[PATCH] D102478: [Matrix] Emit assumption that matrix indices are valid.

fhahn updated this revision to Diff 345703. fhahn added a comment. Undo unrelated changes. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D102478/new/ https://reviews.llvm.org/D102478 Files: clang/lib/CodeGen/CGExpr.cpp clang/lib/CodeGen/CGExprS

[PATCH] D102489: [Clang,Driver] Add -fveclib=Darwin_libsystem_m support.

fhahn added a comment. In D102489#2761195 , @dyung wrote: > The test added in this commit was failing on the PS4 bot > (https://lab.llvm.org/buildbot/#/builders/139/builds/4059). I've reverted the > change in 59b419adc6e608db8d7c31efcc37f34c0b57b7d0 >

[PATCH] D102760: [llvm] Let SmallVector construct from any Iterable

fhahn added inline comments. Comment at: clang/include/clang/Basic/Module.h:98 public: + Module(const Module &) = default; + how is this related? Comment at: llvm/tools/llvm-xray/xray-converter.cpp:161 struct StackIdData { + StackIdData(con

[PATCH] D102494: [Clang, Driver] Default to Darwin_libsystem_m veclib on iOS based targets.

fhahn planned changes to this revision. fhahn added a comment. We are still evaluating whether it is safe to use this as default. I'll mark it as 'changes planned' for now. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D102494/new/ https://reviews.

[PATCH] D101191: [InstCombine] Fully disable select to and/or i1 folding

fhahn added a comment. Herald added a subscriber: foad. Looks like this is causing an infinite loop in instcombine: https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=34661 Reproducer test case: https://oss-fuzz.com/download?testcase_id=6383789942112256 , hangs with `opt -instcombine`, term

[PATCH] D100834: Bug 49739 - [Matrix] Support #pragma clang fp

fhahn added a comment. @effective-light just FYI, if you are interested in improving fast-math flags handling during matrix lowering on the LLVM side, there's https://bugs.llvm.org/show_bug.cgi?id=49738 Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.or

[PATCH] D103163: [Matrix] Skip matrix casts checks for class or struct types in C++.

fhahn created this revision. fhahn added reviewers: rjmccall, anemet, erichkeane. Herald added a subscriber: tschuett. fhahn requested review of this revision. Herald added a project: clang. Herald added a subscriber: cfe-commits. At the moment, the matrix support in CheckCXXCStyleCast (added in D

[PATCH] D103163: [Matrix] Skip matrix casts checks for class or struct types in C++.

fhahn updated this revision to Diff 348106. fhahn added a comment. Fix IR checks, which broke after adding fields to the struct/class used in the tests. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D103163/new/ https://reviews.llvm.org/D103163 Fi

[PATCH] D102478: [Matrix] Emit assumption that matrix indices are valid.

fhahn updated this revision to Diff 348210. fhahn added a comment. Thanks for the comments! Updates: - Only emit assumptions for optimized builds; the assumes are not helpful for unoptimized builds - Use builder::CreateAssumption instead of manually constructing assume calls - Add assert that c

[PATCH] D102478: [Matrix] Emit assumption that matrix indices are valid.

fhahn marked an inline comment as done. fhahn added inline comments. Comment at: llvm/include/llvm/IR/MatrixBuilder.h:242 +auto *Cmp = B.CreateICmpULT(Idx, NumElts); +if (!isa(Cmp)) { + Function *TheFn = rjmccall wrote: > xbolva00 wrote: > > Prefer e

[PATCH] D102478: [Matrix] Emit assumption that matrix indices are valid.

fhahn updated this revision to Diff 348212. fhahn marked an inline comment as done. fhahn added a comment. Fix clang-tidy warnings. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D102478/new/ https://reviews.llvm.org/D102478 Files: clang/lib/CodeG

[PATCH] D102478: [Matrix] Emit assumption that matrix indices are valid.

fhahn updated this revision to Diff 348357. fhahn added a comment. Fix failing clang/test/CodeGenObjC/matrix-type-operators.m. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D102478/new/ https://reviews.llvm.org/D102478 Files: clang/lib/CodeGen/CG

[PATCH] D103163: [Matrix] Skip matrix casts checks for class or struct types in C++.

fhahn updated this revision to Diff 348360. fhahn added a comment. I discussed the problem offline with John and he suggested to handle matrix casts in TryStaticCast. This allows us the handle both static_cast and C-style casts with the same code. As a consequence, the default error messages for

[PATCH] D103163: [Matrix] Skip matrix casts checks for class or struct types in C++.

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG5bccdde070d2: [Matrix] Move C++ matrix cast checks to TryStaticCast. (authored by fhahn). Changed prior to commit: https://reviews.llvm.org/D10316

[PATCH] D92808: [ObjC][ARC] Annotate calls with attributes instead of emitting retainRV or claimRV calls in the IR

fhahn added inline comments. Comment at: llvm/include/llvm/Analysis/ObjCARCRVAttr.h:28 + +static inline const char *getRVAttrValStr(bool Retain) { + return Retain ? "retain" : "claim"; Is there a place the attributes could be documented? Comme

[PATCH] D92808: [ObjC][ARC] Annotate calls with attributes instead of emitting retainRV or claimRV calls in the IR

fhahn added inline comments. Comment at: llvm/lib/IR/Instruction.cpp:580 +if (auto *CB = dyn_cast(this)) + return objcarc::hasRetainRVOrClaimRVAttr(CB); +return false; ahatanak wrote: > fhahn wrote: > > rjmccall wrote: > > > nikic wrote: > > > > This

[PATCH] D92808: [ObjC][ARC] Use operand bundle 'clang.arc.rv' instead of explicitly emitting retainRV or claimRV calls in the IR

fhahn added a comment. Thanks for the update! The approach with operand bundles and the no-op uses looks cleaner than the original version. The special handling in the inliner seems a bit unfortunate, but I guess there's no way around that? (as a side note, it might be easier to submit if this

[PATCH] D92808: [ObjC][ARC] Use operand bundle 'clang.arc.rv' instead of explicitly emitting retainRV or claimRV calls in the IR

fhahn added a comment. Thanks for the updates! The LLVM side of things now looks good to me. Could you also add a description of the new bundle to LangRef https://llvm.org/docs/LangRef.html#operand-bundles? In D92808#2535593 , @ahatanak wrote: > We coul

[PATCH] D92808: [ObjC][ARC] Use operand bundle 'clang.arc.rv' instead of explicitly emitting retainRV or claimRV calls in the IR

fhahn accepted this revision. fhahn added a comment. This revision is now accepted and ready to land. Core LLVM parts LGTM, thanks! I'm going to mark this as accepted, given that the other parts have been reviewed earlier already Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION

[PATCH] D94367: [Clang][Driver] Add -ffinite-loops flags

fhahn added a comment. Herald added a reviewer: jansvoboda11. @atmnpatel reverse ping. Just curious if you think you'll have time to follow up on this patch soonish? Otherwise I'm probably going to add at least the `-fno-finite-loops` option, because we have users which code breaks due to the `

[PATCH] D94367: [Clang][Driver] Add -ffinite-loops flags

fhahn added a comment. In D94367#2542809 , @atmnpatel wrote: > Wait actually, we're gonna need D94366 > anyways, I'll get address @xbolva00's comments by eod. FWIW I think we should sort this one (D94367

[PATCH] D108832: [Builtins] Support ext_vector_type args for __builtin_fminf.

fhahn planned changes to this revision. fhahn added a comment. In D108832#2969538 , @erichkeane wrote: > Codewise I'm ok, but I don't feel comfortable making the direction decision. > I'd like to see some level of 'state of things' on the vector types

[PATCH] D102478: [Matrix] Emit assumption that matrix indices are valid.

fhahn added a comment. All required changes to make use of this have recently landed or are ready to land. So I am going to commit this momentarily. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D102478/new/ https://reviews.llvm.org/D102478 __

[PATCH] D102478: [Matrix] Emit assumption that matrix indices are valid.

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGea21d688dc0a: [Matrix] Emit assumption that matrix indices are valid. (authored by fhahn). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST

[PATCH] D96418: [clang] Refactor mustprogress handling, add it to all loops in c++11+.

fhahn created this revision. fhahn added reviewers: jdoerfert, rjmccall, xbolva00, atmnpatel, aaron.ballman, rsmith. fhahn requested review of this revision. Herald added a project: clang. Currently Clang does not add mustprogress to inifinite loops with a known constant condition, matching C11 b

[PATCH] D96419: [clang] Add -ffinite-loops & -fno-finite-loops options.

fhahn created this revision. fhahn added reviewers: jdoerfert, rjmccall, xbolva00, atmnpatel, aaron.ballman, rsmith. Herald added subscribers: dexonsmith, dang. Herald added a reviewer: jansvoboda11. fhahn requested review of this revision. Herald added a project: clang. Herald added a subscriber:

[PATCH] D96419: [clang] Add -ffinite-loops & -fno-finite-loops options.

fhahn updated this revision to Diff 322679. fhahn added a comment. Add description for options. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D96419/new/ https://reviews.llvm.org/D96419 Files: clang/include/clang/Basic/CodeGenOptions.def clang/

[PATCH] D96419: [clang] Add -ffinite-loops & -fno-finite-loops options.

fhahn updated this revision to Diff 322683. fhahn added a comment. Remove accidentally copied code. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D96419/new/ https://reviews.llvm.org/D96419 Files: clang/include/clang/Basic/CodeGenOptions.def cl

[PATCH] D96419: [clang] Add -ffinite-loops & -fno-finite-loops options.

fhahn added inline comments. Comment at: clang/lib/Frontend/CompilerInvocation.cpp:1923 + if (Arg *A = Args.getLastArg(OPT_save_temps_EQ)) +Opts.SaveTempsFilePrefix = xbolva00 wrote: > Not needed Removed, thanks! Repository: rG LLVM Github Monorepo CH

[PATCH] D96418: [clang] Refactor mustprogress handling, add it to all loops in c++11+.

fhahn updated this revision to Diff 322791. fhahn added a comment. Herald added a subscriber: jfb. Document reasoning. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D96418/new/ https://reviews.llvm.org/D96418 Files: clang/lib/CodeGen/CGStmt.cpp

[PATCH] D96418: [clang] Refactor mustprogress handling, add it to all loops in c++11+.

fhahn added inline comments. Comment at: clang/lib/CodeGen/CodeGenFunction.h:523 return getLangOpts().C11 || getLangOpts().C17 || getLangOpts().C2x; } jdoerfert wrote: > Can you modify the documentation to talk about what loops must make progress, > th

[PATCH] D96419: [clang] Add -ffinite-loops & -fno-finite-loops options.

fhahn updated this revision to Diff 322803. fhahn added a comment. adjust naming of enum members as suggested, thanks! I think it cannot be an enum class unfortunately because the CodeGenOption machinery seems to use regular enums/unsigneds. Repository: rG LLVM Github Monorepo CHANGES SINCE

[PATCH] D96419: [clang] Add -ffinite-loops & -fno-finite-loops options.

fhahn added inline comments. Comment at: clang/include/clang/Basic/CodeGenOptions.h:147 +No = 2, // No loop is assumed to be finite. + }; + jdoerfert wrote: > Can we have different names? > > `FiniteLoopsKind::None` sounds more like what ``FiniteLoopsKind

[PATCH] D96419: [clang] Add -ffinite-loops & -fno-finite-loops options.

fhahn added a comment. In D96419#213 , @xbolva00 wrote: > Maybe we want these options in llvm 12 as well? > > (+ Release note) That's a good idea. The patch currently depends on D96418 . Not sure if we should pull tha

[PATCH] D96418: [clang] Refactor mustprogress handling, add it to all loops in c++11+.

fhahn added inline comments. Comment at: clang/lib/CodeGen/CodeGenFunction.h:523 return getLangOpts().C11 || getLangOpts().C17 || getLangOpts().C2x; } aaron.ballman wrote: > jdoerfert wrote: > > fhahn wrote: > > > jdoerfert wrote: > > > > Can you modify

[PATCH] D92808: [ObjC][ARC] Use operand bundle 'clang.arc.rv' instead of explicitly emitting retainRV or claimRV calls in the IR

fhahn added a comment. In D92808#2555868 , @ahatanak wrote: > In D92808#2555737 , @dexonsmith > wrote: > >> In D92808#2552354 , @rjmccall wrote: >> >>> The ultimate code-g

[PATCH] D96418: [clang] Refactor mustprogress handling, add it to all loops in c++11+.

fhahn updated this revision to Diff 323003. fhahn added a comment. Add reference to standards. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D96418/new/ https://reviews.llvm.org/D96418 Files: clang/lib/CodeGen/CGStmt.cpp clang/lib/CodeGen/CodeG

[PATCH] D92808: [ObjC][ARC] Use operand bundle 'clang.arc.rv' instead of explicitly emitting retainRV or claimRV calls in the IR

fhahn added a comment. Another thing I noticed that there's verifier support missing. I think we should at least check that only a single `clang.arc.rv` bundle is specified (https://github.com/llvm/llvm-project/blob/main/llvm/lib/IR/Verifier.cpp#L3191). We should probably also enforce that the

[PATCH] D92808: [ObjC][ARC] Use operand bundle 'clang.arc.rv' instead of explicitly emitting retainRV or claimRV calls in the IR

fhahn accepted this revision. fhahn added a comment. This revision is now accepted and ready to land. LGTM as initial step, but it would be good to adjust the name as Duncan suggested before landing. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D92

[PATCH] D96419: [clang] Add -ffinite-loops & -fno-finite-loops options.

fhahn updated this revision to Diff 323373. fhahn added a comment. Rebased the patch to apply wihtout D96418 . Will land soon. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D96419/new/ https://reviews.llvm.org/D964

[PATCH] D96419: [clang] Add -ffinite-loops & -fno-finite-loops options.

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG51bf4c0e6d4c: [clang] Add -ffinite-loops & -fno-finite-loops options. (authored by fhahn). Changed prior to commit: https://reviews.llvm.org/D9641

[PATCH] D96418: [clang] Refactor mustprogress handling, add it to all loops in c++11+.

fhahn updated this revision to Diff 323568. fhahn added a comment. Rebased on top of 51bf4c0e6d4c , which added -ffinite-loops. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D9641

[PATCH] D96761: [clang][cli] Generate -f[no-]finite-loops arguments

fhahn accepted this revision. fhahn added a comment. This revision is now accepted and ready to land. LGTM thanks! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D96761/new/ https://reviews.llvm.org/D96761 ___

[PATCH] D96850: [clang] Add -ffinite-loops & -fno-finite-loops options.

fhahn created this revision. Herald added subscribers: jansvoboda11, dexonsmith, dang. fhahn requested review of this revision. Herald added a project: clang. Herald added a subscriber: cfe-commits. This cherry-picks the following patches on the release branch: 6280bb4cd80e

[PATCH] D96419: [clang] Add -ffinite-loops & -fno-finite-loops options.

fhahn added a comment. In D96419#2556411 , @xbolva00 wrote: > In D96419#2556349 , @fhahn wrote: > >> In D96419#213 , @xbolva00 wrote: >> >>> Maybe we want these options i

[PATCH] D104198: [Matrix] Add documentation for compound assignment and type conversion of matrix types

fhahn added a comment. Thanks for the patch! some comments inline. Comment at: clang/docs/LanguageExtensions.rst:527 +The matrix type extension supports compound assignments for addition, subtraction, and multiplication, provided +their types are consistent. +

[PATCH] D102478: [Matrix] Emit assumption that matrix indices are valid.

fhahn added a comment. ping :) Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D102478/new/ https://reviews.llvm.org/D102478 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cg

[PATCH] D102478: [Matrix] Emit assumption that matrix indices are valid.

fhahn updated this revision to Diff 352108. fhahn added a comment. In D102478#2817768 , @erichkeane wrote: > Just a couple of nits here, basically see how much we can put in the 'cheap > to check' branch. Thanks! I moved the suggested bits inside the t

[PATCH] D104198: [Matrix] Add documentation for compound assignment and type conversion of matrix types

fhahn added inline comments. Comment at: clang/docs/LanguageExtensions.rst:535 + int d = 12; + a = a + c; + return a * d; Is there any reason we use variables for the scalars? If not, it might be good to keep the examples as compact as possible. U

[PATCH] D104082: [CodeGen] Don't create a fake FunctionDecl when generating block/block_byref copy/dispose helper functions

fhahn added a comment. In D104082#2835080 , @zequanwu wrote: > Hi, this caused compiler crash with error "Assertion > `cast(Scope)->describes(&MF->getFunction())' failed." on iOS > platforms. So, I reverted it. > I'm working on a reduced repro. FYI th

[PATCH] D104198: [Matrix] Add documentation for compound assignment and type conversion of matrix types

fhahn added a comment. Thanks for the latest changes! Comment at: clang/docs/LanguageExtensions.rst:526 +The matrix type extension also supports operations between a matrix and a scalar. + I'm not a native speaker, but I am not sure if `between` is the best

[PATCH] D104198: [Matrix] Add documentation for compound assignment and type conversion of matrix types

fhahn added a comment. In D104198#2838416 , @SaurabhJha wrote: > Somehow the builds are failing even though this patch contains no code > changes. There a few reasons as to why the tests may fail unrelated to the patch (flaky tests, infrastructure iss

[PATCH] D104198: [Matrix] Add documentation for compound assignment and type conversion of matrix types

fhahn accepted this revision. fhahn added a comment. This revision is now accepted and ready to land. LGTM,thanks. Comment at: clang/docs/LanguageExtensions.rst:546 + +The matrix type extension supports compound assignments for addition, subtraction, and multiplication between

[PATCH] D104198: [Matrix] Add documentation for compound assignment and type conversion of matrix types

fhahn added a comment. In D104198#2838658 , @SaurabhJha wrote: > Sorry, I committed this without the `Differential Revision: > https://reviews.llvm.org/D104198` line. Is there a way to change the commit > message after it is in main? I could not push a

[PATCH] D107349: [Matrix] Overload stride arg in matrix.columnwise.load/store.

fhahn created this revision. fhahn added reviewers: rjmccall, anemet, thegameg, erichkeane. Herald added subscribers: dexonsmith, jdoerfert, tschuett, hiraditya. fhahn requested review of this revision. Herald added projects: clang, LLVM. This patch adjusts the intrinsics definition of llvm.matrix

[PATCH] D74436: Change clang option -ffp-model=precise to select ffp-contract=on

fhahn added a comment. It looks like this patch is also causing mis-compares in both SPEC2006 and SPEC2017 on AArch64, e.g. in `External/SPEC/CFP2006/453.povray/`. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D74436/new/ https://reviews.llvm.org/D

[PATCH] D107349: [Matrix] Overload stride arg in matrix.columnwise.load/store.

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGa1ef81de35a4: [Matrix] Overload stride arg in matrix.columnwise.load/store. (authored by fhahn). Changed prior to commit: https://reviews.llvm.org

[PATCH] D107349: [Matrix] Overload stride arg in matrix.columnwise.load/store.

fhahn added a comment. In D107349#2922208 , @erichkeane wrote: > This looks fine to me, and seems to fix my problem, thanks! I didn't spot > anything obvious,and proof-read the LangRef and think it is all fine, but am > not really the expert here, so

[PATCH] D102478: [Matrix] Emit assumption that matrix indices are valid.

fhahn updated this revision to Diff 365959. fhahn added a comment. rebased after latest changes Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D102478/new/ https://reviews.llvm.org/D102478 Files: clang/lib/CodeGen/CGExpr.cpp clang/lib/CodeGen/CG

[PATCH] D107349: [Matrix] Overload stride arg in matrix.columnwise.load/store.

fhahn added a comment. In D107349#2941385 , @erichkeane wrote: > In D107349#2941257 , @mehdi_amini > wrote: > >> Revert to unbreak bots (like this one : >> https://lab.llvm.org/buildbot/#/builders/13/builds/109

[PATCH] D94366: [Clang] Emit mustprogress for infinite C++ loops

fhahn requested changes to this revision. fhahn added a comment. This revision now requires changes to proceed. I think this has been addressed by some commits a while ago. Marking as changes requested to clear up review queue. Please feel free to update the patch in case there’s anything I miss

[PATCH] D111986: [Clang] Add elementwise abs builtin.

fhahn updated this revision to Diff 382226. fhahn marked 2 inline comments as done. fhahn added a comment. Rebased after recent changes in D111985 : changes the return type of SemaBuiltinElementwiseMathOneArg to bool and drop the CallResult argument. Also added

[PATCH] D112001: [Clang] Add min/max reduction builtins.

fhahn updated this revision to Diff 382240. fhahn added a comment. Rebased after recent changes in D111985 https://reviews.llvm.org/D111985: changes the return type of SemaBuiltinReduceMath to bool and drop the CallResult argument. Also added tests with const

[PATCH] D112532: [Matrix] Replace some err kinds with err_builtin_invalid_arg_type. (NFC)

fhahn created this revision. fhahn added reviewers: aaron.ballman, erichkeane, rjmccall. Herald added a subscriber: tschuett. fhahn requested review of this revision. Herald added a project: clang. Replace some custom matrix diagnostic kinds with the more generic err_builtin_invalid_arg_type intro

[PATCH] D111985: [Clang] Add elementwise min/max builtins.

fhahn updated this revision to Diff 382300. fhahn marked 3 inline comments as done. fhahn added a comment. Address latest comments, thanks - Added a generic `err_builtin_invalid_arg_type` diagnostic kind, which can also be used for some of the matrix type mismatches (see D112532

[PATCH] D111985: [Clang] Add elementwise min/max builtins.

fhahn marked 3 inline comments as done. fhahn added inline comments. Comment at: clang/include/clang/Basic/DiagnosticSemaKinds.td:8747-8748 +def err_elementwise_math_arg_types_mismatch : Error < + "argument types do not match, %0 != %1">; + aaron.ballman wrote

[PATCH] D111529: Specify Clang vector builtins.

This revision was automatically updated to reflect the committed changes. Closed by commit rG025988ded6b2: Specify Clang vector builtins. (authored by fhahn). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D111529/new/ https://reviews.llvm.org/D111529

[PATCH] D111985: [Clang] Add elementwise min/max builtins.

This revision was automatically updated to reflect the committed changes. fhahn marked 3 inline comments as done. Closed by commit rG1ef25d28c19e: [Clang] Add elementwise min/max builtins. (authored by fhahn). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm

[PATCH] D112532: [Matrix] Replace some err kinds with err_builtin_invalid_arg_type. (NFC)

fhahn updated this revision to Diff 382373. fhahn marked an inline comment as done. fhahn added a comment. Reabsed after landing D111985 . Also turned the custom message parts into a select as suggested. In D112532#3087371

[PATCH] D112532: [Matrix] Replace some err kinds with err_builtin_invalid_arg_type. (NFC)

fhahn added inline comments. Comment at: clang/lib/Sema/SemaChecking.cpp:16654 +Diag(PtrExpr->getBeginLoc(), diag::err_builtin_invalid_arg_type) +<< PtrArgIdx + 1 << "pointer to a valid matrix element type" +<< PtrExpr->getType(); erichkeane w

[PATCH] D111986: [Clang] Add elementwise abs builtin.

fhahn updated this revision to Diff 382376. fhahn marked 2 inline comments as done. fhahn added a comment. Rebased and updated to use `err_builtin_invalid_arg_type`. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D111986/new/ https://reviews.llvm.org

[PATCH] D111986: [Clang] Add elementwise abs builtin.

fhahn added inline comments. Comment at: clang/include/clang/Basic/DiagnosticSemaKinds.td:8754 +def err_elementwise_math_invalid_arg_type_2: Error < + "argument must have a %0 type, but was %1">; + aaron.ballman wrote: > I feel like we must already have a diagno

[PATCH] D112532: [Matrix] Replace some err kinds with err_builtin_invalid_arg_type. (NFC)

fhahn updated this revision to Diff 382413. fhahn added a comment. Add comment to int selecting message as suggested. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D112532/new/ https://reviews.llvm.org/D112532 Files: clang/include/clang/Basic/Dia

[PATCH] D112532: [Matrix] Replace some err kinds with err_builtin_invalid_arg_type. (NFC)

fhahn marked an inline comment as done. fhahn added inline comments. Comment at: clang/lib/Sema/SemaChecking.cpp:16582 +Diag(Matrix->getBeginLoc(), diag::err_builtin_invalid_arg_type) +<< 1 << 1 << Matrix->getType(); return ExprError(); erichkean

[PATCH] D112532: [Matrix] Replace some err kinds with err_builtin_invalid_arg_type. (NFC)

This revision was automatically updated to reflect the committed changes. fhahn marked an inline comment as done. Closed by commit rGd7fbad0dcfc9: [Matrix] Replace some err kinds with err_builtin_invalid_arg_type. (NFC) (authored by fhahn). Repository: rG LLVM Github Monorepo CHANGES SINCE LAS

[PATCH] D111986: [Clang] Add elementwise abs builtin.

fhahn updated this revision to Diff 382443. fhahn added a comment. Apply UsualUnaryConversions to input argument. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D111986/new/ https://reviews.llvm.org/D111986 Files: clang/include/clang/Basic/Builtin

[PATCH] D111986: [Clang] Add elementwise abs builtin.

fhahn marked an inline comment as done. fhahn added inline comments. Comment at: clang/lib/Sema/SemaChecking.cpp:16547 + Expr *A = TheCall->getArg(0); + QualType TyA = A->getType(); + aaron.ballman wrote: > fhahn wrote: > > aaron.ballman wrote: > > > Should thi

[PATCH] D111986: [Clang] Add elementwise abs builtin.

fhahn updated this revision to Diff 382445. fhahn marked an inline comment as done. fhahn added a comment. add comment to arg of Diag message Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D111986/new/ https://reviews.llvm.org/D111986 Files: clang

[PATCH] D111986: [Clang] Add elementwise abs builtin.

This revision was automatically updated to reflect the committed changes. Closed by commit rG01870d51b848: [Clang] Add elementwise abs builtin. (authored by fhahn). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D111986/new/ https://reviews.llvm.org/D

[PATCH] D112001: [Clang] Add min/max reduction builtins.

fhahn updated this revision to Diff 383756. fhahn added a comment. rebase on top of recent changes. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D112001/new/ https://reviews.llvm.org/D112001 Files: clang/include/clang/Basic/Builtins.def clang/

[PATCH] D112001: [Clang] Add min/max reduction builtins.

fhahn updated this revision to Diff 383900. fhahn added a comment. In D112001#3099965 , @aaron.ballman wrote: > Precommit CI looks to be failing. > > > /var/lib/buildkite-agent/builds/llvm-project/clang/test/CodeGen/builtins-reduction-math.c:13:9: >

[PATCH] D112001: [Clang] Add min/max reduction builtins.

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG7999355106fb: [Clang] Add min/max reduction builtins. (authored by fhahn). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https:

[PATCH] D108832: [Builtins] Support ext_vector_type args for __builtin_fminf.

fhahn abandoned this revision. fhahn added a comment. The vector builtin support took a different direction, see D111529 and following Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D108832/new/ https://reviews.ll

[PATCH] D113738: [LTO] Allow passing -Os/-Oz as the optimization level

fhahn added a comment. Could you add tests on the LLVM side for the new options as well? See also D95541 Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D113738/new/ https://reviews.llvm.org/D113738

[PATCH] D113779: [Clang] Add mfp16, mfp16fml and mdotprod flags for ARM target features.

fhahn created this revision. fhahn added reviewers: aemerson, ab, dmgreen, SjoerdMeijer, scanon. Herald added subscribers: dang, kristof.beyls. fhahn requested review of this revision. Herald added a project: clang. This patch adds support for `-m[no]fpf16`, ` `-m[no]fpf16fml` and `-m[no]dotprod`

[PATCH] D92808: [ObjC][ARC] Use operand bundle 'clang.arc.attachedcall' instead of explicitly emitting retainRV or claimRV calls in the IR

fhahn added a comment. In D92808#2600031 , @thakis wrote: >> Thank you. Let me know when you have more information. > > Repro is moving along. > https://bugs.chromium.org/p/chromium/issues/detail?id=1182642#c26 and onward > are getting pretty close. Tha

[PATCH] D97857: [Matrix] Add support for matrix-by-scalar division.

fhahn created this revision. fhahn added reviewers: anemet, rjmccall, rsmith, Bigcheese. Herald added subscribers: dexonsmith, tschuett. fhahn requested review of this revision. Herald added projects: clang, LLVM. Herald added a subscriber: llvm-commits. This patch extends the matrix spec to allow

[PATCH] D92808: [ObjC][ARC] Use operand bundle 'clang.arc.attachedcall' instead of explicitly emitting retainRV or claimRV calls in the IR

fhahn added a comment. In D92808#2600245 , @hans wrote: > Reverted in > https://github.com/llvm/llvm-project/commit/0a5dd067181dac2a8882a139ea3bd19bdea5fa44 > until this can be fixed. Thanks! I pushed a fix for the issue in 75805dce5ff8

[PATCH] D92808: [ObjC][ARC] Use operand bundle 'clang.arc.attachedcall' instead of explicitly emitting retainRV or claimRV calls in the IR

fhahn added a comment. In D92808#2602687 , @hans wrote: > In D92808#2601408 , @fhahn wrote: > >> In D92808#2600245 , @hans wrote: >> >>> Reverted in >>> https://github.com/l

[PATCH] D98055: [ExtVectorType] Support conditional select operator for C++.

fhahn created this revision. fhahn added reviewers: rjmccall, erichkeane, aaron.ballman. fhahn requested review of this revision. Herald added a project: clang. This patch implements the conditional select operator for ext_vector_types in C++. It does so by using the same semantics as for C. D714

[PATCH] D98075: [Matrix] Implement += and -= for MatrixType

fhahn added reviewers: rjmccall, erichkeane. fhahn added a comment. Thanks for the patch! Could you also extend the Sema `matrix-type-operators` tests with checks for `+=` & `-=`? It looks like the handling for multiply is slightly different and already works but is missing test coverage. It wo

[PATCH] D98055: [ExtVectorType] Support conditional select operator for C++.

fhahn updated this revision to Diff 328815. fhahn added a comment. Got rid of the template parameter, greatly simplified the code. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D98055/new/ https://reviews.llvm.org/D98055 Files: clang/include/clan

[PATCH] D98055: [ExtVectorType] Support conditional select operator for C++.

fhahn added inline comments. Comment at: clang/lib/Sema/SemaExprCXX.cpp:5951 const QualType EltTy = - cast(CondTy.getCanonicalType())->getElementType(); + CondTy->isVectorType() + ? cast(CondTy.getCanonicalType())->getElementType() erichkean

[PATCH] D98075: [Matrix] Implement += and -= for MatrixType

fhahn accepted this revision. fhahn added a comment. This revision is now accepted and ready to land. LGTM, thanks! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D98075/new/ https://reviews.llvm.org/D98075 __

[PATCH] D98075: [Matrix] Implement += and -= for MatrixType

This revision was automatically updated to reflect the committed changes. Closed by commit rG63851a701eac: [Matrix] Implement += and -= for MatrixType. (authored by SaurabhJha, committed by fhahn). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D98075/

[PATCH] D98055: [ExtVectorType] Support conditional select operator for C++.

fhahn marked an inline comment as done. fhahn added inline comments. Comment at: clang/lib/Sema/SemaExprCXX.cpp:6172-6174 + if (IsVectorConditional) { +return CheckVectorConditionalTypes(Cond, LHS, RHS, QuestionLoc); + } aaron.ballman wrote: > I'll fix tha

[PATCH] D98055: [ExtVectorType] Support conditional select operator for C++.

This revision was automatically updated to reflect the committed changes. fhahn marked an inline comment as done. Closed by commit rGfc8d3766d721: [ExtVectorType] Support conditional select operator for C++. (authored by fhahn). Changed prior to commit: https://reviews.llvm.org/D98055?vs=328815

[PATCH] D96418: [clang] Refactor mustprogress handling, add it to all loops in c++11+.

fhahn added a comment. ping :) Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D96418/new/ https://reviews.llvm.org/D96418 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-

[PATCH] D97857: [Matrix] Add support for matrix-by-scalar division.

fhahn updated this revision to Diff 329596. fhahn added a comment. rebase & added a matrix_types_scalar_division feature, which allows users to check if the current version of clang is new enough to support it. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.

[PATCH] D97857: [Matrix] Add support for matrix-by-scalar division.

fhahn added inline comments. Comment at: clang/test/CodeGen/matrix-type-operators.c:303 } +// CHECK-LABEL: @divide_double_matrix_scalar_float( everton.constantino wrote: > Shouldn't a test for half floating point be added here as well? Thanks for taking a loo

<    1   2   3   4   5   6   7   8   >