summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* [MIPS GlobalISel] Select float constantsPetar Avramovic2019-03-287-4/+275
| | | | | | | | Select 32 and 64 bit float constants for MIPS32. Differential Revision: https://reviews.llvm.org/D59933 llvm-svn: 357183
* gn build: Add some build files for clangdNico Weber2019-03-287-2/+181
| | | | | | | | | | | | | | Enough to build the clangd binaries, but this is still missing build files for: - fuzzer - indexer - index/dex/dexp - benchmarks - xpc Differential Revision: https://reviews.llvm.org/D59899 llvm-svn: 357182
* [libc++] Remove unnecessary <iostream> #includes in testsLouis Dionne2019-03-2825-46/+4
| | | | | | | Some tests #include <iostream> but they don't use anything from the header. Those are probably artifacts of when the tests were developped. llvm-svn: 357181
* Add "git llvm revert" and "git llvm svn-lookup" subcommandsJordan Rupprecht2019-03-281-3/+117
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Summary: The current git-svnrevert script only works with git-svn repos (e.g. using "git svn find-rev" to find the commit to revert). This adds a similar implementation that works with the llvm git command handler. Usage: ``` // Revert by svn id $ git llvm revert r123456 // See what commands would be run instead of actually reverting $ git llvm revert -n r123456 <full git revert + git commit commands> // Git commit hash also fine $ git llvm revert abc123456 // For convenience, the git->svn method can be used directly: $ git llvm svn-lookup abc123456 r123456 // Push revert upstream (drop the -n when ready) $ git llvm push -n ``` Regardless of how the command is invoked (with a svn revision or git hash), the message is: ``` Revert [LibFoo] Change Foo implementation This reverts r123456 (git commit abc123) ``` Reviewers: jyknight, mehdi_amini, jlebar Reviewed By: jlebar Subscribers: llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D59837 llvm-svn: 357180
* [DAG] Fix Lifetime Node ID hashing.Nirav Dave2019-03-282-2/+10
| | | | llvm-svn: 357179
* [DAGCombiner] fold sext into negationSanjay Patel2019-03-284-49/+49
| | | | | | | | | | | | | | As noted in D59818: %z = zext i8 %x to i32 %neg = sub i32 0, %z %r = sext i32 %neg to i64 => %z2 = zext i8 %x to i64 %r = sub i64 0, %z2 https://rise4fun.com/Alive/KzSR llvm-svn: 357178
* [x86] add vector test for sext of negate; NFCSanjay Patel2019-03-281-0/+86
| | | | llvm-svn: 357177
* [AMDGPU] Clarify Code Object V2/V3 differences in AMDGPUUsageScott Linder2019-03-281-25/+128
| | | | | | | | | | | | Ensure Code Object V2 documentation is complete, but always contains a warning and a link to the equivalent Code Object V3 documentation. Explicitly indicate that any note records present in a code object that are not documented must be considered deprecated and ignored. Differential Revision: https://reviews.llvm.org/D59782 llvm-svn: 357176
* [clangd] Update error message to fix tests after r357173Sam McCall2019-03-281-3/+3
| | | | llvm-svn: 357175
* [Documentation] Proposal to change variable namesMichael Platings2019-03-282-0/+403
| | | | | | Differential Revision: https://reviews.llvm.org/D59251 llvm-svn: 357174
* [clangd] Support UTF-32 (i.e. codepoint) offsets.Sam McCall2019-03-284-54/+171
| | | | | | | | | | | | | | | | Summary: (Changes to UTF-8/UTF-16 here are NFC, moving things around to make the cases more symmetrical) Reviewers: ilya-biryukov Subscribers: ioeric, MaskRay, jkorous, arphaman, kadircet, cfe-commits Tags: #clang Differential Revision: https://reviews.llvm.org/D59927 llvm-svn: 357173
* [x86] avoid cmov in movmsk reductionSanjay Patel2019-03-283-338/+272
| | | | | | | | | | | | | | | | | | | | | | This is probably the least important of our movmsk problems, but I'm starting at the bottom to reduce distractions. We were creating a select_cc which bypasses the select and bitmask codegen optimizations that we have now. If we produce a compare+negate instead, we allow things like neg/sbb carry bit hacks, and in all cases we avoid a cmov. There's no partial register update danger in these sequences because we always produce the zero-register xor ahead of the 'set' if needed. There seems to be a missing fold for sext of a bool bit here: negl %ecx movslq %ecx, %rax ...but that's an independent transform. Differential Revision: https://reviews.llvm.org/D59818 llvm-svn: 357172
* [X86MacroFusion] Handle branch fusion (AMD CPUs).Clement Courbet2019-03-287-81/+161
| | | | | | | | | | | | | | | | | | Summary: This adds a BranchFusion feature to replace the usage of the MacroFusion for AMD CPUs. See D59688 for context. Reviewers: andreadb, lebedev.ri Subscribers: hiraditya, jdoerfert, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D59872 llvm-svn: 357171
* AMDGPU: Make exec mask optimzations more resistant to block splitsMatt Arsenault2019-03-284-31/+200
| | | | | | | Also improve the check for SALU instructions to also ignore implicit_def and other fake instructions. llvm-svn: 357170
* [X86] AMD Piledriver (BdVer2): fine-tune some latenciesRoman Lebedev2019-03-2813-225/+247
| | | | | | | | | | | | | | Based on llvm-exegesis measurements. Now that llvm-exegesis is ~2 magnitudes faster, and is a bit smarter, it is now possible to continue cleanup of the scheduler model. With this, there are no more latency inconsistencies for the opcodes that produce stable measurements, and only a few inconsistencies for unstable measurements (MMX_* opcodes, opcodes that llvm-exegesis measures by chaining - CMP, TEST, BT, SETcc, CVT, MOV, etc.) llvm-svn: 357169
* [NFC] Format InlineFeatureIgnoreList.Clement Courbet2019-03-281-51/+52
| | | | | | To avoid more spurious clang-format changes when adding features (D59872). llvm-svn: 357168
* [Analyzer] Constraint Manager - Calculate Effective Range for DifferencesAdam Balogh2019-03-283-6/+44
| | | | | | | | | | | | | | | | | | | | | Since rL335814, if the constraint manager cannot find a range set for `A - B` (where `A` and `B` are symbols) it looks for a range for `B - A` and returns it negated if it exists. However, if a range set for both `A - B` and `B - A` is stored then it only returns the first one. If we both use `A - B` and `B - A`, these expressions behave as two totally unrelated symbols. This way we miss some useful deductions which may lead to false negatives or false positives. This tiny patch changes this behavior: if the symbolic expression the constraint manager is looking for is a difference `A - B`, it tries to retrieve the range for both `A - B` and `B - A` and if both exists it returns the intersection of range `A - B` and the negated range of `B - A`. This way every time a checker applies new constraints to the symbolic difference or to its negated it always affects both the original difference and its negated. Differential Revision: https://reviews.llvm.org/D55007 llvm-svn: 357167
* - Addressed commentsXing GUO2019-03-281-3/+3
| | | | llvm-svn: 357166
* - Addressed @jhenderson 's commentsXing GUO2019-03-281-12/+11
| | | | | | - Format patch llvm-svn: 357165
* [llvm-readobj] Add new helper function `getSymbolVersionByIndex()`Xing GUO2019-03-281-32/+43
| | | | | | | | | | | | | | | | Summary: When implementing `GNU style` dumper for `.gnu.version` section, we should find symbol version name by `vs_index`. Reviewers: jhenderson, rupprecht Reviewed By: rupprecht Subscribers: arphaman, rupprecht, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D59545 llvm-svn: 357164
* Fix Wdocumentation warning. NFCI.Simon Pilgrim2019-03-281-1/+0
| | | | llvm-svn: 357163
* [PR41247] Fixed parsing of private keyword in C++.Anastasia Stulova2019-03-283-3/+27
| | | | | | | | | Fixed bug in C++ to prevent parsing 'private' as a valid address space qualifier. Differential Revision: https://reviews.llvm.org/D59874 llvm-svn: 357162
* [DAGCombiner] Fold truncate(build_vector(x,y)) -> ↵Simon Pilgrim2019-03-283-22/+29
| | | | | | | | | | | | build_vector(truncate(x),truncate(y)) If scalar truncates are free, attempt to pre-truncate build_vectors source operands. Only attempt to do this before legalization as we often end up with truncations/extensions during build_vector lowering. Differential Revision: https://reviews.llvm.org/D59654 llvm-svn: 357161
* [ELF][ARM] Recommit Redesign of .ARM.exidx handling to use a SyntheticSectionPeter Smith2019-03-2813-176/+498
| | | | | | | | | | | | | | | | | | | | | | Recommit r356666 with fixes for buildbot failure, as well as handling for --emit-relocs, which we decide not to emit any relocation sections as the table is already position independent and an offline tool can deduce the relocations. Instead of creating extra Synthetic .ARM.exidx sections to account for gaps in the table, create a single .ARM.exidx SyntheticSection that can derive the contents of the gaps from a sorted list of the executable InputSections. This has the benefit of moving the ARM specific code for SyntheticSections in SHF_LINK_ORDER processing and the table merging code into the ARM specific SyntheticSection. This also makes it easier to create EXIDX_CANTUNWIND table entries for executable InputSections that don't have an associated .ARM.exidx section. Fixes pr40277 Differential Revision: https://reviews.llvm.org/D59216 llvm-svn: 357160
* [ARM GlobalISel] Run regbankselect test for Thumb. NFCIDiana Picus2019-03-281-1/+2
| | | | | | | This should just work, since ARM mode and Thumb2 mode are at the same level of support now and should map the same to GPR and FPR. llvm-svn: 357159
* [yaml2obj][obj2yaml] - Teach yaml2obj/obj2yaml tools about STB_GNU_UNIQUE ↵George Rimar2019-03-286-20/+77
| | | | | | | | | | | | | symbols. yaml2obj/obj2yaml does not support the symbols with STB_GNU_UNIQUE yet. Currently, obj2yaml fails with llvm_unreachable when met such a symbol. I faced it when investigated the https://bugs.llvm.org/show_bug.cgi?id=41196. Differential revision: https://reviews.llvm.org/D59875 llvm-svn: 357158
* [asan] Add options -asan-detect-invalid-pointer-cmp and ↵Pierre Gousseau2019-03-282-6/+64
| | | | | | | | | | | | | -asan-detect-invalid-pointer-sub options. This is in preparation to a driver patch to add gcc 8's -fsanitize=pointer-compare and -fsanitize=pointer-subtract. Disabled by default as this is still an experimental feature. Reviewed By: morehouse, vitalybuka Differential Revision: https://reviews.llvm.org/D59220 llvm-svn: 357157
* [VPlan] Determine Vector Width programmatically.Florian Hahn2019-03-285-27/+239
| | | | | | | | | | | | | | With this change, the VPlan native path is triggered with the directive: #pragma clang loop vectorize(enable) There is no need to specify the vectorize_width(N) clause. Patch by Francesco Petrogalli <francesco.petrogalli@arm.com> Differential Revision: https://reviews.llvm.org/D57598 llvm-svn: 357156
* [X85][AVX] Add missing vXi16 broadcast fold patternsSimon Pilgrim2019-03-285-35/+37
| | | | | | | | Now that D59484 has landed its easier to add these. Added missing AVX512BW v32i16 equivalents while I was at it. llvm-svn: 357155
* [ARM GlobalISel] Fix G_STORE with s1Diana Picus2019-03-283-7/+62
| | | | | | | G_STORE for 1-bit values uses a STRBi12, which stores the whole byte. Zero out the undefined bits before writing. llvm-svn: 357154
* [ARM GlobalISel] Fix selection of G_SELECTDiana Picus2019-03-284-11/+9
| | | | | | | | | | | G_SELECT uses a 1-bit scalar for the condition, and is currently implemented with a plain CMPri against 0. This means that values such as 0x1110 are interpreted as true, when instead the higher bits should be treated as undefined and therefore ignored. Replace the CMPri with a TSTri against 0x1, which performs an implicit AND, yielding the expected result. llvm-svn: 357153
* [llvm-exegesis] Introduce a 'naive' clustering algorithm (PR40880)Roman Lebedev2019-03-2811-37/+629
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Summary: This is an alternative to D59539. Let's suppose we have measured 4 different opcodes, and got: `0.5`, `1.0`, `1.5`, `2.0`. Let's suppose we are using `-analysis-clustering-epsilon=0.5`. By default now we will start processing the `0.5` point, find that `1.0` is it's neighbor, add them to a new cluster. Then we will notice that `1.5` is a neighbor of `1.0` and add it to that same cluster. Then we will notice that `2.0` is a neighbor of `1.5` and add it to that same cluster. So all these points ended up in the same cluster. This may or may not be a correct implementation of dbscan clustering algorithm. But this is rather horribly broken for the reasons of comparing the clusters with the LLVM sched data. Let's suppose all those opcodes are currently in the same sched cluster. If i specify `-analysis-inconsistency-epsilon=0.5`, then no matter the LLVM values this cluster will **never** match the LLVM values, and thus this cluster will **always** be displayed as inconsistent. The solution is obviously to split off some of these opcodes into different sched cluster. But how do i do that? Out of 4 opcodes displayed in the inconsistency report, which ones are the "bad ones"? Which ones are the most different from the checked-in data? I'd need to go in to the `.yaml` and look it up manually. The trivial solution is to, when creating clusters, don't use the full dbscan algorithm, but instead "pick some unclustered point, pick all unclustered points that are it's neighbor, put them all into a new cluster, repeat". And just so as it happens, we can arrive at that algorithm by not performing the "add neighbors of a neighbor to the cluster" step. But that won't work well once we teach analyze mode to operate in on-1D mode (i.e. on more than a single measurement type at a time), because the clustering would depend on the order of the measurements. Instead, let's just create a single cluster per opcode, and put all the points of that opcode into said cluster. And simultaneously check that every point in that cluster is a neighbor of every other point in the cluster, and if they are not, the cluster (==opcode) is unstable. This is //yet another// step to bring me closer to being able to continue cleanup of bdver2 sched model.. Fixes [[ https://bugs.llvm.org/show_bug.cgi?id=40880 | PR40880 ]]. Reviewers: courbet, gchatelet Reviewed By: courbet Subscribers: tschuett, jdoerfert, RKSimon, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D59820 llvm-svn: 357152
* Fix tests after rC357150Fangrui Song2019-03-282-2/+2
| | | | llvm-svn: 357151
* [Driver] Allow -gsplit-dwarf on ELF OSes other than Linux and FuchsiaFangrui Song2019-03-282-5/+7
| | | | | | | | | | | In gcc, -gsplit-dwarf is handled in gcc/gcc.c as a spec (ASM_FINAL_SPEC): objcopy --extract-dwo + objcopy --strip-dwo. In gcc/opts.c, -gsplit_dwarf has the same semantic of a -g. Except for the availability of the external command 'objcopy', nothing precludes the feature working on other ELF OSes. llvm doesn't use objcopy, so it doesn't have to exclude other OSes. llvm-svn: 357150
* [SelectionDAG] Add 2 tests for selection across basic blocksPiotr Sobczak2019-03-281-0/+39
| | | | | | | | | | | | | | | | | | | | | | | | | | Summary: Add tests for selection across basic block boundary: * one test containing a buffer load, where part of the offset computation is placed in the predecessor of the load * similar test, but containing two buffer loads and shared computations Please note that the behaviour being tested will be updated in a subsequent commit. This commit was extracted from https://reviews.llvm.org/D59535. Reviewers: RKSimon Reviewed By: RKSimon Subscribers: jvesely, nhaehnle, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D59690 llvm-svn: 357149
* SafepointIRVerifier port to new Pass ManagerSerguei Katkov2019-03-281-0/+2
| | | | | | Add missed include. llvm-svn: 357148
* SafepointIRVerifier port to new Pass ManagerSerguei Katkov2019-03-284-0/+23
| | | | | | | | | | | Straightforward port of StatepointIRVerifier pass to new Pass Manager framework. Reviewers: fedor.sergeev, reames Reviewed By: fedor.sergeev Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D59825 llvm-svn: 357147
* Reworked all the utilities/meta tests to use ASSERT_SAME_TYPE instead of ↵Marshall Clow2019-03-2827-153/+120
| | | | | | 'static_assert( is_same<'. Much easier to read. I left two tests alone: is_same.pass.cpp, which should call 'is_same' directly, and common_type.pass.cpp, which Eric is working on. NFC intended llvm-svn: 357146
* [WebAssembly] Rename wasm fixup kindsSam Clegg2019-03-285-18/+13
| | | | | | | | | | | These fixup kinds are not explicitly related to the code section. They are there to signal how to apply the fixup. Also, a couple of other minor wasm cleanups. Differential Revision: https://reviews.llvm.org/D59908 llvm-svn: 357145
* [WebAssembly] Fix typo from rL357143Sam Clegg2019-03-283-5/+5
| | | | llvm-svn: 357144
* [WebAssembly] Improve invalid relocation error messageSam Clegg2019-03-284-14/+21
| | | | | | | | This message now matches the equivalent message in the ELF linker. Differential Revision: https://reviews.llvm.org/D59860 llvm-svn: 357143
* Add reproduction instructions to llvm-objdump's embedded source test.Eric Christopher2019-03-281-0/+10
| | | | llvm-svn: 357142
* Copy the breakpoint site owner's collection so we can dropJim Ingham2019-03-283-2/+29
| | | | | | | | | | | | | | | | | | | | | | | | | the collection lock before we iterate over the owners calling ShouldStop. BreakpointSite::ShouldStop can do a lot of work, and might by chance hit the same breakpoint site again on another thread. So instead of holding the site's owners lock while iterating over them calling ShouldStop, I make a local copy of the list, drop the lock and then iterate over the copy calling BreakpointLocation::ShouldStop. It's actually quite difficult to make this cause problems because usually all the action happens on the private state thread, and the lock is recursive. I have a report where some code hit the ASAN error breakpoint, went to compile the ASAN error gathering expression, in the course of compiling that we went to fetch the ObjC runtime data, but the state of the program was such that the ObjC runtime grubbing function triggered an ASAN error and we were executing that function on another thread. I couldn't figure out a way to reproduce that situation in a test. But this is an NFC change anyway, it just makes the locking strategy more narrowly focused. <rdar://problem/49074093> llvm-svn: 357141
* Inline a trivial function. NFC.Rui Ueyama2019-03-285-12/+10
| | | | | | | I found that hiding this particular actual expression doesn't help readers understand the code. So I remove and inline that function. llvm-svn: 357140
* [libc++] Rename span's as_writeable_bytes to as_writable_bytesLouis Dionne2019-03-284-58/+58
| | | | | | | | | | | | Summary: The Standard says as_writable_bytes. Reviewers: mclow.lists, EricWF Subscribers: christof, jkorous, dexonsmith, libcxx-commits Differential Revision: https://reviews.llvm.org/D59882 llvm-svn: 357139
* Fix typoed variable name.Eric Christopher2019-03-281-4/+4
| | | | | | NFCI. llvm-svn: 357138
* [NewPM] Fix a nasty bug with analysis invalidation in the new PM.Chandler Carruth2019-03-288-184/+298
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The issue here is that we actually allow CGSCC passes to mutate IR (and therefore invalidate analyses) outside of the current SCC. At a minimum, we need to support mutating parent and ancestor SCCs to support the ArgumentPromotion pass which rewrites all calls to a function. However, the analysis invalidation infrastructure is heavily based around not needing to invalidate the same IR-unit at multiple levels. With Loop passes for example, they don't invalidate other Loops. So we need to customize how we handle CGSCC invalidation. Doing this without gratuitously re-running analyses is even harder. I've avoided most of these by using an out-of-band preserved set to accumulate the cross-SCC invalidation, but it still isn't perfect in the case of re-visiting the same SCC repeatedly *but* it coming off the worklist. Unclear how important this use case really is, but I wanted to call it out. Another wrinkle is that in order for this to successfully propagate to function analyses, we have to make sure we have a proxy from the SCC to the Function level. That requires pre-creating the necessary proxy. The motivating test case now works cleanly and is added for ArgumentPromotion. Thanks for the review from Philip and Wei! Differential Revision: https://reviews.llvm.org/D59869 llvm-svn: 357137
* [Process] Reorder declarations and document ReadMemoryFromInferior.Davide Italiano2019-03-281-3/+33
| | | | llvm-svn: 357136
* [ObjectFileMachO] Remove another debugging aid.Davide Italiano2019-03-281-7/+0
| | | | | | Pointed out by Jason. llvm-svn: 357135
* [ObjectFileMachO] Disable memory caching for savecore.Davide Italiano2019-03-281-13/+6
| | | | | | | | | | | | | | | | | Summary: It's not really useful, and largely increases the footprint. <rdar://problem/49293525> Reviewers: jasonmolenda Subscribers: llvm-commits, lldb-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D59896 llvm-svn: 357134
OpenPOWER on IntegriCloud