similar to: [LLVMdev] Possible typo in LoopUnrollPass.cpp

Is it possible to unroll a loop (forcibly if necessary) with llvm (possibly the -loop-unroll pass) a constant number of times. I believe that I read that the -unroll-count=x option was removed, correct? So is there some other way to do this or is this just not possible in llvm? -------------- next part -------------- An HTML attachment was scrubbed... URL:

Could we move this setting to function attributes? We already have OptimizeForSize / MinSize there, but not the other opt levels. We also have OptimizeNone, which seems to be completely unused. This would let us support __attribute__((optimize())) in the future, which is currently ignored. Another example would be an LTO link of objects compiled with different optimization settings. I'm not

Hi all, I'm interested in adding code size optimizations as a GSoC project. This is necessary when compiling for embedded devices, where LLVM should optimize for size even at the expense of speed. I'm still working on my proposal, but I'd like some advice on the technical parts and overall project plan. First, I would add a way to determine which parts of the code should be optimized

Hi Hal, I did opt.exe -S -debug -loop-unroll -unroll-runtime=true -unroll-count=4 csShader.ll and it prints out: Args: opt.exe -S -debug -loop-unroll -unroll-runtime=true -unroll-count=4 csShader.ll Loop Unroll: F[build_cs_5_0] Loop %loop_entry Loop Size = 82 partially unrolling with count: 1 Thanks, Frances On Thu, Oct 15, 2015 at 9:35 PM, Hal Finkel <hfinkel at anl.gov>

Good evening, Kenneth. Thank you to apply (and rewrite my naive code better) and to file the issue to http://llvm.org/bugs/show_bug.cgi?id=7304 I have checked r105528 at this morning. I think the pass must be still cleaned up and rewritten. There are my two proposals for enhancement. 1) To separate Specialization(and rewriting callsites) to other module. It would be better if new module were

Hi, I ran into this issue recently and wanted to know if it was a bug or expected behavior. In the R600 backend's TargetTransformInfo implementation, we were setting UnrollingPreferences::Count = UINT_MAX. This was a mistake as we should have been setting UnrollingPreferences::MaxCount instead. However, as a result of setting Count to UINT_MAX, this loop would be unrolled 15 times: if (b

Hi, Looking at bugzilla PR11087, I'd like to conditionalise a transformation in ARMIselLowering.cpp based on whether we're compiling for codesize or performance. -Os doesn't actually exist for llc, and I can't see an obvious place where that condition would be set. Where do we specify if we're optimizing for codesize or performance? Cheers, James --------------

On 28 November 2013 13:31, David Tweed <david.tweed at gmail.com> wrote: > Indeed, a most of the bugs which > really need a debugger are manifest in big applications where even a > non-debug build can be very "not simple".) My example was a very crude example of simplicity. But the more complex your application is, the simpler you want the compiler to be for a debug

So.. I have investigated more on how a new function attribute to disable optimization on a per-function basis could be implemented. At the current state, with the lack of specific support from the pass managers I found two big problems when trying to implement a prototype implementation of the new attribute. Here are the problems found: 1) It is not safe to disable some transform passes in the

On Thu, Nov 28, 2013 at 1:11 PM, Renato Golin <renato.golin at linaro.org>wrote: > On 28 November 2013 00:00, Robinson, Paul > <Paul_Robinson at playstation.sony.com> wrote: > > In my experience, to a first approximation, anything > > that changes the CFG or that reorders generated code beyond source > > statement boundaries is likely to make things more

Hi, For a certain type of analysis I generate inline assembly which I insert into the LLVM IR code. This inline assembly code contains labels that should not be duplicated. Problem is that the tail duplication pass duplicates code. It checks isNotDuplicatable on a machine instruction however there is no such a flag on an IR instruction that I could set. Is there a way to tell in an IR pass that

I'm running an LSODA to generate some graphs, but I need to stop at a certain point and use those values to generate another LSODA output. This is working fine, but when I try to run the second LSODA, I get the "Error in eval(expr, envir, enclos) : object 'N' not found". Any ideas what can be causing this? I have no object 'N' anywhere in the script. I made an

Hi, I am trying to test if disabling function (entry) alignment has any effect on performance. On my x86 machine, all functions seem to get aligned to 16-bytes. I know that GCC provides a flag (-fno-align-functions) to disable function alignment. However, I could not find the similar flag for clang. According to the following quote from LLVM doc, clang delegates the job to the target in case no

Dear guys, I am having problems to obtain good information from the LoopInfo. I am always getting a trip count of 0, even though I am clearly passing a loop with a constant bound. I am using this pass below: void testLoopInfo(const Function& F) const { const LoopInfo *LI = &getAnalysis<LoopInfo>(); Function::const_iterator BB = F.begin(), E = F.end(); for (; BB !=

Dear All, I am glad to be part of GSoC 2016 with LLVM organization. I am a first year PhD student at IIT Hyderabad, India and my research area is compiler optimizations using polyhedral model. My GSoC 2016 project is to implement Polly as an Analysis pass in LLVM [1]. We have a discussion on Polly-dev mailing list [2] on taking a better approach to implement this project. Based upon this

Hi Zhoulai, I am trying to modify "LoopUnrollPass" in llvm which produces multiple copies of loop equal to the loop unroll factor.Currently, using multicore architecture, say 3 for example and the execution goes like: for 3 cores if there are 9 iterations of loop core instruction 1 0,3,6 2 1,4,7 3 2,5,8 But I want to to

Hi, This is a proposal about implementing an analysis that calculates loop cost based on cache data. The primary motivation for implementing this is to write profitability measures for cache related optimizations like interchange, fusion, fission, pre-fetching and others. I have implemented a prototypical version at http://reviews.llvm.org/D21124. The patch basically creates groups of references

Andrea_DiBiagio at sn.scee.net wrote: > So.. > I have investigated more on how a new function attribute to disable > optimization on a per-function basis could be implemented. > At the current state, with the lack of specific support from the pass > managers I found two big problems when trying to implement a prototype > implementation of the new attribute. > > Here are the

Hi, On Fri, Jun 4, 2010 at 5:18 AM, Pekka Nikander <pekka.nikander at nomadiclab.com> wrote: >  Would the best way be to add an option to -loop-unroll, and hack away at lib/Transforms/Utils/LoopUnroll.cpp? Instead, the better alternative is to write another pass similar to LoopUnrollPass.cpp (say LoopPeelPass.cpp) and add new option -loop-peel. The new pass could use llvm::UnrollLoop()

Hi, Ivan: Sorry for deviating the topic a bit. As I told you before I'm a LLVM newbie, I cannot give you conclusive answer if the proposed interface is ok or not. My personal opinion on these two interface is summarized bellow: - hasZeroCostLoop() pro: it is clearly state the HW support. con: Having zero cost loop doesn't imply the benefit HW loop could achieve.