en Copyright 2015 Java http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#dc5ce72a Append new attributes to the end of an AttributeList.Recommit of r335084 after revert in r335516.... instead of prepending it at the beginning (the original behaviorsince implemented in r122535 2010-12-23). This builds up anAttributeList in the the order in which the attributes appear in thesource.The reverse order caused nodes for attributes in the AST (e.g. LoopHint)to be in the reverse order, and therefore printed in the wrong order in-ast-dump. Some TODO comments mention this. The order was explicitlyreversed for enable_if attribute overload resolution and name mangling,which is not necessary anymore with this patch.The change unfortunately has some secondary effect, especially ondiagnostic output. In the simplest cases, the CHECK lines or expecteddiagnostic were changed to the the new output. If the kind oferror/warning changed, the attributes' order was changed instead.This unfortunately causes some 'previous occurrence here' hints to betextually after the main marker. This typically happens when attributesare merged, but are incompatible to each other. Interchanging the roleof the the main and note SourceLocation will also cause the case wheretwo different declaration's attributes (in contrast to multipleattributes of the same declaration) are merged to be reverse. There isno easy fix because sometimes previous attributes are merged into a newdeclaration's attribute list, sometimes new attributes are added to aprevious declaration's attribute list. Since 'previous occurrence here'pointing to locations after the main marker is not rare, I left themarkers as-is; it is only relevant when the attributes are declared inthe same declaration anyway.Differential Revision: https://reviews.llvm.org/D48100llvm-svn: 338800 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Fri, 03 Aug 2018 01:21:16 +0000 Michael Kruse <llvm@meinersbur.de> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#41dd6ced Revert "Append new attributes to the end of an AttributeList."This reverts commit r335084 as requested by David Jones andEric Christopher because of differences of emitted warnings.llvm-svn: 335516 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Mon, 25 Jun 2018 20:06:13 +0000 Michael Kruse <llvm@meinersbur.de> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#ea31f0e4 Append new attributes to the end of an AttributeList.... instead of prepending it at the beginning (the original behaviorsince implemented in r122535 2010-12-23). This builds up anAttributeList in the the order in which the attributes appear in thesource.The reverse order caused nodes for attributes in the AST (e.g. LoopHint)to be in the reverse, and therefore printed in the wrong order by-ast-dump. Some TODO comments mention this. The order was explicitlyreversed for enable_if attribute overload resolution and name mangling,which is not necessary anymore with this patch.The change unfortunately has some secondary effects, especially fordiagnostic output. In the simplest cases, the CHECK lines or expecteddiagnostic were changed to the the new output. If the kind oferror/warning changed, the attribute's order was changed instead.It also causes some 'previous occurrence here' hints to be textuallyafter the main marker. This typically happens when attributes aremerged, but are incompatible. Interchanging the role of the the mainand note SourceLocation will also cause the case where two differentdeclaration's attributes (in contrast to multiple attributes of thesame declaration) are merged to be reversed. There is no easy fixbecause sometimes previous attributes are merged into a newdeclaration's attribute list, sometimes new attributes are added to aprevious declaration's attribute list. Since 'previous occurrence here'pointing to locations after the main marker is not rare, I left themarkers as-is; it is only relevant when the attributes are declared inthe same declaration anyway, which often is on the same line.Differential Revision: https://reviews.llvm.org/D48100llvm-svn: 335084 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Tue, 19 Jun 2018 23:46:52 +0000 Michael Kruse <llvm@meinersbur.de> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#cbecfdfe [OpenMP] Fix trailing space when printing pragmas, by Joel. E. DennySummary:-ast-print prints omp pragmas with a trailing space. While thisbehavior is likely of little concern to most users, surely it'sunintentional, and it's annoying for some source-level work I'mpursuing. This patch focuses on omp pragmas, but it also fixesinit_seg and loop hint pragmas because they share implementation.The testing strategy here is to add usually just one '{{$}}' perrelevant -ast-print test file. This seems to achieve good codecoverage. However, this strategy is probably easy to forget as thetests evolve. That's probably fine as this fix is far from critical.The main goal of the testing is to aid the initial review.This patch also adds a fixme for "#pragma unroll", which prints as"#pragma unroll (enable)", which is invalid syntax.Reviewers: ABataevReviewed By: ABataevSubscribers: guansong, cfe-commitsDifferential Revision: https://reviews.llvm.org/D43204llvm-svn: 325145 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Wed, 14 Feb 2018 17:38:47 +0000 Alexey Bataev <a.bataev@hotmail.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#2de463ec Add loop pragma for Loop DistributionSummary:This is similar to other loop pragmas like 'vectorize'. Currently itonly has state values: distribute(enable) and distribute(disable). Whenone of these is specified the corresponding loop metadata is generated: !{!"llvm.loop.distribute.enable", i1 true/false}As a result, loop distribution will be attempted on the loop even ifLoop Distribution in not enabled globally. Analogously, with 'disable'distribution can be turned off for an individual loop even when the passis otherwise enabled.There are some slight differences compared to the existing loop pragmas.1. There is no 'assume_safety' variant which makes its handling slightlydifferent from 'vectorize'/'interleave'.2. Unlike the existing loop pragmas, it does not have a correspondingnumeric pragma like 'vectorize' -> 'vectorize_width'. So for theconsistency checks in CheckForIncompatibleAttributes we don't need tocheck it against other pragmas. We just need to check for duplicates ofthe same pragma.Reviewers: rsmith, dexonsmith, aaron.ballmanSubscribers: bob.wilson, cfe-commits, hfinkelDifferential Revision: http://reviews.llvm.org/D19403llvm-svn: 272656 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Tue, 14 Jun 2016 12:04:26 +0000 Adam Nemet <anemet@apple.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#c724a83e Allow constant expressions in pragma loop hints.Previously loop hints such as #pragma loop vectorize_width(#) required a constant. This patch allows a constant expression to be used as well. Such as a non-type template parameter or an expression (2 * c + 1).Reviewed by Richard Smithllvm-svn: 219589 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Sun, 12 Oct 2014 20:46:07 +0000 Tyler Nowicki <tyler.nowicki@gmail.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#c888e41c Add support for #pragma nounroll.llvm-svn: 213885 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Thu, 24 Jul 2014 18:09:38 +0000 Mark Heffernan <meheff@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#450c2384 In unroll pragma syntax and loop hint metadata, change "enable" forms to a new form using the string "full".llvm-svn: 213771 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Wed, 23 Jul 2014 17:31:31 +0000 Mark Heffernan <meheff@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#bd26f5ea Add support for '#pragma unroll'.llvm-svn: 213574 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Mon, 21 Jul 2014 18:08:34 +0000 Mark Heffernan <meheff@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#86483b3a Add loop unroll pragma supporthttp://reviews.llvm.org/D4089Patch by Mark Heffernan.llvm-svn: 210667 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Wed, 11 Jun 2014 17:56:26 +0000 Eli Bendersky <eliben@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#4173f6af *Really* fix DOS newlines introduced in r210330r210369 didn't quite catch all of them.llvm-svn: 210593 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Tue, 10 Jun 2014 21:35:24 +0000 Reid Kleckner <reid@kleckner.net> http://172.16.0.5:8080/history/llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp#b06b15aa Adding a new #pragma for the vectorize and interleave optimization hints.Patch thanks to Tyler Nowicki!llvm-svn: 210330 List of files: /llvm-project-15.0.7/clang/test/PCH/pragma-loop.cpp Fri, 06 Jun 2014 12:40:24 +0000 Aaron Ballman <aaron@aaronballman.com>