and SUBE nodes that actually expose what's going on and allow for
significant simplifications in the targets.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26255 91177308-0d34-0410-b5e6-96231b3b80d8
other small targets that do that can be learned from. They also have
the added advantage of being tested :)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26243 91177308-0d34-0410-b5e6-96231b3b80d8
proves to be worth 20% on Ptrdist/ks. Might be related to dependency
breaking support.
2. Added FsMOVAPSrr and FsMOVAPDrr as aliases to MOVAPSrr and MOVAPDrr. These
are used for FR32 / FR64 reg-to-reg copies.
3. Tell reg-allocator to generate MOVSSrm / MOVSDrm and MOVSSmr / MOVSDmr to
spill / restore FsMOVAPSrr and FsMOVAPDrr.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26241 91177308-0d34-0410-b5e6-96231b3b80d8
and ComputeMaskedBits to match the new improved versions in instcombine.
Tested against all of multisource/benchmarks on ppc.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26238 91177308-0d34-0410-b5e6-96231b3b80d8
We do not want to emit "Loop: ... brcond Out; br Loop", as it adds an extra
instruction in the loop. Instead, invert the condition and emit
"Loop: ... br!cond Loop; br Out.
Generalize the fix by moving it from PPCDAGToDAGISel to SelectionDAGLowering.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26231 91177308-0d34-0410-b5e6-96231b3b80d8
transfer.
According to the Intel P4 Optimization Manual:
Moves that write a portion of a register can introduce unwanted
dependences. The movsd reg, reg instruction writes only the bottom
64 bits of a register, not to all 128 bits. This introduces a dependence on
the preceding instruction that produces the upper 64 bits (even if those
bits are not longer wanted). The dependence inhibits register renaming,
and thereby reduces parallelism.
Not to mention movaps is shorter than movss.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26226 91177308-0d34-0410-b5e6-96231b3b80d8
Turns them into calls to memset / memcpy if 1) buffer(s) are not DWORD aligned,
2) size is not known to be greater or equal to some minimum value (currently 128).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26224 91177308-0d34-0410-b5e6-96231b3b80d8
it, because this:
.bss
X:
.byte 0
results in the assembler warning: "initialization in bss segment". Annoying.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26204 91177308-0d34-0410-b5e6-96231b3b80d8
The ABI specifies that there is a register save area at the bottom of the
stack, which means the actual used pointer needs to be an offset from
the subtracted value.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26202 91177308-0d34-0410-b5e6-96231b3b80d8
emit it (instead of .space), saving a bit of space in the .o file.
For example:
int foo[100];
int bar[100] = {};
when compiled with C++ or -fno-common results in shrinkage from 1160 to 360
bytes of space. The X86 backend can also do this on darwin.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26185 91177308-0d34-0410-b5e6-96231b3b80d8
X86 addressing mode. Currently we do not allow any node whose target node
produces a chain as well as any node that is at the root of the addressing
mode expression tree.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26117 91177308-0d34-0410-b5e6-96231b3b80d8