| #
8d12f40e
|
| 07-Feb-2025 |
Jeremy L Thompson <jeremy@jeremylt.org> |
hip - gen fallback to shared if error
|
| #
ddae5012
|
| 07-Feb-2025 |
Jeremy L Thompson <jeremy@jeremylt.org> |
cuda - gen fallback to shared if error
|
| #
de84fe53
|
| 25-Oct-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
Merge pull request #1698 from CEED/jeremy/debug-cleanup
Truncate jit output
|
| #
26ef7cda
|
| 22-Oct-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
debug - truncate jit output
|
| #
e036be40
|
| 22-Oct-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
Merge pull request #1699 from CEED/jeremy/set-jit-defines
Add CeedAddJitDefine
|
| #
4753b775
|
| 21-Oct-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
jit - add CeedAddJitDefine
|
| #
1dc8b1e6
|
| 21-Oct-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
Merge pull request #1696 from CEED/jeremy/jit-include
JiT include update
|
| #
91adc9c8
|
| 16-Oct-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
cuda/hip - include *-jit.h via #include
|
| #
a491a57e
|
| 16-Oct-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
jit - redirect ceed.h to ceed/types.h over ceed/ceed.h for jit
|
| #
b13efd58
|
| 16-Oct-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
jit - I include JiT source dirs set
|
| #
daaf13a4
|
| 16-Oct-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
jit - allow <> includes other than math, std*
|
| #
9e9230d4
|
| 23-May-2024 |
Jed Brown <jed@jedbrown.org> |
Merge pull request #1588 from CEED/jed/nvrtc-cubin
backends/cuda: NVRTC compile to CUBIN when supported (resolve #1587)
|
| #
29ec485e
|
| 22-May-2024 |
Jed Brown <jed@jedbrown.org> |
backends/cuda: NVRTC compile to CUBIN when supported (resolve #1587)
This allows using a newer CUDA runtime with an older driver, and seems
to have no downsides.
NVRTC can generate cubins directl
backends/cuda: NVRTC compile to CUBIN when supported (resolve #1587)
This allows using a newer CUDA runtime with an older driver, and seems
to have no downsides.
NVRTC can generate cubins directly starting with CUDA 11.1. [...]
NVRTC used to support only virtual architectures through the option
-arch, since it was only emitting PTX. It will now support actual
architectures as well to emit SASS. The interface is augmented to
retrieve either the PTX or cubin if an actual architecture is
specified.
https://docs.nvidia.com/cuda/cuda-c-best-practices-guide/index.html#dynamic-code-generation
show more ...
|
| #
a171b6ef
|
| 27-Mar-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
Merge pull request #1537 from CEED/jeremy/pragma-once
Use #pragma once for non-JiT headers
|
| #
5aed82e4
|
| 27-Mar-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
minor - update copyright headers
|
| #
b81bcd5a
|
| 06-Mar-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
Merge pull request #1504 from CEED/jeremy/debug-on-error
Write source to debug when CUDA/HIP fails to compile
|
| #
f5a0ec79
|
| 06-Mar-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
fix - need C string for debug printing
|
| #
edaedbd9
|
| 06-Mar-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
fix - print actual source code, not the defs only
|
| #
a4bfdec2
|
| 06-Mar-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
jit - write debug info when CUDA/HIP fails to compile
|
| #
1f70653f
|
| 23-Feb-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
Merge pull request #1488 from CEED/jeremy/const-path
Const for CeedQFunctionGetSourcePath
|
| #
22070f95
|
| 23-Feb-2024 |
Jeremy L Thompson <jeremy@jeremylt.org> |
internal - more updates for const
|
| #
21f16bf6
|
| 13-Sep-2023 |
Jeremy L Thompson <jeremy@jeremylt.org> |
Merge pull request #1315 from CEED/jeremy/queep
Backend Consistency Fixes
|
| #
ca735530
|
| 31-Aug-2023 |
Jeremy L Thompson <jeremy@jeremylt.org> |
style - fixes for CUDA backends
|
| #
3c708d80
|
| 24-Aug-2023 |
Jeremy L Thompson <jeremy@jeremylt.org> |
Merge pull request #1299 from CEED/sjg/cuda-hip-backends-consistency
Style fixes for `cuda` and `hip` backend consistency
|
| #
b2165e7a
|
| 11-Aug-2023 |
Sebastian Grimberg <sjg@amazon.com> |
Whitespace, style, and formatting updates for consistency between CUDA and HIP backends
Adds include guards in JiT header files, even if not strictly necessary, to match the precedent set in cuda-sh
Whitespace, style, and formatting updates for consistency between CUDA and HIP backends
Adds include guards in JiT header files, even if not strictly necessary, to match the precedent set in cuda-shared and hip-shared as well as sycl.
show more ...
|