Commits

Branch selector

 master

User selector

jholewinski

Datepicker

All time

Commit History

Commits on Jul 22, 2013

• [NVPTX] Remove unused prototypes

  Show description for 6303ef1
  jholewinski

  committed
  6303ef1

  Copy full SHA for 6303ef1

• [NVPTX] Use approximate FP ops when unsafe-fp-math is used, and append

  Show description for 3a8ee4f
  jholewinski

  committed
  3a8ee4f

  Copy full SHA for 3a8ee4f

Commits on Jul 1, 2013

• [NVPTX] Add support for module-scope inline asm

  Show description for 728af3d
  jholewinski

  committed
  728af3d

  Copy full SHA for 728af3d

• [NVPTX] We dont use NVBuiltin anymore

  Show description for 1fd0917
  jholewinski

  committed
  1fd0917

  Copy full SHA for 1fd0917

• [NVPTX] Cut down on physical register defs

  Show description for c676590
  jholewinski

  committed
  c676590

  Copy full SHA for c676590

• [NVPTX] 64-bit ADDC/ADDE are not legal

  Show description for fc32eb4
  jholewinski

  committed
  fc32eb4

  Copy full SHA for fc32eb4

• [NVPTX] Fix vector loads from parameters that span multiple loads, and fix some typos

  Show description for be08c60
  jholewinski

  committed
  be08c60

  Copy full SHA for be08c60

• [NVPTX] Handle signext/zeroext attributes properly

  Show description for 03e5bb2
  jholewinski

  committed
  03e5bb2

  Copy full SHA for 03e5bb2

• [NVPTX] Add support for native SIGN_EXTEND_INREG where available

  Show description for 8834184
  jholewinski

  committed
  8834184

  Copy full SHA for 8834184

• [NVPTX] Add isel patterns for [reg+offset] form of ldg/ldu.

  Show description for 9bc8fee
  jholewinski

  committed
  9bc8fee

  Copy full SHA for 9bc8fee

• [NVPTX] Make sure we zero out high-order 24 bits for 8-bit load into 32-bit value

  Show description for 30b13eb
  jholewinski

  committed
  30b13eb

  Copy full SHA for 30b13eb

Commits on Jun 28, 2013

• [NVPTX] Select -1 instead of 1 when anyextend'ing i1 types

  Show description for b206498
  jholewinski

  committed
  b206498

  Copy full SHA for b206498

• [NVPTX] Add (1.0 / sqrt(x)) => rsqrt(x) generation when allowable by FP flags

  Show description for f51a22e
  jholewinski

  committed
  f51a22e

  Copy full SHA for f51a22e

• [NVPTX] Calling conventions fix

  Show description for ac78a06
  jholewinski

  committed
  ac78a06

  Copy full SHA for ac78a06

• [NVPTX] Add support for cttz/ctlz/ctpop

  Show description for 331ba27
  jholewinski

  committed
  331ba27

  Copy full SHA for 331ba27

• [NVPTX] Clean up comparison/select/convert patterns and factor out PTX instructions from their patterns

  Show description for ef0ccc9
  jholewinski

  committed
  ef0ccc9

  Copy full SHA for ef0ccc9

• [NVPTX] Remove i8 register class. PTX support for i8 (.b8, .u8, .s8) is rather poor and we're better off just ignoring it and letting LLVM expand all i8 ops out to i16.

  Show description for 1c07dae
  jholewinski

  committed
  1c07dae

  Copy full SHA for 1c07dae

• [NVPTX] Add support for vectorized function return values

  Show description for bc48ce8
  jholewinski

  committed
  bc48ce8

  Copy full SHA for bc48ce8

• [NVPTX] Clean up handling of formal arguments and enable generation of vector parameter loads

  Show description for b673665
  jholewinski

  committed
  b673665

  Copy full SHA for b673665

• [NVPTX] Add infrastructure for vector loads/stores of parameters

  Show description for 00df125
  jholewinski

  committed
  00df125

  Copy full SHA for 00df125

Commits on Jun 25, 2013

• [NVPTX] Default pointer type doesn't make sense for getParamSymbol()

  Show description for 5df8c1b
  jholewinski

  committed
  5df8c1b

  Copy full SHA for 5df8c1b

Commits on Jun 21, 2013

• [NVPTX] Add support for selecting CUDA vs OCL mode based on triple

  Show description for 0ff4287
  jholewinski

  committed
  0ff4287

  Copy full SHA for 0ff4287

Commits on Jun 10, 2013

• [NVPTX] Remove old CONST_NOT_GEN address space that is not being used anymore and causes constants to be emitted in the global address space

  Show description for 7c32502
  jholewinski

  committed
  7c32502

  Copy full SHA for 7c32502

Commits on May 31, 2013

• [NVPTX] Re-enable support for virtual registers in the final output

  Show description for 5443e7d
  jholewinski

  committed
  5443e7d

  Copy full SHA for 5443e7d

Commits on May 30, 2013

• [NVPTX] Fix case where a sext load of an i1 type may produce an

  Show description for d5c52f1
  jholewinski

  committed
  d5c52f1

  Copy full SHA for d5c52f1

Commits on May 21, 2013

• [NVPTX] Add @llvm.nvvm.sqrt.f() intrinsic

  Show description for b9c26dc
  jholewinski

  committed
  b9c26dc

  Copy full SHA for b9c26dc

• Drop @llvm.annotation and @llvm.ptr.annotation intrinsics during codegen.

  Show description for c2b7f5f
  jholewinski

  committed
  c2b7f5f

  Copy full SHA for c2b7f5f

Commits on May 20, 2013

• [NVPTX] Fix mis-use of CurrentFnSym in NVPTXAsmPrinter. This was causing a symbol name error in the output PTX.

  Show description for 9b39c72
  jholewinski

  committed
  9b39c72

  Copy full SHA for 9b39c72

• [NVPTX] Add programmatic interface to NVVMReflect pass

  Show description for d223675
  jholewinski

  committed
  d223675

  Copy full SHA for d223675

• [NVPTX] Add GenericToNVVM IR converter to better handle idiomatic LLVM IR inputs

  Show description for 7536ecf
  jholewinski

  committed
  7536ecf

  Copy full SHA for 7536ecf

• [NVPTX] Fix i1 kernel parameters and global variables. ABI rules say we need to use .u8 for i1 parameters for kernels.

  Show description for 55fdf53
  jholewinski

  committed
  55fdf53

  Copy full SHA for 55fdf53

Commits on Apr 2, 2013

• [NVPTX] Fix a few style issues in NVVMReflect

  Show description for d4ab192
  jholewinski

  committed
  d4ab192

  Copy full SHA for d4ab192

Commits on Mar 30, 2013

• Add start of user documentation for NVPTX

  Show description for 531ebc8
  jholewinski

  committed
  531ebc8

  Copy full SHA for 531ebc8

• [NVPTX] Remove support for SM < 2.0. This was never fully supported anyway.

  Show description for b24fc1c
  jholewinski

  committed
  b24fc1c

  Copy full SHA for b24fc1c

• [NVPTX] Add NVVMReflect pass to allow compile-time selection of

  Show description for 21fdcb0
  jholewinski

  committed
  21fdcb0

  Copy full SHA for 21fdcb0

Pagination

Previous

Next