OSchip
/
llvm-project
mirror of https://github.com/THU-DSP-LAB/llvm-project.git
13
0
Fork 6
Code Issues Packages Projects Releases Wiki Activity
189,989 Commits 31 Branches 0 Tags 2.2 GiB
Commit Graph

76 Commits

Author SHA1 Message Date
Tom Stellard bf9f76fbe0 Implement log1p builtin 
llvm-svn: 219230
 2014-10-07 20:22:42 +00:00
Jan Vesely 8f64c3d842 Implement fmod 
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Tom Stellard <tom@stellard.net>
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 219087
 2014-10-05 20:24:52 +00:00
Tom Stellard 081e778d22 Implement async_work_group_copy builtin v3 
This is a simple implementation which just copies data synchronously.

v2:
  - Use size_t.

v3:
  - Fix possible race condition by splitting the copy among multiple
    work items.

llvm-svn: 219008
 2014-10-03 19:49:39 +00:00
Tom Stellard ed5bbfdb1b Implement async_work_group_strided_copy builtin v2 
This is a simple implementation which just copies data synchronously.

v2:
  - Use size_t.

llvm-svn: 219007
 2014-10-03 19:49:37 +00:00
Tom Stellard b5064f79ef Implement wait_group_events builtin v2 
This is a simple default implemetation which just calls barrier().

v2:
  - Only call barrier() once.

llvm-svn: 219006
 2014-10-03 19:49:34 +00:00
Aaron Watry 0d976ba497 atomic: Add generic atom[ic]_cmpxchg 
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 217918
 2014-09-16 22:34:49 +00:00
Aaron Watry 025d79ad6c atomic: Implement generic atom[ic]_xchg 
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 217917
 2014-09-16 22:34:45 +00:00
Aaron Watry 7cfa12c2a5 atomic: Add generic atomic_min implementation 
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 217916
 2014-09-16 22:34:41 +00:00
Aaron Watry 3f0a1a4c27 atomic: Add generic atom[ic]_xor 
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 217915
 2014-09-16 22:34:36 +00:00
Aaron Watry 31e67d1cff atomic: Add atom[ic]_or 
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 217914
 2014-09-16 22:34:32 +00:00
Aaron Watry cc68405761 atomics: Add generic atom[ic]_and 
Not used yet.

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 217913
 2014-09-16 22:34:28 +00:00
Aaron Watry 49614fbfd9 atomic: Add generic implementation of atom[ic]_max 
Not used yet...

v2: Correct int/uint behavior

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 217912
 2014-09-16 22:34:24 +00:00
Aaron Watry c9b88d32be atomic: define extension functions for existing atomic implementations 
We were missing the local versions of the atom_* before

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 217911
 2014-09-16 22:34:21 +00:00
Aaron Watry 947bdd059a math: Add tan implementation 
Uses the algorithm:
tan(x) = sin(x) / sqrt(1-sin^2(x))

An alternative is:
tan(x) = sin(x) / cos(x)

Which produces more verbose bitcode and longer assembly.

Either way, the generated bitcode seems pretty nasty and a more optimized
but still precise-enough solution is welcome.

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 217511
 2014-09-10 15:43:35 +00:00
Aaron Watry 951ab64d19 math: Add asin implementation 
asin(x) = atan2(x, sqrt( 1-x^2 ))

alternatively:
asin(x) = PI/2 - acos(x)

Use the atan2 implementation since it produces slightly shorter bitcode and
R600 machine code.

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 217510
 2014-09-10 15:43:32 +00:00
Aaron Watry 268beab921 math: Add acos implementation 
Passes the tests that were submitted to the piglit list

Tested on R600 (Pitcairn)

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 217509
 2014-09-10 15:43:29 +00:00
Jan Vesely 05a60b7ac3 add isordered builtin 
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 217247
 2014-09-05 13:59:15 +00:00
Jan Vesely 63486c1f0e add isunordered builtin 
v2: remove trailing newline

Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 217246
 2014-09-05 13:59:13 +00:00
Jan Vesely 41a0c491de add islessgreater builtin 
v2: remove trailing newline

Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 217245
 2014-09-05 13:59:11 +00:00
Jan Vesely 369e20353c add isnormal builtin 
v2: simplify and remove isnan leftovers
    remove trailing newline

Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 217244
 2014-09-05 13:59:09 +00:00
Jan Vesely a5a3b023b4 add isfinite builtin 
v2: simplify and remove isinf leftovers
    remove trailing newline

Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 217243
 2014-09-05 13:59:06 +00:00
Tom Stellard 7a9e2c6879 Implement isinf builtin 
llvm-svn: 217046
 2014-09-03 15:55:40 +00:00
Tom Stellard d8a73abfc3 Fix implementation of copysign 
This was previously implemented with a macro and we were using
__builtin_copysign(), which takes double inputs for the float
version of copysign().

Reviewed-and-Tested-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 217045
 2014-09-03 15:55:38 +00:00
Jan Vesely ef513d392b Implement generic mad_sat 
v2: Fix trailing whitespace
    Fix signed long overflow
    improve comment

v3: fix typo

Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Tom Stellard <tom@stellard.net>
llvm-svn: 216923
 2014-09-02 17:55:02 +00:00
Aaron Watry 9447097636 Revert "Implement generic mad_sat" 
This reverts commit cf62eded8b623a1c10d3692d25e5882b7939f564.

I didn't mean to commit this...  Jan has a v3 incoming

llvm-svn: 216322
 2014-08-23 14:06:01 +00:00
Aaron Watry 6bfac7ae69 Implement generic mad_sat 
v2: Fix trailing whitespace
    Fix signed long overflow
    improve comment

Signed-off-by: Jan Vesely <jan.vesely at rutgers.edu>
llvm-svn: 216320
 2014-08-23 14:04:33 +00:00
Tom Stellard 2ad4243bf7 Implement prefetch builtin 
The default implementation is a no-op.  Targets should override this
with their own implementations.

llvm-svn: 216127
 2014-08-20 21:23:03 +00:00
Aaron Watry f991505d02 vload/vstore: Use casts instead of scalarizing everything in CLC version 
This generates bitcode which is indistinguishable from what was
hand-written for int32 types in v[load|store]_impl.ll.

v4: Use vec2+scalar for vec3 load/stores to prevent corruption (per Tom)
v3: Also remove unused generic/lib/shared/v[load|store]_impl.ll
v2: (Per Matt Arsenault) Fix alignment issues with vector load stores

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
CC: Matt Arsenault <Matthew.Arsenault@amd.com>
CC: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 216069
 2014-08-20 13:58:57 +00:00
Jan Vesely 12c660827e relational: Add islessequal(floatN) builtin 
v2: remove the initial undef

Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 214568
 2014-08-01 21:50:59 +00:00
Jan Vesely acba2c98eb relational: Add isless(floatN) builtin 
v2: remove the initial undef

Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 214567
 2014-08-01 21:50:55 +00:00
Tom Stellard 903a78b7c6 Implement sin builtin for float types 
This double version still uses @llvm.sin.

llvm-svn: 213762
 2014-07-23 15:16:21 +00:00
Tom Stellard c0ab2f81e3 Implement cos builtin for float types 
The double version still uses @llvm.cos.

llvm-svn: 213761
 2014-07-23 15:16:18 +00:00
Tom Stellard f9caca8b9d Implement atan2 builtin 
llvm-svn: 213760
 2014-07-23 15:16:16 +00:00
Tom Stellard 47882923c7 Implement atan builtin 
llvm-svn: 213759
 2014-07-23 15:16:13 +00:00
Aaron Watry d7f022a582 relational: Implement isnotequal 
v2: Use relational macros instead of hand-rolled ones

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 213320
 2014-07-17 22:07:32 +00:00
Aaron Watry 30102536c0 relational: Implement isgreaterequal 
v2: Use relational macros instead of hand-rolled macros

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 213319
 2014-07-17 22:07:27 +00:00
Aaron Watry 803a992f04 relational: Implement isgreater 
v2: Use relational macros instead of hand-rolled macros

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 213318
 2014-07-17 22:07:19 +00:00
Aaron Watry d9ee196eab relational: Implement signbit 
v2 Changes:
   - use __builtin_signbit instead of shifting by hand
   - significantly improve vector shuffling
   - Works correctly now for signbit(float16) on radeonsi

Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 211696
 2014-06-25 13:29:23 +00:00
Jeroen Ketema 42df5d2a8f Add exp10 
Reviewed-by: Tom Stellard <tom@stellard.net>
llvm-svn: 211680
 2014-06-25 10:06:35 +00:00
Jeroen Ketema 09516fa27d Add pown 
Reviewed-by: Tom Stellard <tom@stellard.net>
llvm-svn: 211211
 2014-06-18 19:42:23 +00:00
Aaron Watry 6af2969a61 math: Implement mix builtin 
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <tom@stellard.net>
llvm-svn: 211047
 2014-06-16 19:53:59 +00:00
Aaron Watry f7f79d2a94 relational: Add isequal(floatN) builtin 
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <tom@stellard.net>
llvm-svn: 211046
 2014-06-16 19:53:57 +00:00
Aaron Watry e167db9238 Add all(igentype) builtin 
Signed-off-by: Aaron Watry <awatry@gmail.com>
Reviewed-by: Tom Stellard <tom@stellard.net>
llvm-svn: 211045
 2014-06-16 19:53:54 +00:00
Jeroen Ketema 82aaa41286 Implementations for exp(float) and exp(double) v2 
Use separate implementations instead of a macro
to ensure the constant multiplied with is of
higher precision.

v2: Use the correct formula, spotted by Dan Liew <daniel.liew@imperial.ac.uk>

Reviewed-by: Aaron Warty <awatry@gmail.com>
Reviewed-by: Tom Stellard <tom@stellard.net>
llvm-svn: 210891
 2014-06-13 09:40:09 +00:00
Tom Stellard 3a12fc6a07 Add sincos 
Patch by: Jeroen Ketema

Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 204478
 2014-03-21 16:22:01 +00:00
Tom Stellard 457e35912e Implement builtins for cl_khr_global_int32_base_atomics extension 
llvm-svn: 195021
 2013-11-18 18:21:23 +00:00
Tom Stellard 436bf70519 Implement sign() builtin 
llvm-svn: 192384
 2013-10-10 19:08:56 +00:00
Tom Stellard 6c7b86c106 Implement nextafter() builtin 
There are two implementations of nextafter():
1. Using clang's __builtin_nextafter.  Clang replaces this builtin with
a call to nextafter which is part of libm.  Therefore, this
implementation will only work for targets with an implementation of
libm (e.g. most CPU targets).

2. The other implementation is written in OpenCL C.  This function is
known internally as __clc_nextafter and can be used by targets that
don't have access to libm.

llvm-svn: 192383
 2013-10-10 19:08:51 +00:00
Tom Stellard e36e9dec65 Implement isnan() builtin 
llvm-svn: 192382
 2013-10-10 19:08:41 +00:00
Aaron Watry 50a7bcbac9 Add atomic_inc and atomic_add builtins 
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 190058
 2013-09-05 16:04:01 +00:00