Home
last modified time | relevance | path

Searched refs:speculative (Results 1 – 25 of 43) sorted by relevance

12

/llvm-project-15.0.7/llvm/test/CodeGen/NVPTX/
H A Dspeculative-execution-divergent-target.ll1 ; Checks that speculative-execution only runs on divergent targets, if you pass
4 ; RUN: opt < %s -S -mtriple=nvptx-nvidia-cuda -speculative-execution | \
6 ; RUN: opt < %s -S -mtriple=nvptx-nvidia-cuda -speculative-execution \
9 ; RUN: opt < %s -S -speculative-execution -spec-exec-only-if-divergent-target | \
/llvm-project-15.0.7/mlir/test/Dialect/OpenMP/
H A Dops.mlir596 // CHECK: omp.critical.declare @mutex4 hint(speculative)
597 omp.critical.declare @mutex4 hint(speculative)
603 omp.critical.declare @mutex7 hint(uncontended, speculative)
605 omp.critical.declare @mutex8 hint(contended, speculative)
804 omp.atomic.update hint(speculative) %x : memref<i32> {
1005 // CHECK: omp.atomic.capture hint(speculative) {
1013 omp.atomic.capture hint(speculative) {
1056 // CHECK: omp.atomic.capture hint(uncontended, speculative) {
1064 omp.atomic.capture hint(uncontended, speculative) {
1073 // CHECK: omp.atomic.capture hint(contended, speculative) {
[all …]
H A Dinvalid.mlir430 omp.critical.declare @mutex hint(nonspeculative, speculative)
501 omp.atomic.read %v = %x hint(speculative, nonspeculative) : memref<i32>
541 omp.atomic.read %v = %x hint(speculative) hint(contended) : memref<i32>
549 omp.atomic.read %x = %x hint(speculative) : memref<i32>
589 omp.atomic.write %addr = %val hint(contended) hint(speculative) : memref<i32>, i32
733 omp.atomic.update hint(nonspeculative, speculative) %x : memref<i32> {
912 omp.atomic.capture hint(nonspeculative, speculative) {
/llvm-project-15.0.7/llvm/test/CodeGen/X86/
H A Dspeculative-load-hardening-no-spill.ll1 ; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -x86-speculative-load-hardening | FileCheck %s
H A DO0-pipeline.ll40 ; CHECK-NEXT: X86 speculative load hardening
H A Dspeculative-load-hardening-call-and-ret.ll2 ; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -x86-speculative-load-hardening | FileCheck %s --…
3 ; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -x86-speculative-load-hardening -code-model mediu…
4 ; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -x86-speculative-load-hardening -relocation-model…
H A Dspeculative-load-hardening-indirect.ll5 ; RUN: llc < %s -debug-entry-values -mtriple=x86_64-unknown-linux-gnu -x86-speculative-load-hardeni…
7 ; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -x86-speculative-load-hardening -data-sections | …
9 ; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -x86-speculative-load-hardening -relocation-model…
10 ; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -x86-speculative-load-hardening -data-sections -m…
/llvm-project-15.0.7/lldb/test/Shell/SymbolFile/DWARF/x86/
H A DDW_AT_const_value.s17 ## Variables specified using string forms. This behavior purely speculative -- I
/llvm-project-15.0.7/llvm/test/Transforms/SpeculativeExecution/
H A Dspec-compares.ll1 ; RUN: opt < %s -S -passes=speculative-execution \
H A Dspec-calls.ll1 ; RUN: opt < %s -S -passes=speculative-execution \
H A Dspec-other.ll1 ; RUN: opt < %s -S -passes=speculative-execution \
H A Dspec-other-inseltpoison.ll1 ; RUN: opt < %s -S -passes=speculative-execution \
H A Dspec-fp.ll1 ; RUN: opt < %s -S -speculative-execution \
H A Dspec-casts.ll1 ; RUN: opt < %s -S -passes=speculative-execution \
H A DPR46267.ll1 ; RUN: opt < %s -S -passes='speculative-execution' | FileCheck %s
H A Dspec.ll1 ; RUN: opt < %s -S -passes=speculative-execution \
/llvm-project-15.0.7/llvm/test/CodeGen/AMDGPU/
H A Dspeculative-execution-freecasts.ll1 ; RUN: opt < %s -S -mtriple=amdgcn-unknown-amdhsa -speculative-execution \
/llvm-project-15.0.7/llvm/docs/
H A DSpeculativeLoadHardening.md10 vulnerabilities by exploiting speculative execution in modern CPUs. These
21 The core problem of GPZ Variant #1 is that speculative execution uses branch
25 speculative execution is unwound due to being incorrect. Mispredicted paths can
130 * It does not completely block speculative execution, and merely prevents
148 non-speculative execution. Code dealing with this, e.g cryptographic
246 speculative store then ends up being used by the CPU during speculative
247 execution of the return, potentially directing speculative execution to
276 speculative, this will often be possible even when these are stored in
311 speculative stores. However, as most stores aren't interesting and don't
920 forming these invalid stack pointer addresses we can prevent speculative
[all …]
H A DORCv2.rst434 compilation, concurrent JIT'd code, and speculative compilation.
950 *speculative* JIT compilation: compilation of code that is not needed yet,
953 example of speculative compilation with ORC has already been developed (see
957 speculative compilation.
/llvm-project-15.0.7/mlir/lib/Dialect/OpenMP/IR/
H A DOpenMPDialect.cpp430 bool speculative = bitn(hint, 3); in printSynchronizationHint() local
439 if (speculative) in printSynchronizationHint()
454 bool speculative = bitn(hint, 3); in verifySynchronizationHint() local
459 if (nonspeculative && speculative) in verifySynchronizationHint()
/llvm-project-15.0.7/llvm/test/Transforms/PGOProfile/
H A Dcounter_promo_exit_merge.ll1 ; RUN: opt < %s --passes=instrprof -do-counter-promotion=true -speculative-counter-promotion-max-ex…
H A Dcounter_promo_mexits.ll1 ; RUN: opt < %s --passes=pgo-instr-gen,instrprof -do-counter-promotion=true -speculative-counter-pr…
/llvm-project-15.0.7/llvm/test/Transforms/SimpleLoopUnswitch/
H A Dmsan.ll63 ; the actual problem is in the speculative branching.
/llvm-project-15.0.7/llvm/docs/HistoricalNotes/
H A D2000-11-18-EarlyDesignIdeas.txt42 aliased, but check at runtime", like speculative execution in
/llvm-project-15.0.7/clang-tools-extra/pseudo/
H A DDesignNotes.md105 By design GLR creates multiple speculative parse heads and lets incorrect heads

12