similar to: [LLVMdev] DominatorTree not updated properly after calling the llvm::SplitBlock.

Hi Andrew, Thanks a lot. But the function "DT->dominate(A,B)" decides the dominance relationship through comparing the DFS numbers, right? At least, in my example, when I check whether the newly split node (i.e., % *for.end.split*) DOMINATES the original node (I.e., %for.end), the answer is true, which is obviously wrong. Paul On Fri, Feb 14, 2014 at 1:59 AM, Andrew Trick

Hi Piotr, On March 31, 2017 at 1:07:12 PM, Piotr Padlewski (piotr.padlewski at gmail.com) wrote: > [snip] > Do I understand it correctly, that it is legal to do the hoist because all > of the instructions above %vtable does not throw? Yes, I think you're right.  HeaderMayThrow is a conservative approximation, and the conservativeness is biting us here. > Are there any plans to

On Nov 19, 2008, at 7:41 AM, Jack Tzu-Han Hung wrote: > Hi, > > I found a problem in CodeExtractor::severSplitPHINodes() > <CodeExtractor.cpp>. > > The algorithm first separates the header block into two, one > containing only PHI nodes and the other containing the remaining non- > PHI nodes. The variable NewBB holds the pointer to the latter half > block.

Hi, I found a problem in CodeExtractor::severSplitPHINodes() <CodeExtractor.cpp>. The algorithm first separates the header block into two, one containing only PHI nodes and the other containing the remaining non-PHI nodes. The variable NewBB holds the pointer to the latter half block. Later, it tries to update DT information. if (DT) DT->splitBlock(NewBB); In splitBlock, it checks

In Dominators.h, the splitBlock(NewBB) method assumes that the NewBB has only one successor. I don't see why it cant have more than one successors. Is there a reason? Or it's just assumed that way to simplify things? Thanks a lot. -- Wei Dang -------------- next part -------------- An HTML attachment was scrubbed... URL:

I'm having an issue with the CodeExtractor. When I try to extract the lone basic block from the following function, I get an assertion error. define i32 @test(i32 %x) { %tmp = call i32 @test3( i32 %x ) ; <i32> [#uses=1] ret i32 %tmp } The assertion error is: lli: Dominators.cpp:71: void llvm::DominatorTree::splitBlock(llvm::BasicBlock*): Assertion

Hi Devang, Thanks for your reply. But if you look at the comment of BasicBlock::splitBasicBlock(), it says that "...an unconditional branch is added to the new BB, and the rest of the instructions in the BB are moved to the newBB, including the old terminator." So, the terminator of the newBB is exactly the same as the terminator of the oldBB. IF the oldBB has multiple successors,

On Tue, Jan 24, 2017 at 1:20 PM, Sanjay Patel via llvm-dev < llvm-dev at lists.llvm.org> wrote: > > I started looking at the log files that you attached, and I'm confused. > The code that is supposedly causing the perf regression is created by the > loop vectorizer, right? Except the bad code is not in the "vector.body", so > is there something peculiar about

Hi Tobias, 1) Daniel and Chandler have for a long time been talking about computing > dominance and post-dominance in one shot to reduce the cost of > post-dominance and make it (freely) available everywhere. Is this > covered by your current (or planned) work? I'm not sure what you exactly mean by one shot; I'll ask around tomorrow. I wanted to play a little bit with your

Hi All, Still I could not figure out how to use Pass* while calling SplitEdge() function. Can anyone provide me some example? Regards, Chayan On Sun, Jul 18, 2010 at 11:49 PM, Nick Lewycky <nicholas at mxc.ca> wrote: > Chayan Sarkar wrote: >> >> Hi, >> >> I have tried to use SplitEdge function, but failed. Actually the third >> parameter is a variable of

Hi, I'm not sure if this is really a problem or not, but I noticed that if you're working with bitcode that has debug metadata, and you split a BasicBlock using llvm::SplitBlock or BasicBlock::splitBasicBlock, the unconditional branch added to the new block will lack debug metadata. This turned out not to be a problem for me, since my analysis pass needs only one instruction in

On Mon, Sep 5, 2011 at 5:13 PM, Rafael Baldiati Parizi <parizi.computacao at gmail.com> wrote: > I have two values and I want see if them are equals. Then, I would a > instruction like > if Value1 != Value2, error > without create a new block, continuing the execution in the same block, > forcing the LLI to execute this. You can't do that; a br instruction must be at the

Chayan Sarkar wrote: > Hi, > > I have tried to use SplitEdge function, but failed. Actually the third > parameter is a variable of type Pass and it need to be non-null. But I > could not figure out how to use it. Please help me out. The only reason it needs a non-NULL Pass* is to call llvm::SplitBlock which uses P->getAnalysisIfAvailable unconditionally. Feel free to wrap

Hello, Here are a few misc. comments on this patch. Would it make sense to mark the ProfileInfo passes as "CFGOnly?" If so, that would let them be automatically preserved by passes which don't modify the CFG (and that call AU.setPreservesCFG()). > + if (ProfileInfo* PI = getAnalysisIfAvailable<ProfileInfo>()) { > + PI->splitEdge(OrigPreHeader, NewHeader,

Hi, I could not figure out, how to declare a Pass*. Can anyone give me an example, how to use SplitEdge() function ? Please help me out. Regards, Chayan On Tue, Jul 20, 2010 at 10:12 AM, Chayan Sarkar <chayan.ju at gmail.com> wrote: > Hi All, > > Still I could not figure out how to use Pass* while calling > SplitEdge() function. Can anyone provide me some example? > >

On 08/05/2010 06:46 AM, Wenbin Zhang wrote: > Hi all, > I'm using postDominatorTree to do some program analysis. My code works > well for small tests, but when I run it on real applications, the > following error occurs: > /Inorder PostDominator Tree: DFSNumbers invalid: 0 slow queries. > [1] <<exit node>> {0,21} > [2] %bb1 {1,2} > [2] %bb {3,4} > [2]

I'll try the trunk, as well as check my code again. If indeed it's not fixed, I'll try to post a triggering case here. Thanks for the advice~ Best, --Wenbin ----- Original Message ----- From: "Tobias Grosser" <grosser at fim.uni-passau.de> To: "Wenbin Zhang" <zhangwen at cse.ohio-state.edu> Cc: <llvmdev at cs.uiuc.edu> Sent: Thursday, August

Hi All, I need to remove successors from every basic block to insert new ones I tried this code, but it doesn't work void RemoveSuccessor(TerminatorInst *TI, unsigned SuccNum) { assert(SuccNum < TI->getNumSuccessors() && "Trying to remove a nonexistant successor!"); // If our old successor block contains any PHI nodes, remove the entry in the //

Hi, I have tried to use SplitEdge function, but failed. Actually the third parameter is a variable of type Pass and it need to be non-null. But I could not figure out how to use it. Please help me out. Regards, Chayan On Sat, Jul 17, 2010 at 10:16 PM, Nick Lewycky <nicholas at mxc.ca> wrote: > Chayan Sarkar wrote: >> >> Hi all, >> >> Suppose in a CFG bb1 has two

Hi all, this (altough a big patch) is actually pretty straight forward: It (tries) to preserve ProfileInfo in all -std-compile-opts passes and all X86-Backend passes. There is still some passes that have corner cases where the ProfileInfo is not correct after the pass. Some passes are still missing... How shall I proceed with this? Andi -------------- next part -------------- A non-text