Gabor Marton [Thu, 24 Jan 2019 15:42:20 +0000 (15:42 +0000)]
Fix failing buildbots
Related commit which caused the buildbots to fail:
rL352050
llvm-svn: 352055
Erich Keane [Thu, 24 Jan 2019 15:28:57 +0000 (15:28 +0000)]
[CPU-Dispatch] Make pentium_iii_no_xmm_regs and pentium_iii alias.
I discovered that in ICC (where this list comes from), that the two
pentium_iii versions were actually identical despite the two different
names (despite them implying a difference). Because of this, they ended
up having identical manglings, which obviously caused problems when used
together.
This patch makes pentium_iii_no_xmm_regs an alias for pentium_iii so
that it can still be used, but has the same meaning as ICC. However, we
still prohibit using the two together which is different (albeit better)
behavior.
Change-Id: I4f3c9a47e48490c81525c8a3d23ed4201921b288
llvm-svn: 352054
Nirav Dave [Thu, 24 Jan 2019 15:15:32 +0000 (15:15 +0000)]
[SelectionDAGBuilder] Fuse inline asm input operand loops passes. NFCI.
llvm-svn: 352053
Michael Platings [Thu, 24 Jan 2019 15:11:26 +0000 (15:11 +0000)]
[Docs] Add information about unit tests to the testing guide
Differential Revision: https://reviews.llvm.org/D57088
llvm-svn: 352052
Nirav Dave [Thu, 24 Jan 2019 15:04:17 +0000 (15:04 +0000)]
[X86] Add missing isReg() guards in FixupSetCCs pass.
llvm-svn: 352051
Gabor Marton [Thu, 24 Jan 2019 14:47:44 +0000 (14:47 +0000)]
[ASTImporter] Fix inequality of functions with different attributes
Summary:
FunctionType::ExtInfo holds such properties of a function which are needed
mostly for code gen. We should not compare these bits when checking for
structural equivalency.
Checking ExtInfo caused false ODR errors during CTU analysis (of tmux).
Reviewers: a_sidorin, a.sidorin, shafik
Subscribers: rnkovacs, dkrupp, Szelethus, cfe-commits
Differential Revision: https://reviews.llvm.org/D53699
llvm-svn: 352050
Haojian Wu [Thu, 24 Jan 2019 14:25:53 +0000 (14:25 +0000)]
[clangd] Clean the cache of file statuses on vscode-clangd when clangd crashes.
Summary:
Clear the cached file statuses, otherwise we will leave some garbage texts on
the status bar when clangd crashes.
Reviewers: ilya-biryukov
Subscribers: ioeric, MaskRay, jkorous, arphaman, kadircet, cfe-commits
Differential Revision: https://reviews.llvm.org/D56540
llvm-svn: 352049
Sanjay Patel [Thu, 24 Jan 2019 14:12:34 +0000 (14:12 +0000)]
[x86] add tests for unpack shuffle lowering; NFC
https://bugs.llvm.org/show_bug.cgi?id=40434
llvm-svn: 352048
Bruno Ricci [Thu, 24 Jan 2019 13:52:47 +0000 (13:52 +0000)]
[Sema] Don't crash when recovering from a misspelled pseudo destructor call to an incomplete type.
When attempting to correct a misspelled pseudo destructor call as in:
struct Foo;
void foo(Foo *p) {
p.~Foo();
}
a call is made in canRecoverDotPseudoDestructorCallsOnPointerObjects
to LookupDestructor without checking that the record has a definition.
This causes an assertion later in LookupSpecialMember which assumes that
the record has a definition.
Patch By Roman Zhikharevich!
Differential Revision: https://reviews.llvm.org/D57111
Reviewed By: riccibruno
llvm-svn: 352047
Simon Pilgrim [Thu, 24 Jan 2019 13:48:20 +0000 (13:48 +0000)]
[CostModel][X86] Add SMUL fixed point cost tests
llvm-svn: 352046
Simon Pilgrim [Thu, 24 Jan 2019 13:36:45 +0000 (13:36 +0000)]
[TTI] Add generic SADDO/SSUBO costs
Added x86 scalar sadd_with_overflow/ssub_with_overflow costs.
llvm-svn: 352045
Simon Pilgrim [Thu, 24 Jan 2019 12:27:10 +0000 (12:27 +0000)]
[TTI] Add generic UADDSAT/USUBSAT costs
Add generic costs calculation for UADDSAT/USUBSAT intrinsics, this fallbacks to using generic costs for uadd_with_overflow/usub_with_overflow + a select.
Differential Revision: https://reviews.llvm.org/D56907
llvm-svn: 352044
Simon Pilgrim [Thu, 24 Jan 2019 12:10:20 +0000 (12:10 +0000)]
[TTI] Add generic UADDO/USUBO costs
Added x86 scalar uadd_with_overflow/usub_with_overflow costs.
Differential Revision: https://reviews.llvm.org/D56907
llvm-svn: 352043
Pierre Gousseau [Thu, 24 Jan 2019 11:44:24 +0000 (11:44 +0000)]
Test commit: fix typo.
llvm-svn: 352042
Florian Hahn [Thu, 24 Jan 2019 11:22:08 +0000 (11:22 +0000)]
Revert "[HotColdSplitting] Get DT and PDT from the pass manager."
This reverts commit
a6982414edf315c39ae93f3c3322476217119e99 (llvm-svn: 352036),
because it causes a memory leak in the pass manager. Failing bot
http://lab.llvm.org:8011/builders/sanitizer-x86_64-linux-bootstrap/builds/10351/steps/check-llvm%20asan/logs/stdio
llvm-svn: 352041
Ilya Biryukov [Thu, 24 Jan 2019 10:41:43 +0000 (10:41 +0000)]
[CodeComplete] [clangd] Fix crash on ValueDecl with a null type
Reviewers: kadircet
Reviewed By: kadircet
Subscribers: ioeric, MaskRay, jkorous, arphaman, cfe-commits
Differential Revision: https://reviews.llvm.org/D57093
llvm-svn: 352040
Serge Guelton [Thu, 24 Jan 2019 10:34:44 +0000 (10:34 +0000)]
Fix python3 compability issue in clang binding
The file contents could be of str type. Should use byte length instead
of str length, otherwise utf-8 encoded files may not get properly parsed
for completion.
Source issue: https://github.com/ncm2/ncm2-pyclang#2
Commited on behalf of `roxma'
Differential Revision: https://reviews.llvm.org/D56429
llvm-svn: 352039
Petar Avramovic [Thu, 24 Jan 2019 10:27:21 +0000 (10:27 +0000)]
[MIPS GlobalISel] Select zero extending and sign extending load
Select zero extending and sign extending load for MIPS32.
Use size from MachineMemOperand to determine number of bytes to load.
Differential Revision: https://reviews.llvm.org/D57099
llvm-svn: 352038
Petar Avramovic [Thu, 24 Jan 2019 10:09:52 +0000 (10:09 +0000)]
[MIPS GlobalISel] Combine extending loads
Use CombinerHelper to combine extending load instructions.
G_LOAD combined with G_ZEXT, G_SEXT or G_ANYEXT gives G_ZEXTLOAD,
G_SEXTLOAD or G_LOAD with same type as def of extending instruction
respectively.
Similarly G_ZEXTLOAD combined with G_ZEXT gives G_ZEXTLOAD and
G_SEXTLOAD combined with G_SEXT gives G_SEXTLOAD with same type
as def of extending instruction.
Differential Revision: https://reviews.llvm.org/D56914
llvm-svn: 352037
Florian Hahn [Thu, 24 Jan 2019 09:44:52 +0000 (09:44 +0000)]
[HotColdSplitting] Get DT and PDT from the pass manager.
Instead of manually computing DT and PDT, we can get the from the pass
manager, which ideally has them already cached. With the new pass
manager, we could even preserve DT/PDT on a per function basis in a
module pass.
I think this also addresses the TODO about re-using the computed DTs for
BFI. IIUC, GetBFI will fetch the DT from the pass manager and when we
will fetch the cached version later.
Reviewers: vsk, hiraditya, tejohnson, thegameg, sebpop
Reviewed By: vsk
Differential Revision: https://reviews.llvm.org/D57092
llvm-svn: 352036
Haojian Wu [Thu, 24 Jan 2019 09:35:08 +0000 (09:35 +0000)]
Fix potential ODR vialation.
llvm-svn: 352035
Simon Atanasyan [Thu, 24 Jan 2019 09:13:14 +0000 (09:13 +0000)]
Reapply: [mips] Handle MipsMCExpr sub-expression for the MEK_DTPREL tag
This reapplies commit r351987 with a failed test fix. Now the test
accepts both DW_OP_GNU_push_tls_address and DW_OP_form_tls_address
opcode.
Original commit message:
```
This is a fix for a regression introduced by the rL348194 commit. In
that change new type (MEK_DTPREL) of MipsMCExpr expression was added,
but in some places of the code this type of expression considered as
unexpected.
This change fixes the bug. The MEK_DTPREL type of expression is used for
marking TLS DIEExpr only and contains a regular sub-expression. Where we
need to handle the expression, we retrieve the sub-expression and
handle it in a common way.
```
llvm-svn: 352034
Martin Liska [Thu, 24 Jan 2019 08:46:06 +0000 (08:46 +0000)]
Revert https://reviews.llvm.org/D56485.
llvm-svn: 352033
Martin Liska [Thu, 24 Jan 2019 08:25:36 +0000 (08:25 +0000)]
Always compare C++ typeinfo (based on libstdc++ implementation).
Differential Revision: https://reviews.llvm.org/D56485.
llvm-svn: 352032
David Carlier [Thu, 24 Jan 2019 07:58:42 +0000 (07:58 +0000)]
[extra] unit tests enable crash-recovery cases on FreeBSD
Seems the previous statement does not hold up anymore.
Reviewers: steveire
Reviewed By: steveire
Differential Revision: https://reviews.llvm.org/D57102
llvm-svn: 352031
Jonas Paulsson [Thu, 24 Jan 2019 07:54:41 +0000 (07:54 +0000)]
[SystemZ] Remember to reset the NoPHIs property on MF in createPHIsForSelects()
After creating new PHI instructions during isel pseudo expansion, the NoPHIs
property of MF should be reset in case it was previously set.
Review: Ulrich Weigand
llvm-svn: 352030
Craig Topper [Thu, 24 Jan 2019 07:51:34 +0000 (07:51 +0000)]
[X86] Update SelectionDAGDumper to print the extension type and expanding flag for masked loads. Add truncating and compressing for masked stores.
llvm-svn: 352029
Craig Topper [Thu, 24 Jan 2019 06:15:03 +0000 (06:15 +0000)]
[X86] Add test cases for opportunities to fold a truncate and a masked store into a truncating masked store.
llvm-svn: 352027
Max Kazantsev [Thu, 24 Jan 2019 05:43:19 +0000 (05:43 +0000)]
[NFC] Add another failing test on LoopSimplifyCFG
llvm-svn: 352026
Hsiangkai Wang [Thu, 24 Jan 2019 05:34:29 +0000 (05:34 +0000)]
Reland r345009 "[DebugInfo] Generate debug information for labels."
Generate DILabel metadata and call llvm.dbg.label after label
statement to associate the metadata with the label.
After fixing PR37395.
After fixing problems in LiveDebugVariables.
After fixing NULL symbol problems in AddressPool when enabling
split-dwarf-file.
After fixing PR39094.
After landing D54199 and D54465 to fix Chromium build failed.
Differential Revision: https://reviews.llvm.org/D45045
llvm-svn: 352025
Max Kazantsev [Thu, 24 Jan 2019 05:20:29 +0000 (05:20 +0000)]
[LoopSimplifyCFG] Fix inconsistency in live blocks markup
When we choose whether or not we should mark block as dead, we have an
inconsistent logic in markup of live blocks.
- We take candidate IF its terminator branches on constant AND it is immediately
in current loop;
- We mark successor live IF its terminator doesn't branch by constant OR it branches
by constant and the successor is its always taken block.
What we are missing here is that when the terminator branches on a constant but is
not taken as a candidate because is it not immediately in the current loop, we will
mark only one (always taken) successor as live. Therefore, we do NOT do the actual
folding but may NOT mark one of the successors as live. So the result of markup is
wrong in this case, and we may then hit various asserts.
Thanks Jordan Rupprech for reporting this!
Differential Revision: https://reviews.llvm.org/D57095
Reviewed By: rupprecht
llvm-svn: 352024
Max Kazantsev [Thu, 24 Jan 2019 05:05:55 +0000 (05:05 +0000)]
[NFC] Add a failing test on live block markup in term folding
llvm-svn: 352023
Pavel Labath [Thu, 24 Jan 2019 04:17:59 +0000 (04:17 +0000)]
BreakpadRecords: Address post-commit feedback
Summary:
This addresses the issues raised in D56844. It removes the accessors from the
breakpad record structures by making the fields public. Also, I refactor the
UUID parsing code to remove hard-coded constants.
Reviewers: lemo
Subscribers: clayborg, lldb-commits
Differential Revision: https://reviews.llvm.org/D57037
llvm-svn: 352021
Julian Lettner [Thu, 24 Jan 2019 03:51:01 +0000 (03:51 +0000)]
Temporarily deactivate tests; it fails on certain bots
llvm-svn: 352020
David Blaikie [Thu, 24 Jan 2019 03:27:57 +0000 (03:27 +0000)]
DebugInfo: Use assembly label arithmetic for address pool size for easier reading/editing
Recommits 350048, 350050 That broke buildbots because of some typos in
the test case.
llvm-svn: 352019
Petr Hosek [Thu, 24 Jan 2019 03:24:37 +0000 (03:24 +0000)]
[CMake][Fuchsia] Enable hermetic static libc++abi for Fuchsia
Similarly to libc++, we want to use hermetic static libc++abi.
Differential Revision: https://reviews.llvm.org/D57136
llvm-svn: 352018
Petr Hosek [Thu, 24 Jan 2019 03:18:29 +0000 (03:18 +0000)]
[libcxxabi] Support building hermetic static library
This is useful when the static libc++abi library is being linked into
shared libraries that may be used in with other shared libraries that
use different C++ library. We want to avoid avoid exporting libc++abi
or libc++ symbols in those cases. This achieved by a new CMake option
which can be enabled by libc++abi vendors as needed.
Differential Revision: https://reviews.llvm.org/D56026
llvm-svn: 352017
Petr Hosek [Thu, 24 Jan 2019 03:04:42 +0000 (03:04 +0000)]
[libunwind] Don't abort if encoutering invalid .eh_frame_hdr
Recent Linux kernel release has introduced a bug as part of the ORC
rollout where the vDSO has a valid .eh_frame section, but it's missing
the .eh_frame_hdr section and GNU_EH_FRAME segment has zero size. This
causes libunwind to abort which breaks programs that use libunwind.
The other unwinder implementation (libgcc, non-gnu) instead silently
bail out unless being compiled as debug. This change modifies libunwind
to use the same strategy.
Differential Revision: https://reviews.llvm.org/D57081
llvm-svn: 352016
Ana Pazos [Thu, 24 Jan 2019 03:00:26 +0000 (03:00 +0000)]
Revert "[RISCV] Set isAsCheapAsAMove for ADDI, ORI, XORI, LUI"
This reverts commit
ccfb060ecb5d7e18ea729455660484d576bde2cc.
Some tests need to to fixed before reapplying this commit.
llvm-svn: 352014
Fangrui Song [Thu, 24 Jan 2019 02:56:34 +0000 (02:56 +0000)]
[ubsan] Fix test when isPICDefault() returns false after rCTE352003
llvm-svn: 352013
Richard Smith [Thu, 24 Jan 2019 02:41:46 +0000 (02:41 +0000)]
Improve diagnostic for -fvisibility mismatch between module/PCH build
and use.
llvm-svn: 352011
Ana Pazos [Thu, 24 Jan 2019 02:41:40 +0000 (02:41 +0000)]
[RISCV] Set isAsCheapAsAMove for ADDI, ORI, XORI, LUI
Summary:
Affected instructions:
PseudoLI simplest form (ADDI with X0)
ALU operations with immediate (they do not set status flag - ADDI, ORI, XORI)
Reviewers: asb
Reviewed By: asb
Subscribers: shiva0217, rkruppe, kito-cheng, asb, rbar, johnrusso, simoncook, sabuasal, niosHD, zzheng, edward-jones, mgrang, rogfer01, MartinMosbeck, brucehoult, the_o, PkmX, jocewei
Differential Revision: https://reviews.llvm.org/D56526
llvm-svn: 352010
Adrian Prantl [Thu, 24 Jan 2019 02:37:28 +0000 (02:37 +0000)]
Skip test on clang <8 instead of 7
llvm-svn: 352009
Ana Pazos [Thu, 24 Jan 2019 02:31:23 +0000 (02:31 +0000)]
[RISCV] Set isReMaterializable for ORI, XORI
Reviewers: asb
Reviewed By: asb
Subscribers: asb, rbar, johnrusso, simoncook, sabuasal, niosHD, kito-cheng, shiva0217, zzheng, edward-jones, rogfer01, MartinMosbeck, brucehoult, the_o, rkruppe, PkmX, jocewei
Differential Revision: https://reviews.llvm.org/D57069
llvm-svn: 352008
Marshall Clow [Thu, 24 Jan 2019 02:02:50 +0000 (02:02 +0000)]
Change a couple of '&' to addressof(). NFC
llvm-svn: 352007
Marshall Clow [Thu, 24 Jan 2019 01:52:56 +0000 (01:52 +0000)]
Uncomment the entire test, but mark as XFAIL on linux-gnu because it uses locales that aren't generally available there, similar to the other regex tests
llvm-svn: 352006
Douglas Yung [Thu, 24 Jan 2019 01:22:32 +0000 (01:22 +0000)]
[docs] Remove extra character from git URL in Getting Started guide.
llvm-svn: 352005
David Blaikie [Thu, 24 Jan 2019 01:19:17 +0000 (01:19 +0000)]
llvm-symbolizer: Extract individual test cases now that it's easier to use directly (without a piped input file)
Pulling out the split-dwarf tests by way of example of how I think
llvm-symbolizer should be tested going forward. Open to
debate/discussion, though.
llvm-svn: 352004
Julian Lettner [Thu, 24 Jan 2019 01:06:19 +0000 (01:06 +0000)]
[Sanitizers] UBSan unreachable incompatible with ASan in the presence of `noreturn` calls
Summary:
UBSan wants to detect when unreachable code is actually reached, so it
adds instrumentation before every `unreachable` instruction. However,
the optimizer will remove code after calls to functions marked with
`noreturn`. To avoid this UBSan removes `noreturn` from both the call
instruction as well as from the function itself. Unfortunately, ASan
relies on this annotation to unpoison the stack by inserting calls to
`_asan_handle_no_return` before `noreturn` functions. This is important
for functions that do not return but access the the stack memory, e.g.,
unwinder functions *like* `longjmp` (`longjmp` itself is actually
"double-proofed" via its interceptor). The result is that when ASan and
UBSan are combined, the `noreturn` attributes are missing and ASan
cannot unpoison the stack, so it has false positives when stack
unwinding is used.
Changes:
# UBSan now adds the `expect_noreturn` attribute whenever it removes
the `noreturn` attribute from a function
# ASan additionally checks for the presence of this attribute
Generated code:
```
call void @__asan_handle_no_return // Additionally inserted to avoid false positives
call void @longjmp
call void @__asan_handle_no_return
call void @__ubsan_handle_builtin_unreachable
unreachable
```
The second call to `__asan_handle_no_return` is redundant. This will be
cleaned up in a follow-up patch.
rdar://problem/
40723397
Reviewers: delcypher, eugenis
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D56624
llvm-svn: 352003
Nico Weber [Thu, 24 Jan 2019 01:00:52 +0000 (01:00 +0000)]
gn build: Merge r351320 (the 9.0.0 version bump)
llvm-svn: 352002
David Callahan [Thu, 24 Jan 2019 00:55:23 +0000 (00:55 +0000)]
Update entry count for cold calls
Summary:
Profile sample files include the number of times each entry or inlined
call site is sampled. This is translated into the entry count metadta
on functions.
When sample data is being read, if a call site that was inlined
in the sample program is considered cold and not inlined, then
the entry count of the out-of-line functions does not reflect
the current compilation.
In this patch, we note call sites where the function was not inlined
and as a last action of the sample profile loading, we update the
called function's entry count to reflect the calls from these
call sites which are not included in the profile file.
Reviewers: danielcdh, wmi, Kader, modocache
Reviewed By: wmi
Subscribers: davidxl, eraman, llvm-commits
Differential Revision: https://reviews.llvm.org/D52845
llvm-svn: 352001
Douglas Yung [Thu, 24 Jan 2019 00:34:09 +0000 (00:34 +0000)]
[llvm-symbolizer] Add support for -i and -inlines as aliases for -inlining
This change adds two options, -i and -inlines as aliases for the -inlining option to llvm-symbolizer to improve compatibility with the GNU addr2line utility which accepts these options.
It also modifies existing tests that use -inlining to exercise these new aliases as well.
This fixes PR40073.
Reviewed by: jhenderson, Quolyk, ruiu
Differential Revision: https://reviews.llvm.org/D57083
llvm-svn: 351999
Amara Emerson [Thu, 24 Jan 2019 00:24:59 +0000 (00:24 +0000)]
Revert "[mips] Handle MipsMCExpr sub-expression for the MEK_DTPREL tag"
This reverts commit r351987 as it broke some bots.
llvm-svn: 351998
Leonard Chan [Thu, 24 Jan 2019 00:11:35 +0000 (00:11 +0000)]
[Sema] Fix Modified Type in address_space AttributedType
This is a fix for https://reviews.llvm.org/D51229 where we pass the
address_space qualified type as the modified type of an AttributedType. This
change now instead wraps the AttributedType with either the address_space
qualifier or a DependentAddressSpaceType.
Differential Revision: https://reviews.llvm.org/D55447
llvm-svn: 351997
Mircea Trofin [Thu, 24 Jan 2019 00:10:25 +0000 (00:10 +0000)]
[llvm] Clarify responsiblity of some of DILocation discriminator APIs
Summary:
Renamed setBaseDiscriminator to cloneWithBaseDiscriminator, to match
similar APIs. Also changed its behavior to copy over the other
discriminator components, instead of eliding them.
Renamed cloneWithDuplicationFactor to
cloneByMultiplyingDuplicationFactor, which more closely matches what
this API does.
Reviewers: dblaikie, wmi
Reviewed By: dblaikie
Subscribers: zzheng, llvm-commits
Differential Revision: https://reviews.llvm.org/D56220
llvm-svn: 351996
Kamil Rytarowski [Wed, 23 Jan 2019 23:24:43 +0000 (23:24 +0000)]
Mark another test as flaky
Reported on the NetBSD 8 buildbot.
llvm-svn: 351995
Marshall Clow [Wed, 23 Jan 2019 23:06:18 +0000 (23:06 +0000)]
Apply D28248: 'Work around GCC PR37804'. Thanks to mdaniels for the patch
llvm-svn: 351993
Reid Kleckner [Wed, 23 Jan 2019 22:59:52 +0000 (22:59 +0000)]
[ADT] Notify ilist traits about in-list transfers
Summary:
Previously no client of ilist traits has needed to know about transfers
of nodes within the same list, so as an optimization, ilist doesn't call
transferNodesFromList in that case. However, now there are clients that
want to use ilist traits to cache instruction ordering information to
optimize dominance queries of instructions in the same basic block.
This change updates the existing ilist traits users to detect in-list
transfers and do nothing in that case.
After this change, we can start caching instruction ordering information
in LLVM IR data structures. There are two main ways to do that:
- by putting an order integer into the Instruction class
- by maintaining order integers in a hash table on BasicBlock
I plan to implement and measure both, but I wanted to commit this change
first to enable other out of tree ilist clients to implement this
optimization as well.
Reviewers: lattner, hfinkel, chandlerc
Subscribers: hiraditya, dexonsmith, llvm-commits
Differential Revision: https://reviews.llvm.org/D57120
llvm-svn: 351992
Casey Carter [Wed, 23 Jan 2019 22:49:44 +0000 (22:49 +0000)]
[test] Define _LIBCPP_AVAILABILITY_THROW_BAD_ANY_CAST in msvc_stdlib_force_include.hpp
...so the tests under test/std/utilities/any continue to
compile with MSVC's standard library.
While we're here, let's test >C++17 features when _HAS_CXX20.
llvm-svn: 351991
Hideki Saito [Wed, 23 Jan 2019 22:43:12 +0000 (22:43 +0000)]
[LV][VPlan] Change to implement VPlan based predication for
VPlan-native path
Context: Patch Series #2 for outer loop vectorization support in LV
using VPlan. (RFC:
http://lists.llvm.org/pipermail/llvm-dev/2017-December/119523.html).
Patch series #2 checks that inner loops are still trivially lock-step
among all vector elements. Non-loop branches are blindly assumed as
divergent.
Changes here implement VPlan based predication algorithm to compute
predicates for blocks that need predication. Predicates are computed
for the VPLoop region in reverse post order. A block's predicate is
computed as OR of the masks of all incoming edges. The mask for an
incoming edge is computed as AND of predecessor block's predicate and
either predecessor's Condition bit or NOT(Condition bit) depending on
whether the edge from predecessor block to the current block is true
or false edge.
Reviewers: fhahn, rengolin, hsaito, dcaballe
Reviewed By: fhahn
Patch by Satish Guggilla, thanks!
Differential Revision: https://reviews.llvm.org/D53349
llvm-svn: 351990
Peter Collingbourne [Wed, 23 Jan 2019 22:39:11 +0000 (22:39 +0000)]
hwasan: Read shadow address from ifunc if we don't need a frame record.
This saves a cbz+cold call in the interceptor ABI, as well as a realign
in both ABIs, trading off a dcache entry against some branch predictor
entries and some code size.
Unfortunately the functionality is hidden behind a flag because ifunc is
known to be broken on static binaries on Android.
Differential Revision: https://reviews.llvm.org/D57084
llvm-svn: 351989
Kamil Rytarowski [Wed, 23 Jan 2019 22:35:57 +0000 (22:35 +0000)]
Mark another test as flaky
Reported on the NetBSD 8 buildbot.
llvm-svn: 351988
Simon Atanasyan [Wed, 23 Jan 2019 22:02:53 +0000 (22:02 +0000)]
[mips] Handle MipsMCExpr sub-expression for the MEK_DTPREL tag
This is a fix for a regression introduced by the rL348194 commit. In
that change new type (MEK_DTPREL) of MipsMCExpr expression was added,
but in some places of the code this type of expression considered as
unexpected.
This change fixes the bug. The MEK_DTPREL type of expression is used for
marking TLS DIEExpr only and contains a regular sub-expression. Where we
need to handle the expression, we retrieve the sub-expression and
handle it in a common way.
llvm-svn: 351987
Sean Fertile [Wed, 23 Jan 2019 21:46:28 +0000 (21:46 +0000)]
Revert "[PPC64] Sort .toc sections accessed with small code model ..."
This reverts commit
ca87c57a3aa4770c9cf0defd4b2feccbc342ee93.
Added test fails on several windows buildbots.
llvm-svn: 351985
Kamil Rytarowski [Wed, 23 Jan 2019 21:45:02 +0000 (21:45 +0000)]
Correct mark for flaky tests
Add missing trailing dot.
llvm-svn: 351983
Reid Kleckner [Wed, 23 Jan 2019 21:10:48 +0000 (21:10 +0000)]
Revert r351938 "[ARM] Alter the register allocation order for minsize on Thumb2"
This change caused fatal backend errors when compiling a file in libvpx
for Android.
llvm-svn: 351979
Sean Fertile [Wed, 23 Jan 2019 21:04:03 +0000 (21:04 +0000)]
[PPC64] Sort .toc sections accessed with small code model relocs close to .got.
Small code model global variable access on PPC64 has a very limited range of
addressing. The instructions the relocations are used on add an offset in the
range [-0x8000, 0x7FFC] to the toc pointer which points to .got +0x8000, giving
an addressable range of [.got, .got + 0xFFFC]. While user code can be recompiled
with medium and large code models when the binary grows too large for small code
model, there are small code model relocations in the crt files and libgcc.a
which are typically shipped with the distros, and the ABI dictates that linkers
must allow linking of relocatable object files using different code models.
To minimze the chance of relocation overflow, any file that contains a small
code model relocation should have its .toc section placed closer to the .got
then any .toc from a file without small code model relocations.
Differential Revision: https://reviews.llvm.org/D56920
llvm-svn: 351978
Eugene Zelenko [Wed, 23 Jan 2019 20:51:06 +0000 (20:51 +0000)]
[Documentation] Fix problem in docs/SafeStack.rst introduced in r351976.
llvm-svn: 351977
Eugene Zelenko [Wed, 23 Jan 2019 20:39:07 +0000 (20:39 +0000)]
[Documentation] Use HTTPS whenever possible
Differential revision: https://reviews.llvm.org/D56946
llvm-svn: 351976
Alexey Bataev [Wed, 23 Jan 2019 18:59:54 +0000 (18:59 +0000)]
[DEBUGINFO, NVPTX] Enable support for the debug info on NVPTX target.
Enable full support for the debug info.
Differential revision: https://reviews.llvm.org/D46189
llvm-svn: 351974
Alexey Bataev [Wed, 23 Jan 2019 18:48:36 +0000 (18:48 +0000)]
Revert "[DEBUGINFO, NVPTX] Enable support for the debug info on NVPTX target."
This reverts commit r351972. Some pieces of the patch was not applied
correctly.
llvm-svn: 351973
Alexey Bataev [Wed, 23 Jan 2019 18:28:59 +0000 (18:28 +0000)]
[DEBUGINFO, NVPTX] Enable support for the debug info on NVPTX target.
Enable full support for the debug info. Recommit to fix the emission of
the not required closing brace.
Differential revision: https://reviews.llvm.org/D46189
llvm-svn: 351972
Marshall Clow [Wed, 23 Jan 2019 18:27:22 +0000 (18:27 +0000)]
Commit D11348: 'Win32 support: wcsnrtombs and mbsnrtowcs don't handle null output buffers correctly' which has been hanging around for a long time
llvm-svn: 351971
Craig Topper [Wed, 23 Jan 2019 18:25:49 +0000 (18:25 +0000)]
[X86] Autogenerate complete checks. NFC
llvm-svn: 351970
Aaron Ballman [Wed, 23 Jan 2019 18:02:17 +0000 (18:02 +0000)]
Merge similar target diagnostics for interrupt attribute into one; NFC
Patch by Kristina Bessonova!
llvm-svn: 351969
James Henderson [Wed, 23 Jan 2019 17:27:48 +0000 (17:27 +0000)]
[llvm-symbolizer] Improve compatibility of --functions with GNU addr2line
This fixes https://bugs.llvm.org/show_bug.cgi?id=40072.
GNU addr2line's --functions switch is off by default, has a short alias
of -f, and does not take an argument. This patch changes llvm-symbolizer
to allow the second and third point (changing the default behaviour may
have negative impacts on users). If the option is missing a value, it
now treats it as "linkage".
This change does cause one previously valid command-line to behave
differently. Before --functions <value> was accepted, but now only
--functions=<value> is allowed (as well as --functions). The old
behaviour will result in the value being treated as a positional
argument.
The previous testing for --functions=short has been pulled out into a
new test that also tests the other accepted values and option formats.
Reviewed by: ruiu
Differential Revision: https://reviews.llvm.org/D57049
llvm-svn: 351968
Haojian Wu [Wed, 23 Jan 2019 16:39:57 +0000 (16:39 +0000)]
Revert "[DEBUGINFO, NVPTX] Enable support for the debug info on NVPTX target."
This reverts commit r351846.
This patch may generate illegal assembly code, see
```
$ ./bin/clang -cc1 -triple nvptx64-nvidia-cuda -aux-triple x86_64-grtev4-linux-gnu -S -disable-free -disable-llvm-verifier -discard-value-names -main-file-name new.cc -mrelocation-model pic -pic-level 2 -mthread-model posix -fmerge-all-constants -mdisable-fp-elim -relaxed-aliasing -no-integrated-as -mpie-copy-relocations -munwind-tables -fcuda-is-device -target-feature +ptx60 -target-cpu sm_35 -dwarf-column-info -debug-info-kind=line-directives-only -dwarf-version=2 -debugger-tuning=gdb -o empty.s -x cuda empty.cc
$ cat empty.s
//
// Generated by LLVM NVPTX Back-End
//
.version 6.0
.target sm_35
.address_size 64
}
```
llvm-svn: 351966
Andrea Di Biagio [Wed, 23 Jan 2019 16:35:07 +0000 (16:35 +0000)]
[MC][X86] Correctly model additional operand latency caused by transfer delays from the integer to the floating point unit.
This patch adds a new ReadAdvance definition named ReadInt2Fpu.
ReadInt2Fpu allows x86 scheduling models to accurately describe delays caused by
data transfers from the integer unit to the floating point unit.
ReadInt2Fpu currently defaults to a delay of zero cycles (i.e. no delay) for all
x86 models excluding BtVer2. That means, this patch is only a functional change
for the Jaguar cpu model only.
Tablegen definitions for instructions (V)PINSR* have been updated to account for
the new ReadInt2Fpu. That read is mapped to the the GPR input operand.
On Jaguar, int-to-fpu transfers are modeled as a +6cy delay. Before this patch,
that extra delay was added to the opcode latency. In practice, the insert opcode
only executes for 1cy. Most of the actual latency is actually contributed by the
so-called operand-latency. According to the AMD SOG for family 16h, (V)PINSR*
latency is defined by expression f+1, where f is defined as a forwarding delay
from the integer unit to the fpu.
When printing instruction latency from MCA (see InstructionInfoView.cpp) and LLC
(only when flag -print-schedule is speified), we now need to account for any
extra forwarding delays. We do this by checking if scheduling classes declare
any negative ReadAdvance entries. Quoting a code comment in TargetSchedule.td:
"A negative advance effectively increases latency, which may be used for
cross-domain stalls". When computing the instruction latency for the purpose of
our scheduling tests, we now add any extra delay to the formula. This avoids
regressing existing codegen and mca schedule tests. It comes with the cost of an
extra (but very simple) hook in MCSchedModel.
Differential Revision: https://reviews.llvm.org/D57056
llvm-svn: 351965
James Henderson [Wed, 23 Jan 2019 16:15:39 +0000 (16:15 +0000)]
[llvm-readelf] Don't suppress static symbol table with --dyn-symbols + --symbols
In r287786, a bug was introduced into llvm-readelf where it didn't print
the static symbol table if both --symbols and --dyn-symbols were
specified, even if there was no dynamic symbol table. This is obviously
incorrect.
This patch fixes this issue, by delegating the decision of which symbol
tables should be printed to the final dumper, rather than trying to
decide in the command-line option handling layer. The decision was made
to follow the approach taken in this patch because the LLVM style dumper
uses a different order to the original GNU style behaviour (and GNU
readelf) for ELF output. Other approaches resulted in behaviour changes
for other dumpers which felt wrong. In particular, I wanted to avoid
changing the order of the output for --symbols --dyn-symbols for LLVM
style, keep what is emitted by --symbols unchanged for all dumpers, and
avoid having different orders of .dynsym and .symtab dumping for GNU
"--symbols" and "--symbols --dyn-symbols".
Reviewed by: grimar, rupprecht
Differential Revision: https://reviews.llvm.org/D57016
llvm-svn: 351960
Simon Pilgrim [Wed, 23 Jan 2019 16:01:19 +0000 (16:01 +0000)]
Fix indentation. NFCI.
llvm-svn: 351958
Simon Pilgrim [Wed, 23 Jan 2019 16:00:22 +0000 (16:00 +0000)]
[IR] Match intrinsic parameter by scalar/vectorwidth
This patch replaces the existing LLVMVectorSameWidth matcher with LLVMScalarOrSameVectorWidth.
The matching args must be either scalars or vectors with the same number of elements, but in either case the scalar/element type can differ, specified by LLVMScalarOrSameVectorWidth.
I've updated the _overflow intrinsics to demonstrate this - allowing it to return a i1 or <N x i1> overflow result, matching the scalar/vectorwidth of the other (add/sub/mul) result type.
The masked load/store/gather/scatter intrinsics have also been updated to use this, although as we specify the reference type to be llvm_anyvector_ty we guarantee the mask will be <N x i1> so no change in behaviour
Differential Revision: https://reviews.llvm.org/D57090
llvm-svn: 351957
Krzysztof Parzyszek [Wed, 23 Jan 2019 15:36:33 +0000 (15:36 +0000)]
[Hexagon] Remove incorrect bit negation
llvm-svn: 351956
Benjamin Kramer [Wed, 23 Jan 2019 14:51:21 +0000 (14:51 +0000)]
[AArch64] Fix out of bounds strlen
CFIInst is not zero-terminated. This is one of more annoying functional
differences between StringRef and ArrayRef.
Found by asan.
llvm-svn: 351955
Clement Courbet [Wed, 23 Jan 2019 14:20:59 +0000 (14:20 +0000)]
Re-land rL322538 "Add a value_type to ArrayRef."
llvm-svn: 351954
Simon Pilgrim [Wed, 23 Jan 2019 13:49:10 +0000 (13:49 +0000)]
Move saturated arithmetic intrinsics to other integer intrinsics. NFCI.
They were in the floating point group.
llvm-svn: 351953
Nico Weber [Wed, 23 Jan 2019 13:43:42 +0000 (13:43 +0000)]
Disable test better.
llvm-svn: 351952
George Rimar [Wed, 23 Jan 2019 13:39:12 +0000 (13:39 +0000)]
[llvm-objdump] - Move common code to a new printRelocation() helper. NFC.
This extracts the common code for printing relocations into a new helper function.
llvm-svn: 351951
Tim Renouf [Wed, 23 Jan 2019 13:38:06 +0000 (13:38 +0000)]
[AMDGPU] With XNACK, cannot clause a load with result coalesced with operand
Summary:
With XNACK, an smem load whose result is coalesced with an operand (thus
it overwrites its own operand) cannot appear in a clause, because some
other instruction might XNACK and restart the whole clause.
The clause breaker already realized that an smem that overwrites an
operand cannot appear in a clause, and broke the clause. The problem
that this commit fixes is that the SIFormMemoryClauses optimization
formed a bundle with early clobber, which caused the earlier code that
set up the coalesced operand to be removed as dead.
Differential Revision: https://reviews.llvm.org/D57008
Change-Id: I703c4d5b0bf7d6060222bec491f45c18bb3c0016
llvm-svn: 351950
Nico Weber [Wed, 23 Jan 2019 12:35:08 +0000 (12:35 +0000)]
Disable test added in r351916.
It doesn't pass on Windows:
http://lab.llvm.org:8011/builders/clang-x64-windows-msvc/builds/3627
FAIL: lld :: ELF/stdout.s (1521 of 1966)
******************** TEST 'lld :: ELF/stdout.s' FAILED ********************
Script:
--
: 'RUN: at line 3'; C:\b\slave\clang-x64-windows-msvc\build\build\stage1\bin\llvm-mc.EXE -filetype=obj -triple=x86_64-unknown-linux C:\b\slave\clang-x64-windows-msvc\build\llvm.src\tools\lld\test\ELF\stdout.s -o C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp.o
: 'RUN: at line 4'; c:\b\slave\clang-x64-windows-msvc\build\build\stage1\bin\ld.lld.EXE C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp.o -o - > C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp1
: 'RUN: at line 5'; C:\b\slave\clang-x64-windows-msvc\build\build\stage1\bin\llvm-objdump.EXE -d C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp1 | C:\b\slave\clang-x64-windows-msvc\build\build\stage1\bin\FileCheck.EXE C:\b\slave\clang-x64-windows-msvc\build\llvm.src\tools\lld\test\ELF\stdout.s
: 'RUN: at line 10'; c:\b\slave\clang-x64-windows-msvc\build\build\stage1\bin\ld.lld.EXE C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp.o -o C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp2
: 'RUN: at line 11'; diff C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp1 C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp2
--
Exit Code: 1
Command Output (stdout):
--
$ ":" "RUN: at line 3"
$ "C:\b\slave\clang-x64-windows-msvc\build\build\stage1\bin\llvm-mc.EXE" "-filetype=obj" "-triple=x86_64-unknown-linux" "C:\b\slave\clang-x64-windows-msvc\build\llvm.src\tools\lld\test\ELF\stdout.s" "-o" "C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp.o"
$ ":" "RUN: at line 4"
$ "c:\b\slave\clang-x64-windows-msvc\build\build\stage1\bin\ld.lld.EXE" "C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp.o" "-o" "-"
$ ":" "RUN: at line 5"
$ "C:\b\slave\clang-x64-windows-msvc\build\build\stage1\bin\llvm-objdump.EXE" "-d" "C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp1"
$ "C:\b\slave\clang-x64-windows-msvc\build\build\stage1\bin\FileCheck.EXE" "C:\b\slave\clang-x64-windows-msvc\build\llvm.src\tools\lld\test\ELF\stdout.s"
$ ":" "RUN: at line 10"
$ "c:\b\slave\clang-x64-windows-msvc\build\build\stage1\bin\ld.lld.EXE" "C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp.o" "-o" "C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp2"
$ ":" "RUN: at line 11"
$ "diff" "C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp1" "C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp2"
# command output:
*** C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp1
--- C:\b\slave\clang-x64-windows-msvc\build\build\stage1\tools\lld\test\ELF\Output\stdout.s.tmp2
***************
*** 1 ****
llvm-svn: 351949
Martin Storsjo [Wed, 23 Jan 2019 11:54:55 +0000 (11:54 +0000)]
[llvm-objcopy] [COFF] Error out on use of unhandled options
Prefer erroring out than silently not doing what was requested.
Differential Revision: https://reviews.llvm.org/D57045
llvm-svn: 351948
Martin Storsjo [Wed, 23 Jan 2019 11:54:51 +0000 (11:54 +0000)]
[llvm-objcopy] [COFF] Fix handling of aux symbols for big objects
The aux symbols were stored in an opaque std::vector<uint8_t>,
with contents interpreted according to the rest of the symbol.
All aux symbol types but one fit in 18 bytes (sizeof(coff_symbol16)),
and if written to a bigobj, two extra padding bytes are written (as
sizeof(coff_symbol32) is 20). In the storage agnostic intermediate
representation, store the aux symbols as a series of coff_symbol16
sized opaque blobs. (In practice, all such aux symbols only consist
of one aux symbol, so this is more flexible than what reality needs.)
The special case is the file aux symbols, which are written in
potentially more than one aux symbol slot, without any padding,
as one single long string. This can't be stored in the same opaque
vector of fixed sized aux symbol entries. The file aux symbols will
occupy a different number of aux symbol slots depending on the type
of output object file. As nothing in the intermediate process needs
to have accurate raw symbol indices, updating that is moved into the
writer class.
Differential Revision: https://reviews.llvm.org/D57009
llvm-svn: 351947
Martin Storsjo [Wed, 23 Jan 2019 11:54:36 +0000 (11:54 +0000)]
[llvm-objcopy] [COFF] Remove testcase debugging lines. NFC.
These are no longer necessary as the testcase now seems to run fine
on the buildbots that previously failed on this case, after SVN r351934.
llvm-svn: 351946
Florian Hahn [Wed, 23 Jan 2019 11:51:38 +0000 (11:51 +0000)]
[HotColdSplitting] Remove unused SSAUpdater.h include (NFC).
llvm-svn: 351945
Kamil Rytarowski [Wed, 23 Jan 2019 11:36:19 +0000 (11:36 +0000)]
Mark more tests flaky
Reported on the NetBSD 8 buildbot
llvm-svn: 351944
Ilya Biryukov [Wed, 23 Jan 2019 11:32:07 +0000 (11:32 +0000)]
[clangd] Workaround a test failure after r351941
This should fix failing buildbots.
llvm-svn: 351943
George Rimar [Wed, 23 Jan 2019 10:52:38 +0000 (10:52 +0000)]
[llvm-objdump] - Move variable. NFC.
It was too far from the place where it is used.
llvm-svn: 351942
Ilya Biryukov [Wed, 23 Jan 2019 10:35:12 +0000 (10:35 +0000)]
[clangd] Fix crash due to ObjCPropertyDecl
With ObjCPropertyDecl, ASTNode.OrigD can be a ObjCPropertyImplDecl
which is not a NamedDecl, leading to a crash since the code
incorrectly assumes ASTNode.OrigD will always be a NamedDecl.
Change by dgoldman (David Goldman)!
Differential Revision: https://reviews.llvm.org/D56916
llvm-svn: 351941
George Rimar [Wed, 23 Jan 2019 10:33:26 +0000 (10:33 +0000)]
[llvm-objdump] - Split disassembleObject() into two methods. NFCI.
Currently, disassembleObject() is a ~550 lines length function.
This patch splits it into two, where first do all helper objects initializations
and calls the second which does all the rest job.
This is a straightforward split.
Differential revision: https://reviews.llvm.org/D57020
llvm-svn: 351940
Jonas Paulsson [Wed, 23 Jan 2019 10:29:12 +0000 (10:29 +0000)]
[SystemZ] Fix test case for buildbot.
llvm-clang-x86_64-expensive-checks-win triggered this assert:
"llvm.dbg.value intrinsic requires a !dbg attachment"
Hopefully, adding reasonable !dbg operands solves this.
llvm-svn: 351939
David Green [Wed, 23 Jan 2019 10:18:30 +0000 (10:18 +0000)]
[ARM] Alter the register allocation order for minsize on Thumb2
Currently in Arm code, we allocate LR first, under the assumption that
it needs to be saved anyway. Unfortunately this has the disadvantage
that it will require any instructions using it to be the longer thumb2
instructions, not the shorter thumb1 ones.
This switches the order when we are optimising for minsize, returning to
the default order so that more lower registers can be used. It can end
up requiring more pushed registers, but on average produces smaller code.
Differential Revision: https://reviews.llvm.org/D56008
llvm-svn: 351938