diff options
Diffstat (limited to 'erts/emulator/beam/erl_sched_spec_pre_alloc.c')
-rw-r--r-- | erts/emulator/beam/erl_sched_spec_pre_alloc.c | 45 |
1 files changed, 33 insertions, 12 deletions
diff --git a/erts/emulator/beam/erl_sched_spec_pre_alloc.c b/erts/emulator/beam/erl_sched_spec_pre_alloc.c index 96238318c9..4a6e02281a 100644 --- a/erts/emulator/beam/erl_sched_spec_pre_alloc.c +++ b/erts/emulator/beam/erl_sched_spec_pre_alloc.c @@ -32,13 +32,12 @@ # include "config.h" #endif -#ifdef ERTS_SMP #include "erl_process.h" #include "erl_thr_progress.h" erts_sspa_data_t * -erts_sspa_create(size_t blk_sz, int pa_size) +erts_sspa_create(size_t blk_sz, int pa_size, int nthreads, const char* name) { erts_sspa_data_t *data; size_t tot_size; @@ -48,23 +47,40 @@ erts_sspa_create(size_t blk_sz, int pa_size) int cix; int no_blocks = pa_size; int no_blocks_per_chunk; + size_t aligned_blk_sz; - if (erts_no_schedulers == 1) +#if !defined(ERTS_STRUCTURE_ALIGNED_ALLOC) + /* Force 64-bit alignment... */ + aligned_blk_sz = ((blk_sz - 1) / 8) * 8 + 8; +#else + /* Alignment of structure is enough... */ + aligned_blk_sz = blk_sz; +#endif + + if (!name) { /* schedulers only variant */ + ASSERT(!nthreads); + nthreads = erts_no_schedulers; + } + else { + ASSERT(nthreads > 0); + } + + if (nthreads == 1) no_blocks_per_chunk = no_blocks; else { int extra = (no_blocks - 1)/4 + 1; if (extra == 0) extra = 1; no_blocks_per_chunk = no_blocks; - no_blocks_per_chunk += extra*erts_no_schedulers; - no_blocks_per_chunk /= erts_no_schedulers; + no_blocks_per_chunk += extra * nthreads; + no_blocks_per_chunk /= nthreads; } - no_blocks = no_blocks_per_chunk * erts_no_schedulers; + no_blocks = no_blocks_per_chunk * nthreads; chunk_mem_size = ERTS_ALC_CACHE_LINE_ALIGN_SIZE(sizeof(erts_sspa_chunk_header_t)); - chunk_mem_size += blk_sz * no_blocks_per_chunk; + chunk_mem_size += aligned_blk_sz * no_blocks_per_chunk; chunk_mem_size = ERTS_ALC_CACHE_LINE_ALIGN_SIZE(chunk_mem_size); tot_size = ERTS_ALC_CACHE_LINE_ALIGN_SIZE(sizeof(erts_sspa_data_t)); - tot_size += chunk_mem_size*erts_no_schedulers; + tot_size += chunk_mem_size * nthreads; p = erts_alloc_permanent_cache_aligned(ERTS_ALC_T_PRE_ALLOC_DATA, tot_size); data = (erts_sspa_data_t *) p; @@ -73,10 +89,16 @@ erts_sspa_create(size_t blk_sz, int pa_size) data->chunks_mem_size = chunk_mem_size; data->start = chunk_start; - data->end = chunk_start + chunk_mem_size*erts_no_schedulers; + data->end = chunk_start + chunk_mem_size * nthreads; + data->nthreads = nthreads; + + if (name) { /* thread variant */ + erts_tsd_key_create(&data->tsd_key, (char*)name); + erts_atomic_init_nob(&data->id_generator, 0); + } /* Initialize all chunks */ - for (cix = 0; cix < erts_no_schedulers; cix++) { + for (cix = 0; cix < nthreads; cix++) { erts_sspa_chunk_t *chnk = erts_sspa_cix2chunk(data, cix); erts_sspa_chunk_header_t *chdr = &chnk->aligned.header; erts_sspa_blk_t *blk; @@ -102,7 +124,7 @@ erts_sspa_create(size_t blk_sz, int pa_size) blk = (erts_sspa_blk_t *) p; for (i = 0; i < no_blocks_per_chunk; i++) { blk = (erts_sspa_blk_t *) p; - p += blk_sz; + p += aligned_blk_sz; blk->next_ptr = (erts_sspa_blk_t *) p; } @@ -325,4 +347,3 @@ erts_sspa_process_remote_frees(erts_sspa_chunk_header_t *chdr, return res; } -#endif /* ERTS_SMP */ |