Tobias Grosser
2011-Nov-14 13:59 UTC
[LLVMdev] How to make Polly ignore some non-affine memory accesses
On 11/14/2011 02:45 PM, Marcello Maggioni wrote:> 2011/11/14 Tobias Grosser<tobias at grosser.es>: >> On 11/14/2011 01:24 AM, Marcello Maggioni wrote: >>> >>> Hi Tobias. >>> >>> I worked on enabling Polly accepting non affine memory accesses and I >>> produced a patch. >> >> Great. >> >>> I saw that there were a lot of updates in Polly recently, so I had to >>> redo a lot of the work I did and that slowed me quite a bit. >> >> Ups, sorry! However, I believe without these changes detecting non-affine >> memory access functions would have been a lot more complicated. > > Indeed! The patch almost shrinked to half the code it was before those > changes! Great work.Thanks. (You motivated me to fix this, when you asked about non-affine access functions)>> this should be >> } else { >> >>> + Type = MayWrite; >> >> You are right, we should use may-accesses here. But why always setting the >> type to may-write? A read should remain a read (as there is no difference >> between read and may-read). > > You are right, in the patch for the old version that was handled > correctly, I don't know how this reverted back to this way in the > patch for the new version , I'll get it fixed.Wonderful. Very cool that you realized I added MayWrite exactly for this case.>>> + Functions.push_back(std::make_pair(IRAccess(Type, >>> + >>> BasePointer->getValue(), >>> + AccessFunction, Size, >>> false), >>> +&Inst)); >>> + } >>> + } >>> } >>> >>> if (Functions.empty()) > > Yeah, there are quite a few stylistic problems actually, my bad!! I'll > get all the style problems fixed as fast as possible! Sorry for that.Alright. I failed on this myself, which is why Polly is not everywhere perfect. Still, we should try to follow the LLVM style guide. BTW, if you find any problems in patches I commit or in Polly itself or if your find anything that you don't understand, please go ahead and let me know. Even if it is correct and you just did not get it, I think this is a documentation bug and should be fixed.>> Besides the comments above, the patch looks great. It is a nice improvement >> to Polly. Can you repost it after the comments are addressed? Also could you >> include a minimal test case in the patch, that verifies this features is >> working correctly. > > Thank you, for your time Tobias explaining all the issues with the > patch so throughly and in a clear manner .You did a good job working on the patch. It is great if people go ahead, work for themselves and send in a patch that works. Even better if the right approach was taken and just some smaller remarks are necessary.> Should I make a specific subdirectory for the test case?No. I think you can put one test case in test/ScopInfo that runs -polly-scops -analyze and one in test/CodeGeneration that ensures that the correct code is generated.> >> Thanks >> Tobi >> >> P.S.: Please post patches to the mailing lists.> What do you mean with post patches to the mailing lists? Do you mean > in llvm-dev or llvm-commit? > The previous patch has been posted on llvm-dev, was that wrong?Normally patches should either be posted on llvm-commit or polly-dev at googlegroups.com. In this case it seems you sent two separate emails, one to me and another to Sebastian and llvmdev. I just saw the one to me and did not see that another one was sent to llvmdev. Just send next time one email with all of us copied. Cheers Tobi
Marcello Maggioni
2011-Nov-15 20:41 UTC
[LLVMdev] How to make Polly ignore some non-affine memory accesses
2011/11/14 Tobias Grosser <tobias at grosser.es>:> On 11/14/2011 02:45 PM, Marcello Maggioni wrote: >> >> 2011/11/14 Tobias Grosser<tobias at grosser.es>: >>> >>> On 11/14/2011 01:24 AM, Marcello Maggioni wrote: >>>> >>>> Hi Tobias. >>>> >>>> I worked on enabling Polly accepting non affine memory accesses and I >>>> produced a patch. >>> >>> Great. >>> >>>> I saw that there were a lot of updates in Polly recently, so I had to >>>> redo a lot of the work I did and that slowed me quite a bit. >>> >>> Ups, sorry! However, I believe without these changes detecting non-affine >>> memory access functions would have been a lot more complicated. >> >> Indeed! The patch almost shrinked to half the code it was before those >> changes! Great work. > > Thanks. (You motivated me to fix this, when you asked about non-affine > access functions) > >>> this should be >>> } else { >>> >>>> + Type = MayWrite; >>> >>> You are right, we should use may-accesses here. But why always setting >>> the >>> type to may-write? A read should remain a read (as there is no difference >>> between read and may-read). >> >> You are right, in the patch for the old version that was handled >> correctly, I don't know how this reverted back to this way in the >> patch for the new version , I'll get it fixed. > > Wonderful. Very cool that you realized I added MayWrite exactly for this > case. > >>>> + Functions.push_back(std::make_pair(IRAccess(Type, >>>> + >>>> BasePointer->getValue(), >>>> + AccessFunction, Size, >>>> false), >>>> +&Inst)); >>>> + } >>>> + } >>>> } >>>> >>>> if (Functions.empty()) >> >> Yeah, there are quite a few stylistic problems actually, my bad!! I'll >> get all the style problems fixed as fast as possible! Sorry for that. > > Alright. I failed on this myself, which is why Polly is not everywhere > perfect. Still, we should try to follow the LLVM style guide. > > BTW, if you find any problems in patches I commit or in Polly itself or if > your find anything that you don't understand, please go ahead and let me > know. Even if it is correct and you just did not get it, I think this is a > documentation bug and should be fixed. > >>> Besides the comments above, the patch looks great. It is a nice >>> improvement >>> to Polly. Can you repost it after the comments are addressed? Also could >>> you >>> include a minimal test case in the patch, that verifies this features is >>> working correctly. >> >> Thank you, for your time Tobias explaining all the issues with the >> patch so throughly and in a clear manner . > > You did a good job working on the patch. It is great if people go ahead, > work for themselves and send in a patch that works. Even better if the right > approach was taken and just some smaller remarks are necessary. > >> Should I make a specific subdirectory for the test case? > > No. I think you can put one test case in test/ScopInfo that runs > -polly-scops -analyze and one in test/CodeGeneration that ensures that the > correct code is generated. > >> >> Thanks >>> >>> Tobi >>> >>> P.S.: Please post patches to the mailing lists. > >> What do you mean with post patches to the mailing lists? Do you mean >> in llvm-dev or llvm-commit? >> The previous patch has been posted on llvm-dev, was that wrong? > > Normally patches should either be posted on llvm-commit or > polly-dev at googlegroups.com. > > In this case it seems you sent two separate emails, one to me and another to > Sebastian and llvmdev. I just saw the one to me and did not see that another > one was sent to llvmdev. Just send next time one email with all of us > copied. > > Cheers > Tobi >Hi, I implemented the changes you required, I hope this I solved all the stylistic issues now and I also added the flag you required. I added a test case to the patch , but I'm not very expert with the llvm testing framework and I had difficulties testing running the test. I tried running the polly test suite compiling polly with CMAKE, but I have linking problems with CMAKE while linking the LLVMPolly.dylib library and can't figure out why (lots of linker errors). The autotools build works perfectly but polly-tests doesn't work with autotools. I tried running directly the tests with lit , but it gives me lit.py: lit.cfg:74: fatal: No site specific configuration available! Do you know how can I run the tests manually to check if they are correct? I tried searching around , but couldn't find a reliable way to check if the test works. The preliminary patch is attached to the mail. Marcello -------------- next part -------------- A non-text attachment was scrubbed... Name: accept_affine3.diff Type: application/octet-stream Size: 10584 bytes Desc: not available URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20111115/d1d4444b/attachment.obj>
Marcello Maggioni
2011-Nov-18 21:34 UTC
[LLVMdev] How to make Polly ignore some non-affine memory accesses
2011/11/15 Marcello Maggioni <hayarms at gmail.com>:> 2011/11/14 Tobias Grosser <tobias at grosser.es>: >> On 11/14/2011 02:45 PM, Marcello Maggioni wrote: >>> >>> 2011/11/14 Tobias Grosser<tobias at grosser.es>: >>>> >>>> On 11/14/2011 01:24 AM, Marcello Maggioni wrote: >>>>> >>>>> Hi Tobias. >>>>> >>>>> I worked on enabling Polly accepting non affine memory accesses and I >>>>> produced a patch. >>>> >>>> Great. >>>> >>>>> I saw that there were a lot of updates in Polly recently, so I had to >>>>> redo a lot of the work I did and that slowed me quite a bit. >>>> >>>> Ups, sorry! However, I believe without these changes detecting non-affine >>>> memory access functions would have been a lot more complicated. >>> >>> Indeed! The patch almost shrinked to half the code it was before those >>> changes! Great work. >> >> Thanks. (You motivated me to fix this, when you asked about non-affine >> access functions) >> >>>> this should be >>>> } else { >>>> >>>>> + Type = MayWrite; >>>> >>>> You are right, we should use may-accesses here. But why always setting >>>> the >>>> type to may-write? A read should remain a read (as there is no difference >>>> between read and may-read). >>> >>> You are right, in the patch for the old version that was handled >>> correctly, I don't know how this reverted back to this way in the >>> patch for the new version , I'll get it fixed. >> >> Wonderful. Very cool that you realized I added MayWrite exactly for this >> case. >> >>>>> + Functions.push_back(std::make_pair(IRAccess(Type, >>>>> + >>>>> BasePointer->getValue(), >>>>> + AccessFunction, Size, >>>>> false), >>>>> +&Inst)); >>>>> + } >>>>> + } >>>>> } >>>>> >>>>> if (Functions.empty()) >>> >>> Yeah, there are quite a few stylistic problems actually, my bad!! I'll >>> get all the style problems fixed as fast as possible! Sorry for that. >> >> Alright. I failed on this myself, which is why Polly is not everywhere >> perfect. Still, we should try to follow the LLVM style guide. >> >> BTW, if you find any problems in patches I commit or in Polly itself or if >> your find anything that you don't understand, please go ahead and let me >> know. Even if it is correct and you just did not get it, I think this is a >> documentation bug and should be fixed. >> >>>> Besides the comments above, the patch looks great. It is a nice >>>> improvement >>>> to Polly. Can you repost it after the comments are addressed? Also could >>>> you >>>> include a minimal test case in the patch, that verifies this features is >>>> working correctly. >>> >>> Thank you, for your time Tobias explaining all the issues with the >>> patch so throughly and in a clear manner . >> >> You did a good job working on the patch. It is great if people go ahead, >> work for themselves and send in a patch that works. Even better if the right >> approach was taken and just some smaller remarks are necessary. >> >>> Should I make a specific subdirectory for the test case? >> >> No. I think you can put one test case in test/ScopInfo that runs >> -polly-scops -analyze and one in test/CodeGeneration that ensures that the >> correct code is generated. >> >>> >> Thanks >>>> >>>> Tobi >>>> >>>> P.S.: Please post patches to the mailing lists. >> >>> What do you mean with post patches to the mailing lists? Do you mean >>> in llvm-dev or llvm-commit? >>> The previous patch has been posted on llvm-dev, was that wrong? >> >> Normally patches should either be posted on llvm-commit or >> polly-dev at googlegroups.com. >> >> In this case it seems you sent two separate emails, one to me and another to >> Sebastian and llvmdev. I just saw the one to me and did not see that another >> one was sent to llvmdev. Just send next time one email with all of us >> copied. >> >> Cheers >> Tobi >> > > Hi, I implemented the changes you required, I hope this I solved all > the stylistic issues now and I also added the flag you required. > > I added a test case to the patch , but I'm not very expert with the > llvm testing framework and I had difficulties testing running the > test. I tried running the polly test suite compiling polly with CMAKE, > but I have linking problems with CMAKE while linking the > LLVMPolly.dylib library and can't figure out why (lots of linker > errors). The autotools build works perfectly but polly-tests doesn't > work with autotools. > I tried running directly the tests with lit , but it gives me > > lit.py: lit.cfg:74: fatal: No site specific configuration available! > > Do you know how can I run the tests manually to check if they are > correct? I tried searching around , but couldn't find a reliable way > to check if the test works. > > The preliminary patch is attached to the mail. > > Marcello >Ok , this is what I believe is the final patch that adds the non-affine accept functionality to Polly, this should have no issues. I added three tests, two in ScopInfo (two simple tests, one expected fail and one success based on the same source) and one in CodeGen that verifies that the code is generated. The patch is attached. The patch to correct the test runs on OSX will be posted shortly after this one (I preferred to separate the two so that a problem with either one of the two wouldn't give problems to the other and also to give more clarity on what patch solves what problem). Marcello -------------- next part -------------- A non-text attachment was scrubbed... Name: accept_affine_final.diff Type: application/octet-stream Size: 8004 bytes Desc: not available URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20111118/e9319a42/attachment.obj>
Seemingly Similar Threads
- [LLVMdev] How to make Polly ignore some non-affine memory accesses
- [LLVMdev] How to make Polly ignore some non-affine memory accesses
- [LLVMdev] How to make Polly ignore some non-affine memory accesses
- [LLVMdev] How to make Polly ignore some non-affine memory accesses
- [LLVMdev] How to make Polly ignore some non-affine memory accesses