Commit Graph

326681 Commits

Author SHA1 Message Date
Clement Courbet 664d9d2da2 [ExpandMemCmp] Add lit.local.cfg
To prevent AArch64 tests from running when the target is not compiled.

Fixes r371502:

/home/buildslave/ps4-buildslave4/llvm-clang-lld-x86_64-scei-ps4-ubuntu-fast/llvm.src/test/Transforms/ExpandMemCmp/AArch64/memcmp.ll:11:15: error: CHECK-NEXT: expected string not found in input
; CHECK-NEXT: [[TMP0:%.*]] = bitcast i8* [[S1:%.*]] to i64*

llvm-svn: 371503
2019-09-10 10:00:15 +00:00
Clement Courbet 612c260ec3 Reland "r364412 [ExpandMemCmp][MergeICmps] Move passes out of CodeGen into opt pipeline."
With a fix for sanitizer breakage (see explanation in D60318).

llvm-svn: 371502
2019-09-10 09:18:00 +00:00
Fangrui Song 1da4f47195 [yaml2obj] Set p_align to the maximum sh_addralign of contained sections
The address difference between two sections in a PT_LOAD is a constant.
Consider a hypothetical case (pagesize can be very small, say, 4).

```
.text     sh_addralign=4
.text.hot sh_addralign=16
```

If we set p_align to 4, the PT_LOAD will be loaded at an address which
is a multiple of 4. The address of .text.hot is guaranteed to be a
multiple of 4, but not necessarily a multiple of 16.

This patch deletes the constraint

  if (SHeader->sh_offset == PHeader.p_offset)

Reviewed By: grimar, jhenderson

Differential Revision: https://reviews.llvm.org/D67260

llvm-svn: 371501
2019-09-10 09:16:34 +00:00
Igor Kudrin 3565c6b15c [DWARF] Add a unit test for DWARFUnit::getLength().
This is a follow-up of rL369529, where the return value of
DWARFUnit::getLength() was changed from uint32_t to uint64_t.
The test checks that a unit header with Length > 4G can be successfully
parsed and the value of the Length field is not truncated.

Differential Revision: https://reviews.llvm.org/D67276

llvm-svn: 371499
2019-09-10 09:03:24 +00:00
Guillaume Chatelet b6722af068 [Alignment] Use Align for TargetLowering::MinStackArgumentAlignment
Summary:
This is patch is part of a series to introduce an Alignment type.
See this thread for context: http://lists.llvm.org/pipermail/llvm-dev/2019-July/133851.html
See this patch for the introduction of the type: https://reviews.llvm.org/D64790

Reviewers: courbet

Subscribers: sdardis, nemanjai, hiraditya, kbarton, jrtc27, MaskRay, atanasyan, jsji, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D67288

llvm-svn: 371498
2019-09-10 09:01:18 +00:00
James Henderson 3d7e9ab7b9 Revert Remove REQUIRES:shell from tests that pass for me on Windows
This reverts r371478 (git commit a9980f60ce)

llvm-svn: 371497
2019-09-10 08:48:33 +00:00
Roger Ferrer Ibanez 93c4d53b0a [RISCV] Make -march=rv{32,64}gc the default in RISC-V Linux
This is the logical follow-up of D65634.

Differential Revision: https://reviews.llvm.org/D66003

llvm-svn: 371496
2019-09-10 08:16:24 +00:00
Haojian Wu 2fa2d45963 [clangd] some tweaks on the vscode readme, NFC
llvm-svn: 371495
2019-09-10 08:07:12 +00:00
Roger Ferrer Ibanez 8e87396307 [RISCV] Default to ilp32d/lp64d in RISC-V Linux
When running clang as a native compiler in RISC-V Linux the flag
-mabi=ilp32d / -mabi=lp64d is always mandatory. This change makes it the
default there.

Differential Revision: https://reviews.llvm.org/D65634

llvm-svn: 371494
2019-09-10 07:57:36 +00:00
Craig Topper e8b432fa0e [LegalizeTypes] Teach SoftenFloatOp_SELECT_CC to handle operand 2 or 3 being softened.
This can only happen on X86 when fp128 is a legal type, but we
go through softening to generate libcalls. This causes fp128 to
be softened to fp128 instead of an integer type. This can be
removed if D67128 lands.

llvm-svn: 371493
2019-09-10 07:56:02 +00:00
Roger Ferrer Ibanez 60f0a6f6ff [RISCV] Move architecture parsing code into its own function
I plan to reuse it in a later patch.

This is almost NFC except a small change in control flow when diagnosing
+d without +f.

Differential Revision: https://reviews.llvm.org/D66002

llvm-svn: 371492
2019-09-10 07:47:34 +00:00
David Carlier c190890c29 [LLDB] FreeBSD fix new SetFile call.
llvm-svn: 371491
2019-09-10 07:33:39 +00:00
Nico Weber 88d6783f70 gn build: Merge r371488
llvm-svn: 371489
2019-09-10 06:31:59 +00:00
Petr Hosek 7d1757aba8 Revert "clang-misexpect: Profile Guided Validation of Performance Annotations in LLVM"
This reverts commit r371484: this broke sanitizer-x86_64-linux-fast bot.

llvm-svn: 371488
2019-09-10 06:25:13 +00:00
Craig Topper 0e533ca4bb [X86] Add broadcast load unfolding support for VCMPPS/PD.
llvm-svn: 371487
2019-09-10 05:49:53 +00:00
Craig Topper 7c2fdf2779 [X86] Add broadcast load unfold tests for VCMPPS/PD.
llvm-svn: 371486
2019-09-10 05:49:48 +00:00
Nico Weber a6e5a7b637 gn build: Merge r371484
llvm-svn: 371485
2019-09-10 03:18:25 +00:00
Petr Hosek a10802fd73 clang-misexpect: Profile Guided Validation of Performance Annotations in LLVM
This patch contains the basic functionality for reporting potentially
incorrect usage of __builtin_expect() by comparing the developer's
annotation against a collected PGO profile. A more detailed proposal and
discussion appears on the CFE-dev mailing list
(http://lists.llvm.org/pipermail/cfe-dev/2019-July/062971.html) and a
prototype of the initial frontend changes appear here in D65300

We revised the work in D65300 by moving the misexpect check into the
LLVM backend, and adding support for IR and sampling based profiles, in
addition to frontend instrumentation.

We add new misexpect metadata tags to those instructions directly
influenced by the llvm.expect intrinsic (branch, switch, and select)
when lowering the intrinsics. The misexpect metadata contains
information about the expected target of the intrinsic so that we can
check against the correct PGO counter when emitting diagnostics, and the
compiler's values for the LikelyBranchWeight and UnlikelyBranchWeight.
We use these branch weight values to determine when to emit the
diagnostic to the user.

A future patch should address the comment at the top of
LowerExpectIntrisic.cpp to hoist the LikelyBranchWeight and
UnlikelyBranchWeight values into a shared space that can be accessed
outside of the LowerExpectIntrinsic pass. Once that is done, the
misexpect metadata can be updated to be smaller.

In the long term, it is possible to reconstruct portions of the
misexpect metadata from the existing profile data. However, we have
avoided this to keep the code simple, and because some kind of metadata
tag will be required to identify which branch/switch/select instructions
are influenced by the use of llvm.expect

Patch By: paulkirth
Differential Revision: https://reviews.llvm.org/D66324

llvm-svn: 371484
2019-09-10 03:11:39 +00:00
Kai Luo 73da43aeb3 [PowerPC][NFC] Update test assertions using update_llc_test_checks.py
Summary:
This patch is made due to https://reviews.llvm.org/rL371289 where typo
fixes failed.

Differential Revision: https://reviews.llvm.org/D67317

llvm-svn: 371483
2019-09-10 02:28:24 +00:00
Mehdi Amini daa79c5338 Revert [git-llvm] Do not reinvent `@{upstream}`
This reverts r371290 (git commit 7faffd544b)

The change wasnt NFC and broke some users' workflow. Reverting while figuring
out the best alternative to move forward.

llvm-svn: 371480
2019-09-10 01:26:36 +00:00
Nico Weber 9396143412 gn build: Merge r371466
llvm-svn: 371479
2019-09-10 01:11:30 +00:00
Reid Kleckner a9980f60ce Remove REQUIRES:shell from tests that pass for me on Windows
I see in the history for some of these tests REQUIRES:shell was used as
a way to disable tests on Windows because they are flaky there. I tried
not to re-enable such tests, but it's possible that I missed some and
this will re-enable flaky tests on Windows. If so, we should disable
them with UNSUPPORTED:system-windows and add a comment that they are
flaky there. So far as I can tell, the lit internal shell is capable of
running all of these tests, and we shouldn't use REQUIRES:shell as a
proxy for Windows.

llvm-svn: 371478
2019-09-10 00:50:32 +00:00
Nico Weber fcbc512f80 gn build: (manually) merge r371429
llvm-svn: 371477
2019-09-10 00:48:20 +00:00
Richard Smith ae6f7bcb36 Fix crash mangling an explicit lambda non-type template parameter pack
that is not a pack expansion.

llvm-svn: 371476
2019-09-10 00:39:53 +00:00
Jan Korous 79707ecd7d [llvm][ADT][NFC] Add test for makeArrayRef(std::array)
llvm-svn: 371475
2019-09-10 00:29:35 +00:00
Jonas Devlieghere e0ea8d87eb [Utility] Replace `lldb_private::CleanUp` by `llvm::scope_exit`
This removes the CleanUp class and replaces its usages with llvm's
ScopeExit, which has similar semantics.

Differential revision: https://reviews.llvm.org/D67378

llvm-svn: 371474
2019-09-10 00:20:50 +00:00
Reid Kleckner 87d47cb7c4 Remove some unnecessary REQUIRES: shell lines
This means these tests will run on Windows. Replace one with
UNSUPPORTED: system-windows.

llvm-svn: 371473
2019-09-10 00:06:52 +00:00
Alex Langford 1dbee8f043 [Expression] Remove unused header from LLVMUserExpression
llvm-svn: 371472
2019-09-09 23:59:54 +00:00
Matt Arsenault a91f017ae3 AMDGPU/GlobalISel: Fix insert point when lowering fminnum/fmaxnum
llvm-svn: 371471
2019-09-09 23:30:11 +00:00
Alex Langford 9e86561878 [Symbol] Give ClangASTContext a PersistentExpressionState instead of a ClangPersistentVariables
ClangASTContext doesn't use m_persistent_variables in a way specific to
ClangPersistentVariables. Therefore, it should hold a unique pointer to
PersistentExpressionState instead of a ClangPersistentVariablesUP.
This also prevents you from pulling in a plugin header when including
ClangASTContext.h

Doing this exposed an implicit dependency in ObjCLanguage that was
corrected by including ClangModulesDeclVendor.h

llvm-svn: 371470
2019-09-09 23:11:43 +00:00
Richard Smith 865697f9b9 Fix incorrect demangling of call operator of lambda with explicit
template parameters due to registering template parameters twice.

llvm-svn: 371469
2019-09-09 23:07:25 +00:00
Richard Smith 245ba2c25f PR43242: Fix crash when typo-correcting to an operator() that should not
have been visible.

llvm-svn: 371468
2019-09-09 23:07:22 +00:00
Austin Kerbow 06c8cb03ca AMDGPU/GlobalISel: Rename MIRBuilder to B. NFC
Reviewers: arsenm

Reviewed By: arsenm

Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, rovka, dstuttard, tpr, t-tye, hiraditya, Petar.Avramovic, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D67374

llvm-svn: 371467
2019-09-09 23:06:13 +00:00
Reid Kleckner bf02399a85 [Windows] Replace TrapUnreachable with an int3 insertion pass
This is an alternative to D66980, which was reverted. Instead of
inserting a pseudo instruction that optionally expands to nothing, add a
pass that inserts int3 when appropriate after basic block layout.

Reviewers: hans

Differential Revision: https://reviews.llvm.org/D67201

llvm-svn: 371466
2019-09-09 23:04:25 +00:00
Aditya Nandakumar 5112b71126 [GlobalISel]: Fix a bug where we could dereference None
getConstantVRegVal returns None when dealing with constants > 64 bits.
Don't assume we always have a value in GISelKnownBits.

llvm-svn: 371465
2019-09-09 22:51:41 +00:00
Richard Smith 2ca7370183 Simplify demangler rule for lambda-expressions to match discussion on
cxx-abi list.

llvm-svn: 371462
2019-09-09 22:26:04 +00:00
Evgeniy Stepanov f0e2755b45 LangRef: mention MSan's problem with speculative conditional branches.
Summary:
This short blurb aims to disallow optimizations like we had to revert
(under MSan) in
  https://reviews.llvm.org/D21165
  https://bugs.llvm.org/show_bug.cgi?id=28054
  https://reviews.llvm.org/D67205

Reviewers: vitalybuka, efriedma

Subscribers: llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D67244

llvm-svn: 371461
2019-09-09 22:24:57 +00:00
Jonas Devlieghere e0bce4e1c2 Revert "[Reproducer] Add a `cont` to ModuleCXX.test"
This should no longer be necessary after r371459.

llvm-svn: 371460
2019-09-09 22:07:45 +00:00
Jonas Devlieghere 9b961cc604 [Reproducer] Disconnect when the replay server is out of packets.
This is a fix for the issue described in r371144.

> On more than one occasion I've found this test got stuck during replay
> while waiting for a packet from debugserver when the debugger was in
> the process of being destroyed.

When the replay server is out of packets we should just disconnect so
the debugger doesn't have to do any cleanup that it wouldn't do during
capture.

llvm-svn: 371459
2019-09-09 22:05:48 +00:00
Simon Atanasyan 2439b8b0c5 [mips] Make another set of test cases more tolerant to exact symbol addresses. NFC
llvm-svn: 371458
2019-09-09 22:04:20 +00:00
Greg Clayton 4f68c226a5 Fix ELF core file memory reading for PT_LOAD program headers with no p_filesz
Prior to this fix, ELF files might contain PT_LOAD program headers that had a valid p_vaddr, and a valid file p_offset, but the p_filesz would be zero. For example in llvm-project/lldb/test/testcases/functionalities/postmortem/elf-core/thread_crash/linux-i386.core we see:

Program Headers:
Index   p_type           p_flags    p_offset           p_vaddr            p_paddr            p_filesz           p_memsz            p_align
======= ---------------- ---------- ------------------ ------------------ ------------------ ------------------ ------------------ ------------------
[    0] PT_NOTE          0x00000000 0x0000000000000474 0x0000000000000000 0x0000000000000000 0x0000000000001940 0x0000000000000000 0x0000000000000000
[    1] PT_LOAD          0x00000005 0x0000000000002000 0x0000000008048000 0x0000000000000000 0x0000000000000000 0x0000000000003000 0x0000000000001000
[    2] PT_LOAD          0x00000004 0x0000000000002000 0x000000000804b000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[    3] PT_LOAD          0x00000006 0x0000000000002000 0x000000000804c000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[    4] PT_LOAD          0x00000006 0x0000000000002000 0x0000000009036000 0x0000000000000000 0x0000000000000000 0x0000000000025000 0x0000000000001000
[    5] PT_LOAD          0x00000000 0x0000000000002000 0x00000000f63a1000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[    6] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f63a2000 0x0000000000000000 0x0000000000000000 0x0000000000800000 0x0000000000001000
[    7] PT_LOAD          0x00000000 0x0000000000002000 0x00000000f6ba2000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[    8] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f6ba3000 0x0000000000000000 0x0000000000000000 0x0000000000804000 0x0000000000001000
[    9] PT_LOAD          0x00000005 0x0000000000002000 0x00000000f73a7000 0x0000000000000000 0x0000000000000000 0x00000000001b1000 0x0000000000001000
[   10] PT_LOAD          0x00000004 0x0000000000002000 0x00000000f7558000 0x0000000000000000 0x0000000000000000 0x0000000000002000 0x0000000000001000
[   11] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f755a000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[   12] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f755b000 0x0000000000000000 0x0000000000000000 0x0000000000003000 0x0000000000001000
[   13] PT_LOAD          0x00000005 0x0000000000002000 0x00000000f755e000 0x0000000000000000 0x0000000000000000 0x0000000000019000 0x0000000000001000
[   14] PT_LOAD          0x00000004 0x0000000000002000 0x00000000f7577000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[   15] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f7578000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[   16] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f7579000 0x0000000000000000 0x0000000000000000 0x0000000000002000 0x0000000000001000
[   17] PT_LOAD          0x00000005 0x0000000000002000 0x00000000f757b000 0x0000000000000000 0x0000000000000000 0x000000000001c000 0x0000000000001000
[   18] PT_LOAD          0x00000004 0x0000000000002000 0x00000000f7597000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[   19] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f7598000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[   20] PT_LOAD          0x00000005 0x0000000000002000 0x00000000f7599000 0x0000000000000000 0x0000000000000000 0x0000000000053000 0x0000000000001000
[   21] PT_LOAD          0x00000004 0x0000000000002000 0x00000000f75ec000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[   22] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f75ed000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[   23] PT_LOAD          0x00000005 0x0000000000002000 0x00000000f75ee000 0x0000000000000000 0x0000000000000000 0x0000000000176000 0x0000000000001000
[   24] PT_LOAD          0x00000004 0x0000000000002000 0x00000000f7764000 0x0000000000000000 0x0000000000000000 0x0000000000006000 0x0000000000001000
[   25] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f776a000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[   26] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f776b000 0x0000000000000000 0x0000000000000000 0x0000000000003000 0x0000000000001000
[   27] PT_LOAD          0x00000006 0x0000000000002000 0x00000000f778a000 0x0000000000000000 0x0000000000000000 0x0000000000002000 0x0000000000001000
[   28] PT_LOAD          0x00000004 0x0000000000002000 0x00000000f778c000 0x0000000000000000 0x0000000000002000 0x0000000000002000 0x0000000000001000
[   29] PT_LOAD          0x00000005 0x0000000000004000 0x00000000f778e000 0x0000000000000000 0x0000000000002000 0x0000000000002000 0x0000000000001000
[   30] PT_LOAD          0x00000005 0x0000000000006000 0x00000000f7790000 0x0000000000000000 0x0000000000000000 0x0000000000022000 0x0000000000001000
[   31] PT_LOAD          0x00000004 0x0000000000006000 0x00000000f77b3000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[   32] PT_LOAD          0x00000006 0x0000000000006000 0x00000000f77b4000 0x0000000000000000 0x0000000000000000 0x0000000000001000 0x0000000000001000
[   33] PT_LOAD          0x00000006 0x0000000000006000 0x00000000ffa25000 0x0000000000000000 0x0000000000000000 0x0000000000022000 0x0000000000001000
Prior to this fix if users tried to read memory from one of these addresses like 0x8048000, they would end up incorrectly reading from the next memory region that actually had a p_filesz which would be 0x00000000f778c000 in this case. This fix correctly doesn't include program headers with zero p_filesz in the ProcessELFCore::m_core_aranges that is used to read memory. I found two cores files that have this same issue and added tests.

Differential Revision: https://reviews.llvm.org/D67370

llvm-svn: 371457
2019-09-09 21:45:49 +00:00
Philip Reames b8cddb7611 [Tests] Fix a typo in a test
llvm-svn: 371456
2019-09-09 21:33:59 +00:00
Philip Reames 847fbf7013 [Tests] Precommit test case for D67372
llvm-svn: 371455
2019-09-09 21:32:16 +00:00
Simon Pilgrim 7f37d9a714 Fix MSVC "not all control paths return a value" warning. NFCI.
llvm-svn: 371454
2019-09-09 21:30:11 +00:00
Max Moroz ac3dce595c [UBSan] Follow up fix for r371442.
Reviewers: vitalybuka, hctim, Dor1s

Reviewed By: Dor1s

Subscribers: delcypher, #sanitizers, llvm-commits

Tags: #llvm, #sanitizers

Differential Revision: https://reviews.llvm.org/D67371

llvm-svn: 371453
2019-09-09 21:00:25 +00:00
Philip Reames 7403569be7 [LoopVectorize] Leverage speculation safety to avoid masked.loads
If we're vectorizing a load in a predicated block, check to see if the load can be speculated rather than predicated.  This allows us to generate a normal vector load instead of a masked.load.

To do so, we must prove that all bytes accessed on any iteration of the original loop are dereferenceable, and that all loads (across all iterations) are properly aligned.  This is equivelent to proving that hoisting the load into the loop header in the original scalar loop is safe.

Note: There are a couple of code motion todos in the code.  My intention is to wait about a day - to be sure this sticks - and then perform the NFC motion without furthe review.

Differential Revision: https://reviews.llvm.org/D66688

llvm-svn: 371452
2019-09-09 20:54:13 +00:00
Artem Dergachev 589273bebd [analyzer] NFC: Simplify bug report equivalence classes to not be ilists.
Use a vector of unique pointers instead.

Differential Revision: https://reviews.llvm.org/D67024

llvm-svn: 371451
2019-09-09 20:34:44 +00:00
Artem Dergachev 2f169e7cdd [analyzer] NFC: Introduce sub-classes for path-sensitive and basic reports.
Checkers are now required to specify whether they're creating a
path-sensitive report or a path-insensitive report by constructing an
object of the respective type.

This makes BugReporter more independent from the rest of the Static Analyzer
because all Analyzer-specific code is now in sub-classes.

Differential Revision: https://reviews.llvm.org/D66572

llvm-svn: 371450
2019-09-09 20:34:40 +00:00
Philip Reames 48453bb8ed [Tests] Add anyextend tests for unordered atomics
Motivated by work on changing our representation of unordered atomics in SelectionDAG, but as an aside, all our lowerings for O3 are terrible.  Even the ones which ignore the atomicity.  

llvm-svn: 371449
2019-09-09 20:26:52 +00:00
Douglas Yung 4bd6eb8ff2 Relax opcode checks in test to check for only a number instead of a specific number.
llvm-svn: 371447
2019-09-09 20:12:29 +00:00