Koakuma [Sat, 12 Aug 2023 22:28:26 +0000 (18:28 -0400)]
[SPARC][IAS] Add support for v9 DONE, RETRY, SAVED, & RESTORED
Add support for DONE, RETRY, SAVED, and RESTORED (v9 Section A.11 & Section A.47).
Those instructions are used for low-level interrupt handling and register window
management by OS kernels.
Reviewed By: barannikov88
Differential Revision: https://reviews.llvm.org/D144936
(cherry picked from commit
bf499ec2b96c98633a8b2bc2113b19ffd437f647)
Simon Pilgrim [Mon, 21 Aug 2023 10:21:59 +0000 (11:21 +0100)]
[DAG] replaceStoreOfInsertLoad - don't fold if the inserted element is implicitly truncated
D152276 wasn't handling the case where the inserted element is implicitly truncated into the vector - resulting in a i1 element (implicitly truncated from i8) overwriting 8 bits instead of 1 bit.
This patch is intended to be merged into 17.x so I've just disallowed any vector element vs inserted element type mismatch - technically we could be more elegant and permit truncated stores (as long as the store is still byte sized), but the use cases for that are so limited I'd prefer to play it safe for now.
Candidate patch for #64655 17.x merge
Differential Revision: https://reviews.llvm.org/D158366
(cherry picked from commit
ba818c4019c550e1a413e1563a05b241b508defd)
Simon Pilgrim [Sun, 20 Aug 2023 14:34:47 +0000 (15:34 +0100)]
[X86] Add test case for Issue #64655
(cherry picked from commit
2c090e9e67d306578dbeb966a34d9fb85b9f4121)
wangpc [Mon, 21 Aug 2023 10:56:18 +0000 (18:56 +0800)]
[RISCV] Support global address as inline asm memory operand of `m`
In D146245, we have supported lowering inline asm `m` with offset
to `register+imm`, but we didn't handle the case that the offset
is the low part of global address.
This patch will emit `%lo(g)` when `g` is a global address.
Fixes #64656
Reviewed By: asb
Differential Revision: https://reviews.llvm.org/D157839
(cherry picked from commit
dc60003ec8b2faf595b528a39f64b697a589da06)
wangpc [Mon, 21 Aug 2023 10:56:00 +0000 (18:56 +0800)]
[RISCV][NFC] Move tests of inline asm memory constraints to separate file
We will need to check the output of medium code model.
Reviewed By: wangpc
Differential Revision: https://reviews.llvm.org/D157965
(cherry picked from commit
a3b11ce78680ef39e4dfbc72085ae3e1814e3cba)
Martin Storsjö [Fri, 18 Aug 2023 12:23:44 +0000 (15:23 +0300)]
[AArch64] [GlobalISel] Fix clobbered callee saved registers with win64 varargs
This fixes a regression since
1c10d5b175992a9d056a2d763a932e5652386fc1
/ https://reviews.llvm.org/D130903 by applying the same fix from
SelectionDAG from
8cb3667541a94c4fa11b06e19020f753414c1d03 /
https://reviews.llvm.org/D35720.
This could possibly have been detected if the existing testcases
in win64_vararg.ll had been tested with GlobalISel too, but all
the IR snippets there fail to be translated with GlobalISel.
This adds a separate testcase based on real world LLVM IR (instead of
hand-reduced IR), which GlobalISel does translate happily - tested
with both SelectionDAG and GlobalISel.
Before this change, the stack object locations (visible in MIR
with "llc -print-after-all") didn't match with what the prologue
emitted by AArch64FrameLowering actually looked like, which caused
clobbered callee saved registers when function local stack objects
aliased the actual location of the callee saved registers.
This fixes https://github.com/llvm/llvm-project/issues/64740.
Differential Revision: https://reviews.llvm.org/D158272
(cherry picked from commit
955d7615bd7563cc78a5106215daf9e6e47ffb5e)
Aaron Ballman [Mon, 21 Aug 2023 12:25:58 +0000 (08:25 -0400)]
Remove a release note that should not have snuck in
Aaron Ballman [Mon, 21 Aug 2023 11:47:02 +0000 (07:47 -0400)]
Remove a merge conflict marker that was missed; NFC
Aaron Ballman [Thu, 17 Aug 2023 11:10:00 +0000 (07:10 -0400)]
No longer hang on typeof of a function type
We were calling `isFunctionProtoType()` on a `ParsedType` rather than
creating a valid semantic type first and calling the function on that.
The call to `isFunctionProtoType()` would eventually call
`getUnqualifiedDesugaredType()`, which loops indefinitely until we get
a desugared type and a `ParsedType` will never finish desugaring.
Fixes https://github.com/llvm/llvm-project/issues/64713
Eduard Zingerman [Sun, 13 Aug 2023 00:11:37 +0000 (03:11 +0300)]
[BPF] Reset machine register kill mark in BPFMISimplifyPatchable
When LLVM is build with `LLVM_ENABLE_EXPENSIVE_CHECKS=ON` option
the following C code snippet:
struct t {
unsigned long a;
} __attribute__((preserve_access_index));
void foo(volatile struct t *t, volatile unsigned long *p) {
*p = t->a;
*p = t->a;
}
Causes an assertion:
$ clang -g -O2 -c --target=bpf -mcpu=v2 t2.c -o /dev/null
# After BPF PreEmit SimplifyPatchable
# Machine code for function foo: IsSSA, TracksLiveness
Function Live Ins: $r1 in %0, $r2 in %1
bb.0.entry:
liveins: $r1, $r2
DBG_VALUE $r1, $noreg, !"t", !DIExpression()
DBG_VALUE $r2, $noreg, !"p", !DIExpression()
%1:gpr = COPY $r2
DBG_VALUE %1:gpr, $noreg, !"p", !DIExpression()
%0:gpr = COPY $r1
DBG_VALUE %0:gpr, $noreg, !"t", !DIExpression()
%2:gpr = LD_imm64 @"llvm.t:0:0$0:0"
%4:gpr = ADD_rr %0:gpr(tied-def 0), killed %2:gpr
%5:gpr = CORE_LD 344, %0:gpr, @"llvm.t:0:0$0:0"
STD killed %5:gpr, %1:gpr, 0
%7:gpr = ADD_rr %0:gpr(tied-def 0), killed %2:gpr
%8:gpr = CORE_LD 344, %0:gpr, @"llvm.t:0:0$0:0"
STD killed %8:gpr, %1:gpr, 0
RET
# End machine code for function foo.
*** Bad machine code: Using a killed virtual register ***
- function: foo
- basic block: %bb.0 entry (0x6210000e6690)
- instruction: %7:gpr = ADD_rr %0:gpr(tied-def 0), killed %2:gpr
- operand 2: killed %2:gpr
This happens because of the way
BPFMISimplifyPatchable::processDstReg() updates second operand of the
`ADD_rr` instruction. Code before `BPFMISimplifyPatchable`:
.-> %2:gpr = LD_imm64 @"llvm.t:0:0$0:0"
|
|`----------------.
| %3:gpr = LDD %2:gpr, 0
| %4:gpr = ADD_rr %0:gpr(tied-def 0), killed %3:gpr <--- (1)
| %5:gpr = LDD killed %4:gpr, 0 ^^^^^^^^^^^^^
| STD killed %5:gpr, %1:gpr, 0 this is updated
`----------------.
%6:gpr = LDD %2:gpr, 0
%7:gpr = ADD_rr %0:gpr(tied-def 0), killed %6:gpr <--- (2)
%8:gpr = LDD killed %7:gpr, 0 ^^^^^^^^^^^^^
STD killed %8:gpr, %1:gpr, 0 this is updated
Instructions (1) and (2) would be updated to:
ADD_rr %0:gpr(tied-def 0), killed %2:gpr
The `killed` mark is inherited from machine operands `killed %3:gpr`
and `killed %6:gpr` which are updated inplace by `processDstReg()`.
This commit updates `processDstReg()` reset kill marks for updated
machine operands to keep liveness information conservatively correct.
Differential Revision: https://reviews.llvm.org/D157805
(cherry picked from commit
27026fe5633b546ed647efd99eccdfc598686535)
XinWang10 [Fri, 18 Aug 2023 06:01:50 +0000 (23:01 -0700)]
[X86]Support options -mno-gather -mno-scatter
Gather instructions could lead to security issues, details please refer to https://www.intel.com/content/www/us/en/developer/articles/technical/software-security-guidance/technical-documentation/gather-data-sampling.html.
This supported options -mno-gather and -mno-scatter, which could avoid generating gather/scatter instructions in backend except using intrinsics or inline asms.
Reviewed By: pengfei
Differential Revision: https://reviews.llvm.org/D157680
Josh Stone [Wed, 26 Jul 2023 22:00:58 +0000 (15:00 -0700)]
[SystemZ] Avoid type legalization on structs
In SystemZTTIImpl::getMemoryOpCost, the call to getNumberOfParts will
run type legalization, which can't handle structs. So before that, we
check for an unknown value type and forward to BaseT, just like many
other targets do in this situation.
https://bugzilla.redhat.com/show_bug.cgi?id=2224885
Reviewed By: uweigand
Differential Revision: https://reviews.llvm.org/D156379
(cherry picked from commit
85e4ee15d32ae0344755d11d4ca90a15a6e005cd)
Rainer Orth [Wed, 16 Aug 2023 08:37:14 +0000 (10:37 +0200)]
[Driver][unittest] Unbreak ToolChainTest.cpp compilation with -Werror
D157275 broke some buildbots that run with -Werror:
https://lab.llvm.org/buildbot#builders/36/builds/36604
https://lab.llvm.org/buildbot#builders/57/builds/29201
Avoid this by using an overlong line rather than appeasing `clang-format`.
(cherry picked from commit
18252e6c8e2b1bb9901b5cca9e9a75edfeda2c4e)
Rainer Orth [Wed, 16 Aug 2023 07:57:59 +0000 (09:57 +0200)]
[Driver] Select newest GCC installation on Solaris
As described in Issue #53709
<https://github.com/llvm/llvm-project/issues/53709>, since
28d58d8fe2094af6902dee7b4d68ec30a3e9d737
<https://reviews.llvm.org/rG28d58d8fe2094af6902dee7b4d68ec30a3e9d737>
`clang` doesn't find the latest of several parallel GCC installations on
Solaris, but only the first in directory order, which is pretty random.
This patch sorts GCC installations in reverse version order so the latest
is picked.
Tested on `sparcv9-sun-solaris2.11`, `amd64-pc-solaris2.11`, and
`x86_64-pc-linux-gnu`.
Differential Revision: https://reviews.llvm.org/D157275
(cherry picked from commit
ae84ad15efd7ef7da95146e900ec72ceadf98058)
usama hameed [Wed, 16 Aug 2023 22:08:15 +0000 (15:08 -0700)]
Add Requires x86 target to test due to failure in clang-armv8-quick
bot
(cherry picked from commit
bb5f64a6fc059fcf81b4b9ffe035f5063956ddd8)
usama hameed [Tue, 8 Aug 2023 21:49:48 +0000 (14:49 -0700)]
[CodeGen][UBSan] Handle sugared QualTypes correctly in
getUBSanFunctionTypeHash.
getUBSanFunctionTypeHash checks if a Type is a FunctionNoPrototype
by calling isa<FunctionNoProtoType>(). This does not work correctly when
the Type is wrapped in a sugar type such as an AttributedType. This
patch fixes this by using isFunctionNoProtoType() function which removes
sugar and returns the expected result.
The added test is a sanity check that the compiler no longer crashes
during compilation. It also compares the hash with and without the
function attribute for both FunctionNoProtoType and FunctionProtoType.
The hash remains the same for FunctionNoProtoType even with the addition
of an attribute.
rdar://
113144087
Differential Revision: https://reviews.llvm.org/D157445
(cherry picked from commit
9afc57dcb2e5cd36ca1ddf0fee3efa958bfd4c2a)
Fangrui Song [Mon, 14 Aug 2023 17:30:13 +0000 (10:30 -0700)]
[gcov] Ignore blocks from another file
Constructs such as inline variables, #line, and #include can create
lexical blocks with a different filename.
GCOVProfiling and llvm-cov gcov currently don't handle such cases (see
GCOVLines::writeOut and GCOVFile::readGCNO) and would incorrectly
attribute the line number to the current file.
For now, ignore such blocks. Missing line execution counts is better
than wrong ones.
---
As a workaround that Apple targets don't use -mconstructor-aliases yet,
allow line execution count 4 on the A::A line (
1f34e282e8066281eb1447e21e44a2a2e9983e79).
(cherry picked from commit
406e81b79d26dae6838cc69d10a3e22635da09ef)
Mark de Wever [Sat, 19 Aug 2023 18:02:03 +0000 (20:02 +0200)]
[libc++][doc] Marks LLVM-17 release notes as final.
Reviewed By: #libc, philnik
Differential Revision: https://reviews.llvm.org/D158340
(cherry picked from commit
71a7192d6927b422c224f5d231300774cb02348a)
Fangrui Song [Fri, 18 Aug 2023 21:58:53 +0000 (14:58 -0700)]
LowLevelType: replace a reachable llvm_unreachable with assert
assert is more appropriate here and fixes
`runtime error: execution reached an unreachable program point` in a
-DLLVM_USE_SANITIZER=Undefined build (-fno-sanitize-recover=all causes
llc to exit instead of crash (report_fatal_error)) when testing
MachineVerifier/test_g_assert_[sz]ext.mir.
(cherry picked from commit
a7e20dd664bbce6e87b1fdad88d719e497902a42)
Fangrui Song [Fri, 18 Aug 2023 20:36:29 +0000 (13:36 -0700)]
[ARM] Fix abs overflow when encoding instructions like strb r1, [r0], #-0
Tested by llvm/test/MC/ARM/basic-thumb2-instructions.s.
Caught by newer -fsanitize=signed-integer-overflow (D156821).
(cherry picked from commit
d8900f661a6e451be0ca253df3065254008dd93a)
Fangrui Song [Fri, 18 Aug 2023 20:21:36 +0000 (13:21 -0700)]
[llvm-readobj] Fix std::abs overflow after D156821
Tested by llvm/test/tools/llvm-readobj/ELF/relocations.test
(cherry picked from commit
07bb66784c17ef6b98ce38f9cc74d1d92f1aac08)
Craig Topper [Fri, 18 Aug 2023 19:42:45 +0000 (12:42 -0700)]
[RISCV] Fix a UBSAN failure for passing INT64_MIN to std::abs.
clang recently started checking for INT64_MIN being passed to 64-bit std::abs.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D158304
(cherry picked from commit
078eb4bd85ddaac530bf2919bb32c6daa6c3e36f)
Paul Robinson [Fri, 18 Aug 2023 14:33:27 +0000 (07:33 -0700)]
Reapply "[DebugInfo] Alternate (more efficient) MD5 fix"
D155991 changed the file lookup to do a full string compare on the
filename; however, this added ~0.5% to compile time with -g.
Go back to the previous pointer-based lookup, but capture the main
file's checksum as well as its name to use when creating the extra
DIFile entry. This causes all entries to be consistent and also
avoids computing the checksum twice.
This reverts commit
5956648fc3ba11dd6b0d0f2d1d9b923e7f80f247.
There was a string lifetime issue that is now corrected.
Differential Revision: https://reviews.llvm.org/D156571
(cherry picked from commit
1fcc2bc31bb9352a13445ff4a5cccb0bebb8ea5b)
Jonas Hahnfeld [Fri, 28 Jul 2023 13:43:11 +0000 (15:43 +0200)]
[CodeGen] Keep track of eagerly emitted globals
An inline virtual function must be emitted, but we need to remember
it and emit the same definition again in the future in case later
LLVM optimizations stripped it from the Module. The added test case
shows the problem; before this patch, it would fail with:
Symbols not found: [ _ZN1AD0Ev, _ZN1AD1Ev ]
This reapplies commit
f8dadefd4a, reverted in commit
0e17372b38, but
disables RTTI in the test to avoid problems on Windows.
Differential Revision: https://reviews.llvm.org/D156537
(cherry picked from commit
c861d32d7c2791bdc058d9d9fbaecc1c2f07b8c7)
Jonas Hahnfeld [Tue, 8 Aug 2023 09:25:17 +0000 (11:25 +0200)]
[CodeGen] Restrict addEmittedDeferredDecl to incremental extensions
Reemission is only needed in incremental mode. With this early return,
we avoid overhead from addEmittedDeferredDecl in non-incremental mode.
Differential Revision: https://reviews.llvm.org/D157379
(cherry picked from commit
d43a3d6346961e639e29b8083b262416889e78ec)
Jonas Hahnfeld [Wed, 2 Aug 2023 14:06:53 +0000 (16:06 +0200)]
[CodeGen] Clean up access to EmittedDeferredDecls, NFCI.
GlobalDecls should only be added to EmittedDeferredDecls if they
need reemission. This is checked in addEmittedDeferredDecl, which
is called via addDeferredDeclToEmit. Extend these checks to also
handle VarDecls (for lambdas, as tested in Interpreter/lambda.cpp)
and remove the direct access of EmittedDeferredDecls in EmitGlobal
that may actually end up duplicating FunctionDecls.
Differential Revision: https://reviews.llvm.org/D156897
(cherry picked from commit
b719e410781ce9f3a2b316afea31c156bf99e036)
Jonas Hahnfeld [Wed, 2 Aug 2023 13:52:23 +0000 (15:52 +0200)]
[CodeGen] Remove Constant arguments from linkage functions, NFCI.
This was unused since commit
dd2362a8ba last year.
Differential Revision: https://reviews.llvm.org/D156891
(cherry picked from commit
2f3fe3ed97bcf778d73eda5aae9f166810a4af02)
Jonas Hahnfeld [Wed, 9 Aug 2023 08:40:17 +0000 (10:40 +0200)]
[clang-repl] Disambiguate global namespace identifiers
A double colon starts an identifier name in the global namespace and
must be tentatively parsed as such.
Differential Revision: https://reviews.llvm.org/D157480
(cherry picked from commit
ba475a4a3440a21e5de6a39a16215836f2614f04)
Stefan Gränitz [Fri, 18 Aug 2023 11:55:25 +0000 (13:55 +0200)]
[lldb] Fix performance regression after adding GNUstep ObjC runtime
We added support for the GNUstep ObjC runtime in
0b6264738f3d. In order to check if the target process uses
GNUstep we run an expensive symbol lookup in `CreateInstance()`. This turned out to cause a heavy performance
regression for non-GNUstep inferiors.
This patch puts a cheaper check in front, so that the vast majority of requests should return early. This
should fix the symptom for the moment. The conceptual question remains: Why does `LanguageRuntime::FindPlugin`
invoke `create_callback` for each available runtime unconditionally in every `Process::ModulesDidLoad`?
Reviewed By: jasonmolenda, jingham, bulbazord
Differential Revision: https://reviews.llvm.org/D158205
(cherry picked from commit
af2eb838309d88f046d34bca6055f1de6078fa3b)
Nathan Ridge [Mon, 10 Apr 2023 05:42:20 +0000 (01:42 -0400)]
[clangd] Avoid passing -xobjective-c++-header to the system include extractor
Fixes https://github.com/clangd/clangd/issues/1568
Differential Revision: https://reviews.llvm.org/D147905
(cherry picked from commit
bd74186f1a0831b2c596e87ea056419379a223c1)
Paul Kirth [Thu, 17 Aug 2023 17:53:07 +0000 (17:53 +0000)]
Revert "[clang] Preliminary fat-lto-object support"
FatLTO support is only half complete, and shouldn't be included in the
17.x release. https://reviews.llvm.org/D152973.
This reverts commit
5784c47806cf276e8fcc1311201f437cb8169261.
This reverts commit
610fc5cbcc8b68879c562f6458608afe2473ab7f.
Louis Dionne [Mon, 14 Aug 2023 22:14:02 +0000 (18:14 -0400)]
[libc++] Fix a few incorrect CMake configuration options
This patch fixes a few CMake options that were set using incorrect
mechanisms.
CMake's man page for the -D <var>=<value> option states: If a command in
the project sets the type to PATH or FILEPATH, then the <value> will be
converted to an absolute path. That's not what we want for most of the
paths we have as configuration options. Otherwise, using -D to set the
configuration option results in an absolute path being used, which
breaks things.
option() denotes a boolean variable, but what was desired was a
string/list variable. Fix this to prevent cmake from changing any
non-empty user provided values to 'ON'.
Differential Revision: https://reviews.llvm.org/D157926
(cherry picked from commit
760261a3daf98882ccbd177e3133fb4a058f47ad)
Piotr Fusik [Thu, 10 Aug 2023 11:47:43 +0000 (13:47 +0200)]
[libc++] Work around dynamic linking of stringstream::str() on Windows
Also `istringstream::str()` and `ostringstrem::str()`.
https://github.com/llvm/llvm-project/issues/40363 causes `str()`
to be dllimport'ed despite _LIBCPP_HIDE_FROM_ABI.
This is a temporary solution until #40363 is fixed.
Reviewed By: #libc, thakis, philnik
Differential Revision: https://reviews.llvm.org/D157602
(cherry picked from commit
090996f4a1186b1522e1225b6779d74ce9da250c)
Matt Arsenault [Tue, 15 Aug 2023 14:23:33 +0000 (10:23 -0400)]
AMDGPU: Fix fast f32 exp2
Mirror of the previous log changes, OpenCL conformance doesn't like
interpreting afn as ignore denormal handling but was previously hidden
by flag dropping.
(cherry picked from commit
81b278e613fb3bcdf8053f7a6b5dd2f3881898d2)
Matt Arsenault [Mon, 14 Aug 2023 23:38:55 +0000 (19:38 -0400)]
AMDGPU: Fix fast f32 log/log10
OpenCL conformance didn't like interpreting afn as ignore the denormal
handling.
https://reviews.llvm.org/D157940
(cherry picked from commit
4b7b4b945856c46b91ade7b6c4264fe8e258367e)
Matt Arsenault [Mon, 14 Aug 2023 23:37:32 +0000 (19:37 -0400)]
AMDGPU: Fix fast math log2 f32
Apparently afn doesn't allow you to drop the denormal handling
according to OpenCL conformance. This was hidden by losing the flags
during the library linking process. Fast log is still broken and needs
more work.
https://reviews.llvm.org/D157936
(cherry picked from commit
e09b3593ba64d004a9d2b3fa41be2ba84f968a88)
Nikita Popov [Mon, 14 Aug 2023 08:39:05 +0000 (10:39 +0200)]
[PreISelIntrinsicLowering] Use TLI for correct function
We should query the subtarget of the calling function, not of the
intrinsic.
This probably makes no functional difference (as libcalls are
unlikely to vary across subtargets), but fixes minor compile-time
regressions from unnecessary subtarget instantiations.
Followup to D157567.
Differential Revision: https://reviews.llvm.org/D157848
(cherry picked from commit
66bb7521625826cab6fcc24662c053deee1ef90c)
Matt Arsenault [Thu, 10 Aug 2023 00:26:11 +0000 (20:26 -0400)]
PreISelIntrinsicLowering: Check RuntimeLibcalls instead of TLI for memory functions
We need a better mechanism for expressing which calls you are allowed
to emit and which calls are recognized. This should be applied to the
17 branch.
(cherry picked from commit
c8cac156135de8e211b800ea5abe9666803b2875)
Konstantin Varlamov [Fri, 4 Aug 2023 22:11:31 +0000 (15:11 -0700)]
[libc++][PSTL] Disable `-Wpass-failed`.
PSTL contains many pragmas that request loop vectorization, which would
produce a warning when the compiler is unable to fulfill the request (if
`-Wpass-failed` is enabled). This is normal and expected in some cases,
and we don't want `-Werror` to turn that into a compilation failure.
Differential Revision: https://reviews.llvm.org/D157145
(cherry picked from commit
1d340250a894b6ec956e2145b2fc7babbf83e61b)
Paul Walker [Mon, 14 Aug 2023 13:26:35 +0000 (13:26 +0000)]
[SelectionDAG] Use TypeSize variant of ComputeValueVTs to compute correct offsets for scalable aggregate types.
Differential Revision: https://reviews.llvm.org/D157872
Paul Walker [Fri, 11 Aug 2023 15:45:18 +0000 (16:45 +0100)]
[NFC][SelectionDAGBuilder] Use getObjectPtrOffset in place of discrete nodes.
Some prep work to make aggregate loads and stores TypeSize aware.
Paul Walker [Fri, 11 Aug 2023 12:05:08 +0000 (12:05 +0000)]
[SVE] Add test to show incorrect code generation for scalable vector struct loads and stores.
Patch also includes a minor fix to AArch64::isLegalAddressingMode
to ensure all scalable types have a suitable bailout.
Fraser Cormack [Thu, 17 Aug 2023 13:01:08 +0000 (14:01 +0100)]
[AArch64] Fix crash when neither Neon nor SVE are enabled
The subtarget was unconditionally reporting that SVE was to be used to
lower vectors when Neon was unavailable, even when SVE itself was
unavailable. This decision leads other parts of the compiler to crash,
e.g., when querying SVE vector sizes.
Reviewed By: sdesmalen
Differential Revision: https://reviews.llvm.org/D158179
(cherry picked from commit
c058eb998aa6fa89d412984d6cbf9e3f826249f3)
Derek Schuff [Tue, 25 Jul 2023 21:13:05 +0000 (14:13 -0700)]
[WebAssembly][Objcopy] Write output section headers identically to inputs
Previously when objcopy generated section headers, it padded the LEB
that encodes the section size out to 5 bytes, matching the behavior of
clang. This is correct, but results in a binary that differs from the
input. This can sometimes have undesirable consequences (e.g. breaking
source maps).
This change makes the object reader remember the size of the LEB
encoding in the section header, so that llvm-objcopy can reproduce it
exactly. For sections not read from an object file (e.g. that
llvm-objcopy is adding itself), pad to 5 bytes.
Reviewed By: jhenderson
Differential Revision: https://reviews.llvm.org/D155535
(cherry picked from commit
1b21067cf247c62c2442daa7ee2d3915249d1ee2)
Craig Topper [Thu, 10 Aug 2023 16:10:52 +0000 (09:10 -0700)]
[RISCV] Stop performFP_TO_INTCombine from folding with ISD::FRINT.
FRINT was added to matchRoundingOp after this function was written.
So FRINT was not tested originally.
For vectors, folding this causes us to create a CSR swap that tries
to write 7 to FRM. This is an illegal value and will cause the CSR
write to fail.
While this might be a legal fold we could do, I'm disabling it for
now so we can backport to LLVM 17 with the least risk.
Differential Revision: https://reviews.llvm.org/D157583
(cherry picked from commit
2df9328fe3a0575fa415964a3c3544b0912b168e)
Nikita Popov [Thu, 10 Aug 2023 14:30:17 +0000 (16:30 +0200)]
Reapply [CHR] Fix up phi nodes with unreachable predecessors (PR64594)
Relative to the previous attempt, this also adjusts RegionInfo
verification to allow unreachable predecessors.
-----
If a block in the CHR region has an unreachable predecessor, then
there will be no edge from that predecessor to the newly cloned
block. However, a phi node entry for it will be left behind. Make
sure that these incoming blocks get dropped as well.
Fixes https://github.com/llvm/llvm-project/issues/64594.
Differential Revision: https://reviews.llvm.org/D157621
(cherry picked from commit
7e2f1ae7e0ebc7e71ffaa865175aef27fae3b034)
Zequan Wu [Thu, 3 Aug 2023 17:08:53 +0000 (13:08 -0400)]
Revert "[Profile] Remove duplicate file locks when enabled continuous mode and online merging."
This reverts commit
985486dca48bffd9e991d9f5ac32e1d109ae000f.
This seems to cause malformed profdata: crbug.com/1469456
(cherry picked from commit
4b08be77c98a4c15b8b3d634fab4ffc24bf4ef38)
Ryan Burns [Tue, 15 Aug 2023 21:11:58 +0000 (14:11 -0700)]
Fix truncated __OPENMP_NVPTX__ preprocessor condition
I don't have a test case for this but noticed this warning when including system headers with `-I` rather than `-isystem`.
```
In file included from <built-in>:1:
In file included from /nix/store/jq6bpm0xmhnbffhs5rkxq4n88g5xi2zg-clang-wrapper-11.0.1/resource-root/include/__clang_cuda_runtime_wrapper.h:157:
/nix/store/jq6bpm0xmhnbffhs5rkxq4n88g5xi2zg-clang-wrapper-11.0.1/resource-root/include/__clang_cuda_math.h:39:25: warning: extra tokens at end of #ifdef directive [-Wextra-tokens]
^
```
Reviewed By: tra
Differential Revision: https://reviews.llvm.org/D95299
(cherry picked from commit
c49142e4f5c8645a4d741d233f0cb55ef1ef87a2)
Tulio Magno Quites Machado Filho [Mon, 14 Aug 2023 13:22:26 +0000 (10:22 -0300)]
flang: Fix module-debug-file-loc-linux.f90 in standalone
In a standalone build, there is no guarantee that flang code would be
saved in a directory named flang. Check only the path under flang's root
directory.
Reviewed By: #flang, kiranchandramohan
Differential Revision: https://reviews.llvm.org/D157642
(cherry picked from commit
94f377d880376fa6d5653430cad2777e51759d48)
Martin Storsjö [Mon, 7 Aug 2023 06:59:39 +0000 (09:59 +0300)]
[clang] Skip stores in init for fields that are empty structs
An empty struct is handled as a struct with a dummy i8, on all targets.
Most targets treat an empty struct return value as essentially
void - but some don't. (Currently, at least x86_64-windows-* and
powerpc64le-* don't treat it as void.)
When intializing a struct with such a no_unique_address member,
make sure we don't write the dummy i8 into the struct where there's
no space allocated for it.
Previously it would clobber the actual valid data of the struct.
Fixes https://github.com/llvm/llvm-project/issues/64253, and
possibly https://github.com/llvm/llvm-project/issues/64077
and https://github.com/llvm/llvm-project/issues/64427 as well.
We should omit the store for any empty record (not only ones
declared with no_unique_address); we can have a situation where a
class doesn't have the no_unique_address attribute, but is embedded
in an outer struct with the no_unique_address attribute - like this:
struct S {};
S f();
struct S2 : public S { S2();};
S2::S2() : S(f()) {}
struct S3 { int x; [[no_unique_address]] S2 y; S3(); };
S3::S3() : x(1), y() {}
Here, the problematic store (which this patch omits) is in
the constructor of S2. In the case of S3, S2 has no valid storage
and aliases x - thus the constructor of S2 should omit the dummy
store.
Differential Revision: https://reviews.llvm.org/D157332
(cherry picked from commit
d60c3d08e78dfbb4b180776b83e910d810e1f36a)
Anmol P. Paralkar [Mon, 14 Aug 2023 21:53:54 +0000 (14:53 -0700)]
[RISCV] Add bounds check before use on returned iterator.
Check iterator validity before use; fixes a crash seen in the RISC-V
Zcmp Push/Pop optimization pass when compiling an internal benchmark.
Reviewed By: asb, wangpc
Differential Revision: https://reviews.llvm.org/D157674
(cherry picked from commit
53e89f5e3f2cefa432f7262343654c2a582dcbe1)
Fangrui Song [Mon, 14 Aug 2023 15:26:52 +0000 (08:26 -0700)]
[Driver] Default riscv*- triples to -fdebug-default-version=4
This adds a RISC-V special case to ToolChain::GetDefaultDwarfVersion,
affecting Linux/Haiku/RISCVToolChain.
DWARF v5 .debug_loclists/.debug_rnglists's
DW_LLE_offset_pair/DW_RLE_offset_pair entry kinds utilitize `.uleb128 A-B`
directives where A and B reference local labels in code sections.
When A and B are separated by a RISC-V linker-relaxable instruction,
A-B is incorrectly folded without a relocation, causing incorrect debug
information.
```
void ext(void);
int foo(int x) {ext(); return 0;}
// DW_AT_location [DW_FORM_loclistx] of a DW_TAG_formal_parameter references a DW_LLE_offset_pair that can be incorrect after linker relaxation.
int ext(void);
void foo() { {
int ret = ext();
if (__builtin_expect(ret, 0))
ext();
} }
// DW_AT_ranges [DW_FORM_rnglistx] of a DW_TAG_lexical_block references a DW_RLE_offset_pair that can be incorrect after linker relaxation.
```
D157657 will implement R_RISCV_SET_ULEB128/R_RISCV_SUB_ULEB128
relocations, fixing the issue, but the relocation is only supported by
bleeding-edge binutils 2.41 and not by lld/ELF yet.
The goal is to make the emitted DWARF correct after linking.
Many users don't care about the default DWARF version, but a linker
error will be unacceptable. Let's just downgrade the default DWARF
version, before binutils>=2.41 is more widely available.
An alternative compatibility option is to add a toggle to DwarfDebug.cpp,
but that doesn't seem like a good idea.
Reviewed By: asb, kito-cheng
Differential Revision: https://reviews.llvm.org/D157663
(cherry picked from commit
bbc0f99f3bc96f1db16f649fc21dd18e5b0918f6)
(with a release note)
Piotr Zegar [Fri, 11 Aug 2023 15:25:43 +0000 (15:25 +0000)]
[clang-tidy] Fix crash when diagnostic is emit with invalid location
Fix crash when diagnostic is emit with invalid location,
but with attached valid ranges. Diagnostic can contain
invalid location, but SourceManager attached to it still
can be valid, use it in such case or fallback to known
SourceManager.
Fixes: #64602
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D157649
(cherry picked from commit
efd44f80a5a8194b9fe26ff3244ce702cd8dab73)
Kadir Cetinkaya [Thu, 10 Aug 2023 11:45:40 +0000 (13:45 +0200)]
Revert "[Sema] Fix handling of functions that hide classes"
This reverts commit
d031ff38779bd688c514136dbdcce3169ee82b6e.
See https://reviews.llvm.org/D154503#4576393 for a reproducer and
details.
(cherry picked from commit
7d259b36d2e8148d13087844e6494ad3a5c63edf)
Valentin Clement [Tue, 1 Aug 2023 16:47:43 +0000 (09:47 -0700)]
[flang] Add missing dependency on tablegen files
This issue was raised on https://github.com/llvm/llvm-project/issues/64268.
`flang/lib/Optimizer/Transforms/SimplifyIntrinsics.cpp` includes
`flang/Optimizer/HLFIR/HLFIRDialect.h` and might fails if the HLFIR related
tablegen files have not been generated.
Reviewed By: vzakhari
Differential Revision: https://reviews.llvm.org/D156751
(cherry picked from commit
103907bc5fcf656b1551e191af8176cd7bfb91da)
Nikita Popov [Fri, 11 Aug 2023 08:06:35 +0000 (10:06 +0200)]
[SDAG] Don't transfer !range metadata without !noundef to SDAG (PR64589)
D141386 changed the semantics of !range metadata to return poison
on violation. If !range is combined with !noundef, violation is
immediate UB instead, matching the old semantics.
In theory, these IR semantics should also carry over into SDAG.
In practice, DAGCombine has at least one key transform that is
invalid in the presence of poison, namely the conversion of logical
and/or to bitwise and/or (https://github.com/llvm/llvm-project/blob/
c7b537bf0923df05254f9fa4722b298eb8f4790d/llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp#L11252).
Ideally, we would fix this transform, but this will require
substantial work to avoid codegen regressions.
In the meantime, avoid transferring !range metadata without
!noundef, effectively restoring the old !range metadata semantics
on the SDAG layer.
Fixes https://github.com/llvm/llvm-project/issues/64589.
Differential Revision: https://reviews.llvm.org/D157685
(cherry picked from commit
9deee6bffa9c331f46c68e5dd4cb4abf93dc0716)
Nikita Popov [Fri, 11 Aug 2023 07:52:25 +0000 (09:52 +0200)]
[X86] Add test for PR64589 (NFC)
(cherry picked from commit
59d558a37850611f9b63a5c1a8bfe87a9606c81e)
Paul Walker [Mon, 7 Aug 2023 23:28:27 +0000 (00:28 +0100)]
[SelectionDAG] Fix problematic call to EVT::changeVectorElementType().
The function changeVectorElementType assumes MVT input types will
result in MVT output types. There's no gurantee this is possible
during early code generation and so this patch converts an instance
used during initial DAG construction to instead explicitly create a
new EVT.
NOTE: I could have added more MVTs, but that seemed unscalable as
you can either have MVTs with 100% element count coverage or 100%
bitwidth coverage, but not both.
Differential Revision: https://reviews.llvm.org/D157392
(cherry picked from commit
b7e6e568b4f28124060b868e433f36af18c510db)
Nikita Popov [Fri, 11 Aug 2023 07:39:27 +0000 (09:39 +0200)]
Revert "[AggressiveInstCombine] Fold strcmp for short string literals"
This reverts commit
8981520b19f2d2fe3d2bc80cf26318ee6b5b7473.
Tobias Hieta [Mon, 14 Aug 2023 08:37:48 +0000 (10:37 +0200)]
[CMake] Set version suffix to `rc` during the RC phase
See discussion in https://github.com/llvm/llvm-project/issues/64225
Julian Schmidt [Sat, 12 Aug 2023 16:17:54 +0000 (16:17 +0000)]
[clang-tidy] fix None tmpdir when exporting fixes in run-clang-tidy
Differential https://reviews.llvm.org/D145477 removed the check for `(yaml and args.export_fixes)` in line 303 to skip looking for the `clang-apply-replacements` binary. However, the `tmpdir` variable was set in this true branch when exporting fixes and therefore is `None` when invoking run-clang-tidy with `run-clang-tidy -p . -export-fixes fixes.yaml`.
Reviewed By: PiotrZSL
Differential Revision: https://reviews.llvm.org/D157773
(cherry picked from commit
c3da99275a520b73235d975017502876e07e3e8e)
Tamir Duberstein [Thu, 27 Jul 2023 14:26:58 +0000 (10:26 -0400)]
[BPF] Don't crash on missing line info
When compiling Rust code we may end up with calls to functions provided
by other code units. Presently this code crashes on a null pointer
dereference - this patch avoids that crash and adds a test.
Reviewed By: ast
Differential Revision: https://reviews.llvm.org/D156446
(cherry picked from commit
055893beacb34441467eb997a270a620c57c138f)
Jay Foad [Thu, 10 Aug 2023 10:19:41 +0000 (11:19 +0100)]
[AMDGPU] Do not release VGPRs at -O0
This was an oversight when the GFX11 early release VGPRs optimization
was reimplemented in D153279.
Sending the DEALLOC_VGPRS message is a performance optimization so there
is no need to do it at -O0. In addition it makes some kinds of post
mortem debugging hard or impossible, since VGPR values are no longer
available to inspect at the s_endpgm instruction.
Differential Revision: https://reviews.llvm.org/D157599
(cherry picked from commit
3091bdb86d55e404866823b64d21fd87c247d893)
wangpc [Thu, 10 Aug 2023 12:40:28 +0000 (20:40 +0800)]
[RISCV] Fix wrong type prototype of RVVSlideOneBuiltinSet
We need unsigned integer here.
Fixes #64534
Reviewed By: eopXD
Differential Revision: https://reviews.llvm.org/D157476
(cherry picked from commit
5a3753f35632f8fd7920c5c99b1237113b5c435e)
Rainer Orth [Thu, 3 Aug 2023 18:01:13 +0000 (20:01 +0200)]
[Driver] Allow for sparcv8plus subdir with Solaris/SPARC GCC
Since GCC 11, the bundled Solaris/SPARC GCC uses the `sparcv8plus`
subdirectory for 32-bit objects, just like upstream GCC. Before that, it
used `32` instead from a local patch.
Since `clang` doesn't know about that `sparcv8plus` subdirectory, it
wouldn't properly use GCC 11+ installations.
The new `solaris-sparc-gcc-search.test` testcase wasn't run initially (like
the existing `crash-report-null.test`) because the `.test` suffix wasn't
handled.
Tested on `sparcv9-sun-solaris2.11`, `amd64-pc-solaris2.11`, and
`x86_64-pc-linux-gnu`.
Differential Revision: https://reviews.llvm.org/D157013
(cherry picked from commit
43dfe0f08ecaf50f986512d0548bd3ac84d1813b)
Corentin Jabot [Sat, 29 Jul 2023 14:03:44 +0000 (16:03 +0200)]
[Clang] Backport static_assert messages fixes
* 4d494e7: Handle static_assert messages with an expression started by a literal
* 49e0495 Produce a warning instead of an error in unevaluated strings before C++26
Emiting an error on unexpected encoding prefix - which was allowed before C++26 -
caused build errors for a few users.
This downgrade the error to a warning on older language modes and C
Weining Lu [Wed, 9 Aug 2023 01:54:41 +0000 (09:54 +0800)]
[lld][LoongArch] Support the R_LARCH_PCREL20_S2 relocation type
`R_LARCH_PCREL20_S2` is a new added relocation type in LoongArch ELF
psABI v2.10 [1] which is not corvered by D138135 except `R_LARCH_64_PCREL`.
A motivation to support `R_LARCH_PCREL20_S2` in lld is to build the
runtime of .NET core (a.k.a `CoreCLR`) in which strict PC-relative
semantics need to be guaranteed [2]. The normal `pcalau12i + addi.d`
approach doesn't work because the code will be copied to other places
with different "page" and offsets. To achieve this, we can use `pcaddi`
with explicit `R_LARCH_PCREL20_S2` reloc to address +-2MB PC-relative
range with 4-bytes aligned.
[1]: https://github.com/loongson/la-abi-specs/releases/tag/v2.10
[2]: https://github.com/dotnet/runtime/blob/release/7.0/src/coreclr/vm/loongarch64/asmhelpers.S#L307
Reviewed By: xen0n, MaskRay
Differential Revision: https://reviews.llvm.org/D156772
(cherry picked from commit
8a31f7ddb8436fa2a8ad754eb51618139cf63415)
Oliver Stannard [Thu, 27 Jul 2023 13:36:14 +0000 (14:36 +0100)]
[AArch64][PtrAuth] Fix unwind state for tail calls
When generating unwind tables for code which uses return-address
signing, we need to toggle the RA_SIGN_STATE DWARF register around any
tail-calls, because these require the return address to be authenticated
before the call, and could throw an exception. This is done using the
.cfi_negate_ra_state directive before the call, and .cfi_restore_state
at the start of the next basic block.
However, since D153098, the .cfi_restore_state isn't being inserted,
because the CFIFixup pass isn't being run. This re-enables that pass
when return-adress signing is enabled.
Reviewed By: ikudrin, MaskRay
Differential Revision: https://reviews.llvm.org/D156428
(cherry picked from commit
f2e7285b03fbfc263c8a2eaa2b5e2e1cfafc6abd)
Anatoly Trosinenko [Sat, 29 Jul 2023 08:37:51 +0000 (11:37 +0300)]
[AArch64] Refactor checks in sign-return-address.ll test
Using implicit CHECK prefix in one FileCheck invocation and explicit
CHECK-V83A in the other one seems to misguide to use CHECK: lines as
a common matching prefix at various places. Also note that
; CHECK, CHECK-V83A: ...
line only matches the "CHECK-V83A" prefix.
This commit explicitly splits the checks into common ones (CHECK) and
invocation-specific ones (COMPAT and V83A) and updates the assertions
with the update_llc_test_checks.py script.
Reviewed By: efriedma, MaskRay
Differential Revision: https://reviews.llvm.org/D156327
(cherry picked from commit
4210204f521be3caa0e60bd596af3444cbd44d04)
Sean Fertile [Fri, 21 Jul 2023 14:16:26 +0000 (10:16 -0400)]
[PPC][AIX] Fix toc-data peephole bug and some related cleanup.
Set the ReplaceFlags variable to false, since there is code meant only
for the ADDItocHi/ADDItocL nodes. This has the side effect of disabling
the peephole when the load/store instruction has a non-zero offset.
This patch also fixes retrieving the `ImmOpnd` node from the AIX small
code model pseduos and does the same for the register operand node.
This allows cleaning up the later calls to replaceOperands.
Finally move calculating the MaxOffset into the code guarded by
ReplaceFlags as it is only used there and the comment is specific to the ELF
ABI.
Fixes https://github.com/llvm/llvm-project/issues/63927
Differential Revision: https://reviews.llvm.org/D155957
(cherry picked from commit
b37c7ed0c95c7f24758b1532f04275b4bb65d3c1)
Craig Topper [Mon, 7 Aug 2023 20:00:57 +0000 (13:00 -0700)]
[RISCV] Add back handling of X > -1 to ISD::SETCC lowering.
There are cases where the -1 doesn't become visible until lowering
so the folding doesn't have a chance to run.
I think in these cases there is a missed DAGCombine for truncate (undef),
which I may fix separately, but RISC-V backend should protect itself.
Fixes #64503.
Reviewed By: asb
Differential Revision: https://reviews.llvm.org/D157314
(cherry picked from commit
7cc615413fd7c93421052a193bc3e114465747c9)
Amir Ayupov [Thu, 3 Aug 2023 17:47:18 +0000 (10:47 -0700)]
BOLT Release Notes
Martin Storsjö [Sun, 6 Aug 2023 21:17:01 +0000 (00:17 +0300)]
[llvm-rc] Resolve the executable path if not present in Argv[0]
The llvm-rc tool tries to locate a suitable Clang executable to
use for preprocessing. For this purpose, it first checks within
the same directory as the llvm-rc tool, checking with a couple
different names, followed by checking all of $PATH for another
couple names.
On Windows, the InitLLVM() function always sets up Argv[0] with the
full path to the executable, while on Unix, Argv[0] is kept as is.
Therefore, call getMainExecutable to try to resolve the directory of
the executable before looking for colocated Clang executables.
This makes
282744a9ce18120dc0a6eceb02693b36980d9498 actually have
the desired effect.
Differential Revision: https://reviews.llvm.org/D157241
(cherry picked from commit
8c6a0c8bf50bca6c3a0c4de1b84f21466ee31655)
Mark de Wever [Sat, 22 Jul 2023 15:47:04 +0000 (17:47 +0200)]
[libc++][doc] Updates format related release notes.
As suggested by @ldionne.
Reviewed By: ldionne, #libc, avogelsgesang
Differential Revision: https://reviews.llvm.org/D156030
(cherry picked from commit
cf83cfb3a569632c0049f11d0f56b851aaf21393)
Mehdi Amini [Sat, 5 Aug 2023 21:47:39 +0000 (14:47 -0700)]
[MLIR] Make the `ConversionTarget` const ref in the DialectConversion (NFC)
It isn't mutated during the conversion already, communicate this through the API.
Differential Revision: https://reviews.llvm.org/D157199
Mehdi Amini [Sat, 5 Aug 2023 21:53:03 +0000 (14:53 -0700)]
Clarify the invariant of the MLIR pass pipeline around `Pass::initialize()`
This method should not load new dialect or affect the context itself.
Differential Revision: https://reviews.llvm.org/D157198
Benjamin Maxwell [Fri, 4 Aug 2023 13:32:33 +0000 (13:32 +0000)]
[mlir][VectorOps] Fix folding of vector.extract from stretch vector.broadcast
Previously, foldExtractFromBroadcast() would incorrectly fold:
func.func @extract_from_stretch_broadcast(%src: vector<3x1x2xf32>) -> f32 {
%0 = vector.broadcast %src : vector<3x1x2xf32> to vector<3x4x2xf32>
%1 = vector.extract %0[0, 2, 0] : vector<3x4x2xf32>
return %1: f32
}
to:
func.func @extract_from_stretch_broadcast(%src: vector<3x1x2xf32>) -> f32 {
%0 = vector.extract %src[0, 2, 0] : vector<3x1x2xf32>
return %0: f32
}
This was due to the wrong offset being used when zeroing the "dim-1"
broadcasted dims. It should use the difference in rank across the
broadcast as the starting offset, as the ranks after that are the ones
that could have been stretched.
Reviewed By: awarzynski, dcaballe
Differential Revision: https://reviews.llvm.org/D157003
Tobias Gysi [Thu, 3 Aug 2023 07:55:58 +0000 (07:55 +0000)]
[mlir] Store segment sizes in std::array
This revision uses std::array instead of normal c arrays to store the
operand and result segment sizes. This is a follow up to
https://reviews.llvm.org/D155919, which converted the operand and result
segment sizes to properties. Its use of c arrays triggered warnings in
downstream projects due to the direct comparison of c arrays. This
revision fixes the warnings using std::arrays that implement a
proper comparison operator, which compares the array elements rather
that the array pointers.
Note: it seems the comparison operator is effectively dead code for now.
It still seems useful to fix the warning and ensure the comparison works
as expected assume someone starts using it at some point in time.
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D156888
Mehdi Amini [Thu, 3 Aug 2023 05:50:05 +0000 (22:50 -0700)]
[MLIR][Bytecode] Add missing field initializer in constructor initializer list
Leaving this field unitialized could led to crashes when it'll diverge from the
IRNumbering phase.
Differential Revision: https://reviews.llvm.org/D156965
Valentin Clement [Tue, 1 Aug 2023 21:57:10 +0000 (14:57 -0700)]
[mlir] Reduce warnings for bad assertion in generated code
When the operation has no attributes, the generated assertion is
always false and triggers lots of warnings in the build.
```
warning: comparison of unsigned expression < 0 is always false
```
Just return a StringAttr when there is no attribute.
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D156819
Mogball [Fri, 28 Jul 2023 22:37:17 +0000 (15:37 -0700)]
[mlir][llvm] Fix export of 64-bit integer function attributes
The `allocsize` attribute is weird because it packs two 32-bit values
into a 64-bit value. It also turns out that the passthrough attribute
exporter was using `int`, which is incorrectly handling 64-bit integers.
Reviewed By: rriddle
Differential Revision: https://reviews.llvm.org/D156574
Nicolas Vasilache [Wed, 26 Jul 2023 18:05:13 +0000 (18:05 +0000)]
[mlir][gpu] NFC - Fail gracefully when type conversion fails instead of crashing
Thomas Raoux [Fri, 28 Jul 2023 16:35:36 +0000 (09:35 -0700)]
[mlir] Fix arith verifier for tensor with encoding
The verifier for some arith ops were not considering that ranked
tensor types can have encodings.
Differential Revision: https://reviews.llvm.org/D156557
Alexis Engelke [Fri, 28 Jul 2023 10:53:09 +0000 (12:53 +0200)]
[mlir][LLVMIR] Fix identified structs with same name
Different identified struct types may have the same name ("").
Previously, these were deduplicated based on their name, which caused
an assertion failure when nesting identified structs:
%0 = type { %1 }
%1 = type { i8 }
declare void @fn(%0)
Reviewed By: gysit
Differential Revision: https://reviews.llvm.org/D156531
River Riddle [Wed, 26 Jul 2023 07:08:25 +0000 (00:08 -0700)]
[mlir-lsp] Guard writing output to JSONTransport with mutex
This allows for users of the lsp transport libraries to process replies
in parallel, without overlapping/clobbering the output.
Differential Revision: https://reviews.llvm.org/D156295
River Riddle [Tue, 25 Jul 2023 23:06:02 +0000 (16:06 -0700)]
[mlir:bytecode] Only visit the all regions path if the op has regions
Zero region operations return true for both isBeforeAllRegions and
isAfterAllRegions when using WalkStage. The bytecode walk only
expects region holding operations in the after regions path, so
guard against that.
River Riddle [Tue, 25 Jul 2023 02:58:59 +0000 (19:58 -0700)]
[mlir:bytecode] Support lazy loading dynamically isolated regions
We currently only support lazy loading for regions that
statically implement the IsolatedFromAbove trait, but that
limits the amount of operations that can be lazily loaded. This review
lifts that restriction by computing which operations have isolated
regions when numbering, allowing any operation to be lazily loaded
as long as it doesn't use values defined above.
Differential Revision: https://reviews.llvm.org/D156199
River Riddle [Tue, 25 Jul 2023 02:49:38 +0000 (19:49 -0700)]
[mlir:bytecode] Fix bytecode lazy loading for ops with multiple regions
We currently encode each region as a separate section, but
the reader expects all of the regions to be in the same section.
This updates the writer to match the behavior that the reader
expects.
Differential Revision: https://reviews.llvm.org/D156198
Mehdi Amini [Tue, 25 Jul 2023 18:14:16 +0000 (11:14 -0700)]
Add release notes for MLIR
Differential Revision: https://reviews.llvm.org/D156253
Oleg Shyshkov [Tue, 25 Jul 2023 15:39:03 +0000 (17:39 +0200)]
[mlir] Fix assembly format parser generator after
9ea6b30ac20f8223fb6aeae853e5c73691850a8d.
Jakub Kuderski [Tue, 25 Jul 2023 15:30:17 +0000 (11:30 -0400)]
[mlir][spirv] Do not introduce vector<1xT> in UnifyAliasedResource
1-element vectors are not valid in SPIR-V and fail `Bitcast` op verification.
Reviewed By: antiagainst
Differential Revision: https://reviews.llvm.org/D156207
Igor Kirillov [Fri, 4 Aug 2023 12:05:27 +0000 (12:05 +0000)]
[CodeGen] Disable FP LD1RX instructions generation for Neoverse-V1
These instructions show worse performance on Neoverse-V1 compared
to pair of LDR(LDP)/MOV instructions.
This patch adds `no-sve-fp-ld1r` sub-target feature, which is enabled
only on Neoverse-V1.
Fixes https://github.com/llvm/llvm-project/issues/64498
Differential Revision: https://reviews.llvm.org/D157279
(cherry picked from commit
60e2a849b0a537f96ca12fb032c4a0e32e07b4ae)
Amy Huang [Fri, 23 Jun 2023 18:41:44 +0000 (11:41 -0700)]
Reland "Try to implement lambdas with inalloca parameters by forwarding without use of inallocas."t
This reverts commit
8ed7aa59f489715d39d32e72a787b8e75cfda151.
Differential Revision: https://reviews.llvm.org/D154007
(cherry picked from commit
27dab4d305acb6e0935e014c061c5317016ae2b3)
Guillaume Chatelet [Mon, 7 Aug 2023 07:31:53 +0000 (07:31 +0000)]
[llvm-exegesis] Don't try to use SYS_rseq if it's not defined.
When compiling against recent glibc (>= 2.35) but old kernel headers (< 4.18), `SYS_rseq` is not defined and thus llvm-exegesis fails to build. So also check that `SYS_rseq` is defined before trying to use it.
Fixes https://github.com/llvm/llvm-project/issues/64456
Reviewed By: MaskRay, gchatelet
Differential Revision: https://reviews.llvm.org/D157189
(cherry picked from commit
f70e83af7a708a22fdde8c644ac5810223090cd4)
Igor Kirillov [Thu, 3 Aug 2023 15:57:12 +0000 (15:57 +0000)]
[CodeGen] Fix incorrect pattern FMLA_* pseudo instructions
* Remove the incorrect patterns from AArch64fmla_p/AArch64fmls_p
* Add correct patterns to AArch64fmla_m1/AArch64fmls_m1
* Refactor fma_patfrags for the sake of PatFrags
Fixes https://github.com/llvm/llvm-project/issues/64419
Differential Revision: https://reviews.llvm.org/D157095
(cherry picked from commit
84d444f90900d1b9d6c08be61f8d62090df28042)
Igor Kirillov [Tue, 8 Aug 2023 11:30:32 +0000 (11:30 +0000)]
[CodeGen] Precommit tests for D157095
(cherry picked from commit
7542477d5d6e10848ac9ba5dd5421afc7e4947d2)
Igor Kirillov [Fri, 4 Aug 2023 13:45:03 +0000 (13:45 +0000)]
[CodeGen] Pre-commit tests showing incorrect pattern FMLA_* pseudo instructions
Differential Revision: https://reviews.llvm.org/D157094
(cherry picked from commit
b560d5c7e380c1c412b892a3e22f8ee15a522381)
David Spickett [Fri, 28 Jul 2023 08:03:40 +0000 (08:03 +0000)]
[lldb][AArch64] Save/restore TLS registers around expressions
This was cherry-picked from
6239227172cdc92f3bb72131333f50f83a6439cf and has been
modified to remove references to the tpidr2 register that is not supported on
the 17 branch.
Previously lldb was storing them but not restoring them. Meaning that this function:
```
void expr(uint64_t value) {
__asm__ volatile("msr tpidr_el0, %0" ::"r"(value));
}
```
When run from lldb:
```
(lldb) expression expr()
```
Would leave tpidr as `value` instead of the original value of the register.
A check for this scenario has been added to TestAArch64LinuxTLSRegister.py,
Reviewed By: omjavaid
Differential Revision: https://reviews.llvm.org/D156512
David Spickett [Mon, 24 Jul 2023 12:38:27 +0000 (12:38 +0000)]
[lldb][AArch64] Add reading of TLS tpidr register from core files
7e229217f4215b519b886e7881bae4da3742a7d2 did live processes, this does
core files. Pretty simple, there is an NT_ARM_TLS note that contains
at least tpidr, and on systems with the Scalable Matrix Extension (SME), tpidr2
as well.
tpidr2 will be handled in future patches for SME support.
This NT_ARM_TLS note has always been present but it seems convenient to
handle it as "optional" inside of LLDB. We'll probably want the flexibility
when supporting tpidr2.
Normally the C library would set tpidr but all our test sources build
without it. So I've updated the neon test program to write to tpidr
and regenerated the corefile.
I've removed the LLDB_PTRACE_NT_ARM_TLS that was unused, we get
what we need from llvm's defs instead.
Reviewed By: omjavaid
Differential Revision: https://reviews.llvm.org/D156118
Shivam Gupta [Tue, 8 Aug 2023 01:50:36 +0000 (07:20 +0530)]
Revert "[Clang] Fix -Wconstant-logical-operand when LHS is a constant"
This reverts commit
dfdfd306cfaf54fbc43e2d5eb36489dac3eb9976.
An issue is reported for wrong warning, this has to be reconsidered.
Differential Revision: https://reviews.llvm.org/D157352