Home
last modified time | relevance | path

Searched refs:GEPs (Results 1 – 25 of 27) sorted by relevance

12

/external/llvm/test/Transforms/LoopStrengthReduce/
D2011-10-06-ReusePhi.ll4 ; Verify that scaled GEPs are not reused. rdar://5064068
17 ; GEPs are ugly
Daddress-space-loop.ll4 ; should be able to form pretty GEPs.
Duglygep-address-space.ll4 ; should be able to form pretty GEPs.
Duglygep.ll4 ; should be able to form pretty GEPs.
/external/llvm/test/Assembler/
Dgetelementptr.ll11 ; Verify that constant expression GEPs work with i84 indices.
30 ; Verify that constant expression vector GEPs work.
/external/llvm/test/Transforms/MergeFunc/
Dvector-GEP-crash.ll2 ; This used to cause a crash when compairing the GEPs
/external/llvm/test/Transforms/InstCombine/
D2011-05-13-InBoundsGEP.ll4 ; must not discard GEPs that might overflow at runtime (aren't inbounds)
Dgepphigep.ll60 ; Check that instcombine doesn't insert GEPs before landingpad.
103 ; Check that instcombine doesn't fold GEPs into themselves through a loop
Dalloca.ll54 ; Allocas which are only used by GEPs, bitcasts, and stores (transitively)
Dphi.ll136 ;; Suck GEPs into phi
/external/llvm/test/Analysis/BasicAA/
Dconstant-over-index.ll10 ; for GEPs to ignore TBAA.
/external/llvm/test/CodeGen/Generic/
DbadFoldGEP.ll7 ;; The code is legal but not type-safe and the two GEPs should not be folded.
D2003-05-30-BadFoldGEP.ll7 ;; Error: A sequence of GEPs is folded incorrectly by llc during selection
Dconstindices.ll26 ;; Test that a sequence of GEPs with constant indices are folded right
/external/llvm/docs/
DGetElementPtr.rst153 0`` index is needed. However, a closer inspection of how globals and GEPs work
266 What effect do address spaces have on GEPs?
311 How does VLA addressing work with GEPs?
314 GEPs don't natively support VLAs. LLVM's type system is entirely static, and GEP
374 Can I compare two values computed with GEPs?
445 As such, there are some ramifications of this for inbounds GEPs: scales implied
449 is logically treated as an unsigned value. This means that GEPs have an
499 reduces the range of possibilities for cases where two GEPs are effectively the
505 Some LLVM optimizers operate on GEPs by internally lowering them into more
DMergeFunctions.rst601 greater than other instructions, if both instructions are *GEPs* use ``cmpGEP``
629 Compares two GEPs (``getelementptr`` instructions).
634 So, if we get constant offset for both left and right *GEPs*, then compare it as
/external/llvm/test/CodeGen/PowerPC/
Dppc64-gep-opt.ll9 ; into GEPs with a single index, otherwise it will lower it into a
14 ; Check that when two complex GEPs are used in two basic blocks, LLVM can
/external/llvm/test/CodeGen/AArch64/
Daarch64-gep-opt.ll9 ; into GEPs with a single index, otherwise it will lower it into a
14 ; Check that when two complex GEPs are used in two basic blocks, LLVM can
/external/llvm/test/Transforms/InstSimplify/
Dptr_diff.ll36 ; Don't bother with non-inbounds GEPs.
Dcompare.ll755 ; Equality compares with non-inbounds GEPs can be folded.
767 ; Equality compares with non-inbounds GEPs can be folded.
/external/llvm/test/Instrumentation/BoundsChecking/
Dsimple.ll159 ; Self-refential GEPs can occur in dead code.
/external/llvm/test/Transforms/SROA/
Dphi-and-select.ll62 ; Note that we build redundant GEPs here to ensure that having different GEPs
Dbasictest.ll658 ; also gain enough data to prove they must be dead allocas due to GEPs that walk
723 ; different GEPs off the same alloca. We should only delete it once.
983 ; This test case tries to form GEPs into the empty leading struct members, and
/external/llvm/docs/Frontend/
DPerformanceTips.rst77 analysis), prefer GEPs
/external/llvm/lib/Target/X86/
DX86FastISel.cpp643 SmallVector<const Value *, 32> GEPs; in X86SelectAddress() local
770 GEPs.push_back(V); in X86SelectAddress()
787 I = GEPs.rbegin(), E = GEPs.rend(); I != E; ++I) in X86SelectAddress()

12