[PATCH] D34407: Switch scalarize global loads ON by default

Alexander via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Mon Jul 3 07:55:24 PDT 2017 

This revision was automatically updated to reflect the committed changes.
Closed by commit rL307026: [AMDGPU] Switch scalarize global loads ON by default (authored by alex-t).

Changed prior to commit:
  https://reviews.llvm.org/D34407?vs=104850&id=105075#toc

Repository:
  rL LLVM

https://reviews.llvm.org/D34407

Files:
  llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
  llvm/trunk/test/CodeGen/AMDGPU/add.ll
  llvm/trunk/test/CodeGen/AMDGPU/add_i128.ll
  llvm/trunk/test/CodeGen/AMDGPU/add_i64.ll
  llvm/trunk/test/CodeGen/AMDGPU/and-gcn.ll
  llvm/trunk/test/CodeGen/AMDGPU/and.ll
  llvm/trunk/test/CodeGen/AMDGPU/any_extend_vector_inreg.ll
  llvm/trunk/test/CodeGen/AMDGPU/bitreverse.ll
  llvm/trunk/test/CodeGen/AMDGPU/bswap.ll
  llvm/trunk/test/CodeGen/AMDGPU/cgp-addressing-modes-flat.ll
  llvm/trunk/test/CodeGen/AMDGPU/cgp-addressing-modes.ll
  llvm/trunk/test/CodeGen/AMDGPU/coalescer_remat.ll
  llvm/trunk/test/CodeGen/AMDGPU/copy-illegal-type.ll
  llvm/trunk/test/CodeGen/AMDGPU/ctlz.ll
  llvm/trunk/test/CodeGen/AMDGPU/ctlz_zero_undef.ll
  llvm/trunk/test/CodeGen/AMDGPU/ctpop.ll
  llvm/trunk/test/CodeGen/AMDGPU/ctpop64.ll
  llvm/trunk/test/CodeGen/AMDGPU/cttz_zero_undef.ll
  llvm/trunk/test/CodeGen/AMDGPU/cvt_f32_ubyte.ll
  llvm/trunk/test/CodeGen/AMDGPU/early-if-convert-cost.ll
  llvm/trunk/test/CodeGen/AMDGPU/early-if-convert.ll
  llvm/trunk/test/CodeGen/AMDGPU/enable-no-signed-zeros-fp-math.ll
  llvm/trunk/test/CodeGen/AMDGPU/extractelt-to-trunc.ll
  llvm/trunk/test/CodeGen/AMDGPU/fabs.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fadd.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fadd64.ll
  llvm/trunk/test/CodeGen/AMDGPU/fcanonicalize.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fcanonicalize.ll
  llvm/trunk/test/CodeGen/AMDGPU/fcmp.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fcmp64.ll
  llvm/trunk/test/CodeGen/AMDGPU/fconst64.ll
  llvm/trunk/test/CodeGen/AMDGPU/fcopysign.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fma-combine.ll
  llvm/trunk/test/CodeGen/AMDGPU/fma.f64.ll
  llvm/trunk/test/CodeGen/AMDGPU/fma.ll
  llvm/trunk/test/CodeGen/AMDGPU/fmul.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fmul64.ll
  llvm/trunk/test/CodeGen/AMDGPU/fmuladd.f32.ll
  llvm/trunk/test/CodeGen/AMDGPU/fmuladd.f64.ll
  llvm/trunk/test/CodeGen/AMDGPU/fmuladd.v2f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fneg-fabs.ll
  llvm/trunk/test/CodeGen/AMDGPU/fneg.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fp32_to_fp16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fpext.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fptosi.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fptoui.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fptrunc.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fract.f64.ll
  llvm/trunk/test/CodeGen/AMDGPU/fract.ll
  llvm/trunk/test/CodeGen/AMDGPU/frem.ll
  llvm/trunk/test/CodeGen/AMDGPU/fsqrt.f64.ll
  llvm/trunk/test/CodeGen/AMDGPU/fsqrt.ll
  llvm/trunk/test/CodeGen/AMDGPU/fsub.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/fsub.ll
  llvm/trunk/test/CodeGen/AMDGPU/fsub64.ll
  llvm/trunk/test/CodeGen/AMDGPU/ftrunc.f64.ll
  llvm/trunk/test/CodeGen/AMDGPU/global-extload-i16.ll
  llvm/trunk/test/CodeGen/AMDGPU/half.ll
  llvm/trunk/test/CodeGen/AMDGPU/imm.ll
  llvm/trunk/test/CodeGen/AMDGPU/immv216.ll
  llvm/trunk/test/CodeGen/AMDGPU/indirect-addressing-si.ll
  llvm/trunk/test/CodeGen/AMDGPU/invariant-load-no-alias-store.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.class.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.class.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.div.fmas.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.ldexp.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.sbfe.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.sffbh.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.trig.preop.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.ubfe.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.ceil.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.cos.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.exp2.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.floor.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.fma.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.fmuladd.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.log2.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.maxnum.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.minnum.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.rint.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.sin.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.sqrt.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/llvm.trunc.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/load-global-f32.ll
  llvm/trunk/test/CodeGen/AMDGPU/load-global-f64.ll
  llvm/trunk/test/CodeGen/AMDGPU/load-global-i16.ll
  llvm/trunk/test/CodeGen/AMDGPU/load-global-i32.ll
  llvm/trunk/test/CodeGen/AMDGPU/load-global-i64.ll
  llvm/trunk/test/CodeGen/AMDGPU/load-global-i8.ll
  llvm/trunk/test/CodeGen/AMDGPU/load-weird-sizes.ll
  llvm/trunk/test/CodeGen/AMDGPU/max.ll
  llvm/trunk/test/CodeGen/AMDGPU/merge-stores.ll
  llvm/trunk/test/CodeGen/AMDGPU/mubuf.ll
  llvm/trunk/test/CodeGen/AMDGPU/mul.ll
  llvm/trunk/test/CodeGen/AMDGPU/no-shrink-extloads.ll
  llvm/trunk/test/CodeGen/AMDGPU/or.ll
  llvm/trunk/test/CodeGen/AMDGPU/promote-alloca-invariant-markers.ll
  llvm/trunk/test/CodeGen/AMDGPU/reduce-load-width-alignment.ll
  llvm/trunk/test/CodeGen/AMDGPU/reorder-stores.ll
  llvm/trunk/test/CodeGen/AMDGPU/rotl.i64.ll
  llvm/trunk/test/CodeGen/AMDGPU/rotr.i64.ll
  llvm/trunk/test/CodeGen/AMDGPU/rsq.ll
  llvm/trunk/test/CodeGen/AMDGPU/s_movk_i32.ll
  llvm/trunk/test/CodeGen/AMDGPU/saddo.ll
  llvm/trunk/test/CodeGen/AMDGPU/salu-to-valu.ll
  llvm/trunk/test/CodeGen/AMDGPU/scalar_to_vector.ll
  llvm/trunk/test/CodeGen/AMDGPU/schedule-global-loads.ll
  llvm/trunk/test/CodeGen/AMDGPU/scratch-buffer.ll
  llvm/trunk/test/CodeGen/AMDGPU/sdiv.ll
  llvm/trunk/test/CodeGen/AMDGPU/sdwa-peephole.ll
  llvm/trunk/test/CodeGen/AMDGPU/select-vectors.ll
  llvm/trunk/test/CodeGen/AMDGPU/select.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/sext-in-reg.ll
  llvm/trunk/test/CodeGen/AMDGPU/sgpr-copy-duplicate-operand.ll
  llvm/trunk/test/CodeGen/AMDGPU/sgpr-copy.ll
  llvm/trunk/test/CodeGen/AMDGPU/shift-and-i128-ubfe.ll
  llvm/trunk/test/CodeGen/AMDGPU/shift-and-i64-ubfe.ll
  llvm/trunk/test/CodeGen/AMDGPU/shift-i64-opts.ll
  llvm/trunk/test/CodeGen/AMDGPU/shl.ll
  llvm/trunk/test/CodeGen/AMDGPU/si-triv-disjoint-mem-access.ll
  llvm/trunk/test/CodeGen/AMDGPU/sign_extend.ll
  llvm/trunk/test/CodeGen/AMDGPU/sitofp.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/sminmax.ll
  llvm/trunk/test/CodeGen/AMDGPU/sminmax.v2i16.ll
  llvm/trunk/test/CodeGen/AMDGPU/spill-cfg-position.ll
  llvm/trunk/test/CodeGen/AMDGPU/sra.ll
  llvm/trunk/test/CodeGen/AMDGPU/srem.ll
  llvm/trunk/test/CodeGen/AMDGPU/srl.ll
  llvm/trunk/test/CodeGen/AMDGPU/ssubo.ll
  llvm/trunk/test/CodeGen/AMDGPU/sub.ll
  llvm/trunk/test/CodeGen/AMDGPU/trunc-bitcast-vector.ll
  llvm/trunk/test/CodeGen/AMDGPU/trunc.ll
  llvm/trunk/test/CodeGen/AMDGPU/uaddo.ll
  llvm/trunk/test/CodeGen/AMDGPU/udiv.ll
  llvm/trunk/test/CodeGen/AMDGPU/uitofp.f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/urem.ll
  llvm/trunk/test/CodeGen/AMDGPU/usubo.ll
  llvm/trunk/test/CodeGen/AMDGPU/v_mac.ll
  llvm/trunk/test/CodeGen/AMDGPU/v_mac_f16.ll
  llvm/trunk/test/CodeGen/AMDGPU/vectorize-global-local.ll
  llvm/trunk/test/CodeGen/AMDGPU/xor.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D34407.105075.patch
Type: text/x-patch
Size: 228758 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170703/90fc9c05/attachment-0001.bin>

More information about the llvm-commits mailing list