[llvm-dev] Loop Unrolling Fail in Simple Vectorized loop (original) (raw)

Ehsan Amiri via llvm-dev llvm-dev at lists.llvm.org
Thu Oct 13 14:54:05 PDT 2016

In my example change the type of i (induction variable of the loop) to unsigned. It works, but when the type is "int" it can be controversial as int overflow is undefined behavior.

On Thu, Oct 13, 2016 at 5:50 PM, Ehsan Amiri <ehsanamiri at gmail.com> wrote:

On Thu, Oct 13, 2016 at 1:57 PM, Charith Mendis via llvm-dev <_ _llvm-dev at lists.llvm.org> wrote: Oh I see, the original loop may end normally, but by unrolling it may induce an infinite loop.

No. The problem is that step of the original loop is not 1. for(unsigned i=0; i<count; i+=4){_ _a[i] = b[i]*c[i-1];_ _}_ _Let's assume unsigned is a 32 bit integer. Then maximum unsigned number_ _will be 2^32 - 1. Let count = 2^32 - 1. When the loop iterates at some_ _point we will have i = 2^32 - 4. What is the value of i in the next_ _iteration? 2^32 cannot be represented in a 32 bit integer, so what happens_ _is that you will wrap around and in the next iteration you will have i = 0._ _So your original loop may be infinte. You can confirm this by compiling and_ _running the following program_ _#include #include using namespace std; int main() { for (int i = 0; i < UINTMAX; i += 4) { if (i == 0) cout << "i == 0!" << endl; } return 0; }

On Thursday, October 13, 2016, Alexandre Isoard <_ _alexandre.isoard at gmail.com> wrote: If count > MAXUINT-4 your loop loops indefinitely with an increment of 4, I think. On Thu, Oct 13, 2016 at 4:42 PM, Charith Mendis via llvm-dev <_ _llvm-dev at lists.llvm.org> wrote: So, I tried unrolling the following simple loop. int unroll(unsigned * a, unsigned * b, unsigned *c, unsigned count){ for(unsigned i=0; i<count; i++){_ _a[i] = b[i]*c[i-1];_ _}_ _return 0;_ _}_ _Then, the unroller is able to unroll it by 2 even though it doesn't_ _know about the range of count. SCEV of backedge taken count is (-1 + %count)_ _But, when I change the increment to 4, as in_ _int unroll(unsigned * a, unsigned * b, unsigned *c, unsigned count){_ _for(unsigned i=0; i<count; i+=4){_ _a[i] = b[i]*c[i-1];_ _}_ _return 0;_ _}_ _The unroller cannot compute the backedge taken count. Therefore, it_ _seems like the problem is not with the range of "count", can't the unroller_ _compute it as (- 1 + %count / 4)?_ _On Wed, Oct 12, 2016 at 11:28 PM, Charith Mendis <_ _char.mendis1989 at gmail.com> wrote: Thanks for the explanation. But I am a little confused with the following fact. Can't LLVM keep vectorizableelements as a symbolic value and convert the loop to say; for(unsigned i = 0; i < vectorizableelements ; i += 2){_ _//main loop_ _}_ _for(unsigned i=0 ; i < vectorizableelements % 2; i++){_ _//fix up_ _}_ _Why does it have to reason about the range of vectorizableelements?_ _Even if vectorizableelements == SIZEMAX the above decomposition would_ _work?_ _On Wed, Oct 12, 2016 at 8:25 PM, Friedman, Eli <_ _efriedma at codeaurora.org> wrote: On 10/12/2016 4:35 PM, Charith Mendis via llvm-dev wrote: Hi all, Attached herewith is a simple vectorized function with loops performing a simple shuffle. I want all loops (inner and outer) to be unrolled by 2 and as such used -unroll-count=2 The inner loops(with k as the induction variable and having constant trip counts) unroll fully, but the outer loop with (j) fails to unroll. The llvm code is also attached with inner loops fully unrolled. To inspect further, I added the following to the PassManagerBuilder.cpp to run some canonicalization routines and redo unrolling again. I have set partial unrolling on + have a huge threshold + allows expensive loop trip counts. Still it didn't unroll by 2. MPM.add(createLoopUnrollPass()); MPM.add(createCFGSimplificationPass()); MPM.add(createLoopSimplifyPass()); MPM.add(createLoopRotatePass(SizeLevel == 2 ? 0 : -1)); MPM.add(createLCSSAPass()); MPM.add(createIndVarSimplifyPass()); // Canonicalize indvars MPM.add(createLoopUnrollPass());

Digging deeper I found, that it fails in UnrollRuntimeLoopRemainder function, where it is unable to calculate the BackEdge taken amount. Can anybody explain what is need to get the outer loop unrolled by 2? It would be a great help. Well, I can at least explain what is happening... runtime unrolling needs to be able to symbolically compute the trip count to avoid inserting a branch after every iteration. SCEV isn't able to prove that your loop isn't an infinite loop (consider the case of vectorizableelements==SIZEMAX), therefore it can't compute the trip count. Therefore, we don't unroll. There's a few different angles you could use to attack this: you could teach the unroller to unroll loops with an uncomputable trip count, or you can make the trip count of your loop computable somehow. Changing the unroller is probably straightforward (see the recently committed r284044). Making the trip count computable is more complicated... it's probably possible to teach SCEV to reason about the overflow in the pointer computation, or maybe you could version the loop. -Eli -- Employee of Qualcomm Innovation Center, Inc. Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project

-- Kind regards, Charith Mendis Graduate Student, CSAIL, Massachusetts Institute of Technology -- Kind regards, Charith Mendis Graduate Student, CSAIL, Massachusetts Institute of Technology

LLVM Developers mailing list llvm-dev at lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev

-- Alexandre Isoard

LLVM Developers mailing list llvm-dev at lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20161013/d5d523c6/attachment.html>

More information about the llvm-dev mailing list