diff options
| author | Tom de Vries <tdevries@suse.de> | 2019-01-11 11:46:31 +0000 |
|---|---|---|
| committer | Tom de Vries <vries@gcc.gnu.org> | 2019-01-11 11:46:31 +0000 |
| commit | 69b09a587dfddff912bfd1e5d69d3cb0d67c3895 (patch) | |
| tree | bea69be757dd0df5525f7a823608a9911ac3a767 | |
| parent | 9390f91687a994c2ea28b3058b606f67d683a6ab (diff) | |
| download | gcc-69b09a587dfddff912bfd1e5d69d3cb0d67c3895.zip gcc-69b09a587dfddff912bfd1e5d69d3cb0d67c3895.tar.gz gcc-69b09a587dfddff912bfd1e5d69d3cb0d67c3895.tar.bz2 | |
[nvptx] Move PTX_CTA_SIZE up
Move the defition of PTX_CTA_SIZE up in nvptx.c.
2019-01-11 Tom de Vries <tdevries@suse.de>
* config/nvptx/nvptx.c (PTX_CTA_SIZE): Move up.
From-SVN: r267837
| -rw-r--r-- | gcc/ChangeLog | 4 | ||||
| -rw-r--r-- | gcc/config/nvptx/nvptx.c | 11 |
2 files changed, 10 insertions, 5 deletions
diff --git a/gcc/ChangeLog b/gcc/ChangeLog index 0808254..5495b9a 100644 --- a/gcc/ChangeLog +++ b/gcc/ChangeLog @@ -1,3 +1,7 @@ +2019-01-11 Tom de Vries <tdevries@suse.de> + + * config/nvptx/nvptx.c (PTX_CTA_SIZE): Move up. + 2019-01-11 Jan Beulich <jbeulich@suse.com> * config/i386/i386.md (rex64suffix): Add L suffix for SI. diff --git a/gcc/config/nvptx/nvptx.c b/gcc/config/nvptx/nvptx.c index 7fdc285..643f5e8 100644 --- a/gcc/config/nvptx/nvptx.c +++ b/gcc/config/nvptx/nvptx.c @@ -82,17 +82,18 @@ #define WORKAROUND_PTXJIT_BUG_2 1 #define WORKAROUND_PTXJIT_BUG_3 1 +/* The PTX concept CTA (Concurrent Thread Array) maps on the CUDA concept thread + block, which has had a maximum number of threads of 1024 since CUDA version + 2.x. */ +#define PTX_CTA_SIZE 1024 + #define PTX_WARP_SIZE 32 + #define PTX_DEFAULT_VECTOR_LENGTH PTX_WARP_SIZE #define PTX_MAX_VECTOR_LENGTH PTX_WARP_SIZE #define PTX_WORKER_LENGTH 32 #define PTX_DEFAULT_RUNTIME_DIM 0 /* Defer to runtime. */ -/* The PTX concept CTA (Concurrent Thread Array) maps on the CUDA concept thread - block, which has had a maximum number of threads of 1024 since CUDA version - 2.x. */ -#define PTX_CTA_SIZE 1024 - /* The various PTX memory areas an object might reside in. */ enum nvptx_data_area { |