diff options
| -rw-r--r-- | docs/ReleaseNotes.html | 24 |
1 files changed, 24 insertions, 0 deletions
diff --git a/docs/ReleaseNotes.html b/docs/ReleaseNotes.html index ab83daab87..036a0d2a4e 100644 --- a/docs/ReleaseNotes.html +++ b/docs/ReleaseNotes.html @@ -995,6 +995,7 @@ Builder.CreateResume(UnwindData); --> </ul> </div> + <!--=========================================================================--> <h3> @@ -1016,7 +1017,30 @@ Builder.CreateResume(UnwindData); <li>Initial support for MIPS64 has been added.</li> </ul> </div> + +<!--=========================================================================--> +<h3> + <a name="PTX">PTX Target Improvements</a> +</h3> + +<div> + + <p> + The PTX back-end is still experimental, but is fairly usable for compute kernels + in LLVM 3.0. Most scalar arithmetic is implemented, as well as intrinsics to + access the special PTX registers and sync instructions. The major missing + pieces are texture/sampler support and some vector operations.</p> + + <p>That said, the backend is already being used for domain-specific languages + and works well with the <a href="http://www.pcc.me.uk/~peter/libclc/">libclc + library</a> to supply OpenCL built-ins. With it, you can use Clang to compile + OpenCL code into PTX and execute it by loading the resulting PTX as a binary + blob using the nVidia OpenCL library. It has been tested with several OpenCL + programs, including some from the nVidia GPU Computing SDK, and the performance + is on par with the nVidia compiler.</p> +</div> + <!--=========================================================================--> <h3> <a name="OtherTS">Other Target Specific Improvements</a> |