Adrien Guinet via llvm-dev
2017-Jun-17 22:41 UTC
[llvm-dev] LoopVectorize fails to vectorize loops with induction variables with PtrToInt/IntToPtr conversions
Hello all, There is a missing vectorization opportunity issue with clang 4.0 with the file attached. Indeed, when compiled with -O2, the "op_distance" function get vectorized, but not the "op" one. For information, this test case has been reduced from a file generated by the Pythran compiler (https://github.com/serge-sans-paille/pythran). If we take a look at the generated IR without vectorization (using the -fno-vectorize clang flag), we get:> $ clang -O2 -S -emit-llvm op_zip_iterator.cpp -std=c++11 -o - -fno-vectorize> ; Function Attrs: norecurse uwtable > define void @_Z11op_distancePi16add_zip_iteratorS0_(i32* nocapture, i32*, i32* nocapture readonly, i32*, i32* nocapture readnone) local_unnamed_addr #0 { > ; This one is vectorized! > %6 = ptrtoint i32* %1 to i64 > %7 = ptrtoint i32* %3 to i64 > %8 = sub i64 %7, %6 > %9 = icmp sgt i64 %8, 0 > br i1 %9, label %10, label %26 > > ; <label>:10: ; preds = %5 > %11 = lshr exact i64 %8, 2 > br label %12 > > ; <label>:12: ; preds = %12, %10 > %13 = phi i64 [ %23, %12 ], [ %11, %10 ] > %14 = phi i32* [ %22, %12 ], [ %0, %10 ] > %15 = phi i32* [ %21, %12 ], [ %2, %10 ] > %16 = phi i32* [ %20, %12 ], [ %1, %10 ] > %17 = load i32, i32* %16, align 4, !tbaa !1 > %18 = load i32, i32* %15, align 4, !tbaa !1 > %19 = add nsw i32 %18, %17 > store i32 %19, i32* %14, align 4, !tbaa !1 > %20 = getelementptr inbounds i32, i32* %16, i64 1 > %21 = getelementptr inbounds i32, i32* %15, i64 1 > %22 = getelementptr inbounds i32, i32* %14, i64 1 > %23 = add nsw i64 %13, -1 > %24 = icmp sgt i64 %13, 1 > br i1 %24, label %12, label %25 > > ; <label>:25: ; preds = %12 > br label %26 > > ; <label>:26: ; preds = %25, %5 > ret void > } > > ; Function Attrs: norecurse uwtable > define void @_Z2opPi16add_zip_iteratorS0_(i32* nocapture, i32*, i32* nocapture readonly, i32*, i32* nocapture readnone) local_unnamed_addr #0 { > ; This one isn't! > %6 = ptrtoint i32* %1 to i64 > %7 = ptrtoint i32* %3 to i64 > %8 = sub i64 %6, %7 > %9 = icmp sgt i64 %8, 0 > br i1 %9, label %10, label %28 > > ; <label>:10: ; preds = %5 > %11 = lshr exact i64 %8, 2 > br label %12 > > ; <label>:12: ; preds = %12, %10 > %13 = phi i64 [ %25, %12 ], [ %11, %10 ] > %14 = phi i32* [ %24, %12 ], [ %0, %10 ] > %15 = phi i32* [ %23, %12 ], [ %2, %10 ] > %16 = phi i64 [ %22, %12 ], [ %6, %10 ] > %17 = inttoptr i64 %16 to i32* > %18 = load i32, i32* %17, align 4, !tbaa !1 > %19 = load i32, i32* %15, align 4, !tbaa !1 > %20 = add nsw i32 %19, %18 > store i32 %20, i32* %14, align 4, !tbaa !1 > %21 = getelementptr inbounds i32, i32* %17, i64 1 > %22 = ptrtoint i32* %21 to i64 > %23 = getelementptr inbounds i32, i32* %15, i64 1 > %24 = getelementptr inbounds i32, i32* %14, i64 1 > %25 = add nsw i64 %13, -1 > %26 = icmp sgt i64 %13, 1 > br i1 %26, label %12, label %27 > > ; <label>:27: ; preds = %12 > br label %28 > > ; <label>:28: ; preds = %27, %5 > ret void > }If we compile only the "op" function while activation the debug mode, here is the output:> $ clang -O2 -S -emit-llvm op_zip_iterator.cpp -std=c++11 -o - -fno-vectorize |~/dev/epona-llvm/build_debug_shared/bin/opt -debug -debug-only loop-vectorize -O2 -S > > LV: Checking a loop in "_Z2opPi16add_zip_iteratorS0_" from <stdin> > LV: Loop hints: force=? width=0 unroll=0 > LV: Found a loop: > LV: Found an induction variable. > LV: Found an induction variable. > LV: Found an induction variable. > LV: Found an unidentified PHI. %16 = phi i64 [ %22, %12 ], [ %6, %10 ] > LV: Can't vectorize the instructions or CFG > LV: Not vectorizing: Cannot prove legality. > [...]The issue seems to be that the phi node "%16" can't be deduced as an induction variable. If we take a closer look, the cause seems to be in ScalarEvolution, in the createSCEV function (http://llvm.org/docs/doxygen/html/ScalarEvolution_8cpp_source.html#l04770) :> // It's tempting to handle inttoptr and ptrtoint as no-ops, however this can > // lead to pointer expressions which cannot safely be expanded to GEPs, > // because ScalarEvolution doesn't respect the GEP aliasing rules when > // simplifying integer expressions.Indeed, SCEV does not (legitimately) consider inttoptr/ptrtoint as no-op, and does not handle them. The thing is that, in our case, the GEP in %23 is thus not analyzed by SCEV, and the PHI %16 is thus not considered as an induction variable. To confirm this hypothesis, I created a small out-of-tree pass (https://github.com/aguinet/llvm-intptrcleanup) which registers before loop vectorization and does the following: * first, it search for phi nodes who have those properties: - every incoming value of the phi node is a ptrtoint instruction. The original pointer type of every ptrtoint instruction must be the same type T. - every user of this PHI node is an inttoptr instruction of the previous type T * for each of these PHI nodes, it creates a new PHI node which takes the original pointers as incoming values, and replace the uses of the inttoptr instructions that uses the original PHI node by the new one * it then removes the previous inttoptr instructions and the original PHI node The way I understand inttoptr and ptrtoint, this transformation should be valid (but I might have missed something!). Please note that this is a quick'n'dirty pass, which hasn't been heavily tested. Using this pass, the previous example is now vectorized correctly by the loop vectorizer. This can be seen by looking at the output of:> $ clang -Xclang -load -Xclang IntToPtrCleanup.so -O2 ./example/op_zip_operator.cpp -S -emit-llvm -o - -std=c++11The question that remains to me is how this should be correctly fixed: 1) Making SCEV support these no-op (in this case) inttoptr/ptrtoint conversions 2) insert the above transformation at some point in the optimization pipeline 3) clean the pass(es?) that somehow generated this case. I have to admit I'm not really sure which options is the best. 3) seems to be the way to go but might require some tedious work, and does not prevent the issue to come again in the future. 2) seems to be a quick patch that could be inserted in some "canonicalization" pass, let it be a valid transformation in the first place. I don't know SCEV enough to judge of the difficulty/faisability of 1). This mail is thus to discuss this issue and how to fix this properly :) Thanks everyone :) -- Adrien Guinet. -------------- next part -------------- A non-text attachment was scrubbed... Name: op_zip_iterator.cpp Type: text/x-c++src Size: 1930 bytes Desc: not available URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170618/d4c740e5/attachment.cpp>
Davide Italiano via llvm-dev
2017-Jun-17 23:07 UTC
[llvm-dev] LoopVectorize fails to vectorize loops with induction variables with PtrToInt/IntToPtr conversions
FYI. On Sat, Jun 17, 2017 at 3:41 PM, Adrien Guinet via llvm-dev <llvm-dev at lists.llvm.org> wrote:> Hello all, > > There is a missing vectorization opportunity issue with clang 4.0 with > the file attached. > > Indeed, when compiled with -O2, the "op_distance" function get > vectorized, but not the "op" one. > > For information, this test case has been reduced from a file generated > by the Pythran compiler (https://github.com/serge-sans-paille/pythran). > > If we take a look at the generated IR without vectorization (using the > -fno-vectorize clang flag), we get: > >> $ clang -O2 -S -emit-llvm op_zip_iterator.cpp -std=c++11 -o - -fno-vectorize > >> ; Function Attrs: norecurse uwtable >> define void @_Z11op_distancePi16add_zip_iteratorS0_(i32* nocapture, i32*, i32* nocapture readonly, i32*, i32* nocapture readnone) local_unnamed_addr #0 { >> ; This one is vectorized! >> %6 = ptrtoint i32* %1 to i64 >> %7 = ptrtoint i32* %3 to i64 >> %8 = sub i64 %7, %6 >> %9 = icmp sgt i64 %8, 0 >> br i1 %9, label %10, label %26 >> >> ; <label>:10: ; preds = %5 >> %11 = lshr exact i64 %8, 2 >> br label %12 >> >> ; <label>:12: ; preds = %12, %10 >> %13 = phi i64 [ %23, %12 ], [ %11, %10 ] >> %14 = phi i32* [ %22, %12 ], [ %0, %10 ] >> %15 = phi i32* [ %21, %12 ], [ %2, %10 ] >> %16 = phi i32* [ %20, %12 ], [ %1, %10 ] >> %17 = load i32, i32* %16, align 4, !tbaa !1 >> %18 = load i32, i32* %15, align 4, !tbaa !1 >> %19 = add nsw i32 %18, %17 >> store i32 %19, i32* %14, align 4, !tbaa !1 >> %20 = getelementptr inbounds i32, i32* %16, i64 1 >> %21 = getelementptr inbounds i32, i32* %15, i64 1 >> %22 = getelementptr inbounds i32, i32* %14, i64 1 >> %23 = add nsw i64 %13, -1 >> %24 = icmp sgt i64 %13, 1 >> br i1 %24, label %12, label %25 >> >> ; <label>:25: ; preds = %12 >> br label %26 >> >> ; <label>:26: ; preds = %25, %5 >> ret void >> } >> >> ; Function Attrs: norecurse uwtable >> define void @_Z2opPi16add_zip_iteratorS0_(i32* nocapture, i32*, i32* nocapture readonly, i32*, i32* nocapture readnone) local_unnamed_addr #0 { >> ; This one isn't! >> %6 = ptrtoint i32* %1 to i64 >> %7 = ptrtoint i32* %3 to i64 >> %8 = sub i64 %6, %7 >> %9 = icmp sgt i64 %8, 0 >> br i1 %9, label %10, label %28 >> >> ; <label>:10: ; preds = %5 >> %11 = lshr exact i64 %8, 2 >> br label %12 >> >> ; <label>:12: ; preds = %12, %10 >> %13 = phi i64 [ %25, %12 ], [ %11, %10 ] >> %14 = phi i32* [ %24, %12 ], [ %0, %10 ] >> %15 = phi i32* [ %23, %12 ], [ %2, %10 ] >> %16 = phi i64 [ %22, %12 ], [ %6, %10 ] >> %17 = inttoptr i64 %16 to i32* >> %18 = load i32, i32* %17, align 4, !tbaa !1 >> %19 = load i32, i32* %15, align 4, !tbaa !1 >> %20 = add nsw i32 %19, %18 >> store i32 %20, i32* %14, align 4, !tbaa !1 >> %21 = getelementptr inbounds i32, i32* %17, i64 1 >> %22 = ptrtoint i32* %21 to i64 >> %23 = getelementptr inbounds i32, i32* %15, i64 1 >> %24 = getelementptr inbounds i32, i32* %14, i64 1 >> %25 = add nsw i64 %13, -1 >> %26 = icmp sgt i64 %13, 1 >> br i1 %26, label %12, label %27 >> >> ; <label>:27: ; preds = %12 >> br label %28 >> >> ; <label>:28: ; preds = %27, %5 >> ret void >> } > > If we compile only the "op" function while activation the debug mode, > here is the output: > >> $ clang -O2 -S -emit-llvm op_zip_iterator.cpp -std=c++11 -o - -fno-vectorize |~/dev/epona-llvm/build_debug_shared/bin/opt -debug -debug-only loop-vectorize -O2 -S >> >> LV: Checking a loop in "_Z2opPi16add_zip_iteratorS0_" from <stdin> >> LV: Loop hints: force=? width=0 unroll=0 >> LV: Found a loop: >> LV: Found an induction variable. >> LV: Found an induction variable. >> LV: Found an induction variable. >> LV: Found an unidentified PHI. %16 = phi i64 [ %22, %12 ], [ %6, %10 ] >> LV: Can't vectorize the instructions or CFG >> LV: Not vectorizing: Cannot prove legality. >> [...] > > The issue seems to be that the phi node "%16" can't be deduced as an > induction variable. If we take a closer look, the cause seems to be in > ScalarEvolution, in the createSCEV function > (http://llvm.org/docs/doxygen/html/ScalarEvolution_8cpp_source.html#l04770) > : > >> // It's tempting to handle inttoptr and ptrtoint as no-ops, however this can >> // lead to pointer expressions which cannot safely be expanded to GEPs, >> // because ScalarEvolution doesn't respect the GEP aliasing rules when >> // simplifying integer expressions. > > Indeed, SCEV does not (legitimately) consider inttoptr/ptrtoint as > no-op, and does not handle them. The thing is that, in our case, the GEP > in %23 is thus not analyzed by SCEV, and the PHI %16 is thus not > considered as an induction variable. > > To confirm this hypothesis, I created a small out-of-tree pass > (https://github.com/aguinet/llvm-intptrcleanup) which registers before > loop vectorization and does the following: > > * first, it search for phi nodes who have those properties: > - every incoming value of the phi node is a ptrtoint instruction. The > original pointer type of every ptrtoint instruction must be the same type T. > - every user of this PHI node is an inttoptr instruction of the > previous type T > * for each of these PHI nodes, it creates a new PHI node which takes the > original pointers as incoming values, and replace the uses of the > inttoptr instructions that uses the original PHI node by the new one > * it then removes the previous inttoptr instructions and the original > PHI node > > The way I understand inttoptr and ptrtoint, this transformation should > be valid (but I might have missed something!). Please note that this is > a quick'n'dirty pass, which hasn't been heavily tested. Using this pass, > the previous example is now vectorized correctly by the loop vectorizer. > This can be seen by looking at the output of: > >> $ clang -Xclang -load -Xclang IntToPtrCleanup.so -O2 ./example/op_zip_operator.cpp -S -emit-llvm -o - -std=c++11 > > The question that remains to me is how this should be correctly fixed: > > 1) Making SCEV support these no-op (in this case) inttoptr/ptrtoint > conversions > 2) insert the above transformation at some point in the optimization > pipeline > 3) clean the pass(es?) that somehow generated this case. > > I have to admit I'm not really sure which options is the best. 3) seems > to be the way to go but might require some tedious work, and does not > prevent the issue to come again in the future. 2) seems to be a quick > patch that could be inserted in some "canonicalization" pass, let it be > a valid transformation in the first place. I don't know SCEV enough to > judge of the difficulty/faisability of 1). > > This mail is thus to discuss this issue and how to fix this properly :) > > Thanks everyone :) > > -- > Adrien Guinet. > > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev >-- Davide "There are no solved problems; there are only problems that are more or less solved" -- Henri Poincare
Davide Italiano via llvm-dev
2017-Jun-17 23:07 UTC
[llvm-dev] LoopVectorize fails to vectorize loops with induction variables with PtrToInt/IntToPtr conversions
Sorry, hit reply instead of forward :) On Sat, Jun 17, 2017 at 4:07 PM, Davide Italiano <davide at freebsd.org> wrote:> FYI. > > On Sat, Jun 17, 2017 at 3:41 PM, Adrien Guinet via llvm-dev > <llvm-dev at lists.llvm.org> wrote: >> Hello all, >> >> There is a missing vectorization opportunity issue with clang 4.0 with >> the file attached. >> >> Indeed, when compiled with -O2, the "op_distance" function get >> vectorized, but not the "op" one. >> >> For information, this test case has been reduced from a file generated >> by the Pythran compiler (https://github.com/serge-sans-paille/pythran). >> >> If we take a look at the generated IR without vectorization (using the >> -fno-vectorize clang flag), we get: >> >>> $ clang -O2 -S -emit-llvm op_zip_iterator.cpp -std=c++11 -o - -fno-vectorize >> >>> ; Function Attrs: norecurse uwtable >>> define void @_Z11op_distancePi16add_zip_iteratorS0_(i32* nocapture, i32*, i32* nocapture readonly, i32*, i32* nocapture readnone) local_unnamed_addr #0 { >>> ; This one is vectorized! >>> %6 = ptrtoint i32* %1 to i64 >>> %7 = ptrtoint i32* %3 to i64 >>> %8 = sub i64 %7, %6 >>> %9 = icmp sgt i64 %8, 0 >>> br i1 %9, label %10, label %26 >>> >>> ; <label>:10: ; preds = %5 >>> %11 = lshr exact i64 %8, 2 >>> br label %12 >>> >>> ; <label>:12: ; preds = %12, %10 >>> %13 = phi i64 [ %23, %12 ], [ %11, %10 ] >>> %14 = phi i32* [ %22, %12 ], [ %0, %10 ] >>> %15 = phi i32* [ %21, %12 ], [ %2, %10 ] >>> %16 = phi i32* [ %20, %12 ], [ %1, %10 ] >>> %17 = load i32, i32* %16, align 4, !tbaa !1 >>> %18 = load i32, i32* %15, align 4, !tbaa !1 >>> %19 = add nsw i32 %18, %17 >>> store i32 %19, i32* %14, align 4, !tbaa !1 >>> %20 = getelementptr inbounds i32, i32* %16, i64 1 >>> %21 = getelementptr inbounds i32, i32* %15, i64 1 >>> %22 = getelementptr inbounds i32, i32* %14, i64 1 >>> %23 = add nsw i64 %13, -1 >>> %24 = icmp sgt i64 %13, 1 >>> br i1 %24, label %12, label %25 >>> >>> ; <label>:25: ; preds = %12 >>> br label %26 >>> >>> ; <label>:26: ; preds = %25, %5 >>> ret void >>> } >>> >>> ; Function Attrs: norecurse uwtable >>> define void @_Z2opPi16add_zip_iteratorS0_(i32* nocapture, i32*, i32* nocapture readonly, i32*, i32* nocapture readnone) local_unnamed_addr #0 { >>> ; This one isn't! >>> %6 = ptrtoint i32* %1 to i64 >>> %7 = ptrtoint i32* %3 to i64 >>> %8 = sub i64 %6, %7 >>> %9 = icmp sgt i64 %8, 0 >>> br i1 %9, label %10, label %28 >>> >>> ; <label>:10: ; preds = %5 >>> %11 = lshr exact i64 %8, 2 >>> br label %12 >>> >>> ; <label>:12: ; preds = %12, %10 >>> %13 = phi i64 [ %25, %12 ], [ %11, %10 ] >>> %14 = phi i32* [ %24, %12 ], [ %0, %10 ] >>> %15 = phi i32* [ %23, %12 ], [ %2, %10 ] >>> %16 = phi i64 [ %22, %12 ], [ %6, %10 ] >>> %17 = inttoptr i64 %16 to i32* >>> %18 = load i32, i32* %17, align 4, !tbaa !1 >>> %19 = load i32, i32* %15, align 4, !tbaa !1 >>> %20 = add nsw i32 %19, %18 >>> store i32 %20, i32* %14, align 4, !tbaa !1 >>> %21 = getelementptr inbounds i32, i32* %17, i64 1 >>> %22 = ptrtoint i32* %21 to i64 >>> %23 = getelementptr inbounds i32, i32* %15, i64 1 >>> %24 = getelementptr inbounds i32, i32* %14, i64 1 >>> %25 = add nsw i64 %13, -1 >>> %26 = icmp sgt i64 %13, 1 >>> br i1 %26, label %12, label %27 >>> >>> ; <label>:27: ; preds = %12 >>> br label %28 >>> >>> ; <label>:28: ; preds = %27, %5 >>> ret void >>> } >> >> If we compile only the "op" function while activation the debug mode, >> here is the output: >> >>> $ clang -O2 -S -emit-llvm op_zip_iterator.cpp -std=c++11 -o - -fno-vectorize |~/dev/epona-llvm/build_debug_shared/bin/opt -debug -debug-only loop-vectorize -O2 -S >>> >>> LV: Checking a loop in "_Z2opPi16add_zip_iteratorS0_" from <stdin> >>> LV: Loop hints: force=? width=0 unroll=0 >>> LV: Found a loop: >>> LV: Found an induction variable. >>> LV: Found an induction variable. >>> LV: Found an induction variable. >>> LV: Found an unidentified PHI. %16 = phi i64 [ %22, %12 ], [ %6, %10 ] >>> LV: Can't vectorize the instructions or CFG >>> LV: Not vectorizing: Cannot prove legality. >>> [...] >> >> The issue seems to be that the phi node "%16" can't be deduced as an >> induction variable. If we take a closer look, the cause seems to be in >> ScalarEvolution, in the createSCEV function >> (http://llvm.org/docs/doxygen/html/ScalarEvolution_8cpp_source.html#l04770) >> : >> >>> // It's tempting to handle inttoptr and ptrtoint as no-ops, however this can >>> // lead to pointer expressions which cannot safely be expanded to GEPs, >>> // because ScalarEvolution doesn't respect the GEP aliasing rules when >>> // simplifying integer expressions. >> >> Indeed, SCEV does not (legitimately) consider inttoptr/ptrtoint as >> no-op, and does not handle them. The thing is that, in our case, the GEP >> in %23 is thus not analyzed by SCEV, and the PHI %16 is thus not >> considered as an induction variable. >> >> To confirm this hypothesis, I created a small out-of-tree pass >> (https://github.com/aguinet/llvm-intptrcleanup) which registers before >> loop vectorization and does the following: >> >> * first, it search for phi nodes who have those properties: >> - every incoming value of the phi node is a ptrtoint instruction. The >> original pointer type of every ptrtoint instruction must be the same type T. >> - every user of this PHI node is an inttoptr instruction of the >> previous type T >> * for each of these PHI nodes, it creates a new PHI node which takes the >> original pointers as incoming values, and replace the uses of the >> inttoptr instructions that uses the original PHI node by the new one >> * it then removes the previous inttoptr instructions and the original >> PHI node >> >> The way I understand inttoptr and ptrtoint, this transformation should >> be valid (but I might have missed something!). Please note that this is >> a quick'n'dirty pass, which hasn't been heavily tested. Using this pass, >> the previous example is now vectorized correctly by the loop vectorizer. >> This can be seen by looking at the output of: >> >>> $ clang -Xclang -load -Xclang IntToPtrCleanup.so -O2 ./example/op_zip_operator.cpp -S -emit-llvm -o - -std=c++11 >> >> The question that remains to me is how this should be correctly fixed: >> >> 1) Making SCEV support these no-op (in this case) inttoptr/ptrtoint >> conversions >> 2) insert the above transformation at some point in the optimization >> pipeline >> 3) clean the pass(es?) that somehow generated this case. >> >> I have to admit I'm not really sure which options is the best. 3) seems >> to be the way to go but might require some tedious work, and does not >> prevent the issue to come again in the future. 2) seems to be a quick >> patch that could be inserted in some "canonicalization" pass, let it be >> a valid transformation in the first place. I don't know SCEV enough to >> judge of the difficulty/faisability of 1). >> >> This mail is thus to discuss this issue and how to fix this properly :) >> >> Thanks everyone :) >> >> -- >> Adrien Guinet. >> >> _______________________________________________ >> LLVM Developers mailing list >> llvm-dev at lists.llvm.org >> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev >> > > > > -- > Davide > > "There are no solved problems; there are only problems that are more > or less solved" -- Henri Poincare-- Davide "There are no solved problems; there are only problems that are more or less solved" -- Henri Poincare
Adrien Guinet via llvm-dev
2017-Jun-17 23:08 UTC
[llvm-dev] LoopVectorize fails to vectorize loops with induction variables with PtrToInt/IntToPtr conversions
For the record, I tried with the latest snapshot available on apt.llvm.org, and the issue is still present.> $ clang-5.0 --version > clang version 5.0.0-svn305634-1~exp1 (trunk)
Hal Finkel via llvm-dev
2017-Jun-20 01:26 UTC
[llvm-dev] LoopVectorize fails to vectorize loops with induction variables with PtrToInt/IntToPtr conversions
Hi, Adrien, Thanks for reporting this. I recommend that you file a bug report at https://bugs.llvm.org/ Whenever I see reports of missed optimization opportunities in the face of ptrtoint/inttoptr, my first question is: why are these instructions present in the first place? At the IR level, use of inttoptr is highly discouraged. Our aliasing analysis, for example, does not look through them, and so you'll generally see a lot of missed optimizations when they're around. In this case, inttoptr seems to be introduced by SROA. SROA should not be introducing inttoptr, but rather should be using GEPs on i8* (at least), to avoid introducing pointers that our AA can't analyze. Beyond that, if we need to handle inttoptr/ptrtoint in SCEV, then maybe there's a way to make it smarter about the expressions with which it can deal. I'm actually not sure to what "aliasing rules" the comment you quote below is referring. I can certainly understand not being able to place "inbounds" on some generated GEPs, but otherwise this seems non-obvious to me (i.e. either the expander can identify a base pointer from which to generate the GEP, or it can't, in which case it needs to generate a inttoptr). Sanjoy, thoughts? -Hal On 06/17/2017 05:41 PM, Adrien Guinet via llvm-dev wrote:> Hello all, > > There is a missing vectorization opportunity issue with clang 4.0 with > the file attached. > > Indeed, when compiled with -O2, the "op_distance" function get > vectorized, but not the "op" one. > > For information, this test case has been reduced from a file generated > by the Pythran compiler (https://github.com/serge-sans-paille/pythran). > > If we take a look at the generated IR without vectorization (using the > -fno-vectorize clang flag), we get: > >> $ clang -O2 -S -emit-llvm op_zip_iterator.cpp -std=c++11 -o - -fno-vectorize >> ; Function Attrs: norecurse uwtable >> define void @_Z11op_distancePi16add_zip_iteratorS0_(i32* nocapture, i32*, i32* nocapture readonly, i32*, i32* nocapture readnone) local_unnamed_addr #0 { >> ; This one is vectorized! >> %6 = ptrtoint i32* %1 to i64 >> %7 = ptrtoint i32* %3 to i64 >> %8 = sub i64 %7, %6 >> %9 = icmp sgt i64 %8, 0 >> br i1 %9, label %10, label %26 >> >> ; <label>:10: ; preds = %5 >> %11 = lshr exact i64 %8, 2 >> br label %12 >> >> ; <label>:12: ; preds = %12, %10 >> %13 = phi i64 [ %23, %12 ], [ %11, %10 ] >> %14 = phi i32* [ %22, %12 ], [ %0, %10 ] >> %15 = phi i32* [ %21, %12 ], [ %2, %10 ] >> %16 = phi i32* [ %20, %12 ], [ %1, %10 ] >> %17 = load i32, i32* %16, align 4, !tbaa !1 >> %18 = load i32, i32* %15, align 4, !tbaa !1 >> %19 = add nsw i32 %18, %17 >> store i32 %19, i32* %14, align 4, !tbaa !1 >> %20 = getelementptr inbounds i32, i32* %16, i64 1 >> %21 = getelementptr inbounds i32, i32* %15, i64 1 >> %22 = getelementptr inbounds i32, i32* %14, i64 1 >> %23 = add nsw i64 %13, -1 >> %24 = icmp sgt i64 %13, 1 >> br i1 %24, label %12, label %25 >> >> ; <label>:25: ; preds = %12 >> br label %26 >> >> ; <label>:26: ; preds = %25, %5 >> ret void >> } >> >> ; Function Attrs: norecurse uwtable >> define void @_Z2opPi16add_zip_iteratorS0_(i32* nocapture, i32*, i32* nocapture readonly, i32*, i32* nocapture readnone) local_unnamed_addr #0 { >> ; This one isn't! >> %6 = ptrtoint i32* %1 to i64 >> %7 = ptrtoint i32* %3 to i64 >> %8 = sub i64 %6, %7 >> %9 = icmp sgt i64 %8, 0 >> br i1 %9, label %10, label %28 >> >> ; <label>:10: ; preds = %5 >> %11 = lshr exact i64 %8, 2 >> br label %12 >> >> ; <label>:12: ; preds = %12, %10 >> %13 = phi i64 [ %25, %12 ], [ %11, %10 ] >> %14 = phi i32* [ %24, %12 ], [ %0, %10 ] >> %15 = phi i32* [ %23, %12 ], [ %2, %10 ] >> %16 = phi i64 [ %22, %12 ], [ %6, %10 ] >> %17 = inttoptr i64 %16 to i32* >> %18 = load i32, i32* %17, align 4, !tbaa !1 >> %19 = load i32, i32* %15, align 4, !tbaa !1 >> %20 = add nsw i32 %19, %18 >> store i32 %20, i32* %14, align 4, !tbaa !1 >> %21 = getelementptr inbounds i32, i32* %17, i64 1 >> %22 = ptrtoint i32* %21 to i64 >> %23 = getelementptr inbounds i32, i32* %15, i64 1 >> %24 = getelementptr inbounds i32, i32* %14, i64 1 >> %25 = add nsw i64 %13, -1 >> %26 = icmp sgt i64 %13, 1 >> br i1 %26, label %12, label %27 >> >> ; <label>:27: ; preds = %12 >> br label %28 >> >> ; <label>:28: ; preds = %27, %5 >> ret void >> } > If we compile only the "op" function while activation the debug mode, > here is the output: > >> $ clang -O2 -S -emit-llvm op_zip_iterator.cpp -std=c++11 -o - -fno-vectorize |~/dev/epona-llvm/build_debug_shared/bin/opt -debug -debug-only loop-vectorize -O2 -S >> >> LV: Checking a loop in "_Z2opPi16add_zip_iteratorS0_" from <stdin> >> LV: Loop hints: force=? width=0 unroll=0 >> LV: Found a loop: >> LV: Found an induction variable. >> LV: Found an induction variable. >> LV: Found an induction variable. >> LV: Found an unidentified PHI. %16 = phi i64 [ %22, %12 ], [ %6, %10 ] >> LV: Can't vectorize the instructions or CFG >> LV: Not vectorizing: Cannot prove legality. >> [...] > The issue seems to be that the phi node "%16" can't be deduced as an > induction variable. If we take a closer look, the cause seems to be in > ScalarEvolution, in the createSCEV function > (http://llvm.org/docs/doxygen/html/ScalarEvolution_8cpp_source.html#l04770) > : > >> // It's tempting to handle inttoptr and ptrtoint as no-ops, however this can >> // lead to pointer expressions which cannot safely be expanded to GEPs, >> // because ScalarEvolution doesn't respect the GEP aliasing rules when >> // simplifying integer expressions. > Indeed, SCEV does not (legitimately) consider inttoptr/ptrtoint as > no-op, and does not handle them. The thing is that, in our case, the GEP > in %23 is thus not analyzed by SCEV, and the PHI %16 is thus not > considered as an induction variable. > > To confirm this hypothesis, I created a small out-of-tree pass > (https://github.com/aguinet/llvm-intptrcleanup) which registers before > loop vectorization and does the following: > > * first, it search for phi nodes who have those properties: > - every incoming value of the phi node is a ptrtoint instruction. The > original pointer type of every ptrtoint instruction must be the same type T. > - every user of this PHI node is an inttoptr instruction of the > previous type T > * for each of these PHI nodes, it creates a new PHI node which takes the > original pointers as incoming values, and replace the uses of the > inttoptr instructions that uses the original PHI node by the new one > * it then removes the previous inttoptr instructions and the original > PHI node > > The way I understand inttoptr and ptrtoint, this transformation should > be valid (but I might have missed something!). Please note that this is > a quick'n'dirty pass, which hasn't been heavily tested. Using this pass, > the previous example is now vectorized correctly by the loop vectorizer. > This can be seen by looking at the output of: > >> $ clang -Xclang -load -Xclang IntToPtrCleanup.so -O2 ./example/op_zip_operator.cpp -S -emit-llvm -o - -std=c++11 > The question that remains to me is how this should be correctly fixed: > > 1) Making SCEV support these no-op (in this case) inttoptr/ptrtoint > conversions > 2) insert the above transformation at some point in the optimization > pipeline > 3) clean the pass(es?) that somehow generated this case. > > I have to admit I'm not really sure which options is the best. 3) seems > to be the way to go but might require some tedious work, and does not > prevent the issue to come again in the future. 2) seems to be a quick > patch that could be inserted in some "canonicalization" pass, let it be > a valid transformation in the first place. I don't know SCEV enough to > judge of the difficulty/faisability of 1). > > This mail is thus to discuss this issue and how to fix this properly :) > > Thanks everyone :) > > > > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev-- Hal Finkel Lead, Compiler Technology and Programming Languages Leadership Computing Facility Argonne National Laboratory -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170619/be284b2b/attachment.html>
Adrien Guinet via llvm-dev
2017-Jun-20 18:41 UTC
[llvm-dev] LoopVectorize fails to vectorize loops with induction variables with PtrToInt/IntToPtr conversions
On 06/20/2017 03:26 AM, Hal Finkel wrote:> Hi, Adrien,Hello Hal! Thanks for your answer!> Thanks for reporting this. I recommend that you file a bug report at > https://bugs.llvm.org/Will do!> Whenever I see reports of missed optimization opportunities in the face > of ptrtoint/inttoptr, my first question is: why are these instructions > present in the first place? At the IR level, use of inttoptr is highly > discouraged. Our aliasing analysis, for example, does not look through > them, and so you'll generally see a lot of missed optimizations when > they're around. > In this case, inttoptr seems to be introduced by SROA. SROA should not > be introducing inttoptr, but rather should be using GEPs on i8* (at > least), to avoid introducing pointers that our AA can't analyze.It looks so indeed!> Beyond that, if we need to handle inttoptr/ptrtoint in SCEV, then maybe > there's a way to make it smarter about the expressions with which it can > deal.One of the solution I was thinking of was making SCEV "understand" inttoptr(ptrtoint()), as this is clearly a no-op (and what is happening here), but it seems not that straightforward to do (hence this discussion to discuss what should be done :)).> I'm actually not sure to what "aliasing rules" the comment you > quote below is referring. I can certainly understand not being able to > place "inbounds" on some generated GEPs, but otherwise this seems > non-obvious to me (i.e. either the expander can identify a base pointer > from which to generate the GEP, or it can't, in which case it needs to > generate a inttoptr).The way I understand it is that, for a pointer to a fixed-size array P, by using I=ptrtoint(P)/arithmetic operations(I)/inttoptr(I) we can potentially end-up with a pointer which is not in the original values that could be computed from the original array (like an out-of-bound pointer), and thus not "reconstruct" a GEP from it. That's why SCEV wouldn't consider them. IIRC, I tried to add ptrtoint/inttoptr as "no-op" in SCEV, but then the cost-model seemed to be broken (a very high value was given at some point), and thus vectorization still didn't happen. I was also thinking: even if SROA is fixed, nothing prevents another pass from introducing such constructions in the future, so maybe the solution is to have a pass (like the one I wrote) that try and cleanup before any usage of SCEV? (or done by InstCombine for instance, as this basically removes the inttoptr(ptrtoint()) combination). Or having the "inttoptr(ptrtoint())" composition understood by SCEV as a no-op might also be a cleaner "long-term" solution... What do you think?
Reasonably Related Threads
- LoopVectorize fails to vectorize loops with induction variables with PtrToInt/IntToPtr conversions
- Redundant ptrtoint/inttoptr instructions
- Redundant ptrtoint/inttoptr instructions
- [LLVMdev] GEP vs IntToPtr/PtrToInt
- [LLVMdev] Canonicalization of ptrtoint/inttoptr and getelementptr