Commit Graph

355699 Commits

Author SHA1 Message Date
Louis Dionne f46bb9dd5c [NFC] Reformat TEST_FOO macros in test_macros.h
To make them easier to read and to make it easier to add new ones.
2020-05-27 16:54:43 -04:00
Jonas Devlieghere f9bea9bc4a [lldb/Reproducers] Skip & add FIXME to tests failing with unexpected packet.
Add skip decorator to tests failing with an unexpected packet during
passive replay.
2020-05-27 13:52:48 -07:00
Jonas Devlieghere 5f97a540ad [lldb/Reproducers] Differentiate active and passive replay unexpected packet. 2020-05-27 13:52:38 -07:00
Sean Silva 25132b36a8 [mlir][shape] Use IndexElementsAttr in Shape dialect.
Summary:
Index is the proper type for storing shapes when constant folding, so
this fixes the previous code (which was using i64).

Differential Revision: https://reviews.llvm.org/D80600
2020-05-27 13:39:49 -07:00
Sean Silva 9546d8b108 [mlir][core] Add IndexElementsAttr helpers.
Summary:
In a follow-up, I'll update the Shape dialect to use this instead of
I64ElementsAttr.

Differential Revision: https://reviews.llvm.org/D80601
2020-05-27 13:39:48 -07:00
Mircea Trofin 98ef93eabd [llvm] Add function feature extraction analysis
Summary:
This patch introduces an analysis pass to extract function features,
which will be needed by the ML InlineAdvisor.

RFC: http://lists.llvm.org/pipermail/llvm-dev/2020-April/140763.html

Reviewers: davidxl, dblaikie, jdoerfert

Subscribers: mgorny, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D80579
2020-05-27 13:38:50 -07:00
Michael Liao fa342b5c80 Enable `align <n>` to be used in the intrinsic definition.
- This allow us to specify the (minimal) alignment on an intrinsic's
  arguments and, more importantly, the return value.

Differential Revision: https://reviews.llvm.org/D80422
2020-05-27 16:38:18 -04:00
Michael Liao 03481287ca Refactor argument attribute specification in intrinsic definition. NFC.
- Argument attribute needs specifiying through `ArgIndex<n>`
  (corresponding to `FirstArgIndex`) to distinguish explicitly from the
  index number from the overloaded type list.
- In addition, `RetIndex` (corresponding to `ReturnIndex`) and
  `FuncIndex` (corresponding to `FunctionIndex`) are introduced for us
  to associate attributes on the return value and potentially function
  itself.

Differential Revision: https://reviews.llvm.org/D80422
2020-05-27 16:37:53 -04:00
Vitaly Buka 804a39a201 [NFC,StackSafety] Rename some variables 2020-05-27 13:33:28 -07:00
Vitaly Buka 14f3357586 [StackSafety] Bailout more aggressively
Many edge cases, e.g. wrapped ranges, can be processed
precisely without bailout. However it's very unlikely that
memory access with min/max integer offsets will be
classified as safe anyway.
Early bailout may help with ThinLTO where we can
drop unsafe parameters from summaries.
2020-05-27 13:33:28 -07:00
Sean Silva b277382311 Remove error-prone mlir::ExecutionEngine::invoke overload.
I just spent a bunch of time debugging a mysterious bug that ended being due to my SmallVector getting passed to the Args&... overload instead of the MutableArrayRef overload, with disastrous results.

I appreciate the intent of this API, but for a function that does a bunch of unsafe casts, adding in potential overload confusion is just too much C++ footgun. If we end up needing this functionality, having something like a separate `packArgs(Args&...) -> SmallVector` overload would be preferable.

Turns out this API is unused and untested (even out of tree as far as I can tell, modulo the optional passing of no args to the other invoke as I fixed in this patch), so it's an easy fix -- just delete it and touch up the other overload.

Differential Revision: https://reviews.llvm.org/D80607
2020-05-27 13:26:03 -07:00
Juneyoung Lee 54b6457240 [TargetPassConfig] Add CanonicalizeFreezeInLoops before LSR
Summary:
This patch adds CanonicalizeFreezeInLoops before LSR.
Relevant patch: https://reviews.llvm.org/D77523

Reviewers: spatel, efriedma, jdoerfert, fhahn, nikic, reames, xbolva00

Reviewed By: nikic

Subscribers: xbolva00, nikic, lebedev.ri, hiraditya, llvm-commits, sanwou01, nlopes

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D77524
2020-05-28 05:21:12 +09:00
Nicolas Vasilache 79aa9bfdb8 [mlir] Fix RunnerUtils template specialization
Undoing a spurious change that broke SFINAE for some out of core use
cases.
2020-05-27 16:14:43 -04:00
MaheshRavishankar c6fa2efd48 [mlir][Linalg] Fix build failure from D80188
Differential Revision: https://reviews.llvm.org/D80657
2020-05-27 13:06:43 -07:00
Michael Liao 49688b3c30 Fix `-Wpedantic` warning. NFC. 2020-05-27 15:57:03 -04:00
Jessica Paquette c593bf5342 [GlobalISel] Don't combine instructions which are fed by memory instructions.
If we have a memory instruction (e.g. a load), we shouldn't combine it away in
some trivial combine.

It's possible that, say, a call lives between the instructions. This could
modify the value loaded, making the load instructions not safe to fold.

Differential Revision: https://reviews.llvm.org/D80053
2020-05-27 12:48:58 -07:00
Dmitry Vyukov d24dd2b279 tsan: fix test in debug mode
sanitizer-x86_64-linux-autoconf has failed after the previous tsan commit:

FAIL: ThreadSanitizer-x86_64 :: java_finalizer2.cpp (245 of 403)
******************** TEST 'ThreadSanitizer-x86_64 :: java_finalizer2.cpp' FAILED ********************
Script:
--
: 'RUN: at line 1';      /b/sanitizer-x86_64-linux-autoconf/build/tsan_debug_build/./bin/clang  --driver-mode=g++ -fsanitize=thread -Wall  -m64   -gline-tables-only -I/b/sanitizer-x86_64-linux-autoconf/build/llvm-project/compiler-rt/test/tsan/../ -std=c++11 -I/b/sanitizer-x86_64-linux-autoconf/build/llvm-project/compiler-rt/test/tsan/../ -nostdinc++ -I/b/sanitizer-x86_64-linux-autoconf/build/tsan_debug_build/tools/clang/runtime/compiler-rt-bins/lib/tsan/libcxx_tsan_x86_64/include/c++/v1 -O1 /b/sanitizer-x86_64-linux-autoconf/build/llvm-project/compiler-rt/test/tsan/java_finalizer2.cpp -o /b/sanitizer-x86_64-linux-autoconf/build/tsan_debug_build/tools/clang/runtime/compiler-rt-bins/test/tsan/X86_64Config/Output/java_finalizer2.cpp.tmp &&  /b/sanitizer-x86_64-linux-autoconf/build/tsan_debug_build/tools/clang/runtime/compiler-rt-bins/test/tsan/X86_64Config/Output/java_finalizer2.cpp.tmp 2>&1 | FileCheck /b/sanitizer-x86_64-linux-autoconf/build/llvm-project/compiler-rt/test/tsan/java_finalizer2.cpp
--
Exit Code: 1

Command Output (stderr):
--
/b/sanitizer-x86_64-linux-autoconf/build/llvm-project/compiler-rt/test/tsan/java_finalizer2.cpp:82:11: error: CHECK: expected string not found in input
// CHECK: DONE
          ^
<stdin>:1:1: note: scanning from here
FATAL: ThreadSanitizer CHECK failed: /b/sanitizer-x86_64-linux-autoconf/build/llvm-project/compiler-rt/lib/tsan/rtl/tsan_sync.cpp:69 "((*meta)) == ((0))" (0x4000003e, 0x0)
^
<stdin>:5:12: note: possible intended match here
 #3 __tsan::OnUserAlloc(__tsan::ThreadState*, unsigned long, unsigned long, unsigned long, bool) /b/sanitizer-x86_64-linux-autoconf/build/llvm-project/compiler-rt/lib/tsan/rtl/tsan_mman.cpp:225:16 (java_finalizer2.cpp.tmp+0x4af407)
           ^

http://lab.llvm.org:8011/builders/sanitizer-x86_64-linux-autoconf/builds/51143/steps/test%20tsan%20in%20debug%20compiler-rt%20build/logs/stdio

Fix heap object overlap by offsetting java heap as other tests are doing.
2020-05-27 21:48:39 +02:00
alex-t eb1092ada3 [AMDGPU] Fix for the lost CarryOut/CarryIn register operands in S_ADD/SUB_CO_PSEUDO.
Summary: This fixes the 5b898bddff bug when the carry-in and carry-out registers became lost in lowering S_ADD/SUB_CO_PSEUDO.

Reviewers: rampitec, arsenm

Reviewed By: arsenm

Subscribers: msearles, arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, kerbowa, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D80158
2020-05-27 22:41:04 +03:00
Adrian Prantl 3345521507 Also cache negative results in GetXcodeSDKPath (NFC)
This fixes a performance issue in the failure case.

rdar://63547920

Differential Revision: https://reviews.llvm.org/D80595
2020-05-27 12:26:04 -07:00
Jonas Devlieghere c30c2368c7 [lldb/Reproducers] Skip tests relying on timeouts
The reproducer don't model timeouts so tests that rely on them end up
with unexpected packets during replay. Skip them until we can handle
this scenario.
2020-05-27 12:08:41 -07:00
Jonas Devlieghere fe9d8442e0 [lldb/Test] Generate YAML binary in build directory
Although it's not entirely clear to me why, this test was generating its
binary in the source directory instead of the build directory. This
patch fixes that following the same approach as other tests.
2020-05-27 12:08:41 -07:00
Craig Topper 8e7e6a8d6b [X86] Restore selection of MULX on BMI2 targets.
Looking back over gcc and icc behavior it looks like icc does
use mulx32 on 32-bit targets and mulx64 on 64-bit targets. It's
also used when dividing i32 by constant on 32-bit targets and
i64 by constant on 64-bit targets.

gcc uses it multiplies producing a 64 bit result on 32-bit targets
and 128-bit results on a 64-bit target. gcc does not appear to use
it for division by constant.

After this patch clang is closer to the icc behavior. This
basically reverts d1c61861dd, but
there were no strong feelings at the time.

Fixes PR45518.

Differential Revision: https://reviews.llvm.org/D80498
2020-05-27 12:01:18 -07:00
Mircea Trofin fa3b587196 [llvm]NFC] Simplify ProfileSummaryInfo state transitions
ProfileSummaryInfo is updated seldom, as result of very specific
triggers. This patch clearly demarcates state updates from read-only uses.
This, arguably, improves readability and maintainability.
2020-05-27 11:58:37 -07:00
Sanjay Patel 48cb380abd [InstCombine] add tests for vector demanded elements of select condition; NFC 2020-05-27 14:49:36 -04:00
Matt Arsenault 4b4496312e AMDGPU: Start adding MODE register uses to instructions
This is the groundwork required to implement strictfp. For now, this
should be NFC for regular instructoins (many instructions just gain an
extra use of a reserved register). Regalloc won't rematerialize
instructions with reads of physical registers, but we were suffering
from that anyway with the exec reads.

Should add it for all the related FP uses (possibly with some
extras). I did not add it to either the gpr index mode instructions
(or every single VALU instruction) since it's a ridiculous feature
already modeled as an arbitrary side effect.

Also work towards marking instructions with FP exceptions. This
doesn't actually set the bit yet since this would start to change
codegen. It seems nofpexcept is currently not implied from the regular
IR FP operations. Add it to some MIR tests where I think it might
matter.
2020-05-27 14:47:00 -04:00
John Fastabend 13f6c81c5d [BPF] simplify zero extension with MOV_32_64
The current pattern matching for zext results in the following code snippet
being produced,

  w1 = w0
  r1 <<= 32
  r1 >>= 32

Because BPF implementations require zero extension on 32bit loads this
both adds a few extra unneeded instructions but also makes it a bit
harder for the verifier to track the r1 register bounds. For example in
this verifier trace we see at the end of the snippet R2 offset is unknown.
However, if we track this correctly we see w1 should have the same bounds
as r8. R8 smax is less than U32 max value so a zero extend load should keep
the same value. Adding a max value of 800 (R8=inv(id=0,smax_value=800)) to
an off=0, as seen in R7 should create a max offset of 800. However at the
end of the snippet we note the R2 max offset is 0xffffFFFF.

  R0=inv(id=0,smax_value=800)
  R1_w=inv(id=0,umax_value=2147483647,var_off=(0x0; 0x7fffffff))
  R6=ctx(id=0,off=0,imm=0) R7=map_value(id=0,off=0,ks=4,vs=1600,imm=0)
  R8_w=inv(id=0,smax_value=800,umax_value=4294967295,var_off=(0x0; 0xffffffff))
  R9=inv800 R10=fp0 fp-8=mmmm????
 58: (1c) w9 -= w8
 59: (bc) w1 = w8
 60: (67) r1 <<= 32
 61: (77) r1 >>= 32
 62: (bf) r2 = r7
 63: (0f) r2 += r1
 64: (bf) r1 = r6
 65: (bc) w3 = w9
 66: (b7) r4 = 0
 67: (85) call bpf_get_stack#67
  R0=inv(id=0,smax_value=800)
  R1_w=ctx(id=0,off=0,imm=0)
  R2_w=map_value(id=0,off=0,ks=4,vs=1600,umax_value=4294967295,var_off=(0x0; 0xffffffff))
  R3_w=inv(id=0,umax_value=800,var_off=(0x0; 0x3ff))
  R4_w=inv0 R6=ctx(id=0,off=0,imm=0)
  R7=map_value(id=0,off=0,ks=4,vs=1600,imm=0)
  R8_w=inv(id=0,smax_value=800,umax_value=4294967295,var_off=(0x0; 0xffffffff))
  R9_w=inv(id=0,umax_value=800,var_off=(0x0; 0x3ff))
  R10=fp0 fp-8=mmmm????

After this patch R1 bounds are not smashed by the <<=32 >>=32 shift and we
get correct bounds on R2 umax_value=800.

Further it reduces 3 insns to 1.

Signed-off-by: John Fastabend <john.fastabend@gmail.com>

Differential Revision: https://reviews.llvm.org/D73985
2020-05-27 11:26:39 -07:00
Lei Huang 2368bf52cd [PowerPC] Add support for -mcpu=pwr10 in both clang and llvm
Summary:
This patch simply adds support for the new CPU in anticipation of
Power10. There isn't really any functionality added so there are no
associated test cases at this time.

Reviewers: stefanp, nemanjai, amyk, hfinkel, power-llvm-team, #powerpc

Reviewed By: stefanp, nemanjai, amyk, #powerpc

Subscribers: NeHuang, steven.zhang, hiraditya, llvm-commits, wuzish, shchenz, cfe-commits, kbarton, echristo

Tags: #clang, #powerpc, #llvm

Differential Revision: https://reviews.llvm.org/D80020
2020-05-27 13:14:25 -05:00
aartbik c295a65da4 [mlir] [VectorOps] Add 'vector.flat_transpose' operation
Summary:
Provides a representation of the linearized LLVM instrinsic.
With tests and lowering implementation to LLVM IR dialect.
Prepares better lowering for 2-D vector.transpose.

Reviewers: nicolasvasilache, ftynse, reidtatge, bkramer, dcaballe

Reviewed By: ftynse, dcaballe

Subscribers: mehdi_amini, rriddle, jpienaar, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, liufengdb, stephenneuendorffer, Joonsoo, grosul1, frgossen, Kayjukh, jurahul, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D80419
2020-05-27 11:09:48 -07:00
Rithik Sharma eadf295956 [CodeMoverUtils] Use dominator tree level to decide the direction of
code motion

Summary: Currently isSafeToMoveBefore uses DFS numbering for determining
the relative position of instruction and insert point which is not
always correct. This PR proposes the use of Dominator Tree depth for the
same. If a node is at a higher level than the insert point then it is
safe to say that we want to move in the forward direction.
Authored By: RithikSharma
Reviewer: Whitney, nikic, bmahjour, etiotto, fhahn
Reviewed By: Whitney
Subscribers: fhahn, hiraditya, llvm-commits
Tag: LLVM
Differential Revision: https://reviews.llvm.org/D80084
2020-05-27 18:02:06 +00:00
Fangrui Song b9c6871a95 [Driver] Support -fsanitize=shadow-call-stack and cfi-icall on aarch64_be
D80647 did not fix https://bugs.llvm.org/show_bug.cgi?id=46076
This is the fix.
2020-05-27 10:55:05 -07:00
jasonliu 8d9ff23185 [NFC][XCOFF][AIX] Return function entry point symbol with dedicate function
Use getFunctionEntryPointSymbol whenever possible to enclose the
implementation detail and reduce duplicate logic.

Differential Revision: https://reviews.llvm.org/D80402
2020-05-27 17:54:22 +00:00
Matt Arsenault d37ce53ad3 AMDGPU: Set StackPointerRegisterToSaveRestore
This will enable selecting non-entry block allocas. Skip the SP write
check in the base isSchedulingBoundary implementation to preserve the
previous scheduling behavior and avoid test churn. It's apparently for
compile time reasons, but if we were to use this more work would be
needed since in some of the failing tests, we seem to incorrectly get
hazard nops inserted.
2020-05-27 13:44:05 -04:00
Fangrui Song a2a3e9f0a6 [Driver] Support -fsanitize=shadow-call-stack on aarch64_be
Fixes https://bugs.llvm.org/show_bug.cgi?id=46076

Reviewed By: nickdesaulniers, pcc

Differential Revision: https://reviews.llvm.org/D80647
2020-05-27 10:37:39 -07:00
Daniel Martín 6407aa9d2e
[clangd] Add access specifier information to hover contents
Summary:
For https://github.com/clangd/clangd/issues/382

This commit adds access specifier information to the hover
contents. For example, the hover information of a class field or
member function will now indicate if the field or member is private,
public, or protected. This can be particularly useful when a developer
is in the implementation file and wants to know if a particular member
definition is public or private.

Reviewers: kadircet

Reviewed By: kadircet

Subscribers: ilya-biryukov, MaskRay, jkorous, arphaman, kadircet, usaxena95, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D80472
2020-05-27 19:36:46 +02:00
Jonas Devlieghere e7f1067ad6 [lldb/Reproducers] Skip API logging in the DUMMY macro
The purpose of the LLDB_RECORD_DUMMY macro is twofold: it is used in
functions that take arguments that we don't know how to serialize (e.g.
void*) and it's used by function where we want to avoid doing excessive
work because they can be called from a signal handler (e.g.
setTerminalWidth).

To support the latter case, I've disabled API logging form the Recorder
ctor used by the DUMMY macro. This ensures we don't allocate memory when
called from a signal handler.
2020-05-27 10:35:43 -07:00
Matt Arsenault 07cd19efa2 AMDGPU: Fix dropping MI flags when rewriting instructions
All 3 passes that change instruction encodings were dropping MI
flags. This avoids scheduling regressions caused by setting
mayRaiseFPExceptions on FP instructions for non-strictfp functions.
2020-05-27 13:27:06 -04:00
Raphael Isemann 74a51753a6 [lldb] Make order of completions for expressions deterministic and sorted by Clang's priority values.
Summary:

It turns out that the order in which we provide completions for expressions is
nondeterministic. This leads to confusing user experience and also breaks the
reproducer tests (as two LLDB tests can go out of sync due to the
non-determinism in the completion lists)

The reason for the non-determinism is that the CompletionConsumer informs us
about decls in the order in which it finds declarations in the lookup store of
the DeclContexts it visits (mainly this snippet in SemaLookup.cpp):

``` lang=c++
    // Enumerate all of the results in this context.
    for (DeclContextLookupResult R :
         Load ? Ctx->lookups()
              : Ctx->noload_lookups(/*PreserveInternalState=*/false)) {
       [...]
```

This storage of the lookup is sorted by pointer values (see the hash of
`DeclarationName`) and can therefore be non-deterministic. The LLDB code
completion consumer that receives these calls originally expected that the order
of declarations is defined by Clang, but it seems the API expects the client to
provide an order to the completions.

This patch fixes the issue as follows:

* We sort the completions we get from Clang alphabetically and also by the
priority value we get from Clang (with priority value sorting having precedence
over the alphabetical sorting)

* We make all the functions/variables that touch a completion before the sorting
const-qualified. The idea is that this should prevent that we never have
observable side-effect from touching these declarations in a non-deterministic
order (e.g., we don't try to complete the type by accident).

This way we behave like the other parts of Clang which also sort the results by
some deterministic value (usually the name or something computed from a name,
e.g., edit distance to a given string).

We most likely also need to fix the Clang code to make the loop I listed above
deterministic to prevent these issues in the future (tracked in rdar://63442513
). This wouldn't replace the functionality provided in this patch though as we
would still need the priority and overall alphabetical sorting.

Note: I had to increase the lldb-vscode completion limit to 100 as the tests
look for strings that aren't in the first 50 results anymore due to variable
names starting with letters like 'v' (which are now always shown much further
down in the list due to the alphabetical sorting).

Fixes rdar://63200995

Reviewers: JDevlieghere, clayborg

Reviewed By: JDevlieghere

Subscribers: mgrang, abidh

Differential Revision: https://reviews.llvm.org/D80292
2020-05-27 19:22:01 +02:00
Fangrui Song 5b4cd2d4c4 [X86] Assemble movzb 1280(%rbx, %r12), %r12 after D80608
ffmpeg/libavcodec/x86/h264_cabac.c inline assembly may produce
movzb 1280(%rbx, %r12), %r12

After D80608, llvm-mc errors:

error: unknown use of instruction mnemonic without a size suffix
2020-05-27 09:55:55 -07:00
MaheshRavishankar 4d6f44f5f0 [mlir][spirv] Lower allocation/deallocations of workgroup memory.
This allocation of a workgroup memory is lowered to a
spv.globalVariable. Only static size allocation with element type
being int or float is handled. The lowering does account for the
element type that are not supported in the lowered spv.module based on
the extensions/capabilities and adjusts the number of elements to get
the same byte length.

Differential Revision: https://reviews.llvm.org/D80411
2020-05-27 09:53:16 -07:00
Ties Stuij 29f8056b54 [CodeGen] fix typo `def nxv1bf32` -> `def nxv1f32`
The `Add bfloat MVT type` patch introduced a typo in the nxv1f32 definition
in llvm/include/llvm/CodeGen/ValueTypes.td:
https://reviews.llvm.org/D79706/new/#inline-740433

This patch fixes that.
2020-05-27 17:52:07 +01:00
LLVM GN Syncbot bed78845e5 [gn build] Port 0d20ed664f 2020-05-27 16:41:00 +00:00
Bardia Mahjour 0d20ed664f [DDG] Data Dependence Graph - Add query function for memory dependencies between two nodes
Summary:
When working with the DDG it's useful to be able to query details of the
memory dependencies between two nodes connected by a memory edge. The DDG
does not hold a copy of the dependencies, but it contains a reference to a
DependenceInfo object through which dependence information can be queried.
This patch adds a query function to the DDG to obtain all the Dependence
objects that exist between instructions of two nodes.

Authored By: bmahjour

Reviewers: Meinersbur, Whitney, etiotto

Reviewed By: Whitney

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D80529
2020-05-27 12:40:14 -04:00
Nico Weber 4f0eba28eb [gn build] (manually) port dedaf3a2ac 2020-05-27 12:31:59 -04:00
David Truby 5ba874e472 [MLIR] [OpenMP] Add basic OpenMP parallel operation
Summary:
This includes a basic implementation for the OpenMP parallel
operation without a custom pretty-printer and parser.
The if, num_threads, private, shared, first_private, last_private,
proc_bind and default clauses are included in this implementation.

Currently the reduction clause is omitted as it is more complex and
requires analysis to see if we can share implementation with the loop
dialect. The allocate clause is also omitted.

A discussion about the design of this operation can be found here:
https://llvm.discourse.group/t/openmp-parallel-operation-design-issues/686

The current OpenMP Specification can be found here:
https://www.openmp.org/wp-content/uploads/OpenMP-API-Specification-5.0.pdf

Co-authored-by: Kiran Chandramohan <kiran.chandramohan@arm.com>

Reviewers: jdoerfert

Subscribers: mgorny, yaxunl, kristof.beyls, guansong, mehdi_amini, rriddle, jpienaar, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, Joonsoo, grosul1, frgossen, Kayjukh, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D79410
2020-05-27 17:16:44 +01:00
Philip Reames 1af3705c7f Start migrating away from statepoint's inline length prefixed argument bundles
In the current statepoint design, we have four distinct groups of operands to the call: call args, gc transition args, deopt args, and gc args. This format prexisted the support in IR for operand bundles and was in fact one of the inspirations for the extension. However, we never went back and rearchitected statepoints to fully leverage bundles.

This change is the first in a small sequence to do so. All this does is extend the SelectionDAG lowering code to allow deopt and gc transition operands to be specified in either inline argument bundles or operand bundles.

Differential Revision: https://reviews.llvm.org/D8059
2020-05-27 09:16:10 -07:00
Paul Walker 495f18292b [VFABI] Fix parsing of uniform parameters that shouldn't expect step or positional data.
Subscribers: hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D80575
2020-05-27 16:07:45 +00:00
Michael Liao b040468117 Fix warning `-Wpedantic`. NFC. 2020-05-27 12:05:55 -04:00
James Y Knight aca3d067ef Fix Darwin 'constinit thread_local' variables.
Unlike other platforms using ItaniumCXXABI, Darwin does not allow the
creation of a thread-wrapper function for a variable in the TU of
users. Because of this, it can set the linkage of the thread-local
symbol to internal, with the assumption that no TUs other than the one
defining the variable will need it.

However, constinit thread_local variables do not require the use of
the thread-wrapper call, so users reference the variable
directly. Thus, it must not be converted to internal, or users will
get a link failure.

This was a regression introduced by the optimization in
00223827a9.

Differential Revision: https://reviews.llvm.org/D80417
2020-05-27 11:59:30 -04:00
Simon Pilgrim 6022efb0e9 CoverageFilters.h - reduce unnecessary includes to forward declarations. NFC. 2020-05-27 16:57:47 +01:00
Jacques Pienaar 31f40f603d [mlir] Add simple generator for return types
Take advantage of equality constrains to generate the type inference interface.
This is used for equality and trivially built types. The type inference method
is only generated when no type inference trait is specified already.

This reorders verification that changes some test error messages.

Differential Revision: https://reviews.llvm.org/D80484
2020-05-27 08:45:55 -07:00