llvm-project/libclc
Aaron Watry 8548725f29 Add rhadd builtin
rhadd = (x+y+1)>>1

Implemented as:
(x>>1) + (y>>1) + ((x&1)|(y&1))

This prevents us having to do assembly addition and overflow detection

Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 188477
2013-08-15 19:21:10 +00:00
..
build configure.py: Add an install rule. 2012-06-01 17:29:59 +00:00
generic Add rhadd builtin 2013-08-15 19:21:10 +00:00
ptx/lib libclc: Rename [add|sub]_sat.ll to [add|sub]_sat_if.ll 2013-06-26 18:21:31 +00:00
ptx-nvidiacl/lib Add barrier.cl to SOURCES, spotted by Jin Wang. 2012-09-05 18:13:55 +00:00
r600/lib Enable assembly vload3 int/uint constant/global for R600 2013-08-12 14:42:50 +00:00
test Add rsqrt builtin. Based on patch by Cassie Epps! 2012-08-21 10:48:35 +00:00
utils Fix build with LLVM 3.4 2013-07-18 21:24:35 +00:00
www Switch to BSD/MIT dual license. 2012-02-22 04:47:39 +00:00
CREDITS.TXT
LICENSE.TXT Update the copyright coredits -- Happy new year 2013! 2013-01-01 10:00:19 +00:00
README.TXT
compile-test.sh Fix build against recent versions of Clang. Based on patch by Alastair Donaldson! 2012-12-05 07:39:02 +00:00
configure.py configure: Fix build when clang is installed to a non-standard prefix 2013-08-10 03:40:26 +00:00

README.TXT

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

libclc
------

libclc is an open source, BSD licensed implementation of the library
requirements of the OpenCL C programming language, as specified by the
OpenCL 1.1 Specification. The following sections of the specification
impose library requirements:

  * 6.1: Supported Data Types
  * 6.2.3: Explicit Conversions
  * 6.2.4.2: Reinterpreting Types Using as_type() and as_typen()
  * 6.9: Preprocessor Directives and Macros
  * 6.11: Built-in Functions
  * 9.3: Double Precision Floating-Point
  * 9.4: 64-bit Atomics
  * 9.5: Writing to 3D image memory objects
  * 9.6: Half Precision Floating-Point

libclc is intended to be used with the Clang compiler's OpenCL frontend.

libclc is designed to be portable and extensible. To this end, it provides
generic implementations of most library requirements, allowing the target
to override the generic implementation at the granularity of individual
functions.

libclc currently only supports the PTX target, but support for more
targets is welcome.

Compiling
---------

./configure.py --with-llvm-config=/path/to/llvm-config && make

Website
-------

http://www.pcc.me.uk/~peter/libclc/