The two constants should be equal only if their derived types
are the same. This fixes regression caused by D150380.

Differential Revision: https://reviews.llvm.org/D150634

Previously we added the SP operands when an immediate operand was added
to certain opcodes.

This patch moves it to a post processing step using the information
in MCInstrDesc. This avoids an explicit opcode list in RISCVDisassembler.cpp.

In considered using a custom DecoderMethod, but the bit swizzling we
need to do for the immediates on these instructions made that
unattractive.

Reviewed By: asb

Differential Revision: https://reviews.llvm.org/D149931

Whether assertions are enabled or not is orthogonal to the build type
which could lead to surprising behavior for lldbassert. Previously, when
doing a debug build with assertions disabled, lldbassert would become a
NOOP, rather than printing an error like it does in a release build. By
definining lldbassert in terms of NDEBUG, it behaves like a regular
assert when assertions are enabled, and like a soft assert.

Differential revision: https://reviews.llvm.org/D150639

If a symbol needs both JUMP_SLOT and GLOB_DAT relocations, there is a
minor linker optimization to keep just GLOB_DAT. This optimization
is only implemented by GNU ld's x86 port and mold.
https://maskray.me/blog/2021-08-29-all-about-global-offset-table#combining-.got-and-.got.plt

With the optimizing, the PLT entry is placed in .plt.got and the
associated GOTPLT entry is placed in .got (ld.bfd -z now) or .got.plt (ld.bfd -z lazy).
The relocation is in .rel[a].dyn.

This patch synthesizes `symbol@plt` labels for these .plt.got entries.

Example:
```
cat > a.s <<e
.globl _start; _start:
mov combined0@gotpcrel(%rip), %rax; mov combined1@gotpcrel(%rip), %rax
call combined0@plt; call combined1@plt
call foo0@plt; call foo1@plt
e
cat > b.s <<e
.globl foo0, foo1, combined0, combined1
foo0: foo1: combined0: combined1:
e
gcc -fuse-ld=bfd -shared b.s -o b.so
gcc -fuse-ld=bfd -pie -nostdlib a.s b.so -o a
```

```
Disassembly of section .plt:

0000000000001000 <.plt>:
    1000: ff 35 ea 1f 00 00             pushq   0x1fea(%rip)            # 0x2ff0 <_GLOBAL_OFFSET_TABLE_+0x8>
    1006: ff 25 ec 1f 00 00             jmpq    *0x1fec(%rip)           # 0x2ff8 <_GLOBAL_OFFSET_TABLE_+0x10>
    100c: 0f 1f 40 00                   nopl    (%rax)

0000000000001010 <foo1@plt>:
    1010: ff 25 ea 1f 00 00             jmpq    *0x1fea(%rip)           # 0x3000 <_GLOBAL_OFFSET_TABLE_+0x18>
    1016: 68 00 00 00 00                pushq   $0x0
    101b: e9 e0 ff ff ff                jmp     0x1000 <.plt>

0000000000001020 <foo0@plt>:
    1020: ff 25 e2 1f 00 00             jmpq    *0x1fe2(%rip)           # 0x3008 <_GLOBAL_OFFSET_TABLE_+0x20>
    1026: 68 01 00 00 00                pushq   $0x1
    102b: e9 d0 ff ff ff                jmp     0x1000 <.plt>

Disassembly of section .plt.got:

0000000000001030 <combined0@plt>:
    1030: ff 25 a2 1f 00 00             jmpq    *0x1fa2(%rip)           # 0x2fd8 <foo1+0x2fd8>
    1036: 66 90                         nop

0000000000001038 <combined1@plt>:
    1038: ff 25 a2 1f 00 00             jmpq    *0x1fa2(%rip)           # 0x2fe0 <foo1+0x2fe0>
    103e: 66 90                         nop
```

For x86-32, with -z now, if we remove `foo0` and `foo1`, the absence of regular
PLT will cause GNU ld to omit .got.plt, and our code cannot synthesize @plt
labels. This is an extreme corner case that almost never happens in practice (to
trigger the case, ensure every PLT symbol has been taken address). To fix it, we
can get the `_GLOBAL_OFFSET_TABLE_` symbol value, but the complexity is not
worth it.

Close https://github.com/llvm/llvm-project/issues/62537

Reviewed By: bd1976llvm

Differential Revision: https://reviews.llvm.org/D149817

These need to use the IRBuilder provided by InstCombine for proper
worklist management.

Differential Revision: https://reviews.llvm.org/D150668

This is a follow-up to D150608.

This reverts commit b6639930.

This caused a regression on aarch64:
https://lab.llvm.org/buildbot#builders/138/builds/43983

This patch is similar to 86fe88c8 and adds several explicit
constructor calls (bool(...), uint64_t(...), uint8_t(...)) that are
needed when we use UInt<T> (in my case UInt<128> in riscv32).

This patch also adds two operators to UInt<T>:
* operator/= required by printf_core/float_hex_converter.h:148
* operator-- required by FPUtil/ManipulationFunctions.h:166

Reviewed By: sivachandra, lntue

Differential Revision: https://reviews.llvm.org/D149594

Split Linalg vectorization tests from "vectorization.mlir" across more
specialised test files:
  * vectorize-tensor-extract.mlir - tests for tensor.extract with no
    masking,
  * vectorize-tensor-extract-masked.mlir - tests for tensor.extract with
    masking,
  * vectorization-masked.mlir - all other tests that use masking,
  * vectorisation.mlir - the remaining tests.

Differential Revision: https://reviews.llvm.org/D149843

Reviewed By: rnk

Differential Revision: https://reviews.llvm.org/D150645

Just add the trait as acc.yield is a return like op.

Reviewed By: razvanlupusoru, jeanPerier

Differential Revision: https://reviews.llvm.org/D150617

[AMDGPU][InferAddressSpaces] Only rewrite address-spaces that can be trivially casted to flat for llvm.amdgcn.flat.atomic.{fadd,fmax,fmin}

The intrinsic @llvm.amdgcn.flat.atomic.{fadd,fmax,fmin} can only be
selected for flat address spaces (constant, flat and global).

This patch restricts the cases over which GCNTTIImpl::rewriteIntrinsicWithAddressSpace
rewrites the intrinsic.

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D149938

Make it clearer minnum(+0, +0) cannot return -0. Also remove
a note about the result always being quiet which is directly
contradicted by the following paragraph.

Handle addrspacecast when looking at uses.

AllUsesOfValueWillTrapIfNull could handle addrspacecast, but currently
doesn't.

This reverts commit e2917311.

This caused some problems with lldb testing the diagnostic output:
https://lab.llvm.org/buildbot/#/builders/68/builds/52754

Breakage:
https://lab.llvm.org/buildbot/#/builders/92/builds/44222

Show line numbers to the left of diagnostic code snippets and increase
the numbers of lines shown from 1 to 16.

Differential Revision: https://reviews.llvm.org/D147875

MS link accepts *.obj with ehcont bit set only. LLD should match this
behavoir too.

Reviewed By: rnk

Differential Revision: https://reviews.llvm.org/D150508

This patch adds support for DWARFv5 .debug_line_str table.
It replaces code generating line table. Instead of copying original
table and patching certain places this patch implements full line table
generation.

Differential Revision: https://reviews.llvm.org/D150554

Make sure the code comments in processMemSetMemCpyDependence match
with the actual transform. They indicated that the memset being
rewritten was sunk to after a memcpy, while it actually is inserted
just before the memcpy.

Also make sure we use the debug location of the original memset
when creating the new simplified memset. In the past we've been
using the debug location for the memcpy which could be a bit
confusing.

Differential Revision: https://reviews.llvm.org/D135574

RegScavenger::backward is preferred because it does not rely on accurate
kill flags.

Differential Revision: https://reviews.llvm.org/D150571

This reverts commit 5cafecf9.

Buildbots are not happy with the patch.

Lots of crashes and assertion failures such as

  llvm::LegalizeRuleSet &llvm::LegalizerInfo::getActionDefinitionsBuilder(
  std::initializer_list<unsigned int>): Assertion `Opcodes.size() >= 2 &&
  "Initializer list must have at least two opcodes"' failed.

Summary:

1. we use the unsigned type for NextOffset,PrevOffset ,GlobalSymbolOffset , MemberTableSize, it will caused a malform big archive when the archive file size is large than 4G.
2. also fix a NFC comment on https://reviews.llvm.org/D142479#inline-1443927

Reviewers: James Henderson
Differential Revision: https://reviews.llvm.org/D150462

This remove a bunch of #include statements in Scalar.cpp. I do not
think those should be needed any longer (assuming that they once
upon a time possibly were needed for legacy PM C bindings, but
that is not supported any longer).

Also removing some other #include statements not needed any longer
due to deprecation of legacy PM.

Differential Revision: https://reviews.llvm.org/D149438

These intrinsics are used to implement the pext intrinsics that extract
two predicates (mask) from a predicate-as-counter value, e.g.

  __attribute__((arm_streaming))
   svboolx2_t svpext_lane_c8_x2(svcount_t pnn, uint64_t imm);

As described in https://github.com/ARM-software/acle/pull/217

Reviewed By: kmclaughlin

Differential Revision: https://reviews.llvm.org/D150442

These intrinsics are used to implement the pext intrinsics that extract
a predicate (mask) from a predicate-as-counter value, e.g.

   __attribute__((arm_streaming))
   svbool_t svpext_lane_c8(svcount_t pnn, uint64_t imm);

As described in https://github.com/ARM-software/acle/pull/217

Reviewed By: david-arm

Differential Revision: https://reviews.llvm.org/D150441

remark: unable to legalize instruction: %95:_(s16) = G_BSWAP %94:_ (in function: _ZNK4llvm13DataExtractor6getU16EPyPtj) [-Rpass-missed=gisel-legalize]

check plan: ninja check-llvm-codegen-x86

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D150667

https://reviews.llvm.org/D150586 removed a problematic test cases that caused failures on non-ppc buildbots. This patch revises the test case and adds it back.

Reviewed By: hubert.reinterpretcast

Differential Revision: https://reviews.llvm.org/D150597

Fixes #39713

fferential Revision: https://reviews.llvm.org/D150552

In D148674, the negative condition was weakened to only
checking isKnownNever(fcNegative), instead of cannotBeOrderedLessThanZero().

This avoids a regression when CannotBeOrderedLessThanZero is
replaced with computeKnownFPClass.

Copied from CannotBeNegativeZero and extended to cover fsub.