Age | Commit message (Expand) | Author | Files | Lines |
2020-06-17 | amdgcn: Switch to HSACO v3 binary format | Andrew Stubbs | 1 | -230/+5 |
2020-04-23 | amdgcn: Check HSA return codes [PR94629] | Andrew Stubbs | 1 | -0/+8 |
2020-02-13 | openmp: ignore nowait if async execution is unsupported [PR93481] | Frederik Harwath | 1 | -6/+1 |
2020-02-03 | Remove gfx801 "carrizo" support | Andrew Stubbs | 1 | -9/+0 |
2020-01-31 | [amdgcn] Scale number of threads/workers with VGPR usage | Kwok Cheung Yeung | 1 | -0/+23 |
2020-01-30 | Skip plugin-{gcn,hsa} for (-m)x32 (PR bootstrap/93409) | Tobias Burnus | 1 | -2/+2 |
2020-01-29 | Add OpenACC acc_get_property support for AMD GCN | Frederik Harwath | 1 | -10/+63 |
2020-01-23 | Fix libgomp plugin-gcn bug | Andrew Stubbs | 1 | -1/+2 |
2020-01-21 | Add runtime ISA check for amdgcn offloading | Frederik Harwath | 1 | -4/+127 |
2020-01-10 | OpenACC 'acc_get_property' cleanup | Thomas Schwinge | 3 | -105/+81 |
2020-01-01 | Update copyright years. | Jakub Jelinek | 7 | -7/+7 |
2019-12-28 | re PR bootstrap/93074 (build FAIL with --enable-offload-targets=nvptx-none) | Jakub Jelinek | 1 | -0/+6 |
2019-12-22 | Add OpenACC 2.6 `acc_get_property' support | Maciej W. Rozycki | 4 | -1/+127 |
2019-12-11 | libgomp – spelling fixes, incl. omp_lib.h.in | Tobias Burnus | 1 | -3/+3 |
2019-11-20 | AMD GCN libgomp plugin queue-full condition locking fix | Julian Brown | 1 | -13/+10 |
2019-11-20 | Fix host-to-device copies from rodata for AMD GCN | Julian Brown | 1 | -4/+34 |
2019-11-13 | GCN Libgomp Plugin | Andrew Stubbs | 3 | -0/+4034 |
2019-11-13 | Add device number to GOMP_OFFLOAD_openacc_async_construct | Andrew Stubbs | 1 | -1/+1 |
2019-10-01 | configure.ac: Remove GCC_HEADER_STDINT(gstdint.h). | Jakub Jelinek | 2 | -2/+1 |
2019-09-13 | libgomp plugin - init string | Tobias Burnus | 1 | -3/+3 |
2019-05-24 | re PR libgomp/90585 (libgomp hsa plugin ftbfs in the x32 multilib variant) | Jakub Jelinek | 2 | -9/+20 |
2019-05-17 | OpenACC Profiling Interface (incomplete) | Thomas Schwinge | 1 | -6/+190 |
2019-05-13 | 2019-05-13 Chung-Lin Tang <cltang@codesourcery.com> | Chung-Lin Tang | 3 | -1023/+317 |
2019-02-22 | [libgomp] In OpenACC testing, cycle though all offload targets | Thomas Schwinge | 1 | -0/+5 |
2019-02-22 | [libgomp] Clarify difference between offload target, offload plugin, and Open... | Thomas Schwinge | 1 | -17/+19 |
2019-01-24 | [nvptx, libgomp] Fix memleak in GOMP_OFFLOAD_fini_device | Tom de Vries | 1 | -0/+6 |
2019-01-23 | [nvptx, libgomp] Fix cuMemAlloc with size zero | Tom de Vries | 1 | -20/+28 |
2019-01-23 | [nvptx, libgomp] Fix assert (!s->map->active) in map_fini | Tom de Vries | 1 | -2/+25 |
2019-01-23 | [nvptx, libgomp] Fix map_push | Tom de Vries | 1 | -18/+29 |
2019-01-12 | [nvptx] Enable setting vector length using -fopenacc-dim | Tom de Vries | 1 | -1/+1 |
2019-01-12 | [nvptx] Update insufficient launch message for variable vector_length | Tom de Vries | 1 | -8/+10 |
2019-01-11 | [nvptx] Don't allow vector_length 64 with num_workers 16 | Tom de Vries | 1 | -0/+22 |
2019-01-09 | [nvptx, libgomp] Don't launch with num_workers == 0 | Tom de Vries | 1 | -0/+1 |
2019-01-01 | Update copyright years. | Jakub Jelinek | 6 | -6/+6 |
2018-12-14 | [PR88495] An OpenACC async queue is always synchronized with itself | Thomas Schwinge | 1 | -1/+2 |
2018-12-14 | [PR88407] [OpenACC] Correctly handle unseen async-arguments | Thomas Schwinge | 1 | -7/+6 |
2018-12-14 | [PR88370] acc_get_cuda_stream/acc_set_cuda_stream: acc_async_sync, acc_async_... | Thomas Schwinge | 1 | -2/+8 |
2018-09-18 | [nvptx] Remove use of CUDA unified memory in libgomp | Cesar Philippidis | 1 | -79/+91 |
2018-08-13 | [nvptx] Use CUDA driver API to select default runtime launch geometry | Cesar Philippidis | 3 | -12/+75 |
2018-08-08 | [libgomp, nvptx] Fall back to cuLinkAddData/cuLinkCreate if _v2 not found | Tom de Vries | 2 | -4/+26 |
2018-08-08 | [libgomp, nvptx] Allow cuGetErrorString to be NULL | Tom de Vries | 2 | -4/+8 |
2018-08-08 | [libgomp, nvptx] Remove hard-coded const in nvptx_open_device | Tom de Vries | 1 | -2/+5 |
2018-08-08 | [libgomp, nvptx] Note that cuGetErrorString is in CUDA_VERSION >= 6000 | Tom de Vries | 1 | -5/+4 |
2018-08-06 | [libgomp, nvptx] Handle CUDA_ONE_CALL_MAYBE_NULL | Tom de Vries | 1 | -3/+20 |
2018-08-06 | [libgomp, nvptx] Minimize lifetime of CUDA_ONE_CALL defines | Tom de Vries | 1 | -5/+9 |
2018-08-04 | [libgomp, nvptx, --without-cuda-driver] Don't use system cuda driver | Tom de Vries | 1 | -22/+33 |
2018-08-01 | [PATCH] Remove use of 'struct map' from plugin (nvptx) | Cesar Philippidis | 1 | -27/+6 |
2018-08-01 | [libgomp, nvptx] Add cuda-lib.def | Tom de Vries | 2 | -49/+48 |
2018-07-30 | [libgomp, nvptx] Handle per-function max-threads-per-block in default dims | Tom de Vries | 1 | -4/+25 |
2018-07-30 | [libgomp, nvptx] Calculate default dims per device | Tom de Vries | 1 | -7/+21 |