- Jun 22, 2016
-
-
Jonathan Peyton authored
llvm-svn: 273438
-
Tobias Grosser authored
This cleanup was suggested by Eugene Zelenko <eugene.zelenko@gmail.com> in http://reviews.llvm.org/D21488 and was split out to increase readability. llvm-svn: 273437
-
Tobias Grosser authored
This cleanup was suggested by Eugene Zelenko <eugene.zelenko@gmail.com> in http://reviews.llvm.org/D21488 and was split out to increase readability. llvm-svn: 273436
-
Tobias Grosser authored
Instead of using 0 or NULL use the C++11 nullptr symbol when referencing null pointers. This cleanup was suggested by Eugene Zelenko <eugene.zelenko@gmail.com> in http://reviews.llvm.org/D21488 and was split out to increase readability. llvm-svn: 273435
-
Hans Wennborg authored
It currently only takes 2048 gotos to overflow the FixupDepth bitfield, causing silent miscompilation. Apparently some parser generators run into this (see PR). I don't know that that data structure is terribly size sensitive anyway, and since there's no room to widen the bitfield, let's just use a separate word in EHCatchScope for it. Differential Revision: http://reviews.llvm.org/D21566 llvm-svn: 273434
-
Zhan Jun Liau authored
Summary: Recognize RISBG opportunities where the end result is narrower than the original input - where a truncate separates the shift/and operations. The motivating case is some code in postgres which looks like: srlg %r2, %r0, 11 nilh %r2, 255 Reviewers: uweigand Author: RolandF Differential Revision: http://reviews.llvm.org/D21452 llvm-svn: 273433
-
Krzysztof Parzyszek authored
llvm-svn: 273432
-
Reid Kleckner authored
From a design perspective, complete record type emission should not depend on information from other complete record types. Currently this map is unused, and needlessly accumulates data throughout compilation. llvm-svn: 273431
-
Xinliang David Li authored
Differential Revision: http://reviews.llvm.org/D21596 llvm-svn: 273430
-
Saleem Abdulrasool authored
Improve the previous change by using a local function to reduce the duplication of the object file scanning. NFC. llvm-svn: 273429
-
Artur Pilipenko authored
We no longer have corresponding code in autoupgrade and the vast majority of the tests were fixed long time ago. Fix the remaining few. One of the verifier test cases is marked as XFAIL because it was passing only because the signature was incorrect. llvm-svn: 273428
-
George Rimar authored
llvm-svn: 273427
-
George Rimar authored
Since we do not have plans to support it in closest future, it is better than common script parsing error. llvm-svn: 273426
-
Yaxun Liu authored
This patch uses function metadata to represent reqd_work_group_size, work_group_size_hint and vector_type_hint kernel attributes and kernel argument info. Differential Revision: http://reviews.llvm.org/D20979 llvm-svn: 273425
-
Artur Pilipenko authored
Move Verifier::verifyIntrinsicType to Intrinsics::matchIntrinsicsType. Will be used to accumulate overloaded types of a given intrinsic by the upcoming patch to fix intrinsics names when overloaded types are renamed. Reviewed By: reames Differential Revision: http://reviews.llvm.org/D19372 llvm-svn: 273424
-
George Rimar authored
Patch implements hierarchies for version scripts. This allows to handle script files with dependencies, like next one has: LIBSAMPLE_1.0{ global: a; }; LIBSAMPLE_2.0 { global: b; }LIBSAMPLE_1.0; Differential revision: http://reviews.llvm.org/D21556 llvm-svn: 273423
-
Martin Probst authored
Summary: 'as' is a pseudo operator, so automatic semicolon insertion kicks in and the code fails to part. Reviewers: djasper Subscribers: klimek Differential Revision: http://reviews.llvm.org/D21576 llvm-svn: 273422
-
Sanjay Patel authored
llvm-svn: 273421
-
Simon Pilgrim authored
llvm-svn: 273420
-
Sanjay Patel authored
llvm-svn: 273419
-
George Rimar authored
llvm-svn: 273418
-
George Rimar authored
[llvm-readobj] - Teach llvm-readobj to print dependencies of SHT_GNU_verdef and refactor dumping method. This patch changes single method of llvm-readobj. It teaches SHT_GNU_verdef dumper to print version dependencies, also it removes few fields from output that can be dumped with other keys and slightly refactors code. Testcase was also modified to match the changes. Change is required for testcases of upcoming lld patches. Differential revision: http://reviews.llvm.org/D21552 llvm-svn: 273417
-
Simon Pilgrim authored
llvm-svn: 273404
-
Krzysztof Parzyszek authored
The setCallee function will set the number of fixed arguments based on the size of the argument list. The FixedArgs parameter was often explicitly set to 0, leading to a lack of consistent value for non- vararg functions. Differential Revision: http://reviews.llvm.org/D20376 llvm-svn: 273403
-
Rafael Espindola authored
Found by gcc 6. llvm-svn: 273402
-
Michael Zuckerman authored
Differential Revision: http://reviews.llvm.org/D21373 llvm-svn: 273401
-
Roman Gareev authored
llvm-svn: 273400
-
George Rimar authored
llvm-svn: 273399
-
George Rimar authored
Found that during investigation of FreeBsd scripts. Unclosed quote just crashed lld. llvm-svn: 273398
-
Roman Gareev authored
This is the first patch to apply the BLIS matmul optimization pattern on matmul kernels (http://www.cs.utexas.edu/users/flame/pubs/TOMS-BLIS-Analytical.pdf ). BLIS implements gemm as three nested loops around a macro-kernel, plus two packing routines. The macro-kernel is implemented in terms of two additional loops around a micro-kernel. The micro-kernel is a loop around a rank-1 (i.e., outer product) update. In this change we create the BLIS micro-kernel by applying a combination of tiling and unrolling. In subsequent changes we will add the extraction of the BLIS macro-kernel and implement the packing transformation. Contributed-by:
Roman Gareev <gareevroman@gmail.com> Reviewed-by:
Tobias Grosser <tobias@grosser.es> Differential Revision: http://reviews.llvm.org/D21140 llvm-svn: 273397
-
George Rimar authored
For next version script: VER1{ global: a; }; VER2{ global: a; }; gold would produce warning like: "warning: using 'VER1' as version for 'a' which is also named in version 'VER2' in script." Documentation also says we do not want this duplications (https://people.freebsd.org/~deischen/symver/library_versioning.txt): "Note that you do not want to duplicate symbols in the map file. The .symver directives are all that is required to add compatibility symbols into old versions." This patch restricts such mixing and makes lld to produce error in this case. Differential revision: http://reviews.llvm.org/D21555 llvm-svn: 273396
-
Eric Fiselier authored
llvm-svn: 273395
-
Eric Fiselier authored
llvm-svn: 273394
-
Eric Fiselier authored
llvm-svn: 273393
-
Eric Fiselier authored
llvm-svn: 273392
-
Eric Fiselier authored
llvm-svn: 273391
-
Eric Fiselier authored
llvm-svn: 273390
-
Craig Topper authored
[AVX512] Use a __v8hi vector inside of _mm_setzero_hi to match its name. Probably no real functional change. llvm-svn: 273389
-
Craig Topper authored
[AVX512] Fix _mm_setzero_di to not require avx512vl since its used by the avx512dqintrin.h. Also update the avx512dq test to not enable avx512vl feature so we can ensure correct dependencies. llvm-svn: 273388
-
Craig Topper authored
llvm-svn: 273387
-