summaryrefslogtreecommitdiffstats
path: root/llvm/lib/Transforms/InstCombine/InstCombineSelect.cpp
Commit message (Collapse)AuthorAgeFilesLines
* add optional param to copy metadata when creating selects; NFCSanjay Patel2016-11-261-7/+3
| | | | | | | | | | | There are other spots where we can use this; we're currently dropping metadata in some places, and there are proposed changes where we will want to propagate metadata. IRBuilder's CreateSelect() already has a parameter like this, so this change makes the regular 'Create' API line up with that. llvm-svn: 287976
* [InstCombine] canonicalize min/max constant to select's false valueSanjay Patel2016-11-211-0/+42
| | | | | | | | | | | | | | | | | | | | This is a first step towards canonicalization and improved folding/codegen for integer min/max as discussed here: http://lists.llvm.org/pipermail/llvm-dev/2016-November/106868.html Here, we're just matching the simplest min/max patterns and adjusting the icmp predicate while swapping the select operands. I've included FIXME tests in test/Transforms/InstCombine/select_meta.ll so it's easier to see how this might be extended (corresponds to the TODO comment in the code). That's also why I'm using matchSelectPattern() rather than a simpler check; once the backend is patched, we can just remove some of the restrictions to allow the obfuscated min/max patterns in the FIXME tests to be matched. Differential Revision: https://reviews.llvm.org/D26525 llvm-svn: 287585
* [InstCombine] use dyn_cast rather isa+cast; NFCSanjay Patel2016-11-111-2/+2
| | | | | | Follow-up to r286664 cleanup as suggested by Eli. Thanks! llvm-svn: 286671
* [InstCombine] clean up foldSelectOpOp(); NFCSanjay Patel2016-11-111-10/+4
| | | | llvm-svn: 286664
* [InstCombine] fix profitability equation for max-of-nots transformSanjay Patel2016-11-091-7/+6
| | | | | | | | | | As the test change shows, we can increase the critical path by adding a 'not' instruction, so make sure that we're actually removing an instruction if we do this transform. This transform could also cause us to miss folds of min/max pairs. llvm-svn: 286315
* [InstCombine] reduce indentation; NFCSanjay Patel2016-11-081-23/+20
| | | | llvm-svn: 286314
* [InstCombine] allow splat vector folds in adjustMinMax() (retry r285732)Sanjay Patel2016-11-071-14/+12
| | | | | | | | This was reverted at r285866 because there was a crash handling a scalar select of vectors. I added a check for that pattern and a test case based on the example provided in the post-commit thread for r285732. llvm-svn: 286113
* Revert "[InstCombine] allow splat vector folds in adjustMinMax()"Greg Bedwell2016-11-021-10/+14
| | | | | | | | | | | | | | | | | This reverts commit r285732. This change introduced a new assertion failure in the following testcase at -O2: typedef short __v8hi __attribute__((__vector_size__(16))); __v8hi foo(__v8hi &V1, __v8hi &V2, unsigned mask) { __v8hi Result = V1; if (mask & 0x80) Result[0] = V2[0]; return Result; } llvm-svn: 285866
* [InstCombine] allow splat vector folds in adjustMinMax()Sanjay Patel2016-11-011-14/+10
| | | | llvm-svn: 285732
* [InstCombine] clean up adjustMinMax(); NFCISanjay Patel2016-11-011-92/+87
| | | | | | | | | 1. Change param names for readability 2. Change pointer param to ref 3. Early exit to reduce indent 4. Change switch to if/else llvm-svn: 285718
* [InstCombine] add helper function for adjustMinMax(); NFCISanjay Patel2016-11-011-6/+19
| | | | | | This is just a cut and paste; clean-up and enhancements to follow. llvm-svn: 285715
* [InstCombine] re-use bitcasted compare operands in selects (PR28001)Sanjay Patel2016-10-291-0/+50
| | | | | | | | | | | These mixed bitcast patterns show up with SSE/AVX intrinsics because we bitcast function parameters to <2 x i64>. The bitcasts obfuscate the expected min/max forms as shown in PR28001: https://llvm.org/bugs/show_bug.cgi?id=28001#c6 Differential Revision: https://reviews.llvm.org/D25943 llvm-svn: 285495
* [InstCombine] fix foldSPFofSPF() to handle vector splatsSanjay Patel2016-10-271-22/+18
| | | | llvm-svn: 285345
* fix formatting; NFCSanjay Patel2016-10-251-13/+13
| | | | llvm-svn: 285078
* [InstCombine] fold select X, (ext X), CSanjay Patel2016-10-071-1/+21
| | | | | | | | | | | | | | | If we're going to canonicalize IR towards select of constants, try harder to create those. Also, don't lose the metadata. This is actually 4 related transforms in one patch: // select X, (sext X), C --> select X, -1, C // select X, (zext X), C --> select X, 1, C // select X, C, (sext X) --> select X, C, 0 // select X, C, (zext X) --> select X, C, 0 Differential Revision: https://reviews.llvm.org/D25126 llvm-svn: 283575
* [InstCombine] allow non-splat folds of select cond (ext X), CSanjay Patel2016-09-301-38/+33
| | | | llvm-svn: 282906
* [InstCombine] fix function names; NFCSanjay Patel2016-09-291-38/+38
| | | | | | | | Also, make foldSelectExtConst() a member of InstCombiner, remove unnecessary parameters from its interface, and group visitSelectInst helpers together in the header file. llvm-svn: 282796
* fix formatting; NFCSanjay Patel2016-09-291-11/+9
| | | | llvm-svn: 282737
* [InstCombine] canonicalize vector select with constant vector condition to ↵Sanjay Patel2016-09-161-0/+39
| | | | | | | | | | | | | | | | | | | | shuffle As discussed on llvm-dev ( http://lists.llvm.org/pipermail/llvm-dev/2016-August/104210.html ): turn a vector select with constant condition operand into a shuffle as a canonicalization step. Shuffles may be easier to reason about in conjunction with other shuffles and insert/extract. Possible known (minor?) regressions from this change are filed as: https://llvm.org/bugs/show_bug.cgi?id=28530 https://llvm.org/bugs/show_bug.cgi?id=28531 https://llvm.org/bugs/show_bug.cgi?id=30371 If something terrible happens to perf after this commit, feel free to revert until a backend fix is in place. Differential Revision: https://reviews.llvm.org/D24279 llvm-svn: 281787
* fix formatting; NFCSanjay Patel2016-09-061-19/+14
| | | | llvm-svn: 280727
* [Profile] Propagate branch metadata properly in instcombineXinliang David Li2016-08-251-11/+15
| | | | | | Differential Revision: http://reviews.llvm.org/D23590 llvm-svn: 279693
* [InstCombine] try to fold (select C, (sext A), B) into logical opsNicolai Haehnle2016-08-051-0/+56
| | | | | | | | | | | | | | | | | | | | | | Summary: Turn (select C, (sext A), B) into (sext (select C, A, B')) when A is i1 and B is a compatible constant, also for zext instead of sext. This will then be further folded into logical operations. The transformation would be valid for non-i1 types as well, but other parts of InstCombine prefer to have sext from non-i1 as an operand of select. Motivated by the shader compiler frontend in Mesa for AMDGPU, which emits i32 for boolean operations. With this change, the boolean logic is fully recovered. Reviewers: majnemer, spatel, tstellarAMD Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D22747 llvm-svn: 277801
* InstCombine: Replace some never-null pointers with references. NFCJustin Bogner2016-08-051-1/+1
| | | | llvm-svn: 277792
* [InstSimplify][InstCombine] don't crash when folding vector selects of icmpSanjay Patel2016-07-201-1/+4
| | | | | | Differential Revision: https://reviews.llvm.org/D22602 llvm-svn: 276209
* save type in local var; NFCISanjay Patel2016-07-071-10/+11
| | | | llvm-svn: 274760
* [InstCombine] enhance (select X, C1, C2 --> ext X) to handle vectorsSanjay Patel2016-07-061-22/+28
| | | | | | | | | By replacing dyn_cast of ConstantInt with m_Zero/m_One/m_AllOnes, we allow these transforms for splat vectors. Differential Revision: http://reviews.llvm.org/D21899 llvm-svn: 274696
* [InstCombine] use more specific pattern matchers; NFCISanjay Patel2016-07-061-12/+10
| | | | | | | | Follow-up from r274465: we don't need to capture the value in these cases, so just match the constant that we're looking for. m_One/m_Zero work with vector splats as well as scalars. llvm-svn: 274670
* [InstCombine] enable vector select of bools -> logic foldsSanjay Patel2016-07-031-5/+8
| | | | llvm-svn: 274465
* fix formatting; NFCSanjay Patel2016-07-031-6/+6
| | | | llvm-svn: 274463
* [InstCombine] allow more than one use for vector bitcast folding with selectsSanjay Patel2016-06-171-13/+35
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The motivating example for this transform is similar to D20774 where bitcasts interfere with a single cmp/select sequence, but in this case we have 2 uses of each bitcast to produce min and max ops: define void @minmax_bc_store(<4 x float> %a, <4 x float> %b, <4 x float>* %ptr1, <4 x float>* %ptr2) { %cmp = fcmp olt <4 x float> %a, %b %bc1 = bitcast <4 x float> %a to <4 x i32> %bc2 = bitcast <4 x float> %b to <4 x i32> %sel1 = select <4 x i1> %cmp, <4 x i32> %bc1, <4 x i32> %bc2 %sel2 = select <4 x i1> %cmp, <4 x i32> %bc2, <4 x i32> %bc1 %bc3 = bitcast <4 x float>* %ptr1 to <4 x i32>* store <4 x i32> %sel1, <4 x i32>* %bc3 %bc4 = bitcast <4 x float>* %ptr2 to <4 x i32>* store <4 x i32> %sel2, <4 x i32>* %bc4 ret void } With this patch, we move the selects up to use the input args which allows getting rid of all of the bitcasts: define void @minmax_bc_store(<4 x float> %a, <4 x float> %b, <4 x float>* %ptr1, <4 x float>* %ptr2) { %cmp = fcmp olt <4 x float> %a, %b %sel1.v = select <4 x i1> %cmp, <4 x float> %a, <4 x float> %b %sel2.v = select <4 x i1> %cmp, <4 x float> %b, <4 x float> %a store <4 x float> %sel1.v, <4 x float>* %ptr1, align 16 store <4 x float> %sel2.v, <4 x float>* %ptr2, align 16 ret void } The asm for x86 SSE then improves from: movaps %xmm0, %xmm2 cmpltps %xmm1, %xmm2 movaps %xmm2, %xmm3 andnps %xmm1, %xmm3 movaps %xmm2, %xmm4 andnps %xmm0, %xmm4 andps %xmm2, %xmm0 orps %xmm3, %xmm0 andps %xmm1, %xmm2 orps %xmm4, %xmm2 movaps %xmm0, (%rdi) movaps %xmm2, (%rsi) To: movaps %xmm0, %xmm2 minps %xmm1, %xmm2 maxps %xmm0, %xmm1 movaps %xmm2, (%rdi) movaps %xmm1, (%rsi) The TODO comments show that we're limiting this transform only to vectors and only to bitcasts because we need to improve other transforms or risk creating worse codegen. Differential Revision: http://reviews.llvm.org/D21190 llvm-svn: 273011
* [InstCombine] move fold of select of add/sub to helper function; NFCISanjay Patel2016-06-081-61/+75
| | | | llvm-svn: 272199
* [InstCombine] fix outdated comment, simplify logic; NFCISanjay Patel2016-06-081-16/+13
| | | | llvm-svn: 272196
* [InstCombine] reduce indent; NFCSanjay Patel2016-06-081-63/+64
| | | | llvm-svn: 272193
* [InstCombine] use copyIRFlags() ; NFCISanjay Patel2016-06-081-12/+2
| | | | llvm-svn: 272191
* Avoid copies of std::strings and APInt/APFloats where we only read from itBenjamin Kramer2016-06-081-2/+2
| | | | | | | | As suggested by clang-tidy's performance-unnecessary-copy-initialization. This can easily hit lifetime issues, so I audited every change and ran the tests under asan, which came back clean. llvm-svn: 272126
* [InstCombine] Determine the result of a select based on a dominating condition.Chad Rosier2016-04-291-0/+18
| | | | | | Differential Revision: http://reviews.llvm.org/D19550 llvm-svn: 268104
* [InstCombine] Fix miscompile in FoldSPFofSPFDavid Majnemer2016-04-081-0/+3
| | | | | | | | | | | We had a select of a cast of a select but attempted to replace the outer select with the inner select dispite their incompatible types. Patch by Anton Korobeynikov! This fixes PR27236. llvm-svn: 265805
* Minor code cleanup. NFC.Junmo Park2016-03-231-1/+1
| | | | llvm-svn: 264124
* function names start with a lowercase letter; NFCSanjay Patel2016-02-011-21/+21
| | | | llvm-svn: 259425
* function names start with a lower case letter ; NFCSanjay Patel2016-01-121-3/+3
| | | | llvm-svn: 257496
* [InstCombine] Call getCmpPredicateForMinMax only with a valid SPFSanjoy Das2015-12-051-1/+5
| | | | | | | | | | | | | | | | Summary: There are `SelectPatternFlavor`s that don't represent min or max idioms, and we should not be passing those to `getCmpPredicateForMinMax`. Fixes PR25745. Reviewers: majnemer Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D15249 llvm-svn: 254869
* don't repeat function names in comments; NFCSanjay Patel2015-09-091-19/+16
| | | | llvm-svn: 247154
* Add support for floating-point minnum and maxnumJames Molloy2015-08-111-8/+29
| | | | | | | | | | | | | | | | | The select pattern recognition in ValueTracking (as used by InstCombine and SelectionDAGBuilder) only knew about integer patterns. This teaches it about minimum and maximum operations. matchSelectPattern() has been extended to return a struct containing the existing Flavor and a new enum defining the pattern's behavior when given one NaN operand. C minnum() is defined to return the non-NaN operand in this case, but the idiomatic C "a < b ? a : b" would return the NaN operand. ARM and AArch64 at least have different instructions for these different cases. llvm-svn: 244580
* [InstCombine, InstSimplify] Move xforms from Combine to SimplifyDavid Majnemer2015-06-061-115/+2
| | | | | | | | There were several SelectInst combines that always returned an existing instruction instead of modifying an old one or creating a new one. These are prime candidates for moving to InstSimplify. llvm-svn: 239229
* [InstCombine] Don't miscompile select to poisonDavid Majnemer2015-06-061-0/+13
| | | | | | | | | | | | | | | | If we have (select a, b, c), it is sometimes valid to simplify this to a single select operand. However, doing so is only valid if the computation doesn't inject poison into the computation. It might be helpful to consider the following example: (select (icmp ne %i, INT_MAX), (add nsw %i, 1), INT_MIN) The select is equivalent to (add %i, 1) but not (add nsw %i, 1). Self hosting on x86_64 revealed that this occurs very, very rarely so bailing out is hopefully pretty reasonable. llvm-svn: 239215
* Revert "[InstCombine] Rephrase fix to SimplifyWithOpReplaced"Renato Golin2015-06-051-22/+4
| | | | | | | | | This reverts commit r239141. This commit was an attempt to reintroduce a previous patch that broke many self-hosting bots with clang timeouts, but it still has slowdown issues, at least on ARM, increasing the compilation time (stage 2, clang's) by 5x. llvm-svn: 239175
* [InstCombine] Rephrase fix to SimplifyWithOpReplacedDavid Majnemer2015-06-051-4/+22
| | | | | | | | | | | | | | | | | | I don't have the IR which is causing the build bot breakage but I can postulate as to why they are timing out: 1. SimplifyWithOpReplaced was stripping flags from the simplified value. 2. visitSelectInstWithICmp was overriding SimplifyWithOpReplaced because it's simplification wasn't correct. 3. InstCombine would revisit the add instruction and note that it can rederive the flags. 4. By modifying the value, we chose to revisit instructions which reuse the value. One of the instructions is the original select, causing LLVM to never reach fixpoint. Instead, strip the flags only when we are sure we are going to perform the simplification. llvm-svn: 239141
* Revert "[InstCombine] Don't miscompile safe increment idiom"Daniel Jasper2015-06-051-21/+3
| | | | | | | | | This is breaking a lot of build bots and is causing very long-running compiles (infinite loops)? Likely, we shouldn't return nullptr? llvm-svn: 239139
* [InstCombine] Don't miscompile safe increment idiomDavid Majnemer2015-06-041-3/+21
| | | | | | | | | | | We cleverly handle cases where computation done in one argument of a select instruction is suitable for the other operand, thus obviating the need of the select and the comparison. However, the other operand cannot have flags. This fixes PR23757. llvm-svn: 239115
* Reapply r237539 with a fix for the Chromium build.James Molloy2015-05-201-6/+18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Make sure if we're truncating a constant that would then be sign extended that the sign extension of the truncated constant is the same as the original constant. > Canonicalize min/max expressions correctly. > > This patch introduces a canonical form for min/max idioms where one operand > is extended or truncated. This often happens when the other operand is a > constant. For example: > > %1 = icmp slt i32 %a, i32 0 > %2 = sext i32 %a to i64 > %3 = select i1 %1, i64 %2, i64 0 > > Would now be canonicalized into: > > %1 = icmp slt i32 %a, i32 0 > %2 = select i1 %1, i32 %a, i32 0 > %3 = sext i32 %2 to i64 > > This builds upon a patch posted by David Majenemer > (https://www.marc.info/?l=llvm-commits&m=143008038714141&w=2). That pass > passively stopped instcombine from ruining canonical patterns. This > patch additionally actively makes instcombine canonicalize too. > > Canonicalization of expressions involving a change in type from int->fp > or fp->int are not yet implemented. llvm-svn: 237821
OpenPOWER on IntegriCloud