debug mode, it's a total speedup of 11.7%. This also makes it slightly more aggressive. This could be because we are not judging the size of the functions quite as accurately as before. When we start looking at the performance of the generated code, this can be investigated further. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8893 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-06Avoid doing pointless work. Amazingly, this makes us go faster.Chris Lattner
Running the inliner on 252.eon used to take 48.4763s, now it takes 14.4148s. In release mode, it went from taking 25.8741s to taking 11.5712s. This also fixes a FIXME. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8890 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05This changes the PromoteMemToReg function to create "pruned" SSA form, notChris Lattner
"minimal" SSA form (in other words, it doesn't insert dead PHIs). This speeds up the mem2reg pass very significantly because it doesn't have to do a lot of frivolous work in many common cases. In the 252.eon function I have been playing with, this doesn't even insert the 120 PHI nodes that it used to which were trivially dead (in the process of promoting 356 alloca instructions overall). This speeds up the mem2reg pass from 1.2459s to 0.1284s. More significantly, the DCE pass used to take 2.4138s to remove the 120 dead PHI nodes that mem2reg constructed, now it takes 0.0134s (which is the time to scan the function and decide that there is nothing dead). So overall, on this one function, we speed things up a total of 3.5179s, which is a 24.8x speedup! :) This change is tested by the Mem2Reg/2003-10-05-DeadPHIInsertion.ll test, which now passes. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8884 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05Change the interface to PromoteMemToReg to also take a DominatorTreeChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8883 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05Speed up the mem2reg transform for allocas which are only read/written in a ↵Chris Lattner
single basic block. This is amazingly common in code generated by the C/C++ front-ends. This change makes it not have to insert ANY phi nodes, whereas before it would insert a ton of dead ones which DCE would have to clean up. Thus, this fix improves compile-time performance of these trivial allocas in two ways: 1. It doesn't have to do the walking and book-keeping for renaming 2. It does not insert dead phi nodes for them which would have to subsequently be cleaned up. On my favorite testcase from 252.eon, this special case handles 305 out of 356 promoted allocas in the function. It speeds up the mem2reg pass from 7.5256s to 1.2505s. It inserts 677 fewer dead PHI nodes, which speeds up a subsequent -dce pass from 18.7524s to 2.4806s. There are still 120 trivially dead PHI nodes being inserted for variables used in multiple basic blocks, but they are not handled by this patch. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8881 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05Initial checkin of the LLVM->LLVM transform to support code generators whichChris Lattner
do not support stack unwinding yet git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8869 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05simplify-cfg is really a function passChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8868 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05The first PHI node may be null, scan for the first non-null oneChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8865 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05The VersionNumbers vector is only used during PHI placement. Turn it into ↵Chris Lattner
an argument, allowing us to get rid of the vector. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8864 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05* Update file header commentChris Lattner
*** Revamp the code which handled unreachable code in the function. Now the code is much more efficient for high-degree basic blocks, such as those that occur in the 252.eon SPEC benchmark. For the interested, the time to promote a SINGLE alloca in _ZN7mrScene4ReadERSi function used to be > 3.5s. Now it is < .075s. The function has a LOT of allocas in it, so it appeared to be infinite looping, this should make it much nicer. :) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8863 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05Simplify the loop a bitChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8862 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05There is no need for separate WriteSets and PhiNodeBlocks lists. It is just aChris Lattner
work-list of value definitions. This allows elimination of the explicit 'iterative' step of the algorithm, and also reuses temporary memory better. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8861 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05The PhiNodes 2D vector is only used during PHI node placement. It doesn'tChris Lattner
need to be an instance variable! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8860 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05* Document instance vars betterChris Lattner
* Fuse two parallel loops * Use a more specific type for AllocaLookup git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8859 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05Two small cleanups/speedups:Chris Lattner
* Do not insert a new entry into NewPhiNodes during the rename pass if there are no PHIs in a block. * Do not compute WriteSets in parallel git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8858 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-05* Minor cleanupsChris Lattner
* Eliminate the KillList instance variable, instead, just delete loads and stores as they are "renamed", and delete allocas when they are done * Make the 'visited' set an instance variable to avoid passing it on the stack. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8857 91177308-0d34-0410-b5e6-96231b3b80d8
2003-10-02Implement InstCombine/add.ll:test17 & 18Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8817 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-24Use global *_iteratorChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8703 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-24Do not use BasicBlock::*_iterator, just use *_iterator itself.Chris Lattner
Isn't updating copy and pasted code a joy git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8702 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-23Fix bug: IndVarsSimplify/2003-09-23-NotAtTop.llChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8689 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-22Fix bugs in the last changeChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8667 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-22Fix bug: Inline/2003-09-22-PHINodesInNormalInvokeDest.llChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8666 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-22Fix bug: Inline/2003-09-22-PHINodesInExceptionDest.llChris Lattner
... by making sure to update PHI nodes to take into consideration the extra edges we get if we inline a call instruction through an invoke. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8664 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-22Credits now go into the top-level CREDITS.TXTMisha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8662 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-22Credits are now in top-level CREDITS.TXTMisha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8661 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-22Squelch warningChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8659 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-20Global variables with APPENDING linkage are very important to keep around!Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8632 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-20Rename Function::getEntryNode -> getEntryBlockChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8625 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-20Minor cleanups, no functional changesChris Lattner
Rename Function::getEntryNode -> getEntryBlock git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8623 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-20Fix a really obvious huge gaping bug, add a commentChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8620 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-20Expose the TCE passChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8619 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-20New transformation: tail recursion eliminationChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8618 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-19Implement InstCombine/and.ll:test(15|16)Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8607 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-19pull a large nested conditional out into its own functionChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8605 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-19Implement InstCombine/add.ll:test(15|16)Chris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8604 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-16Fix bug raising allocations whose call sites were invoke instructions.Chris Lattner
Thanks to brg for tracking down the problem so precisely! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8568 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-16This is effectively a complete rewrite of the globaldce algorithm, resultingChris Lattner
in it being both shorter and more effective. It no longer depends on the callgraph, so one FIXME has been fixed. Additionally, this pass was not able to delete recursive (but dead) functions if they were pointed to by global variables which were also dead. In fact this pass had a lot of problems deleting functions which were only pointed to by dead globals and other stuff. Fixing this means that the entire EH library should be stripped away now from programs that don't use sjlj or exceptions. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8567 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-15Minor cleanups, give credit, remove code that should not be necessary, andChris Lattner
was a "major hack" :) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8524 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-15Initial checkin of the Setjmp/Longjmp lowering/transformation pass,Chris Lattner
contributed by the masterful Bill Wendling! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8520 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-15Ugh, a bug fix needed because of the bug in the CallGraph codeChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8519 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-15Fix bug: PruneEH/2003-09-14-ExternalCall.llxChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8516 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-15Fix bug: Inline/2003-09-14-InlineValue.llChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8514 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-12Make sure to cannonicalize loops before running indvar simplifyChris Lattner
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8502 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-12Do not return success after checking only the FIRST USE of a gep instruction.Chris Lattner
Instead, check all uses. This fixes bug: ScalarRepl/2003-09-12-IncorrectPromote.ll This also fixes the miscompilation of Ptrdist/bc git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8493 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-12Minor optimization efficiency improvement:Chris Lattner
- Run mem2reg promotion first - Only rerun passes if the previous thing changed something git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8490 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-11Simplify codeChris Lattner
Implement InstCombine/mul.ll:test9 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8488 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-11Apostrophes are only used for possession and quoting.Misha Brukman
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8473 91177308-0d34-0410-b5e6-96231b3b80d8
2003-09-11Integrate functionality of the mem2reg pass directly into this pass to makeChris Lattner
the combination more effective git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8471 91177308-0d34-0410-b5e6-96231b3b80d8
generated by cgit v1.2.3-18-g5258 (git 2.32.0) at 2025-06-24 15:24:26 +0000