From: Samuel Pitoiset Date: Sun, 4 Oct 2015 16:28:55 +0000 (+0200) Subject: nvc0: move HW queries to nvc0_query_hw.c/h files X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=224fec05eac136d734e6ae06f6aab44d5ba640df;p=mesa.git nvc0: move HW queries to nvc0_query_hw.c/h files Signed-off-by: Samuel Pitoiset --- diff --git a/src/gallium/drivers/nouveau/Makefile.sources b/src/gallium/drivers/nouveau/Makefile.sources index 0e1cb19bed7..e45c564f431 100644 --- a/src/gallium/drivers/nouveau/Makefile.sources +++ b/src/gallium/drivers/nouveau/Makefile.sources @@ -152,6 +152,8 @@ NVC0_C_SOURCES := \ nvc0/nvc0_program.h \ nvc0/nvc0_query.c \ nvc0/nvc0_query.h \ + nvc0/nvc0_query_hw.c \ + nvc0/nvc0_query_hw.h \ nvc0/nvc0_query_sw.c \ nvc0/nvc0_query_sw.h \ nvc0/nvc0_resource.c \ diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_query.c b/src/gallium/drivers/nouveau/nvc0/nvc0_query.c index 9409bdea81f..a0ca3fa533c 100644 --- a/src/gallium/drivers/nouveau/nvc0/nvc0_query.c +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_query.c @@ -27,485 +27,48 @@ #include "nvc0/nvc0_context.h" #include "nvc0/nvc0_query.h" #include "nvc0/nvc0_query_sw.h" - -#include "nv_object.xml.h" -#include "nvc0/nve4_compute.xml.h" -#include "nvc0/nvc0_compute.xml.h" - -#define NVC0_QUERY_STATE_READY 0 -#define NVC0_QUERY_STATE_ACTIVE 1 -#define NVC0_QUERY_STATE_ENDED 2 -#define NVC0_QUERY_STATE_FLUSHED 3 - -#define NVC0_QUERY_ALLOC_SPACE 256 - -static boolean nvc0_hw_sm_query_begin(struct nvc0_context *, - struct nvc0_query *); -static void nvc0_hw_sm_query_end(struct nvc0_context *, struct nvc0_query *); -static boolean nvc0_hw_sm_query_result(struct nvc0_context *, - struct nvc0_query *, void *, boolean); - -static bool -nvc0_query_allocate(struct nvc0_context *nvc0, struct nvc0_query *q, int size) -{ - struct nvc0_screen *screen = nvc0->screen; - int ret; - - if (q->bo) { - nouveau_bo_ref(NULL, &q->bo); - if (q->mm) { - if (q->state == NVC0_QUERY_STATE_READY) - nouveau_mm_free(q->mm); - else - nouveau_fence_work(screen->base.fence.current, - nouveau_mm_free_work, q->mm); - } - } - if (size) { - q->mm = nouveau_mm_allocate(screen->base.mm_GART, size, &q->bo, &q->base); - if (!q->bo) - return false; - q->offset = q->base; - - ret = nouveau_bo_map(q->bo, 0, screen->base.client); - if (ret) { - nvc0_query_allocate(nvc0, q, 0); - return false; - } - q->data = (uint32_t *)((uint8_t *)q->bo->map + q->base); - } - return true; -} - -static void -nvc0_query_destroy(struct pipe_context *pipe, struct pipe_query *pq) -{ - nvc0_query_allocate(nvc0_context(pipe), nvc0_query(pq), 0); - nouveau_fence_ref(NULL, &nvc0_query(pq)->fence); - FREE(nvc0_query(pq)); -} +#include "nvc0/nvc0_query_hw.h" static struct pipe_query * -nvc0_query_create(struct pipe_context *pipe, unsigned type, unsigned index) +nvc0_create_query(struct pipe_context *pipe, unsigned type, unsigned index) { struct nvc0_context *nvc0 = nvc0_context(pipe); struct nvc0_query *q; - unsigned space = NVC0_QUERY_ALLOC_SPACE; q = nvc0_sw_create_query(nvc0, type, index); - if (q) - return (struct pipe_query *)q; - - q = CALLOC_STRUCT(nvc0_query); if (!q) - return NULL; - - switch (type) { - case PIPE_QUERY_OCCLUSION_COUNTER: - case PIPE_QUERY_OCCLUSION_PREDICATE: - q->rotate = 32; - space = NVC0_QUERY_ALLOC_SPACE; - break; - case PIPE_QUERY_PIPELINE_STATISTICS: - q->is64bit = true; - space = 512; - break; - case PIPE_QUERY_SO_STATISTICS: - case PIPE_QUERY_SO_OVERFLOW_PREDICATE: - q->is64bit = true; - space = 64; - break; - case PIPE_QUERY_PRIMITIVES_GENERATED: - case PIPE_QUERY_PRIMITIVES_EMITTED: - q->is64bit = true; - q->index = index; - space = 32; - break; - case PIPE_QUERY_TIME_ELAPSED: - case PIPE_QUERY_TIMESTAMP: - case PIPE_QUERY_TIMESTAMP_DISJOINT: - case PIPE_QUERY_GPU_FINISHED: - space = 32; - break; - case NVC0_QUERY_TFB_BUFFER_OFFSET: - space = 16; - break; - default: - if (nvc0->screen->base.device->drm_version >= 0x01000101) { - if (type >= NVE4_HW_SM_QUERY(0) && type <= NVE4_HW_SM_QUERY_LAST) { - /* for each MP: - * [00] = WS0.C0 - * [04] = WS0.C1 - * [08] = WS0.C2 - * [0c] = WS0.C3 - * [10] = WS1.C0 - * [14] = WS1.C1 - * [18] = WS1.C2 - * [1c] = WS1.C3 - * [20] = WS2.C0 - * [24] = WS2.C1 - * [28] = WS2.C2 - * [2c] = WS2.C3 - * [30] = WS3.C0 - * [34] = WS3.C1 - * [38] = WS3.C2 - * [3c] = WS3.C3 - * [40] = MP.C4 - * [44] = MP.C5 - * [48] = MP.C6 - * [4c] = MP.C7 - * [50] = WS0.sequence - * [54] = WS1.sequence - * [58] = WS2.sequence - * [5c] = WS3.sequence - */ - space = (4 * 4 + 4 + 4) * nvc0->screen->mp_count * sizeof(uint32_t); - break; - } else - if (type >= NVC0_HW_SM_QUERY(0) && type <= NVC0_HW_SM_QUERY_LAST) { - /* for each MP: - * [00] = MP.C0 - * [04] = MP.C1 - * [08] = MP.C2 - * [0c] = MP.C3 - * [10] = MP.C4 - * [14] = MP.C5 - * [18] = MP.C6 - * [1c] = MP.C7 - * [20] = MP.sequence - */ - space = (8 + 1) * nvc0->screen->mp_count * sizeof(uint32_t); - break; - } - } - debug_printf("invalid query type: %u\n", type); - FREE(q); - return NULL; - } - if (!nvc0_query_allocate(nvc0, q, space)) { - FREE(q); - return NULL; - } - - q->type = type; - - if (q->rotate) { - /* we advance before query_begin ! */ - q->offset -= q->rotate; - q->data -= q->rotate / sizeof(*q->data); - } else - if (!q->is64bit) - q->data[0] = 0; /* initialize sequence */ + q = nvc0_hw_create_query(nvc0, type, index); return (struct pipe_query *)q; } static void -nvc0_query_get(struct nouveau_pushbuf *push, struct nvc0_query *q, - unsigned offset, uint32_t get) +nvc0_destroy_query(struct pipe_context *pipe, struct pipe_query *pq) { - offset += q->offset; - - PUSH_SPACE(push, 5); - PUSH_REFN (push, q->bo, NOUVEAU_BO_GART | NOUVEAU_BO_WR); - BEGIN_NVC0(push, NVC0_3D(QUERY_ADDRESS_HIGH), 4); - PUSH_DATAh(push, q->bo->offset + offset); - PUSH_DATA (push, q->bo->offset + offset); - PUSH_DATA (push, q->sequence); - PUSH_DATA (push, get); -} - -static void -nvc0_query_rotate(struct nvc0_context *nvc0, struct nvc0_query *q) -{ - q->offset += q->rotate; - q->data += q->rotate / sizeof(*q->data); - if (q->offset - q->base == NVC0_QUERY_ALLOC_SPACE) - nvc0_query_allocate(nvc0, q, NVC0_QUERY_ALLOC_SPACE); + struct nvc0_query *q = nvc0_query(pq); + q->funcs->destroy_query(nvc0_context(pipe), q); } static boolean -nvc0_query_begin(struct pipe_context *pipe, struct pipe_query *pq) +nvc0_begin_query(struct pipe_context *pipe, struct pipe_query *pq) { - struct nvc0_context *nvc0 = nvc0_context(pipe); - struct nouveau_pushbuf *push = nvc0->base.pushbuf; struct nvc0_query *q = nvc0_query(pq); - bool ret = true; - - /* For occlusion queries we have to change the storage, because a previous - * query might set the initial render conition to false even *after* we re- - * initialized it to true. - */ - if (q->rotate) { - nvc0_query_rotate(nvc0, q); - - /* XXX: can we do this with the GPU, and sync with respect to a previous - * query ? - */ - q->data[0] = q->sequence; /* initialize sequence */ - q->data[1] = 1; /* initial render condition = true */ - q->data[4] = q->sequence + 1; /* for comparison COND_MODE */ - q->data[5] = 0; - } - q->sequence++; - - switch (q->type) { - case PIPE_QUERY_OCCLUSION_COUNTER: - case PIPE_QUERY_OCCLUSION_PREDICATE: - q->nesting = nvc0->screen->num_occlusion_queries_active++; - if (q->nesting) { - nvc0_query_get(push, q, 0x10, 0x0100f002); - } else { - PUSH_SPACE(push, 3); - BEGIN_NVC0(push, NVC0_3D(COUNTER_RESET), 1); - PUSH_DATA (push, NVC0_3D_COUNTER_RESET_SAMPLECNT); - IMMED_NVC0(push, NVC0_3D(SAMPLECNT_ENABLE), 1); - } - break; - case PIPE_QUERY_PRIMITIVES_GENERATED: - nvc0_query_get(push, q, 0x10, 0x09005002 | (q->index << 5)); - break; - case PIPE_QUERY_PRIMITIVES_EMITTED: - nvc0_query_get(push, q, 0x10, 0x05805002 | (q->index << 5)); - break; - case PIPE_QUERY_SO_STATISTICS: - nvc0_query_get(push, q, 0x20, 0x05805002 | (q->index << 5)); - nvc0_query_get(push, q, 0x30, 0x06805002 | (q->index << 5)); - break; - case PIPE_QUERY_SO_OVERFLOW_PREDICATE: - nvc0_query_get(push, q, 0x10, 0x03005002 | (q->index << 5)); - break; - case PIPE_QUERY_TIME_ELAPSED: - nvc0_query_get(push, q, 0x10, 0x00005002); - break; - case PIPE_QUERY_PIPELINE_STATISTICS: - nvc0_query_get(push, q, 0xc0 + 0x00, 0x00801002); /* VFETCH, VERTICES */ - nvc0_query_get(push, q, 0xc0 + 0x10, 0x01801002); /* VFETCH, PRIMS */ - nvc0_query_get(push, q, 0xc0 + 0x20, 0x02802002); /* VP, LAUNCHES */ - nvc0_query_get(push, q, 0xc0 + 0x30, 0x03806002); /* GP, LAUNCHES */ - nvc0_query_get(push, q, 0xc0 + 0x40, 0x04806002); /* GP, PRIMS_OUT */ - nvc0_query_get(push, q, 0xc0 + 0x50, 0x07804002); /* RAST, PRIMS_IN */ - nvc0_query_get(push, q, 0xc0 + 0x60, 0x08804002); /* RAST, PRIMS_OUT */ - nvc0_query_get(push, q, 0xc0 + 0x70, 0x0980a002); /* ROP, PIXELS */ - nvc0_query_get(push, q, 0xc0 + 0x80, 0x0d808002); /* TCP, LAUNCHES */ - nvc0_query_get(push, q, 0xc0 + 0x90, 0x0e809002); /* TEP, LAUNCHES */ - break; - default: -#ifdef NOUVEAU_ENABLE_DRIVER_STATISTICS - if (q->type >= NVC0_SW_QUERY_DRV_STAT(0) && - q->type <= NVC0_SW_QUERY_DRV_STAT_LAST) { - return q->funcs->begin_query(nvc0, q); - } else -#endif - if ((q->type >= NVE4_HW_SM_QUERY(0) && q->type <= NVE4_HW_SM_QUERY_LAST) || - (q->type >= NVC0_HW_SM_QUERY(0) && q->type <= NVC0_HW_SM_QUERY_LAST)) { - ret = nvc0_hw_sm_query_begin(nvc0, q); - } - break; - } - q->state = NVC0_QUERY_STATE_ACTIVE; - return ret; + return q->funcs->begin_query(nvc0_context(pipe), q); } static void -nvc0_query_end(struct pipe_context *pipe, struct pipe_query *pq) +nvc0_end_query(struct pipe_context *pipe, struct pipe_query *pq) { - struct nvc0_context *nvc0 = nvc0_context(pipe); - struct nouveau_pushbuf *push = nvc0->base.pushbuf; struct nvc0_query *q = nvc0_query(pq); - - if (q->state != NVC0_QUERY_STATE_ACTIVE) { - /* some queries don't require 'begin' to be called (e.g. GPU_FINISHED) */ - if (q->rotate) - nvc0_query_rotate(nvc0, q); - q->sequence++; - } - q->state = NVC0_QUERY_STATE_ENDED; - - switch (q->type) { - case PIPE_QUERY_OCCLUSION_COUNTER: - case PIPE_QUERY_OCCLUSION_PREDICATE: - nvc0_query_get(push, q, 0, 0x0100f002); - if (--nvc0->screen->num_occlusion_queries_active == 0) { - PUSH_SPACE(push, 1); - IMMED_NVC0(push, NVC0_3D(SAMPLECNT_ENABLE), 0); - } - break; - case PIPE_QUERY_PRIMITIVES_GENERATED: - nvc0_query_get(push, q, 0, 0x09005002 | (q->index << 5)); - break; - case PIPE_QUERY_PRIMITIVES_EMITTED: - nvc0_query_get(push, q, 0, 0x05805002 | (q->index << 5)); - break; - case PIPE_QUERY_SO_STATISTICS: - nvc0_query_get(push, q, 0x00, 0x05805002 | (q->index << 5)); - nvc0_query_get(push, q, 0x10, 0x06805002 | (q->index << 5)); - break; - case PIPE_QUERY_SO_OVERFLOW_PREDICATE: - /* TODO: How do we sum over all streams for render condition ? */ - /* PRIMS_DROPPED doesn't write sequence, use a ZERO query to sync on */ - nvc0_query_get(push, q, 0x00, 0x03005002 | (q->index << 5)); - nvc0_query_get(push, q, 0x20, 0x00005002); - break; - case PIPE_QUERY_TIMESTAMP: - case PIPE_QUERY_TIME_ELAPSED: - nvc0_query_get(push, q, 0, 0x00005002); - break; - case PIPE_QUERY_GPU_FINISHED: - nvc0_query_get(push, q, 0, 0x1000f010); - break; - case PIPE_QUERY_PIPELINE_STATISTICS: - nvc0_query_get(push, q, 0x00, 0x00801002); /* VFETCH, VERTICES */ - nvc0_query_get(push, q, 0x10, 0x01801002); /* VFETCH, PRIMS */ - nvc0_query_get(push, q, 0x20, 0x02802002); /* VP, LAUNCHES */ - nvc0_query_get(push, q, 0x30, 0x03806002); /* GP, LAUNCHES */ - nvc0_query_get(push, q, 0x40, 0x04806002); /* GP, PRIMS_OUT */ - nvc0_query_get(push, q, 0x50, 0x07804002); /* RAST, PRIMS_IN */ - nvc0_query_get(push, q, 0x60, 0x08804002); /* RAST, PRIMS_OUT */ - nvc0_query_get(push, q, 0x70, 0x0980a002); /* ROP, PIXELS */ - nvc0_query_get(push, q, 0x80, 0x0d808002); /* TCP, LAUNCHES */ - nvc0_query_get(push, q, 0x90, 0x0e809002); /* TEP, LAUNCHES */ - break; - case NVC0_QUERY_TFB_BUFFER_OFFSET: - /* indexed by TFB buffer instead of by vertex stream */ - nvc0_query_get(push, q, 0x00, 0x0d005002 | (q->index << 5)); - break; - case PIPE_QUERY_TIMESTAMP_DISJOINT: - /* This query is not issued on GPU because disjoint is forced to false */ - q->state = NVC0_QUERY_STATE_READY; - break; - default: -#ifdef NOUVEAU_ENABLE_DRIVER_STATISTICS - if (q->type >= NVC0_SW_QUERY_DRV_STAT(0) && - q->type <= NVC0_SW_QUERY_DRV_STAT_LAST) { - q->funcs->end_query(nvc0, q); - return; - } else -#endif - if ((q->type >= NVE4_HW_SM_QUERY(0) && q->type <= NVE4_HW_SM_QUERY_LAST) || - (q->type >= NVC0_HW_SM_QUERY(0) && q->type <= NVC0_HW_SM_QUERY_LAST)) { - nvc0_hw_sm_query_end(nvc0, q); - } - break; - } - if (q->is64bit) - nouveau_fence_ref(nvc0->screen->base.fence.current, &q->fence); -} - -static inline void -nvc0_query_update(struct nouveau_client *cli, struct nvc0_query *q) -{ - if (q->is64bit) { - if (nouveau_fence_signalled(q->fence)) - q->state = NVC0_QUERY_STATE_READY; - } else { - if (q->data[0] == q->sequence) - q->state = NVC0_QUERY_STATE_READY; - } + q->funcs->end_query(nvc0_context(pipe), q); } static boolean -nvc0_query_result(struct pipe_context *pipe, struct pipe_query *pq, - boolean wait, union pipe_query_result *result) +nvc0_get_query_result(struct pipe_context *pipe, struct pipe_query *pq, + boolean wait, union pipe_query_result *result) { - struct nvc0_context *nvc0 = nvc0_context(pipe); struct nvc0_query *q = nvc0_query(pq); - uint64_t *res64 = (uint64_t*)result; - uint32_t *res32 = (uint32_t*)result; - uint8_t *res8 = (uint8_t*)result; - uint64_t *data64 = (uint64_t *)q->data; - unsigned i; - -#ifdef NOUVEAU_ENABLE_DRIVER_STATISTICS - if (q->type >= NVC0_SW_QUERY_DRV_STAT(0) && - q->type <= NVC0_SW_QUERY_DRV_STAT_LAST) { - return q->funcs->get_query_result(nvc0, q, wait, result); - } else -#endif - if ((q->type >= NVE4_HW_SM_QUERY(0) && q->type <= NVE4_HW_SM_QUERY_LAST) || - (q->type >= NVC0_HW_SM_QUERY(0) && q->type <= NVC0_HW_SM_QUERY_LAST)) { - return nvc0_hw_sm_query_result(nvc0, q, result, wait); - } - - if (q->state != NVC0_QUERY_STATE_READY) - nvc0_query_update(nvc0->screen->base.client, q); - - if (q->state != NVC0_QUERY_STATE_READY) { - if (!wait) { - if (q->state != NVC0_QUERY_STATE_FLUSHED) { - q->state = NVC0_QUERY_STATE_FLUSHED; - /* flush for silly apps that spin on GL_QUERY_RESULT_AVAILABLE */ - PUSH_KICK(nvc0->base.pushbuf); - } - return false; - } - if (nouveau_bo_wait(q->bo, NOUVEAU_BO_RD, nvc0->screen->base.client)) - return false; - NOUVEAU_DRV_STAT(&nvc0->screen->base, query_sync_count, 1); - } - q->state = NVC0_QUERY_STATE_READY; - - switch (q->type) { - case PIPE_QUERY_GPU_FINISHED: - res8[0] = true; - break; - case PIPE_QUERY_OCCLUSION_COUNTER: /* u32 sequence, u32 count, u64 time */ - res64[0] = q->data[1] - q->data[5]; - break; - case PIPE_QUERY_OCCLUSION_PREDICATE: - res8[0] = q->data[1] != q->data[5]; - break; - case PIPE_QUERY_PRIMITIVES_GENERATED: /* u64 count, u64 time */ - case PIPE_QUERY_PRIMITIVES_EMITTED: /* u64 count, u64 time */ - res64[0] = data64[0] - data64[2]; - break; - case PIPE_QUERY_SO_STATISTICS: - res64[0] = data64[0] - data64[4]; - res64[1] = data64[2] - data64[6]; - break; - case PIPE_QUERY_SO_OVERFLOW_PREDICATE: - res8[0] = data64[0] != data64[2]; - break; - case PIPE_QUERY_TIMESTAMP: - res64[0] = data64[1]; - break; - case PIPE_QUERY_TIMESTAMP_DISJOINT: - res64[0] = 1000000000; - res8[8] = false; - break; - case PIPE_QUERY_TIME_ELAPSED: - res64[0] = data64[1] - data64[3]; - break; - case PIPE_QUERY_PIPELINE_STATISTICS: - for (i = 0; i < 10; ++i) - res64[i] = data64[i * 2] - data64[24 + i * 2]; - break; - case NVC0_QUERY_TFB_BUFFER_OFFSET: - res32[0] = q->data[1]; - break; - default: - assert(0); /* can't happen, we don't create queries with invalid type */ - return false; - } - - return true; -} - -void -nvc0_query_fifo_wait(struct nouveau_pushbuf *push, struct nvc0_query *q) -{ - unsigned offset = q->offset; - - if (q->type == PIPE_QUERY_SO_OVERFLOW_PREDICATE) offset += 0x20; - - PUSH_SPACE(push, 5); - PUSH_REFN (push, q->bo, NOUVEAU_BO_GART | NOUVEAU_BO_RD); - BEGIN_NVC0(push, SUBC_3D(NV84_SUBCHAN_SEMAPHORE_ADDRESS_HIGH), 4); - PUSH_DATAh(push, q->bo->offset + offset); - PUSH_DATA (push, q->bo->offset + offset); - PUSH_DATA (push, q->sequence); - PUSH_DATA (push, (1 << 12) | - NV84_SUBCHAN_SEMAPHORE_TRIGGER_ACQUIRE_EQUAL); + return q->funcs->get_query_result(nvc0_context(pipe), q, wait, result); } static void @@ -515,7 +78,8 @@ nvc0_render_condition(struct pipe_context *pipe, { struct nvc0_context *nvc0 = nvc0_context(pipe); struct nouveau_pushbuf *push = nvc0->base.pushbuf; - struct nvc0_query *q; + struct nvc0_query *q = nvc0_query(pq); + struct nvc0_hw_query *hq = nvc0_hw_query(q); uint32_t cond; bool wait = mode != PIPE_RENDER_COND_NO_WAIT && @@ -525,7 +89,6 @@ nvc0_render_condition(struct pipe_context *pipe, cond = NVC0_3D_COND_MODE_ALWAYS; } else { - q = nvc0_query(pq); /* NOTE: comparison of 2 queries only works if both have completed */ switch (q->type) { case PIPE_QUERY_SO_OVERFLOW_PREDICATE: @@ -536,7 +99,7 @@ nvc0_render_condition(struct pipe_context *pipe, case PIPE_QUERY_OCCLUSION_COUNTER: case PIPE_QUERY_OCCLUSION_PREDICATE: if (likely(!condition)) { - if (unlikely(q->nesting)) + if (unlikely(hq->nesting)) cond = wait ? NVC0_3D_COND_MODE_NOT_EQUAL : NVC0_3D_COND_MODE_ALWAYS; else @@ -564,29 +127,17 @@ nvc0_render_condition(struct pipe_context *pipe, } if (wait) - nvc0_query_fifo_wait(push, q); + nvc0_hw_query_fifo_wait(push, q); PUSH_SPACE(push, 7); - PUSH_REFN (push, q->bo, NOUVEAU_BO_GART | NOUVEAU_BO_RD); + PUSH_REFN (push, hq->bo, NOUVEAU_BO_GART | NOUVEAU_BO_RD); BEGIN_NVC0(push, NVC0_3D(COND_ADDRESS_HIGH), 3); - PUSH_DATAh(push, q->bo->offset + q->offset); - PUSH_DATA (push, q->bo->offset + q->offset); + PUSH_DATAh(push, hq->bo->offset + hq->offset); + PUSH_DATA (push, hq->bo->offset + hq->offset); PUSH_DATA (push, cond); BEGIN_NVC0(push, NVC0_2D(COND_ADDRESS_HIGH), 2); - PUSH_DATAh(push, q->bo->offset + q->offset); - PUSH_DATA (push, q->bo->offset + q->offset); -} - -void -nvc0_query_pushbuf_submit(struct nouveau_pushbuf *push, - struct nvc0_query *q, unsigned result_offset) -{ -#define NVC0_IB_ENTRY_1_NO_PREFETCH (1 << (31 - 8)) - - PUSH_REFN(push, q->bo, NOUVEAU_BO_RD | NOUVEAU_BO_GART); - nouveau_pushbuf_space(push, 0, 0, 1); - nouveau_pushbuf_data(push, q->bo, q->offset + result_offset, 4 | - NVC0_IB_ENTRY_1_NO_PREFETCH); + PUSH_DATAh(push, hq->bo->offset + hq->offset); + PUSH_DATA (push, hq->bo->offset + hq->offset); } /* === DRIVER STATISTICS === */ @@ -628,95 +179,8 @@ static const char *nvc0_sw_query_drv_stat_names[] = #endif /* NOUVEAU_ENABLE_DRIVER_STATISTICS */ - /* === PERFORMANCE MONITORING COUNTERS for NVE4+ === */ -/* Code to read out MP counters: They are accessible via mmio, too, but let's - * just avoid mapping registers in userspace. We'd have to know which MPs are - * enabled/present, too, and that information is not presently exposed. - * We could add a kernel interface for it, but reading the counters like this - * has the advantage of being async (if get_result isn't called immediately). - */ -static const uint64_t nve4_read_hw_sm_counters_code[] = -{ - /* sched 0x20 0x20 0x20 0x20 0x20 0x20 0x20 - * mov b32 $r8 $tidx - * mov b32 $r12 $physid - * mov b32 $r0 $pm0 - * mov b32 $r1 $pm1 - * mov b32 $r2 $pm2 - * mov b32 $r3 $pm3 - * mov b32 $r4 $pm4 - * sched 0x20 0x20 0x23 0x04 0x20 0x04 0x2b - * mov b32 $r5 $pm5 - * mov b32 $r6 $pm6 - * mov b32 $r7 $pm7 - * set $p0 0x1 eq u32 $r8 0x0 - * mov b32 $r10 c0[0x0] - * ext u32 $r8 $r12 0x414 - * mov b32 $r11 c0[0x4] - * sched 0x04 0x2e 0x04 0x20 0x20 0x28 0x04 - * ext u32 $r9 $r12 0x208 - * (not $p0) exit - * set $p1 0x1 eq u32 $r9 0x0 - * mul $r8 u32 $r8 u32 96 - * mul $r12 u32 $r9 u32 16 - * mul $r13 u32 $r9 u32 4 - * add b32 $r9 $r8 $r13 - * sched 0x28 0x04 0x2c 0x04 0x2c 0x04 0x2c - * add b32 $r8 $r8 $r12 - * mov b32 $r12 $r10 - * add b32 $r10 $c $r10 $r8 - * mov b32 $r13 $r11 - * add b32 $r11 $r11 0x0 $c - * add b32 $r12 $c $r12 $r9 - * st b128 wt g[$r10d] $r0q - * sched 0x4 0x2c 0x20 0x04 0x2e 0x00 0x00 - * mov b32 $r0 c0[0x8] - * add b32 $r13 $r13 0x0 $c - * $p1 st b128 wt g[$r12d+0x40] $r4q - * st b32 wt g[$r12d+0x50] $r0 - * exit */ - 0x2202020202020207ULL, - 0x2c00000084021c04ULL, - 0x2c0000000c031c04ULL, - 0x2c00000010001c04ULL, - 0x2c00000014005c04ULL, - 0x2c00000018009c04ULL, - 0x2c0000001c00dc04ULL, - 0x2c00000020011c04ULL, - 0x22b0420042320207ULL, - 0x2c00000024015c04ULL, - 0x2c00000028019c04ULL, - 0x2c0000002c01dc04ULL, - 0x190e0000fc81dc03ULL, - 0x2800400000029de4ULL, - 0x7000c01050c21c03ULL, - 0x280040001002dde4ULL, - 0x204282020042e047ULL, - 0x7000c00820c25c03ULL, - 0x80000000000021e7ULL, - 0x190e0000fc93dc03ULL, - 0x1000000180821c02ULL, - 0x1000000040931c02ULL, - 0x1000000010935c02ULL, - 0x4800000034825c03ULL, - 0x22c042c042c04287ULL, - 0x4800000030821c03ULL, - 0x2800000028031de4ULL, - 0x4801000020a29c03ULL, - 0x280000002c035de4ULL, - 0x0800000000b2dc42ULL, - 0x4801000024c31c03ULL, - 0x9400000000a01fc5ULL, - 0x200002e04202c047ULL, - 0x2800400020001de4ULL, - 0x0800000000d35c42ULL, - 0x9400000100c107c5ULL, - 0x9400000140c01f85ULL, - 0x8000000000001de7ULL -}; - /* NOTE: intentionally using the same names as NV */ static const char *nve4_pm_query_names[] = { @@ -773,168 +237,7 @@ static const char *nve4_pm_query_names[] = "metric-inst_replay_overhead" /* inst_issued, inst_executed */ }; -/* For simplicity, we will allocate as many group slots as we allocate counter - * slots. This means that a single counter which wants to source from 2 groups - * will have to be declared as using 2 counter slots. This shouldn't really be - * a problem because such queries don't make much sense ... (unless someone is - * really creative). - */ -struct nvc0_mp_counter_cfg -{ - uint32_t func : 16; /* mask or 4-bit logic op (depending on mode) */ - uint32_t mode : 4; /* LOGOP,B6,LOGOP_B6(_PULSE) */ - uint32_t num_src : 3; /* number of sources (1 - 6, only for NVC0:NVE4) */ - uint32_t sig_dom : 1; /* if 0, MP_PM_A (per warp-sched), if 1, MP_PM_B */ - uint32_t sig_sel : 8; /* signal group */ - uint64_t src_sel; /* signal selection for up to 6 sources (48 bit) */ -}; - -#define NVC0_COUNTER_OPn_SUM 0 -#define NVC0_COUNTER_OPn_OR 1 -#define NVC0_COUNTER_OPn_AND 2 -#define NVC0_COUNTER_OP2_REL_SUM_MM 3 /* (sum(ctr0) - sum(ctr1)) / sum(ctr0) */ -#define NVC0_COUNTER_OP2_DIV_SUM_M0 4 /* sum(ctr0) / ctr1 of MP[0]) */ -#define NVC0_COUNTER_OP2_AVG_DIV_MM 5 /* avg(ctr0 / ctr1) */ -#define NVC0_COUNTER_OP2_AVG_DIV_M0 6 /* avg(ctr0) / ctr1 of MP[0]) */ - -struct nvc0_hw_sm_query_cfg -{ - struct nvc0_mp_counter_cfg ctr[4]; - uint8_t num_counters; - uint8_t op; - uint8_t norm[2]; /* normalization num,denom */ -}; - -#define _Q1A(n, f, m, g, s, nu, dn) [NVE4_HW_SM_QUERY_##n] = { { { f, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m, 0, 0, NVE4_COMPUTE_MP_PM_A_SIGSEL_##g, s }, {}, {}, {} }, 1, NVC0_COUNTER_OPn_SUM, { nu, dn } } -#define _Q1B(n, f, m, g, s, nu, dn) [NVE4_HW_SM_QUERY_##n] = { { { f, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m, 0, 1, NVE4_COMPUTE_MP_PM_B_SIGSEL_##g, s }, {}, {}, {} }, 1, NVC0_COUNTER_OPn_SUM, { nu, dn } } -#define _M2A(n, f0, m0, g0, s0, f1, m1, g1, s1, o, nu, dn) [NVE4_HW_SM_QUERY_METRIC_##n] = { { \ - { f0, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m0, 0, 0, NVE4_COMPUTE_MP_PM_A_SIGSEL_##g0, s0 }, \ - { f1, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m1, 0, 0, NVE4_COMPUTE_MP_PM_A_SIGSEL_##g1, s1 }, \ - {}, {}, }, 2, NVC0_COUNTER_OP2_##o, { nu, dn } } -#define _M2B(n, f0, m0, g0, s0, f1, m1, g1, s1, o, nu, dn) [NVE4_HW_SM_QUERY_METRIC_##n] = { { \ - { f0, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m0, 0, 1, NVE4_COMPUTE_MP_PM_B_SIGSEL_##g0, s0 }, \ - { f1, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m1, 0, 1, NVE4_COMPUTE_MP_PM_B_SIGSEL_##g1, s1 }, \ - {}, {}, }, 2, NVC0_COUNTER_OP2_##o, { nu, dn } } -#define _M2AB(n, f0, m0, g0, s0, f1, m1, g1, s1, o, nu, dn) [NVE4_HW_SM_QUERY_METRIC_##n] = { { \ - { f0, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m0, 0, 0, NVE4_COMPUTE_MP_PM_A_SIGSEL_##g0, s0 }, \ - { f1, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m1, 0, 1, NVE4_COMPUTE_MP_PM_B_SIGSEL_##g1, s1 }, \ - {}, {}, }, 2, NVC0_COUNTER_OP2_##o, { nu, dn } } - -/* NOTES: - * active_warps: bit 0 alternates btw 0 and 1 for odd nr of warps - * inst_executed etc.: we only count a single warp scheduler - * metric-ipXc: we simply multiply by 4 to account for the 4 warp schedulers; - * this is inaccurate ! - */ -static const struct nvc0_hw_sm_query_cfg nve4_hw_sm_queries[] = -{ - _Q1B(ACTIVE_CYCLES, 0x0001, B6, WARP, 0x00000000, 1, 1), - _Q1B(ACTIVE_WARPS, 0x003f, B6, WARP, 0x31483104, 2, 1), - _Q1A(ATOM_COUNT, 0x0001, B6, BRANCH, 0x00000000, 1, 1), - _Q1A(BRANCH, 0x0001, B6, BRANCH, 0x0000000c, 1, 1), - _Q1A(DIVERGENT_BRANCH, 0x0001, B6, BRANCH, 0x00000010, 1, 1), - _Q1A(GLD_REQUEST, 0x0001, B6, LDST, 0x00000010, 1, 1), - _Q1B(GLD_MEM_DIV_REPLAY, 0x0001, B6, REPLAY, 0x00000010, 1, 1), - _Q1B(GST_TRANSACTIONS, 0x0001, B6, MEM, 0x00000004, 1, 1), - _Q1B(GST_MEM_DIV_REPLAY, 0x0001, B6, REPLAY, 0x00000014, 1, 1), - _Q1A(GRED_COUNT, 0x0001, B6, BRANCH, 0x00000008, 1, 1), - _Q1A(GST_REQUEST, 0x0001, B6, LDST, 0x00000014, 1, 1), - _Q1A(INST_EXECUTED, 0x0003, B6, EXEC, 0x00000398, 1, 1), - _Q1A(INST_ISSUED, 0x0003, B6, ISSUE, 0x00000104, 1, 1), - _Q1A(INST_ISSUED1, 0x0001, B6, ISSUE, 0x00000004, 1, 1), - _Q1A(INST_ISSUED2, 0x0001, B6, ISSUE, 0x00000008, 1, 1), - _Q1B(L1_GLD_HIT, 0x0001, B6, L1, 0x00000010, 1, 1), - _Q1B(L1_GLD_MISS, 0x0001, B6, L1, 0x00000014, 1, 1), - _Q1B(L1_LOCAL_LD_HIT, 0x0001, B6, L1, 0x00000000, 1, 1), - _Q1B(L1_LOCAL_LD_MISS, 0x0001, B6, L1, 0x00000004, 1, 1), - _Q1B(L1_LOCAL_ST_HIT, 0x0001, B6, L1, 0x00000008, 1, 1), - _Q1B(L1_LOCAL_ST_MISS, 0x0001, B6, L1, 0x0000000c, 1, 1), - _Q1B(L1_SHARED_LD_TRANSACTIONS, 0x0001, B6, TRANSACTION, 0x00000008, 1, 1), - _Q1B(L1_SHARED_ST_TRANSACTIONS, 0x0001, B6, TRANSACTION, 0x0000000c, 1, 1), - _Q1A(LOCAL_LD, 0x0001, B6, LDST, 0x00000008, 1, 1), - _Q1B(LOCAL_LD_TRANSACTIONS, 0x0001, B6, TRANSACTION, 0x00000000, 1, 1), - _Q1A(LOCAL_ST, 0x0001, B6, LDST, 0x0000000c, 1, 1), - _Q1B(LOCAL_ST_TRANSACTIONS, 0x0001, B6, TRANSACTION, 0x00000004, 1, 1), - _Q1A(PROF_TRIGGER_0, 0x0001, B6, USER, 0x00000000, 1, 1), - _Q1A(PROF_TRIGGER_1, 0x0001, B6, USER, 0x00000004, 1, 1), - _Q1A(PROF_TRIGGER_2, 0x0001, B6, USER, 0x00000008, 1, 1), - _Q1A(PROF_TRIGGER_3, 0x0001, B6, USER, 0x0000000c, 1, 1), - _Q1A(PROF_TRIGGER_4, 0x0001, B6, USER, 0x00000010, 1, 1), - _Q1A(PROF_TRIGGER_5, 0x0001, B6, USER, 0x00000014, 1, 1), - _Q1A(PROF_TRIGGER_6, 0x0001, B6, USER, 0x00000018, 1, 1), - _Q1A(PROF_TRIGGER_7, 0x0001, B6, USER, 0x0000001c, 1, 1), - _Q1A(SHARED_LD, 0x0001, B6, LDST, 0x00000000, 1, 1), - _Q1B(SHARED_LD_REPLAY, 0x0001, B6, REPLAY, 0x00000008, 1, 1), - _Q1A(SHARED_ST, 0x0001, B6, LDST, 0x00000004, 1, 1), - _Q1B(SHARED_ST_REPLAY, 0x0001, B6, REPLAY, 0x0000000c, 1, 1), - _Q1B(SM_CTA_LAUNCHED, 0x0001, B6, WARP, 0x0000001c, 1, 1), - _Q1A(THREADS_LAUNCHED, 0x003f, B6, LAUNCH, 0x398a4188, 1, 1), - _Q1B(UNCACHED_GLD_TRANSACTIONS, 0x0001, B6, MEM, 0x00000000, 1, 1), - _Q1A(WARPS_LAUNCHED, 0x0001, B6, LAUNCH, 0x00000004, 1, 1), - _M2AB(IPC, 0x3, B6, EXEC, 0x398, 0xffff, LOGOP, WARP, 0x0, DIV_SUM_M0, 10, 1), - _M2AB(IPAC, 0x3, B6, EXEC, 0x398, 0x1, B6, WARP, 0x0, AVG_DIV_MM, 10, 1), - _M2A(IPEC, 0x3, B6, EXEC, 0x398, 0xe, LOGOP, EXEC, 0x398, AVG_DIV_MM, 10, 1), - _M2A(INST_REPLAY_OHEAD, 0x3, B6, ISSUE, 0x104, 0x3, B6, EXEC, 0x398, REL_SUM_MM, 100, 1), - _M2B(MP_OCCUPANCY, 0x3f, B6, WARP, 0x31483104, 0x01, B6, WARP, 0x0, AVG_DIV_MM, 200, 64), - _M2B(MP_EFFICIENCY, 0x01, B6, WARP, 0x0, 0xffff, LOGOP, WARP, 0x0, AVG_DIV_M0, 100, 1), -}; - -#undef _Q1A -#undef _Q1B -#undef _M2A -#undef _M2B - /* === PERFORMANCE MONITORING COUNTERS for NVC0:NVE4 === */ -static const uint64_t nvc0_read_hw_sm_counters_code[] = -{ - /* mov b32 $r8 $tidx - * mov b32 $r9 $physid - * mov b32 $r0 $pm0 - * mov b32 $r1 $pm1 - * mov b32 $r2 $pm2 - * mov b32 $r3 $pm3 - * mov b32 $r4 $pm4 - * mov b32 $r5 $pm5 - * mov b32 $r6 $pm6 - * mov b32 $r7 $pm7 - * set $p0 0x1 eq u32 $r8 0x0 - * mov b32 $r10 c0[0x0] - * mov b32 $r11 c0[0x4] - * ext u32 $r8 $r9 0x414 - * (not $p0) exit - * mul $r8 u32 $r8 u32 36 - * add b32 $r10 $c $r10 $r8 - * add b32 $r11 $r11 0x0 $c - * mov b32 $r8 c0[0x8] - * st b128 wt g[$r10d+0x00] $r0q - * st b128 wt g[$r10d+0x10] $r4q - * st b32 wt g[$r10d+0x20] $r8 - * exit */ - 0x2c00000084021c04ULL, - 0x2c0000000c025c04ULL, - 0x2c00000010001c04ULL, - 0x2c00000014005c04ULL, - 0x2c00000018009c04ULL, - 0x2c0000001c00dc04ULL, - 0x2c00000020011c04ULL, - 0x2c00000024015c04ULL, - 0x2c00000028019c04ULL, - 0x2c0000002c01dc04ULL, - 0x190e0000fc81dc03ULL, - 0x2800400000029de4ULL, - 0x280040001002dde4ULL, - 0x7000c01050921c03ULL, - 0x80000000000021e7ULL, - 0x1000000090821c02ULL, - 0x4801000020a29c03ULL, - 0x0800000000b2dc42ULL, - 0x2800400020021de4ULL, - 0x9400000000a01fc5ULL, - 0x9400000040a11fc5ULL, - 0x9400000080a21f85ULL, - 0x8000000000001de7ULL -}; - static const char *nvc0_pm_query_names[] = { /* MP counters */ @@ -971,377 +274,6 @@ static const char *nvc0_pm_query_names[] = "warps_launched", }; -#define _Q(n, f, m, g, c, s0, s1, s2, s3, s4, s5) [NVC0_HW_SM_QUERY_##n] = { { { f, NVC0_COMPUTE_MP_PM_OP_MODE_##m, c, 0, g, s0|(s1 << 8)|(s2 << 16)|(s3 << 24)|(s4##ULL << 32)|(s5##ULL << 40) }, {}, {}, {} }, 1, NVC0_COUNTER_OPn_SUM, { 1, 1 } } - -static const struct nvc0_hw_sm_query_cfg nvc0_hw_sm_queries[] = -{ - _Q(ACTIVE_CYCLES, 0xaaaa, LOGOP, 0x11, 1, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(ACTIVE_WARPS, 0xaaaa, LOGOP, 0x24, 6, 0x10, 0x21, 0x32, 0x43, 0x54, 0x65), - _Q(ATOM_COUNT, 0xaaaa, LOGOP, 0x63, 1, 0x30, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(BRANCH, 0xaaaa, LOGOP, 0x1a, 2, 0x00, 0x11, 0x00, 0x00, 0x00, 0x00), - _Q(DIVERGENT_BRANCH, 0xaaaa, LOGOP, 0x19, 2, 0x20, 0x31, 0x00, 0x00, 0x00, 0x00), - _Q(GLD_REQUEST, 0xaaaa, LOGOP, 0x64, 1, 0x30, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(GRED_COUNT, 0xaaaa, LOGOP, 0x63, 1, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(GST_REQUEST, 0xaaaa, LOGOP, 0x64, 1, 0x60, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(INST_EXECUTED, 0xaaaa, LOGOP, 0x2d, 3, 0x00, 0x11, 0x22, 0x00, 0x00, 0x00), - _Q(INST_ISSUED1_0, 0xaaaa, LOGOP, 0x7e, 1, 0x10, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(INST_ISSUED1_1, 0xaaaa, LOGOP, 0x7e, 1, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(INST_ISSUED2_0, 0xaaaa, LOGOP, 0x7e, 1, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(INST_ISSUED2_1, 0xaaaa, LOGOP, 0x7e, 1, 0x50, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(LOCAL_LD, 0xaaaa, LOGOP, 0x64, 1, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(LOCAL_ST, 0xaaaa, LOGOP, 0x64, 1, 0x50, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(PROF_TRIGGER_0, 0xaaaa, LOGOP, 0x01, 1, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(PROF_TRIGGER_1, 0xaaaa, LOGOP, 0x01, 1, 0x10, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(PROF_TRIGGER_2, 0xaaaa, LOGOP, 0x01, 1, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(PROF_TRIGGER_3, 0xaaaa, LOGOP, 0x01, 1, 0x30, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(PROF_TRIGGER_4, 0xaaaa, LOGOP, 0x01, 1, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(PROF_TRIGGER_5, 0xaaaa, LOGOP, 0x01, 1, 0x50, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(PROF_TRIGGER_6, 0xaaaa, LOGOP, 0x01, 1, 0x60, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(PROF_TRIGGER_7, 0xaaaa, LOGOP, 0x01, 1, 0x70, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(SHARED_LD, 0xaaaa, LOGOP, 0x64, 1, 0x10, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(SHARED_ST, 0xaaaa, LOGOP, 0x64, 1, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00), - _Q(THREADS_LAUNCHED, 0xaaaa, LOGOP, 0x26, 6, 0x10, 0x21, 0x32, 0x43, 0x54, 0x65), - _Q(TH_INST_EXECUTED_0, 0xaaaa, LOGOP, 0xa3, 6, 0x00, 0x11, 0x22, 0x33, 0x44, 0x55), - _Q(TH_INST_EXECUTED_1, 0xaaaa, LOGOP, 0xa5, 6, 0x00, 0x11, 0x22, 0x33, 0x44, 0x55), - _Q(TH_INST_EXECUTED_2, 0xaaaa, LOGOP, 0xa4, 6, 0x00, 0x11, 0x22, 0x33, 0x44, 0x55), - _Q(TH_INST_EXECUTED_3, 0xaaaa, LOGOP, 0xa6, 6, 0x00, 0x11, 0x22, 0x33, 0x44, 0x55), - _Q(WARPS_LAUNCHED, 0xaaaa, LOGOP, 0x26, 1, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00), -}; - -#undef _Q - -static const struct nvc0_hw_sm_query_cfg * -nvc0_hw_sm_query_get_cfg(struct nvc0_context *nvc0, struct nvc0_query *q) -{ - struct nvc0_screen *screen = nvc0->screen; - - if (screen->base.class_3d >= NVE4_3D_CLASS) - return &nve4_hw_sm_queries[q->type - PIPE_QUERY_DRIVER_SPECIFIC]; - return &nvc0_hw_sm_queries[q->type - NVC0_HW_SM_QUERY(0)]; -} - -boolean -nvc0_hw_sm_query_begin(struct nvc0_context *nvc0, struct nvc0_query *q) -{ - struct nvc0_screen *screen = nvc0->screen; - struct nouveau_pushbuf *push = nvc0->base.pushbuf; - const bool is_nve4 = screen->base.class_3d >= NVE4_3D_CLASS; - const struct nvc0_hw_sm_query_cfg *cfg; - unsigned i, c; - unsigned num_ab[2] = { 0, 0 }; - - cfg = nvc0_hw_sm_query_get_cfg(nvc0, q); - - /* check if we have enough free counter slots */ - for (i = 0; i < cfg->num_counters; ++i) - num_ab[cfg->ctr[i].sig_dom]++; - - if (screen->pm.num_hw_sm_active[0] + num_ab[0] > 4 || - screen->pm.num_hw_sm_active[1] + num_ab[1] > 4) { - NOUVEAU_ERR("Not enough free MP counter slots !\n"); - return false; - } - - assert(cfg->num_counters <= 4); - PUSH_SPACE(push, 4 * 8 * (is_nve4 ? 1 : 6) + 6); - - if (!screen->pm.mp_counters_enabled) { - screen->pm.mp_counters_enabled = true; - BEGIN_NVC0(push, SUBC_SW(0x06ac), 1); - PUSH_DATA (push, 0x1fcb); - } - - /* set sequence field to 0 (used to check if result is available) */ - for (i = 0; i < screen->mp_count; ++i) - q->data[i * 10 + 10] = 0; - - for (i = 0; i < cfg->num_counters; ++i) { - const unsigned d = cfg->ctr[i].sig_dom; - - if (!screen->pm.num_hw_sm_active[d]) { - uint32_t m = (1 << 22) | (1 << (7 + (8 * !d))); - if (screen->pm.num_hw_sm_active[!d]) - m |= 1 << (7 + (8 * d)); - BEGIN_NVC0(push, SUBC_SW(0x0600), 1); - PUSH_DATA (push, m); - } - screen->pm.num_hw_sm_active[d]++; - - for (c = d * 4; c < (d * 4 + 4); ++c) { - if (!screen->pm.mp_counter[c]) { - q->ctr[i] = c; - screen->pm.mp_counter[c] = (struct pipe_query *)q; - break; - } - } - assert(c <= (d * 4 + 3)); /* must succeed, already checked for space */ - - /* configure and reset the counter(s) */ - if (is_nve4) { - if (d == 0) - BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_A_SIGSEL(c & 3)), 1); - else - BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_B_SIGSEL(c & 3)), 1); - PUSH_DATA (push, cfg->ctr[i].sig_sel); - BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_SRCSEL(c)), 1); - PUSH_DATA (push, cfg->ctr[i].src_sel + 0x2108421 * (c & 3)); - BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_FUNC(c)), 1); - PUSH_DATA (push, (cfg->ctr[i].func << 4) | cfg->ctr[i].mode); - BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_SET(c)), 1); - PUSH_DATA (push, 0); - } else { - unsigned s; - - for (s = 0; s < cfg->ctr[i].num_src; s++) { - BEGIN_NVC0(push, NVC0_COMPUTE(MP_PM_SIGSEL(s)), 1); - PUSH_DATA (push, cfg->ctr[i].sig_sel); - BEGIN_NVC0(push, NVC0_COMPUTE(MP_PM_SRCSEL(s)), 1); - PUSH_DATA (push, (cfg->ctr[i].src_sel >> (s * 8)) & 0xff); - BEGIN_NVC0(push, NVC0_COMPUTE(MP_PM_OP(s)), 1); - PUSH_DATA (push, (cfg->ctr[i].func << 4) | cfg->ctr[i].mode); - BEGIN_NVC0(push, NVC0_COMPUTE(MP_PM_SET(s)), 1); - PUSH_DATA (push, 0); - } - } - } - return true; -} - -static void -nvc0_hw_sm_query_end(struct nvc0_context *nvc0, struct nvc0_query *q) -{ - struct nvc0_screen *screen = nvc0->screen; - struct pipe_context *pipe = &nvc0->base.pipe; - struct nouveau_pushbuf *push = nvc0->base.pushbuf; - const bool is_nve4 = screen->base.class_3d >= NVE4_3D_CLASS; - uint32_t mask; - uint32_t input[3]; - const uint block[3] = { 32, is_nve4 ? 4 : 1, 1 }; - const uint grid[3] = { screen->mp_count, 1, 1 }; - unsigned c; - const struct nvc0_hw_sm_query_cfg *cfg; - - cfg = nvc0_hw_sm_query_get_cfg(nvc0, q); - - if (unlikely(!screen->pm.prog)) { - struct nvc0_program *prog = CALLOC_STRUCT(nvc0_program); - prog->type = PIPE_SHADER_COMPUTE; - prog->translated = true; - prog->num_gprs = 14; - prog->parm_size = 12; - if (is_nve4) { - prog->code = (uint32_t *)nve4_read_hw_sm_counters_code; - prog->code_size = sizeof(nve4_read_hw_sm_counters_code); - } else { - prog->code = (uint32_t *)nvc0_read_hw_sm_counters_code; - prog->code_size = sizeof(nvc0_read_hw_sm_counters_code); - } - screen->pm.prog = prog; - } - - /* disable all counting */ - PUSH_SPACE(push, 8); - for (c = 0; c < 8; ++c) - if (screen->pm.mp_counter[c]) { - if (is_nve4) { - IMMED_NVC0(push, NVE4_COMPUTE(MP_PM_FUNC(c)), 0); - } else { - IMMED_NVC0(push, NVC0_COMPUTE(MP_PM_OP(c)), 0); - } - } - /* release counters for this query */ - for (c = 0; c < 8; ++c) { - if (nvc0_query(screen->pm.mp_counter[c]) == q) { - screen->pm.num_hw_sm_active[c / 4]--; - screen->pm.mp_counter[c] = NULL; - } - } - - BCTX_REFN_bo(nvc0->bufctx_cp, CP_QUERY, NOUVEAU_BO_GART | NOUVEAU_BO_WR, - q->bo); - - PUSH_SPACE(push, 1); - IMMED_NVC0(push, SUBC_COMPUTE(NV50_GRAPH_SERIALIZE), 0); - - pipe->bind_compute_state(pipe, screen->pm.prog); - input[0] = (q->bo->offset + q->base); - input[1] = (q->bo->offset + q->base) >> 32; - input[2] = q->sequence; - pipe->launch_grid(pipe, block, grid, 0, input); - - nouveau_bufctx_reset(nvc0->bufctx_cp, NVC0_BIND_CP_QUERY); - - /* re-activate other counters */ - PUSH_SPACE(push, 16); - mask = 0; - for (c = 0; c < 8; ++c) { - unsigned i; - q = nvc0_query(screen->pm.mp_counter[c]); - if (!q) - continue; - cfg = nvc0_hw_sm_query_get_cfg(nvc0, q); - for (i = 0; i < cfg->num_counters; ++i) { - if (mask & (1 << q->ctr[i])) - break; - mask |= 1 << q->ctr[i]; - if (is_nve4) { - BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_FUNC(q->ctr[i])), 1); - } else { - BEGIN_NVC0(push, NVC0_COMPUTE(MP_PM_OP(q->ctr[i])), 1); - } - PUSH_DATA (push, (cfg->ctr[i].func << 4) | cfg->ctr[i].mode); - } - } -} - -static inline bool -nvc0_hw_sm_query_read_data(uint32_t count[32][4], - struct nvc0_context *nvc0, bool wait, - struct nvc0_query *q, - const struct nvc0_hw_sm_query_cfg *cfg, - unsigned mp_count) -{ - unsigned p, c; - - for (p = 0; p < mp_count; ++p) { - const unsigned b = (0x24 / 4) * p; - - for (c = 0; c < cfg->num_counters; ++c) { - if (q->data[b + 8] != q->sequence) { - if (!wait) - return false; - if (nouveau_bo_wait(q->bo, NOUVEAU_BO_RD, nvc0->base.client)) - return false; - } - count[p][c] = q->data[b + q->ctr[c]]; - } - } - return true; -} - -static inline bool -nve4_hw_sm_query_read_data(uint32_t count[32][4], - struct nvc0_context *nvc0, bool wait, - struct nvc0_query *q, - const struct nvc0_hw_sm_query_cfg *cfg, - unsigned mp_count) -{ - unsigned p, c, d; - - for (p = 0; p < mp_count; ++p) { - const unsigned b = (0x60 / 4) * p; - - for (c = 0; c < cfg->num_counters; ++c) { - count[p][c] = 0; - for (d = 0; d < ((q->ctr[c] & ~3) ? 1 : 4); ++d) { - if (q->data[b + 20 + d] != q->sequence) { - if (!wait) - return false; - if (nouveau_bo_wait(q->bo, NOUVEAU_BO_RD, nvc0->base.client)) - return false; - } - if (q->ctr[c] & ~0x3) - count[p][c] = q->data[b + 16 + (q->ctr[c] & 3)]; - else - count[p][c] += q->data[b + d * 4 + q->ctr[c]]; - } - } - } - return true; -} - -/* Metric calculations: - * sum(x) ... sum of x over all MPs - * avg(x) ... average of x over all MPs - * - * IPC : sum(inst_executed) / clock - * INST_REPLAY_OHEAD: (sum(inst_issued) - sum(inst_executed)) / sum(inst_issued) - * MP_OCCUPANCY : avg((active_warps / 64) / active_cycles) - * MP_EFFICIENCY : avg(active_cycles / clock) - * - * NOTE: Interpretation of IPC requires knowledge of MP count. - */ -static boolean -nvc0_hw_sm_query_result(struct nvc0_context *nvc0, struct nvc0_query *q, - void *result, boolean wait) -{ - uint32_t count[32][4]; - uint64_t value = 0; - unsigned mp_count = MIN2(nvc0->screen->mp_count_compute, 32); - unsigned p, c; - const struct nvc0_hw_sm_query_cfg *cfg; - bool ret; - - cfg = nvc0_hw_sm_query_get_cfg(nvc0, q); - - if (nvc0->screen->base.class_3d >= NVE4_3D_CLASS) - ret = nve4_hw_sm_query_read_data(count, nvc0, wait, q, cfg, mp_count); - else - ret = nvc0_hw_sm_query_read_data(count, nvc0, wait, q, cfg, mp_count); - if (!ret) - return false; - - if (cfg->op == NVC0_COUNTER_OPn_SUM) { - for (c = 0; c < cfg->num_counters; ++c) - for (p = 0; p < mp_count; ++p) - value += count[p][c]; - value = (value * cfg->norm[0]) / cfg->norm[1]; - } else - if (cfg->op == NVC0_COUNTER_OPn_OR) { - uint32_t v = 0; - for (c = 0; c < cfg->num_counters; ++c) - for (p = 0; p < mp_count; ++p) - v |= count[p][c]; - value = ((uint64_t)v * cfg->norm[0]) / cfg->norm[1]; - } else - if (cfg->op == NVC0_COUNTER_OPn_AND) { - uint32_t v = ~0; - for (c = 0; c < cfg->num_counters; ++c) - for (p = 0; p < mp_count; ++p) - v &= count[p][c]; - value = ((uint64_t)v * cfg->norm[0]) / cfg->norm[1]; - } else - if (cfg->op == NVC0_COUNTER_OP2_REL_SUM_MM) { - uint64_t v[2] = { 0, 0 }; - for (p = 0; p < mp_count; ++p) { - v[0] += count[p][0]; - v[1] += count[p][1]; - } - if (v[0]) - value = ((v[0] - v[1]) * cfg->norm[0]) / (v[0] * cfg->norm[1]); - } else - if (cfg->op == NVC0_COUNTER_OP2_DIV_SUM_M0) { - for (p = 0; p < mp_count; ++p) - value += count[p][0]; - if (count[0][1]) - value = (value * cfg->norm[0]) / (count[0][1] * cfg->norm[1]); - else - value = 0; - } else - if (cfg->op == NVC0_COUNTER_OP2_AVG_DIV_MM) { - unsigned mp_used = 0; - for (p = 0; p < mp_count; ++p, mp_used += !!count[p][0]) - if (count[p][1]) - value += (count[p][0] * cfg->norm[0]) / count[p][1]; - if (mp_used) - value /= (uint64_t)mp_used * cfg->norm[1]; - } else - if (cfg->op == NVC0_COUNTER_OP2_AVG_DIV_M0) { - unsigned mp_used = 0; - for (p = 0; p < mp_count; ++p, mp_used += !!count[p][0]) - value += count[p][0]; - if (count[0][1] && mp_used) { - value *= cfg->norm[0]; - value /= (uint64_t)count[0][1] * mp_used * cfg->norm[1]; - } else { - value = 0; - } - } - - *(uint64_t *)result = value; - return true; -} - int nvc0_screen_get_driver_query_info(struct pipe_screen *pscreen, unsigned id, @@ -1481,10 +413,10 @@ nvc0_init_query_functions(struct nvc0_context *nvc0) { struct pipe_context *pipe = &nvc0->base.pipe; - pipe->create_query = nvc0_query_create; - pipe->destroy_query = nvc0_query_destroy; - pipe->begin_query = nvc0_query_begin; - pipe->end_query = nvc0_query_end; - pipe->get_query_result = nvc0_query_result; + pipe->create_query = nvc0_create_query; + pipe->destroy_query = nvc0_destroy_query; + pipe->begin_query = nvc0_begin_query; + pipe->end_query = nvc0_end_query; + pipe->get_query_result = nvc0_get_query_result; pipe->render_condition = nvc0_render_condition; } diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_query.h b/src/gallium/drivers/nouveau/nvc0/nvc0_query.h index dfb2fe3c749..c4f0cb0ec6e 100644 --- a/src/gallium/drivers/nouveau/nvc0/nvc0_query.h +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_query.h @@ -4,9 +4,6 @@ #include "pipe/p_context.h" #include "nouveau_context.h" -#include "nouveau_mm.h" - -#define NVC0_QUERY_TFB_BUFFER_OFFSET (PIPE_QUERY_TYPES + 0) struct nvc0_context; struct nvc0_query; @@ -21,20 +18,8 @@ struct nvc0_query_funcs { struct nvc0_query { const struct nvc0_query_funcs *funcs; - uint32_t *data; uint16_t type; uint16_t index; - int8_t ctr[4]; - uint32_t sequence; - struct nouveau_bo *bo; - uint32_t base; - uint32_t offset; /* base + i * rotate */ - uint8_t state; - boolean is64bit; - uint8_t rotate; - int nesting; /* only used for occlusion queries */ - struct nouveau_mm_allocation *mm; - struct nouveau_fence *fence; }; static inline struct nvc0_query * @@ -49,106 +34,6 @@ nvc0_query(struct pipe_query *pipe) #define NVC0_QUERY_MP_COUNTER_GROUP 0 #define NVC0_SW_QUERY_DRV_STAT_GROUP 1 -/* - * Performance counter queries: - */ -#define NVE4_HW_SM_QUERY(i) (PIPE_QUERY_DRIVER_SPECIFIC + (i)) -#define NVE4_HW_SM_QUERY_LAST NVE4_HW_SM_QUERY(NVE4_HW_SM_QUERY_COUNT - 1) -enum nve4_pm_queries -{ - NVE4_HW_SM_QUERY_ACTIVE_CYCLES = 0, - NVE4_HW_SM_QUERY_ACTIVE_WARPS, - NVE4_HW_SM_QUERY_ATOM_COUNT, - NVE4_HW_SM_QUERY_BRANCH, - NVE4_HW_SM_QUERY_DIVERGENT_BRANCH, - NVE4_HW_SM_QUERY_GLD_REQUEST, - NVE4_HW_SM_QUERY_GLD_MEM_DIV_REPLAY, - NVE4_HW_SM_QUERY_GST_TRANSACTIONS, - NVE4_HW_SM_QUERY_GST_MEM_DIV_REPLAY, - NVE4_HW_SM_QUERY_GRED_COUNT, - NVE4_HW_SM_QUERY_GST_REQUEST, - NVE4_HW_SM_QUERY_INST_EXECUTED, - NVE4_HW_SM_QUERY_INST_ISSUED, - NVE4_HW_SM_QUERY_INST_ISSUED1, - NVE4_HW_SM_QUERY_INST_ISSUED2, - NVE4_HW_SM_QUERY_L1_GLD_HIT, - NVE4_HW_SM_QUERY_L1_GLD_MISS, - NVE4_HW_SM_QUERY_L1_LOCAL_LD_HIT, - NVE4_HW_SM_QUERY_L1_LOCAL_LD_MISS, - NVE4_HW_SM_QUERY_L1_LOCAL_ST_HIT, - NVE4_HW_SM_QUERY_L1_LOCAL_ST_MISS, - NVE4_HW_SM_QUERY_L1_SHARED_LD_TRANSACTIONS, - NVE4_HW_SM_QUERY_L1_SHARED_ST_TRANSACTIONS, - NVE4_HW_SM_QUERY_LOCAL_LD, - NVE4_HW_SM_QUERY_LOCAL_LD_TRANSACTIONS, - NVE4_HW_SM_QUERY_LOCAL_ST, - NVE4_HW_SM_QUERY_LOCAL_ST_TRANSACTIONS, - NVE4_HW_SM_QUERY_PROF_TRIGGER_0, - NVE4_HW_SM_QUERY_PROF_TRIGGER_1, - NVE4_HW_SM_QUERY_PROF_TRIGGER_2, - NVE4_HW_SM_QUERY_PROF_TRIGGER_3, - NVE4_HW_SM_QUERY_PROF_TRIGGER_4, - NVE4_HW_SM_QUERY_PROF_TRIGGER_5, - NVE4_HW_SM_QUERY_PROF_TRIGGER_6, - NVE4_HW_SM_QUERY_PROF_TRIGGER_7, - NVE4_HW_SM_QUERY_SHARED_LD, - NVE4_HW_SM_QUERY_SHARED_LD_REPLAY, - NVE4_HW_SM_QUERY_SHARED_ST, - NVE4_HW_SM_QUERY_SHARED_ST_REPLAY, - NVE4_HW_SM_QUERY_SM_CTA_LAUNCHED, - NVE4_HW_SM_QUERY_THREADS_LAUNCHED, - NVE4_HW_SM_QUERY_UNCACHED_GLD_TRANSACTIONS, - NVE4_HW_SM_QUERY_WARPS_LAUNCHED, - NVE4_HW_SM_QUERY_METRIC_IPC, - NVE4_HW_SM_QUERY_METRIC_IPAC, - NVE4_HW_SM_QUERY_METRIC_IPEC, - NVE4_HW_SM_QUERY_METRIC_MP_OCCUPANCY, - NVE4_HW_SM_QUERY_METRIC_MP_EFFICIENCY, - NVE4_HW_SM_QUERY_METRIC_INST_REPLAY_OHEAD, - NVE4_HW_SM_QUERY_COUNT -}; - -#define NVC0_HW_SM_QUERY(i) (PIPE_QUERY_DRIVER_SPECIFIC + 2048 + (i)) -#define NVC0_HW_SM_QUERY_LAST NVC0_HW_SM_QUERY(NVC0_HW_SM_QUERY_COUNT - 1) -enum nvc0_pm_queries -{ - NVC0_HW_SM_QUERY_ACTIVE_CYCLES = 0, - NVC0_HW_SM_QUERY_ACTIVE_WARPS, - NVC0_HW_SM_QUERY_ATOM_COUNT, - NVC0_HW_SM_QUERY_BRANCH, - NVC0_HW_SM_QUERY_DIVERGENT_BRANCH, - NVC0_HW_SM_QUERY_GLD_REQUEST, - NVC0_HW_SM_QUERY_GRED_COUNT, - NVC0_HW_SM_QUERY_GST_REQUEST, - NVC0_HW_SM_QUERY_INST_EXECUTED, - NVC0_HW_SM_QUERY_INST_ISSUED1_0, - NVC0_HW_SM_QUERY_INST_ISSUED1_1, - NVC0_HW_SM_QUERY_INST_ISSUED2_0, - NVC0_HW_SM_QUERY_INST_ISSUED2_1, - NVC0_HW_SM_QUERY_LOCAL_LD, - NVC0_HW_SM_QUERY_LOCAL_ST, - NVC0_HW_SM_QUERY_PROF_TRIGGER_0, - NVC0_HW_SM_QUERY_PROF_TRIGGER_1, - NVC0_HW_SM_QUERY_PROF_TRIGGER_2, - NVC0_HW_SM_QUERY_PROF_TRIGGER_3, - NVC0_HW_SM_QUERY_PROF_TRIGGER_4, - NVC0_HW_SM_QUERY_PROF_TRIGGER_5, - NVC0_HW_SM_QUERY_PROF_TRIGGER_6, - NVC0_HW_SM_QUERY_PROF_TRIGGER_7, - NVC0_HW_SM_QUERY_SHARED_LD, - NVC0_HW_SM_QUERY_SHARED_ST, - NVC0_HW_SM_QUERY_THREADS_LAUNCHED, - NVC0_HW_SM_QUERY_TH_INST_EXECUTED_0, - NVC0_HW_SM_QUERY_TH_INST_EXECUTED_1, - NVC0_HW_SM_QUERY_TH_INST_EXECUTED_2, - NVC0_HW_SM_QUERY_TH_INST_EXECUTED_3, - NVC0_HW_SM_QUERY_WARPS_LAUNCHED, - NVC0_HW_SM_QUERY_COUNT -}; - void nvc0_init_query_functions(struct nvc0_context *); -void nvc0_query_pushbuf_submit(struct nouveau_pushbuf *, struct nvc0_query *, - unsigned); -void nvc0_query_fifo_wait(struct nouveau_pushbuf *, struct nvc0_query *); #endif diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw.c b/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw.c new file mode 100644 index 00000000000..b7923d549f4 --- /dev/null +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw.c @@ -0,0 +1,1135 @@ +/* + * Copyright 2011 Christoph Bumiller + * Copyright 2015 Samuel Pitoiset + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#define NVC0_PUSH_EXPLICIT_SPACE_CHECKING + +#include "nvc0/nvc0_context.h" +#include "nvc0/nvc0_query_hw.h" + +#include "nv_object.xml.h" +#include "nvc0/nve4_compute.xml.h" +#include "nvc0/nvc0_compute.xml.h" + +#define NVC0_HW_QUERY_STATE_READY 0 +#define NVC0_HW_QUERY_STATE_ACTIVE 1 +#define NVC0_HW_QUERY_STATE_ENDED 2 +#define NVC0_HW_QUERY_STATE_FLUSHED 3 + +#define NVC0_HW_QUERY_ALLOC_SPACE 256 + +/* === PERFORMANCE MONITORING COUNTERS for NVE4+ === */ + +/* Code to read out MP counters: They are accessible via mmio, too, but let's + * just avoid mapping registers in userspace. We'd have to know which MPs are + * enabled/present, too, and that information is not presently exposed. + * We could add a kernel interface for it, but reading the counters like this + * has the advantage of being async (if get_result isn't called immediately). + */ +static const uint64_t nve4_read_hw_sm_counters_code[] = +{ + /* sched 0x20 0x20 0x20 0x20 0x20 0x20 0x20 + * mov b32 $r8 $tidx + * mov b32 $r12 $physid + * mov b32 $r0 $pm0 + * mov b32 $r1 $pm1 + * mov b32 $r2 $pm2 + * mov b32 $r3 $pm3 + * mov b32 $r4 $pm4 + * sched 0x20 0x20 0x23 0x04 0x20 0x04 0x2b + * mov b32 $r5 $pm5 + * mov b32 $r6 $pm6 + * mov b32 $r7 $pm7 + * set $p0 0x1 eq u32 $r8 0x0 + * mov b32 $r10 c0[0x0] + * ext u32 $r8 $r12 0x414 + * mov b32 $r11 c0[0x4] + * sched 0x04 0x2e 0x04 0x20 0x20 0x28 0x04 + * ext u32 $r9 $r12 0x208 + * (not $p0) exit + * set $p1 0x1 eq u32 $r9 0x0 + * mul $r8 u32 $r8 u32 96 + * mul $r12 u32 $r9 u32 16 + * mul $r13 u32 $r9 u32 4 + * add b32 $r9 $r8 $r13 + * sched 0x28 0x04 0x2c 0x04 0x2c 0x04 0x2c + * add b32 $r8 $r8 $r12 + * mov b32 $r12 $r10 + * add b32 $r10 $c $r10 $r8 + * mov b32 $r13 $r11 + * add b32 $r11 $r11 0x0 $c + * add b32 $r12 $c $r12 $r9 + * st b128 wt g[$r10d] $r0q + * sched 0x4 0x2c 0x20 0x04 0x2e 0x00 0x00 + * mov b32 $r0 c0[0x8] + * add b32 $r13 $r13 0x0 $c + * $p1 st b128 wt g[$r12d+0x40] $r4q + * st b32 wt g[$r12d+0x50] $r0 + * exit */ + 0x2202020202020207ULL, + 0x2c00000084021c04ULL, + 0x2c0000000c031c04ULL, + 0x2c00000010001c04ULL, + 0x2c00000014005c04ULL, + 0x2c00000018009c04ULL, + 0x2c0000001c00dc04ULL, + 0x2c00000020011c04ULL, + 0x22b0420042320207ULL, + 0x2c00000024015c04ULL, + 0x2c00000028019c04ULL, + 0x2c0000002c01dc04ULL, + 0x190e0000fc81dc03ULL, + 0x2800400000029de4ULL, + 0x7000c01050c21c03ULL, + 0x280040001002dde4ULL, + 0x204282020042e047ULL, + 0x7000c00820c25c03ULL, + 0x80000000000021e7ULL, + 0x190e0000fc93dc03ULL, + 0x1000000180821c02ULL, + 0x1000000040931c02ULL, + 0x1000000010935c02ULL, + 0x4800000034825c03ULL, + 0x22c042c042c04287ULL, + 0x4800000030821c03ULL, + 0x2800000028031de4ULL, + 0x4801000020a29c03ULL, + 0x280000002c035de4ULL, + 0x0800000000b2dc42ULL, + 0x4801000024c31c03ULL, + 0x9400000000a01fc5ULL, + 0x200002e04202c047ULL, + 0x2800400020001de4ULL, + 0x0800000000d35c42ULL, + 0x9400000100c107c5ULL, + 0x9400000140c01f85ULL, + 0x8000000000001de7ULL +}; + +/* For simplicity, we will allocate as many group slots as we allocate counter + * slots. This means that a single counter which wants to source from 2 groups + * will have to be declared as using 2 counter slots. This shouldn't really be + * a problem because such queries don't make much sense ... (unless someone is + * really creative). + */ +struct nvc0_mp_counter_cfg +{ + uint32_t func : 16; /* mask or 4-bit logic op (depending on mode) */ + uint32_t mode : 4; /* LOGOP,B6,LOGOP_B6(_PULSE) */ + uint32_t num_src : 3; /* number of sources (1 - 6, only for NVC0:NVE4) */ + uint32_t sig_dom : 1; /* if 0, MP_PM_A (per warp-sched), if 1, MP_PM_B */ + uint32_t sig_sel : 8; /* signal group */ + uint64_t src_sel; /* signal selection for up to 6 sources (48 bit) */ +}; + +#define NVC0_COUNTER_OPn_SUM 0 +#define NVC0_COUNTER_OPn_OR 1 +#define NVC0_COUNTER_OPn_AND 2 +#define NVC0_COUNTER_OP2_REL_SUM_MM 3 /* (sum(ctr0) - sum(ctr1)) / sum(ctr0) */ +#define NVC0_COUNTER_OP2_DIV_SUM_M0 4 /* sum(ctr0) / ctr1 of MP[0]) */ +#define NVC0_COUNTER_OP2_AVG_DIV_MM 5 /* avg(ctr0 / ctr1) */ +#define NVC0_COUNTER_OP2_AVG_DIV_M0 6 /* avg(ctr0) / ctr1 of MP[0]) */ + +struct nvc0_hw_sm_query_cfg +{ + struct nvc0_mp_counter_cfg ctr[4]; + uint8_t num_counters; + uint8_t op; + uint8_t norm[2]; /* normalization num,denom */ +}; + +#define _Q1A(n, f, m, g, s, nu, dn) [NVE4_HW_SM_QUERY_##n] = { { { f, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m, 0, 0, NVE4_COMPUTE_MP_PM_A_SIGSEL_##g, s }, {}, {}, {} }, 1, NVC0_COUNTER_OPn_SUM, { nu, dn } } +#define _Q1B(n, f, m, g, s, nu, dn) [NVE4_HW_SM_QUERY_##n] = { { { f, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m, 0, 1, NVE4_COMPUTE_MP_PM_B_SIGSEL_##g, s }, {}, {}, {} }, 1, NVC0_COUNTER_OPn_SUM, { nu, dn } } +#define _M2A(n, f0, m0, g0, s0, f1, m1, g1, s1, o, nu, dn) [NVE4_HW_SM_QUERY_METRIC_##n] = { { \ + { f0, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m0, 0, 0, NVE4_COMPUTE_MP_PM_A_SIGSEL_##g0, s0 }, \ + { f1, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m1, 0, 0, NVE4_COMPUTE_MP_PM_A_SIGSEL_##g1, s1 }, \ + {}, {}, }, 2, NVC0_COUNTER_OP2_##o, { nu, dn } } +#define _M2B(n, f0, m0, g0, s0, f1, m1, g1, s1, o, nu, dn) [NVE4_HW_SM_QUERY_METRIC_##n] = { { \ + { f0, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m0, 0, 1, NVE4_COMPUTE_MP_PM_B_SIGSEL_##g0, s0 }, \ + { f1, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m1, 0, 1, NVE4_COMPUTE_MP_PM_B_SIGSEL_##g1, s1 }, \ + {}, {}, }, 2, NVC0_COUNTER_OP2_##o, { nu, dn } } +#define _M2AB(n, f0, m0, g0, s0, f1, m1, g1, s1, o, nu, dn) [NVE4_HW_SM_QUERY_METRIC_##n] = { { \ + { f0, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m0, 0, 0, NVE4_COMPUTE_MP_PM_A_SIGSEL_##g0, s0 }, \ + { f1, NVE4_COMPUTE_MP_PM_FUNC_MODE_##m1, 0, 1, NVE4_COMPUTE_MP_PM_B_SIGSEL_##g1, s1 }, \ + {}, {}, }, 2, NVC0_COUNTER_OP2_##o, { nu, dn } } + +/* NOTES: + * active_warps: bit 0 alternates btw 0 and 1 for odd nr of warps + * inst_executed etc.: we only count a single warp scheduler + * metric-ipXc: we simply multiply by 4 to account for the 4 warp schedulers; + * this is inaccurate ! + */ +static const struct nvc0_hw_sm_query_cfg nve4_hw_sm_queries[] = +{ + _Q1B(ACTIVE_CYCLES, 0x0001, B6, WARP, 0x00000000, 1, 1), + _Q1B(ACTIVE_WARPS, 0x003f, B6, WARP, 0x31483104, 2, 1), + _Q1A(ATOM_COUNT, 0x0001, B6, BRANCH, 0x00000000, 1, 1), + _Q1A(BRANCH, 0x0001, B6, BRANCH, 0x0000000c, 1, 1), + _Q1A(DIVERGENT_BRANCH, 0x0001, B6, BRANCH, 0x00000010, 1, 1), + _Q1A(GLD_REQUEST, 0x0001, B6, LDST, 0x00000010, 1, 1), + _Q1B(GLD_MEM_DIV_REPLAY, 0x0001, B6, REPLAY, 0x00000010, 1, 1), + _Q1B(GST_TRANSACTIONS, 0x0001, B6, MEM, 0x00000004, 1, 1), + _Q1B(GST_MEM_DIV_REPLAY, 0x0001, B6, REPLAY, 0x00000014, 1, 1), + _Q1A(GRED_COUNT, 0x0001, B6, BRANCH, 0x00000008, 1, 1), + _Q1A(GST_REQUEST, 0x0001, B6, LDST, 0x00000014, 1, 1), + _Q1A(INST_EXECUTED, 0x0003, B6, EXEC, 0x00000398, 1, 1), + _Q1A(INST_ISSUED, 0x0003, B6, ISSUE, 0x00000104, 1, 1), + _Q1A(INST_ISSUED1, 0x0001, B6, ISSUE, 0x00000004, 1, 1), + _Q1A(INST_ISSUED2, 0x0001, B6, ISSUE, 0x00000008, 1, 1), + _Q1B(L1_GLD_HIT, 0x0001, B6, L1, 0x00000010, 1, 1), + _Q1B(L1_GLD_MISS, 0x0001, B6, L1, 0x00000014, 1, 1), + _Q1B(L1_LOCAL_LD_HIT, 0x0001, B6, L1, 0x00000000, 1, 1), + _Q1B(L1_LOCAL_LD_MISS, 0x0001, B6, L1, 0x00000004, 1, 1), + _Q1B(L1_LOCAL_ST_HIT, 0x0001, B6, L1, 0x00000008, 1, 1), + _Q1B(L1_LOCAL_ST_MISS, 0x0001, B6, L1, 0x0000000c, 1, 1), + _Q1B(L1_SHARED_LD_TRANSACTIONS, 0x0001, B6, TRANSACTION, 0x00000008, 1, 1), + _Q1B(L1_SHARED_ST_TRANSACTIONS, 0x0001, B6, TRANSACTION, 0x0000000c, 1, 1), + _Q1A(LOCAL_LD, 0x0001, B6, LDST, 0x00000008, 1, 1), + _Q1B(LOCAL_LD_TRANSACTIONS, 0x0001, B6, TRANSACTION, 0x00000000, 1, 1), + _Q1A(LOCAL_ST, 0x0001, B6, LDST, 0x0000000c, 1, 1), + _Q1B(LOCAL_ST_TRANSACTIONS, 0x0001, B6, TRANSACTION, 0x00000004, 1, 1), + _Q1A(PROF_TRIGGER_0, 0x0001, B6, USER, 0x00000000, 1, 1), + _Q1A(PROF_TRIGGER_1, 0x0001, B6, USER, 0x00000004, 1, 1), + _Q1A(PROF_TRIGGER_2, 0x0001, B6, USER, 0x00000008, 1, 1), + _Q1A(PROF_TRIGGER_3, 0x0001, B6, USER, 0x0000000c, 1, 1), + _Q1A(PROF_TRIGGER_4, 0x0001, B6, USER, 0x00000010, 1, 1), + _Q1A(PROF_TRIGGER_5, 0x0001, B6, USER, 0x00000014, 1, 1), + _Q1A(PROF_TRIGGER_6, 0x0001, B6, USER, 0x00000018, 1, 1), + _Q1A(PROF_TRIGGER_7, 0x0001, B6, USER, 0x0000001c, 1, 1), + _Q1A(SHARED_LD, 0x0001, B6, LDST, 0x00000000, 1, 1), + _Q1B(SHARED_LD_REPLAY, 0x0001, B6, REPLAY, 0x00000008, 1, 1), + _Q1A(SHARED_ST, 0x0001, B6, LDST, 0x00000004, 1, 1), + _Q1B(SHARED_ST_REPLAY, 0x0001, B6, REPLAY, 0x0000000c, 1, 1), + _Q1B(SM_CTA_LAUNCHED, 0x0001, B6, WARP, 0x0000001c, 1, 1), + _Q1A(THREADS_LAUNCHED, 0x003f, B6, LAUNCH, 0x398a4188, 1, 1), + _Q1B(UNCACHED_GLD_TRANSACTIONS, 0x0001, B6, MEM, 0x00000000, 1, 1), + _Q1A(WARPS_LAUNCHED, 0x0001, B6, LAUNCH, 0x00000004, 1, 1), + _M2AB(IPC, 0x3, B6, EXEC, 0x398, 0xffff, LOGOP, WARP, 0x0, DIV_SUM_M0, 10, 1), + _M2AB(IPAC, 0x3, B6, EXEC, 0x398, 0x1, B6, WARP, 0x0, AVG_DIV_MM, 10, 1), + _M2A(IPEC, 0x3, B6, EXEC, 0x398, 0xe, LOGOP, EXEC, 0x398, AVG_DIV_MM, 10, 1), + _M2A(INST_REPLAY_OHEAD, 0x3, B6, ISSUE, 0x104, 0x3, B6, EXEC, 0x398, REL_SUM_MM, 100, 1), + _M2B(MP_OCCUPANCY, 0x3f, B6, WARP, 0x31483104, 0x01, B6, WARP, 0x0, AVG_DIV_MM, 200, 64), + _M2B(MP_EFFICIENCY, 0x01, B6, WARP, 0x0, 0xffff, LOGOP, WARP, 0x0, AVG_DIV_M0, 100, 1), +}; + +#undef _Q1A +#undef _Q1B +#undef _M2A +#undef _M2B + +/* === PERFORMANCE MONITORING COUNTERS for NVC0:NVE4 === */ +static const uint64_t nvc0_read_hw_sm_counters_code[] = +{ + /* mov b32 $r8 $tidx + * mov b32 $r9 $physid + * mov b32 $r0 $pm0 + * mov b32 $r1 $pm1 + * mov b32 $r2 $pm2 + * mov b32 $r3 $pm3 + * mov b32 $r4 $pm4 + * mov b32 $r5 $pm5 + * mov b32 $r6 $pm6 + * mov b32 $r7 $pm7 + * set $p0 0x1 eq u32 $r8 0x0 + * mov b32 $r10 c0[0x0] + * mov b32 $r11 c0[0x4] + * ext u32 $r8 $r9 0x414 + * (not $p0) exit + * mul $r8 u32 $r8 u32 36 + * add b32 $r10 $c $r10 $r8 + * add b32 $r11 $r11 0x0 $c + * mov b32 $r8 c0[0x8] + * st b128 wt g[$r10d+0x00] $r0q + * st b128 wt g[$r10d+0x10] $r4q + * st b32 wt g[$r10d+0x20] $r8 + * exit */ + 0x2c00000084021c04ULL, + 0x2c0000000c025c04ULL, + 0x2c00000010001c04ULL, + 0x2c00000014005c04ULL, + 0x2c00000018009c04ULL, + 0x2c0000001c00dc04ULL, + 0x2c00000020011c04ULL, + 0x2c00000024015c04ULL, + 0x2c00000028019c04ULL, + 0x2c0000002c01dc04ULL, + 0x190e0000fc81dc03ULL, + 0x2800400000029de4ULL, + 0x280040001002dde4ULL, + 0x7000c01050921c03ULL, + 0x80000000000021e7ULL, + 0x1000000090821c02ULL, + 0x4801000020a29c03ULL, + 0x0800000000b2dc42ULL, + 0x2800400020021de4ULL, + 0x9400000000a01fc5ULL, + 0x9400000040a11fc5ULL, + 0x9400000080a21f85ULL, + 0x8000000000001de7ULL +}; + +#define _Q(n, f, m, g, c, s0, s1, s2, s3, s4, s5) [NVC0_HW_SM_QUERY_##n] = { { { f, NVC0_COMPUTE_MP_PM_OP_MODE_##m, c, 0, g, s0|(s1 << 8)|(s2 << 16)|(s3 << 24)|(s4##ULL << 32)|(s5##ULL << 40) }, {}, {}, {} }, 1, NVC0_COUNTER_OPn_SUM, { 1, 1 } } + +static const struct nvc0_hw_sm_query_cfg nvc0_hw_sm_queries[] = +{ + _Q(ACTIVE_CYCLES, 0xaaaa, LOGOP, 0x11, 1, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(ACTIVE_WARPS, 0xaaaa, LOGOP, 0x24, 6, 0x10, 0x21, 0x32, 0x43, 0x54, 0x65), + _Q(ATOM_COUNT, 0xaaaa, LOGOP, 0x63, 1, 0x30, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(BRANCH, 0xaaaa, LOGOP, 0x1a, 2, 0x00, 0x11, 0x00, 0x00, 0x00, 0x00), + _Q(DIVERGENT_BRANCH, 0xaaaa, LOGOP, 0x19, 2, 0x20, 0x31, 0x00, 0x00, 0x00, 0x00), + _Q(GLD_REQUEST, 0xaaaa, LOGOP, 0x64, 1, 0x30, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(GRED_COUNT, 0xaaaa, LOGOP, 0x63, 1, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(GST_REQUEST, 0xaaaa, LOGOP, 0x64, 1, 0x60, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(INST_EXECUTED, 0xaaaa, LOGOP, 0x2d, 3, 0x00, 0x11, 0x22, 0x00, 0x00, 0x00), + _Q(INST_ISSUED1_0, 0xaaaa, LOGOP, 0x7e, 1, 0x10, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(INST_ISSUED1_1, 0xaaaa, LOGOP, 0x7e, 1, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(INST_ISSUED2_0, 0xaaaa, LOGOP, 0x7e, 1, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(INST_ISSUED2_1, 0xaaaa, LOGOP, 0x7e, 1, 0x50, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(LOCAL_LD, 0xaaaa, LOGOP, 0x64, 1, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(LOCAL_ST, 0xaaaa, LOGOP, 0x64, 1, 0x50, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(PROF_TRIGGER_0, 0xaaaa, LOGOP, 0x01, 1, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(PROF_TRIGGER_1, 0xaaaa, LOGOP, 0x01, 1, 0x10, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(PROF_TRIGGER_2, 0xaaaa, LOGOP, 0x01, 1, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(PROF_TRIGGER_3, 0xaaaa, LOGOP, 0x01, 1, 0x30, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(PROF_TRIGGER_4, 0xaaaa, LOGOP, 0x01, 1, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(PROF_TRIGGER_5, 0xaaaa, LOGOP, 0x01, 1, 0x50, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(PROF_TRIGGER_6, 0xaaaa, LOGOP, 0x01, 1, 0x60, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(PROF_TRIGGER_7, 0xaaaa, LOGOP, 0x01, 1, 0x70, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(SHARED_LD, 0xaaaa, LOGOP, 0x64, 1, 0x10, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(SHARED_ST, 0xaaaa, LOGOP, 0x64, 1, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00), + _Q(THREADS_LAUNCHED, 0xaaaa, LOGOP, 0x26, 6, 0x10, 0x21, 0x32, 0x43, 0x54, 0x65), + _Q(TH_INST_EXECUTED_0, 0xaaaa, LOGOP, 0xa3, 6, 0x00, 0x11, 0x22, 0x33, 0x44, 0x55), + _Q(TH_INST_EXECUTED_1, 0xaaaa, LOGOP, 0xa5, 6, 0x00, 0x11, 0x22, 0x33, 0x44, 0x55), + _Q(TH_INST_EXECUTED_2, 0xaaaa, LOGOP, 0xa4, 6, 0x00, 0x11, 0x22, 0x33, 0x44, 0x55), + _Q(TH_INST_EXECUTED_3, 0xaaaa, LOGOP, 0xa6, 6, 0x00, 0x11, 0x22, 0x33, 0x44, 0x55), + _Q(WARPS_LAUNCHED, 0xaaaa, LOGOP, 0x26, 1, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00), +}; + +#undef _Q + +static const struct nvc0_hw_sm_query_cfg * +nvc0_hw_sm_query_get_cfg(struct nvc0_context *nvc0, struct nvc0_query *q) +{ + struct nvc0_screen *screen = nvc0->screen; + + if (screen->base.class_3d >= NVE4_3D_CLASS) + return &nve4_hw_sm_queries[q->type - PIPE_QUERY_DRIVER_SPECIFIC]; + return &nvc0_hw_sm_queries[q->type - NVC0_HW_SM_QUERY(0)]; +} + +static boolean +nvc0_hw_sm_query_begin(struct nvc0_context *nvc0, struct nvc0_query *q) +{ + struct nvc0_screen *screen = nvc0->screen; + struct nouveau_pushbuf *push = nvc0->base.pushbuf; + const bool is_nve4 = screen->base.class_3d >= NVE4_3D_CLASS; + struct nvc0_hw_query *hq = nvc0_hw_query(q); + const struct nvc0_hw_sm_query_cfg *cfg; + unsigned i, c; + unsigned num_ab[2] = { 0, 0 }; + + cfg = nvc0_hw_sm_query_get_cfg(nvc0, q); + + /* check if we have enough free counter slots */ + for (i = 0; i < cfg->num_counters; ++i) + num_ab[cfg->ctr[i].sig_dom]++; + + if (screen->pm.num_hw_sm_active[0] + num_ab[0] > 4 || + screen->pm.num_hw_sm_active[1] + num_ab[1] > 4) { + NOUVEAU_ERR("Not enough free MP counter slots !\n"); + return false; + } + + assert(cfg->num_counters <= 4); + PUSH_SPACE(push, 4 * 8 * (is_nve4 ? 1 : 6) + 6); + + if (!screen->pm.mp_counters_enabled) { + screen->pm.mp_counters_enabled = true; + BEGIN_NVC0(push, SUBC_SW(0x06ac), 1); + PUSH_DATA (push, 0x1fcb); + } + + /* set sequence field to 0 (used to check if result is available) */ + for (i = 0; i < screen->mp_count; ++i) + hq->data[i * 10 + 10] = 0; + + for (i = 0; i < cfg->num_counters; ++i) { + const unsigned d = cfg->ctr[i].sig_dom; + + if (!screen->pm.num_hw_sm_active[d]) { + uint32_t m = (1 << 22) | (1 << (7 + (8 * !d))); + if (screen->pm.num_hw_sm_active[!d]) + m |= 1 << (7 + (8 * d)); + BEGIN_NVC0(push, SUBC_SW(0x0600), 1); + PUSH_DATA (push, m); + } + screen->pm.num_hw_sm_active[d]++; + + for (c = d * 4; c < (d * 4 + 4); ++c) { + if (!screen->pm.mp_counter[c]) { + hq->ctr[i] = c; + screen->pm.mp_counter[c] = (struct pipe_query *)q; + break; + } + } + assert(c <= (d * 4 + 3)); /* must succeed, already checked for space */ + + /* configure and reset the counter(s) */ + if (is_nve4) { + if (d == 0) + BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_A_SIGSEL(c & 3)), 1); + else + BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_B_SIGSEL(c & 3)), 1); + PUSH_DATA (push, cfg->ctr[i].sig_sel); + BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_SRCSEL(c)), 1); + PUSH_DATA (push, cfg->ctr[i].src_sel + 0x2108421 * (c & 3)); + BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_FUNC(c)), 1); + PUSH_DATA (push, (cfg->ctr[i].func << 4) | cfg->ctr[i].mode); + BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_SET(c)), 1); + PUSH_DATA (push, 0); + } else { + unsigned s; + + for (s = 0; s < cfg->ctr[i].num_src; s++) { + BEGIN_NVC0(push, NVC0_COMPUTE(MP_PM_SIGSEL(s)), 1); + PUSH_DATA (push, cfg->ctr[i].sig_sel); + BEGIN_NVC0(push, NVC0_COMPUTE(MP_PM_SRCSEL(s)), 1); + PUSH_DATA (push, (cfg->ctr[i].src_sel >> (s * 8)) & 0xff); + BEGIN_NVC0(push, NVC0_COMPUTE(MP_PM_OP(s)), 1); + PUSH_DATA (push, (cfg->ctr[i].func << 4) | cfg->ctr[i].mode); + BEGIN_NVC0(push, NVC0_COMPUTE(MP_PM_SET(s)), 1); + PUSH_DATA (push, 0); + } + } + } + return true; +} + +static void +nvc0_hw_sm_query_end(struct nvc0_context *nvc0, struct nvc0_query *q) +{ + struct nvc0_screen *screen = nvc0->screen; + struct pipe_context *pipe = &nvc0->base.pipe; + struct nouveau_pushbuf *push = nvc0->base.pushbuf; + const bool is_nve4 = screen->base.class_3d >= NVE4_3D_CLASS; + struct nvc0_hw_query *hq = nvc0_hw_query(q); + uint32_t mask; + uint32_t input[3]; + const uint block[3] = { 32, is_nve4 ? 4 : 1, 1 }; + const uint grid[3] = { screen->mp_count, 1, 1 }; + unsigned c; + const struct nvc0_hw_sm_query_cfg *cfg; + + cfg = nvc0_hw_sm_query_get_cfg(nvc0, q); + + if (unlikely(!screen->pm.prog)) { + struct nvc0_program *prog = CALLOC_STRUCT(nvc0_program); + prog->type = PIPE_SHADER_COMPUTE; + prog->translated = true; + prog->num_gprs = 14; + prog->parm_size = 12; + if (is_nve4) { + prog->code = (uint32_t *)nve4_read_hw_sm_counters_code; + prog->code_size = sizeof(nve4_read_hw_sm_counters_code); + } else { + prog->code = (uint32_t *)nvc0_read_hw_sm_counters_code; + prog->code_size = sizeof(nvc0_read_hw_sm_counters_code); + } + screen->pm.prog = prog; + } + + /* disable all counting */ + PUSH_SPACE(push, 8); + for (c = 0; c < 8; ++c) + if (screen->pm.mp_counter[c]) { + if (is_nve4) { + IMMED_NVC0(push, NVE4_COMPUTE(MP_PM_FUNC(c)), 0); + } else { + IMMED_NVC0(push, NVC0_COMPUTE(MP_PM_OP(c)), 0); + } + } + /* release counters for this query */ + for (c = 0; c < 8; ++c) { + if (nvc0_query(screen->pm.mp_counter[c]) == q) { + screen->pm.num_hw_sm_active[c / 4]--; + screen->pm.mp_counter[c] = NULL; + } + } + + BCTX_REFN_bo(nvc0->bufctx_cp, CP_QUERY, NOUVEAU_BO_GART | NOUVEAU_BO_WR, + hq->bo); + + PUSH_SPACE(push, 1); + IMMED_NVC0(push, SUBC_COMPUTE(NV50_GRAPH_SERIALIZE), 0); + + pipe->bind_compute_state(pipe, screen->pm.prog); + input[0] = (hq->bo->offset + hq->base_offset); + input[1] = (hq->bo->offset + hq->base_offset) >> 32; + input[2] = hq->sequence; + pipe->launch_grid(pipe, block, grid, 0, input); + + nouveau_bufctx_reset(nvc0->bufctx_cp, NVC0_BIND_CP_QUERY); + + /* re-activate other counters */ + PUSH_SPACE(push, 16); + mask = 0; + for (c = 0; c < 8; ++c) { + unsigned i; + q = nvc0_query(screen->pm.mp_counter[c]); + if (!q) + continue; + cfg = nvc0_hw_sm_query_get_cfg(nvc0, q); + for (i = 0; i < cfg->num_counters; ++i) { + if (mask & (1 << hq->ctr[i])) + break; + mask |= 1 << hq->ctr[i]; + if (is_nve4) { + BEGIN_NVC0(push, NVE4_COMPUTE(MP_PM_FUNC(hq->ctr[i])), 1); + } else { + BEGIN_NVC0(push, NVC0_COMPUTE(MP_PM_OP(hq->ctr[i])), 1); + } + PUSH_DATA (push, (cfg->ctr[i].func << 4) | cfg->ctr[i].mode); + } + } +} + +static inline bool +nvc0_hw_sm_query_read_data(uint32_t count[32][4], + struct nvc0_context *nvc0, bool wait, + struct nvc0_query *q, + const struct nvc0_hw_sm_query_cfg *cfg, + unsigned mp_count) +{ + struct nvc0_hw_query *hq = nvc0_hw_query(q); + unsigned p, c; + + for (p = 0; p < mp_count; ++p) { + const unsigned b = (0x24 / 4) * p; + + for (c = 0; c < cfg->num_counters; ++c) { + if (hq->data[b + 8] != hq->sequence) { + if (!wait) + return false; + if (nouveau_bo_wait(hq->bo, NOUVEAU_BO_RD, nvc0->base.client)) + return false; + } + count[p][c] = hq->data[b + hq->ctr[c]]; + } + } + return true; +} + +static inline bool +nve4_hw_sm_query_read_data(uint32_t count[32][4], + struct nvc0_context *nvc0, bool wait, + struct nvc0_query *q, + const struct nvc0_hw_sm_query_cfg *cfg, + unsigned mp_count) +{ + struct nvc0_hw_query *hq = nvc0_hw_query(q); + unsigned p, c, d; + + for (p = 0; p < mp_count; ++p) { + const unsigned b = (0x60 / 4) * p; + + for (c = 0; c < cfg->num_counters; ++c) { + count[p][c] = 0; + for (d = 0; d < ((hq->ctr[c] & ~3) ? 1 : 4); ++d) { + if (hq->data[b + 20 + d] != hq->sequence) { + if (!wait) + return false; + if (nouveau_bo_wait(hq->bo, NOUVEAU_BO_RD, nvc0->base.client)) + return false; + } + if (hq->ctr[c] & ~0x3) + count[p][c] = hq->data[b + 16 + (hq->ctr[c] & 3)]; + else + count[p][c] += hq->data[b + d * 4 + hq->ctr[c]]; + } + } + } + return true; +} + +/* Metric calculations: + * sum(x) ... sum of x over all MPs + * avg(x) ... average of x over all MPs + * + * IPC : sum(inst_executed) / clock + * INST_REPLAY_OHEAD: (sum(inst_issued) - sum(inst_executed)) / sum(inst_issued) + * MP_OCCUPANCY : avg((active_warps / 64) / active_cycles) + * MP_EFFICIENCY : avg(active_cycles / clock) + * + * NOTE: Interpretation of IPC requires knowledge of MP count. + */ +static boolean +nvc0_hw_sm_query_result(struct nvc0_context *nvc0, struct nvc0_query *q, + void *result, boolean wait) +{ + uint32_t count[32][4]; + uint64_t value = 0; + unsigned mp_count = MIN2(nvc0->screen->mp_count_compute, 32); + unsigned p, c; + const struct nvc0_hw_sm_query_cfg *cfg; + bool ret; + + cfg = nvc0_hw_sm_query_get_cfg(nvc0, q); + + if (nvc0->screen->base.class_3d >= NVE4_3D_CLASS) + ret = nve4_hw_sm_query_read_data(count, nvc0, wait, q, cfg, mp_count); + else + ret = nvc0_hw_sm_query_read_data(count, nvc0, wait, q, cfg, mp_count); + if (!ret) + return false; + + if (cfg->op == NVC0_COUNTER_OPn_SUM) { + for (c = 0; c < cfg->num_counters; ++c) + for (p = 0; p < mp_count; ++p) + value += count[p][c]; + value = (value * cfg->norm[0]) / cfg->norm[1]; + } else + if (cfg->op == NVC0_COUNTER_OPn_OR) { + uint32_t v = 0; + for (c = 0; c < cfg->num_counters; ++c) + for (p = 0; p < mp_count; ++p) + v |= count[p][c]; + value = ((uint64_t)v * cfg->norm[0]) / cfg->norm[1]; + } else + if (cfg->op == NVC0_COUNTER_OPn_AND) { + uint32_t v = ~0; + for (c = 0; c < cfg->num_counters; ++c) + for (p = 0; p < mp_count; ++p) + v &= count[p][c]; + value = ((uint64_t)v * cfg->norm[0]) / cfg->norm[1]; + } else + if (cfg->op == NVC0_COUNTER_OP2_REL_SUM_MM) { + uint64_t v[2] = { 0, 0 }; + for (p = 0; p < mp_count; ++p) { + v[0] += count[p][0]; + v[1] += count[p][1]; + } + if (v[0]) + value = ((v[0] - v[1]) * cfg->norm[0]) / (v[0] * cfg->norm[1]); + } else + if (cfg->op == NVC0_COUNTER_OP2_DIV_SUM_M0) { + for (p = 0; p < mp_count; ++p) + value += count[p][0]; + if (count[0][1]) + value = (value * cfg->norm[0]) / (count[0][1] * cfg->norm[1]); + else + value = 0; + } else + if (cfg->op == NVC0_COUNTER_OP2_AVG_DIV_MM) { + unsigned mp_used = 0; + for (p = 0; p < mp_count; ++p, mp_used += !!count[p][0]) + if (count[p][1]) + value += (count[p][0] * cfg->norm[0]) / count[p][1]; + if (mp_used) + value /= (uint64_t)mp_used * cfg->norm[1]; + } else + if (cfg->op == NVC0_COUNTER_OP2_AVG_DIV_M0) { + unsigned mp_used = 0; + for (p = 0; p < mp_count; ++p, mp_used += !!count[p][0]) + value += count[p][0]; + if (count[0][1] && mp_used) { + value *= cfg->norm[0]; + value /= (uint64_t)count[0][1] * mp_used * cfg->norm[1]; + } else { + value = 0; + } + } + + *(uint64_t *)result = value; + return true; +} + +static bool +nvc0_hw_query_allocate(struct nvc0_context *nvc0, struct nvc0_query *q, + int size) +{ + struct nvc0_hw_query *hq = nvc0_hw_query(q); + struct nvc0_screen *screen = nvc0->screen; + int ret; + + if (hq->bo) { + nouveau_bo_ref(NULL, &hq->bo); + if (hq->mm) { + if (hq->state == NVC0_HW_QUERY_STATE_READY) + nouveau_mm_free(hq->mm); + else + nouveau_fence_work(screen->base.fence.current, + nouveau_mm_free_work, hq->mm); + } + } + if (size) { + hq->mm = nouveau_mm_allocate(screen->base.mm_GART, size, &hq->bo, + &hq->base_offset); + if (!hq->bo) + return false; + hq->offset = hq->base_offset; + + ret = nouveau_bo_map(hq->bo, 0, screen->base.client); + if (ret) { + nvc0_hw_query_allocate(nvc0, q, 0); + return false; + } + hq->data = (uint32_t *)((uint8_t *)hq->bo->map + hq->base_offset); + } + return true; +} + +static void +nvc0_hw_query_get(struct nouveau_pushbuf *push, struct nvc0_query *q, + unsigned offset, uint32_t get) +{ + struct nvc0_hw_query *hq = nvc0_hw_query(q); + + offset += hq->offset; + + PUSH_SPACE(push, 5); + PUSH_REFN (push, hq->bo, NOUVEAU_BO_GART | NOUVEAU_BO_WR); + BEGIN_NVC0(push, NVC0_3D(QUERY_ADDRESS_HIGH), 4); + PUSH_DATAh(push, hq->bo->offset + offset); + PUSH_DATA (push, hq->bo->offset + offset); + PUSH_DATA (push, hq->sequence); + PUSH_DATA (push, get); +} + +static void +nvc0_hw_query_rotate(struct nvc0_context *nvc0, struct nvc0_query *q) +{ + struct nvc0_hw_query *hq = nvc0_hw_query(q); + + hq->offset += hq->rotate; + hq->data += hq->rotate / sizeof(*hq->data); + if (hq->offset - hq->base_offset == NVC0_HW_QUERY_ALLOC_SPACE) + nvc0_hw_query_allocate(nvc0, q, NVC0_HW_QUERY_ALLOC_SPACE); +} + +static inline void +nvc0_hw_query_update(struct nouveau_client *cli, struct nvc0_query *q) +{ + struct nvc0_hw_query *hq = nvc0_hw_query(q); + + if (hq->is64bit) { + if (nouveau_fence_signalled(hq->fence)) + hq->state = NVC0_HW_QUERY_STATE_READY; + } else { + if (hq->data[0] == hq->sequence) + hq->state = NVC0_HW_QUERY_STATE_READY; + } +} + +static void +nvc0_hw_destroy_query(struct nvc0_context *nvc0, struct nvc0_query *q) +{ + struct nvc0_hw_query *hq = nvc0_hw_query(q); + nvc0_hw_query_allocate(nvc0, q, 0); + nouveau_fence_ref(NULL, &hq->fence); + FREE(hq); +} + +static boolean +nvc0_hw_begin_query(struct nvc0_context *nvc0, struct nvc0_query *q) +{ + struct nouveau_pushbuf *push = nvc0->base.pushbuf; + struct nvc0_hw_query *hq = nvc0_hw_query(q); + bool ret = true; + + /* For occlusion queries we have to change the storage, because a previous + * query might set the initial render conition to false even *after* we re- + * initialized it to true. + */ + if (hq->rotate) { + nvc0_hw_query_rotate(nvc0, q); + + /* XXX: can we do this with the GPU, and sync with respect to a previous + * query ? + */ + hq->data[0] = hq->sequence; /* initialize sequence */ + hq->data[1] = 1; /* initial render condition = true */ + hq->data[4] = hq->sequence + 1; /* for comparison COND_MODE */ + hq->data[5] = 0; + } + hq->sequence++; + + switch (q->type) { + case PIPE_QUERY_OCCLUSION_COUNTER: + case PIPE_QUERY_OCCLUSION_PREDICATE: + hq->nesting = nvc0->screen->num_occlusion_queries_active++; + if (hq->nesting) { + nvc0_hw_query_get(push, q, 0x10, 0x0100f002); + } else { + PUSH_SPACE(push, 3); + BEGIN_NVC0(push, NVC0_3D(COUNTER_RESET), 1); + PUSH_DATA (push, NVC0_3D_COUNTER_RESET_SAMPLECNT); + IMMED_NVC0(push, NVC0_3D(SAMPLECNT_ENABLE), 1); + } + break; + case PIPE_QUERY_PRIMITIVES_GENERATED: + nvc0_hw_query_get(push, q, 0x10, 0x09005002 | (q->index << 5)); + break; + case PIPE_QUERY_PRIMITIVES_EMITTED: + nvc0_hw_query_get(push, q, 0x10, 0x05805002 | (q->index << 5)); + break; + case PIPE_QUERY_SO_STATISTICS: + nvc0_hw_query_get(push, q, 0x20, 0x05805002 | (q->index << 5)); + nvc0_hw_query_get(push, q, 0x30, 0x06805002 | (q->index << 5)); + break; + case PIPE_QUERY_SO_OVERFLOW_PREDICATE: + nvc0_hw_query_get(push, q, 0x10, 0x03005002 | (q->index << 5)); + break; + case PIPE_QUERY_TIME_ELAPSED: + nvc0_hw_query_get(push, q, 0x10, 0x00005002); + break; + case PIPE_QUERY_PIPELINE_STATISTICS: + nvc0_hw_query_get(push, q, 0xc0 + 0x00, 0x00801002); /* VFETCH, VERTICES */ + nvc0_hw_query_get(push, q, 0xc0 + 0x10, 0x01801002); /* VFETCH, PRIMS */ + nvc0_hw_query_get(push, q, 0xc0 + 0x20, 0x02802002); /* VP, LAUNCHES */ + nvc0_hw_query_get(push, q, 0xc0 + 0x30, 0x03806002); /* GP, LAUNCHES */ + nvc0_hw_query_get(push, q, 0xc0 + 0x40, 0x04806002); /* GP, PRIMS_OUT */ + nvc0_hw_query_get(push, q, 0xc0 + 0x50, 0x07804002); /* RAST, PRIMS_IN */ + nvc0_hw_query_get(push, q, 0xc0 + 0x60, 0x08804002); /* RAST, PRIMS_OUT */ + nvc0_hw_query_get(push, q, 0xc0 + 0x70, 0x0980a002); /* ROP, PIXELS */ + nvc0_hw_query_get(push, q, 0xc0 + 0x80, 0x0d808002); /* TCP, LAUNCHES */ + nvc0_hw_query_get(push, q, 0xc0 + 0x90, 0x0e809002); /* TEP, LAUNCHES */ + break; + default: + if ((q->type >= NVE4_HW_SM_QUERY(0) && q->type <= NVE4_HW_SM_QUERY_LAST) || + (q->type >= NVC0_HW_SM_QUERY(0) && q->type <= NVC0_HW_SM_QUERY_LAST)) { + ret = nvc0_hw_sm_query_begin(nvc0, q); + } + break; + } + hq->state = NVC0_HW_QUERY_STATE_ACTIVE; + return ret; +} + +static void +nvc0_hw_end_query(struct nvc0_context *nvc0, struct nvc0_query *q) +{ + struct nouveau_pushbuf *push = nvc0->base.pushbuf; + struct nvc0_hw_query *hq = nvc0_hw_query(q); + + if (hq->state != NVC0_HW_QUERY_STATE_ACTIVE) { + /* some queries don't require 'begin' to be called (e.g. GPU_FINISHED) */ + if (hq->rotate) + nvc0_hw_query_rotate(nvc0, q); + hq->sequence++; + } + hq->state = NVC0_HW_QUERY_STATE_ENDED; + + switch (q->type) { + case PIPE_QUERY_OCCLUSION_COUNTER: + case PIPE_QUERY_OCCLUSION_PREDICATE: + nvc0_hw_query_get(push, q, 0, 0x0100f002); + if (--nvc0->screen->num_occlusion_queries_active == 0) { + PUSH_SPACE(push, 1); + IMMED_NVC0(push, NVC0_3D(SAMPLECNT_ENABLE), 0); + } + break; + case PIPE_QUERY_PRIMITIVES_GENERATED: + nvc0_hw_query_get(push, q, 0, 0x09005002 | (q->index << 5)); + break; + case PIPE_QUERY_PRIMITIVES_EMITTED: + nvc0_hw_query_get(push, q, 0, 0x05805002 | (q->index << 5)); + break; + case PIPE_QUERY_SO_STATISTICS: + nvc0_hw_query_get(push, q, 0x00, 0x05805002 | (q->index << 5)); + nvc0_hw_query_get(push, q, 0x10, 0x06805002 | (q->index << 5)); + break; + case PIPE_QUERY_SO_OVERFLOW_PREDICATE: + /* TODO: How do we sum over all streams for render condition ? */ + /* PRIMS_DROPPED doesn't write sequence, use a ZERO query to sync on */ + nvc0_hw_query_get(push, q, 0x00, 0x03005002 | (q->index << 5)); + nvc0_hw_query_get(push, q, 0x20, 0x00005002); + break; + case PIPE_QUERY_TIMESTAMP: + case PIPE_QUERY_TIME_ELAPSED: + nvc0_hw_query_get(push, q, 0, 0x00005002); + break; + case PIPE_QUERY_GPU_FINISHED: + nvc0_hw_query_get(push, q, 0, 0x1000f010); + break; + case PIPE_QUERY_PIPELINE_STATISTICS: + nvc0_hw_query_get(push, q, 0x00, 0x00801002); /* VFETCH, VERTICES */ + nvc0_hw_query_get(push, q, 0x10, 0x01801002); /* VFETCH, PRIMS */ + nvc0_hw_query_get(push, q, 0x20, 0x02802002); /* VP, LAUNCHES */ + nvc0_hw_query_get(push, q, 0x30, 0x03806002); /* GP, LAUNCHES */ + nvc0_hw_query_get(push, q, 0x40, 0x04806002); /* GP, PRIMS_OUT */ + nvc0_hw_query_get(push, q, 0x50, 0x07804002); /* RAST, PRIMS_IN */ + nvc0_hw_query_get(push, q, 0x60, 0x08804002); /* RAST, PRIMS_OUT */ + nvc0_hw_query_get(push, q, 0x70, 0x0980a002); /* ROP, PIXELS */ + nvc0_hw_query_get(push, q, 0x80, 0x0d808002); /* TCP, LAUNCHES */ + nvc0_hw_query_get(push, q, 0x90, 0x0e809002); /* TEP, LAUNCHES */ + break; + case PIPE_QUERY_TIMESTAMP_DISJOINT: + /* This query is not issued on GPU because disjoint is forced to false */ + hq->state = NVC0_HW_QUERY_STATE_READY; + break; + case NVC0_HW_QUERY_TFB_BUFFER_OFFSET: + /* indexed by TFB buffer instead of by vertex stream */ + nvc0_hw_query_get(push, q, 0x00, 0x0d005002 | (q->index << 5)); + break; + default: + if ((q->type >= NVE4_HW_SM_QUERY(0) && q->type <= NVE4_HW_SM_QUERY_LAST) || + (q->type >= NVC0_HW_SM_QUERY(0) && q->type <= NVC0_HW_SM_QUERY_LAST)) { + nvc0_hw_sm_query_end(nvc0, q); + } + break; + } + if (hq->is64bit) + nouveau_fence_ref(nvc0->screen->base.fence.current, &hq->fence); +} + +static boolean +nvc0_hw_get_query_result(struct nvc0_context *nvc0, struct nvc0_query *q, + boolean wait, union pipe_query_result *result) +{ + struct nvc0_hw_query *hq = nvc0_hw_query(q); + uint64_t *res64 = (uint64_t*)result; + uint32_t *res32 = (uint32_t*)result; + uint8_t *res8 = (uint8_t*)result; + uint64_t *data64 = (uint64_t *)hq->data; + unsigned i; + + if ((q->type >= NVE4_HW_SM_QUERY(0) && q->type <= NVE4_HW_SM_QUERY_LAST) || + (q->type >= NVC0_HW_SM_QUERY(0) && q->type <= NVC0_HW_SM_QUERY_LAST)) { + return nvc0_hw_sm_query_result(nvc0, q, result, wait); + } + + if (hq->state != NVC0_HW_QUERY_STATE_READY) + nvc0_hw_query_update(nvc0->screen->base.client, q); + + if (hq->state != NVC0_HW_QUERY_STATE_READY) { + if (!wait) { + if (hq->state != NVC0_HW_QUERY_STATE_FLUSHED) { + hq->state = NVC0_HW_QUERY_STATE_FLUSHED; + /* flush for silly apps that spin on GL_QUERY_RESULT_AVAILABLE */ + PUSH_KICK(nvc0->base.pushbuf); + } + return false; + } + if (nouveau_bo_wait(hq->bo, NOUVEAU_BO_RD, nvc0->screen->base.client)) + return false; + NOUVEAU_DRV_STAT(&nvc0->screen->base, query_sync_count, 1); + } + hq->state = NVC0_HW_QUERY_STATE_READY; + + switch (q->type) { + case PIPE_QUERY_GPU_FINISHED: + res8[0] = true; + break; + case PIPE_QUERY_OCCLUSION_COUNTER: /* u32 sequence, u32 count, u64 time */ + res64[0] = hq->data[1] - hq->data[5]; + break; + case PIPE_QUERY_OCCLUSION_PREDICATE: + res8[0] = hq->data[1] != hq->data[5]; + break; + case PIPE_QUERY_PRIMITIVES_GENERATED: /* u64 count, u64 time */ + case PIPE_QUERY_PRIMITIVES_EMITTED: /* u64 count, u64 time */ + res64[0] = data64[0] - data64[2]; + break; + case PIPE_QUERY_SO_STATISTICS: + res64[0] = data64[0] - data64[4]; + res64[1] = data64[2] - data64[6]; + break; + case PIPE_QUERY_SO_OVERFLOW_PREDICATE: + res8[0] = data64[0] != data64[2]; + break; + case PIPE_QUERY_TIMESTAMP: + res64[0] = data64[1]; + break; + case PIPE_QUERY_TIMESTAMP_DISJOINT: + res64[0] = 1000000000; + res8[8] = false; + break; + case PIPE_QUERY_TIME_ELAPSED: + res64[0] = data64[1] - data64[3]; + break; + case PIPE_QUERY_PIPELINE_STATISTICS: + for (i = 0; i < 10; ++i) + res64[i] = data64[i * 2] - data64[24 + i * 2]; + break; + case NVC0_HW_QUERY_TFB_BUFFER_OFFSET: + res32[0] = hq->data[1]; + break; + default: + assert(0); /* can't happen, we don't create queries with invalid type */ + return false; + } + + return true; +} + +static const struct nvc0_query_funcs hw_query_funcs = { + .destroy_query = nvc0_hw_destroy_query, + .begin_query = nvc0_hw_begin_query, + .end_query = nvc0_hw_end_query, + .get_query_result = nvc0_hw_get_query_result, +}; + +struct nvc0_query * +nvc0_hw_create_query(struct nvc0_context *nvc0, unsigned type, unsigned index) +{ + struct nvc0_hw_query *hq; + struct nvc0_query *q; + unsigned space = NVC0_HW_QUERY_ALLOC_SPACE; + + hq = CALLOC_STRUCT(nvc0_hw_query); + if (!hq) + return NULL; + + q = &hq->base; + q->funcs = &hw_query_funcs; + q->type = type; + + switch (q->type) { + case PIPE_QUERY_OCCLUSION_COUNTER: + case PIPE_QUERY_OCCLUSION_PREDICATE: + hq->rotate = 32; + space = NVC0_HW_QUERY_ALLOC_SPACE; + break; + case PIPE_QUERY_PIPELINE_STATISTICS: + hq->is64bit = true; + space = 512; + break; + case PIPE_QUERY_SO_STATISTICS: + case PIPE_QUERY_SO_OVERFLOW_PREDICATE: + hq->is64bit = true; + space = 64; + break; + case PIPE_QUERY_PRIMITIVES_GENERATED: + case PIPE_QUERY_PRIMITIVES_EMITTED: + hq->is64bit = true; + q->index = index; + space = 32; + break; + case PIPE_QUERY_TIME_ELAPSED: + case PIPE_QUERY_TIMESTAMP: + case PIPE_QUERY_TIMESTAMP_DISJOINT: + case PIPE_QUERY_GPU_FINISHED: + space = 32; + break; + case NVC0_HW_QUERY_TFB_BUFFER_OFFSET: + space = 16; + break; + default: + if (nvc0->screen->base.device->drm_version >= 0x01000101) { + if (type >= NVE4_HW_SM_QUERY(0) && type <= NVE4_HW_SM_QUERY_LAST) { + /* for each MP: + * [00] = WS0.C0 + * [04] = WS0.C1 + * [08] = WS0.C2 + * [0c] = WS0.C3 + * [10] = WS1.C0 + * [14] = WS1.C1 + * [18] = WS1.C2 + * [1c] = WS1.C3 + * [20] = WS2.C0 + * [24] = WS2.C1 + * [28] = WS2.C2 + * [2c] = WS2.C3 + * [30] = WS3.C0 + * [34] = WS3.C1 + * [38] = WS3.C2 + * [3c] = WS3.C3 + * [40] = MP.C4 + * [44] = MP.C5 + * [48] = MP.C6 + * [4c] = MP.C7 + * [50] = WS0.sequence + * [54] = WS1.sequence + * [58] = WS2.sequence + * [5c] = WS3.sequence + */ + space = (4 * 4 + 4 + 4) * nvc0->screen->mp_count * sizeof(uint32_t); + break; + } else + if (type >= NVC0_HW_SM_QUERY(0) && type <= NVC0_HW_SM_QUERY_LAST) { + /* for each MP: + * [00] = MP.C0 + * [04] = MP.C1 + * [08] = MP.C2 + * [0c] = MP.C3 + * [10] = MP.C4 + * [14] = MP.C5 + * [18] = MP.C6 + * [1c] = MP.C7 + * [20] = MP.sequence + */ + space = (8 + 1) * nvc0->screen->mp_count * sizeof(uint32_t); + break; + } + } + debug_printf("invalid query type: %u\n", type); + FREE(q); + return NULL; + } + + if (!nvc0_hw_query_allocate(nvc0, q, space)) { + FREE(hq); + return NULL; + } + + if (hq->rotate) { + /* we advance before query_begin ! */ + hq->offset -= hq->rotate; + hq->data -= hq->rotate / sizeof(*hq->data); + } else + if (!hq->is64bit) + hq->data[0] = 0; /* initialize sequence */ + + return q; +} + +void +nvc0_hw_query_pushbuf_submit(struct nouveau_pushbuf *push, + struct nvc0_query *q, unsigned result_offset) +{ + struct nvc0_hw_query *hq = nvc0_hw_query(q); + +#define NVC0_IB_ENTRY_1_NO_PREFETCH (1 << (31 - 8)) + + PUSH_REFN(push, hq->bo, NOUVEAU_BO_RD | NOUVEAU_BO_GART); + nouveau_pushbuf_space(push, 0, 0, 1); + nouveau_pushbuf_data(push, hq->bo, hq->offset + result_offset, 4 | + NVC0_IB_ENTRY_1_NO_PREFETCH); +} + +void +nvc0_hw_query_fifo_wait(struct nouveau_pushbuf *push, struct nvc0_query *q) +{ + struct nvc0_hw_query *hq = nvc0_hw_query(q); + unsigned offset = hq->offset; + + if (q->type == PIPE_QUERY_SO_OVERFLOW_PREDICATE) offset += 0x20; + + PUSH_SPACE(push, 5); + PUSH_REFN (push, hq->bo, NOUVEAU_BO_GART | NOUVEAU_BO_RD); + BEGIN_NVC0(push, SUBC_3D(NV84_SUBCHAN_SEMAPHORE_ADDRESS_HIGH), 4); + PUSH_DATAh(push, hq->bo->offset + offset); + PUSH_DATA (push, hq->bo->offset + offset); + PUSH_DATA (push, hq->sequence); + PUSH_DATA (push, (1 << 12) | + NV84_SUBCHAN_SEMAPHORE_TRIGGER_ACQUIRE_EQUAL); +} diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw.h b/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw.h new file mode 100644 index 00000000000..d998c42c213 --- /dev/null +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw.h @@ -0,0 +1,138 @@ +#ifndef __NVC0_QUERY_HW_H__ +#define __NVC0_QUERY_HW_H__ + +#include "nouveau_fence.h" +#include "nouveau_mm.h" + +#include "nvc0_query.h" + +#define NVC0_HW_QUERY_TFB_BUFFER_OFFSET (PIPE_QUERY_TYPES + 0) + +struct nvc0_hw_query { + struct nvc0_query base; + uint32_t *data; + int8_t ctr[4]; + uint32_t sequence; + struct nouveau_bo *bo; + uint32_t base_offset; + uint32_t offset; /* base_offset + i * rotate */ + uint8_t state; + boolean is64bit; + uint8_t rotate; + int nesting; /* only used for occlusion queries */ + struct nouveau_mm_allocation *mm; + struct nouveau_fence *fence; +}; + +static inline struct nvc0_hw_query * +nvc0_hw_query(struct nvc0_query *q) +{ + return (struct nvc0_hw_query *)q; +} + +/* + * Performance counter queries: + */ +#define NVE4_HW_SM_QUERY(i) (PIPE_QUERY_DRIVER_SPECIFIC + (i)) +#define NVE4_HW_SM_QUERY_LAST NVE4_HW_SM_QUERY(NVE4_HW_SM_QUERY_COUNT - 1) +enum nve4_pm_queries +{ + NVE4_HW_SM_QUERY_ACTIVE_CYCLES = 0, + NVE4_HW_SM_QUERY_ACTIVE_WARPS, + NVE4_HW_SM_QUERY_ATOM_COUNT, + NVE4_HW_SM_QUERY_BRANCH, + NVE4_HW_SM_QUERY_DIVERGENT_BRANCH, + NVE4_HW_SM_QUERY_GLD_REQUEST, + NVE4_HW_SM_QUERY_GLD_MEM_DIV_REPLAY, + NVE4_HW_SM_QUERY_GST_TRANSACTIONS, + NVE4_HW_SM_QUERY_GST_MEM_DIV_REPLAY, + NVE4_HW_SM_QUERY_GRED_COUNT, + NVE4_HW_SM_QUERY_GST_REQUEST, + NVE4_HW_SM_QUERY_INST_EXECUTED, + NVE4_HW_SM_QUERY_INST_ISSUED, + NVE4_HW_SM_QUERY_INST_ISSUED1, + NVE4_HW_SM_QUERY_INST_ISSUED2, + NVE4_HW_SM_QUERY_L1_GLD_HIT, + NVE4_HW_SM_QUERY_L1_GLD_MISS, + NVE4_HW_SM_QUERY_L1_LOCAL_LD_HIT, + NVE4_HW_SM_QUERY_L1_LOCAL_LD_MISS, + NVE4_HW_SM_QUERY_L1_LOCAL_ST_HIT, + NVE4_HW_SM_QUERY_L1_LOCAL_ST_MISS, + NVE4_HW_SM_QUERY_L1_SHARED_LD_TRANSACTIONS, + NVE4_HW_SM_QUERY_L1_SHARED_ST_TRANSACTIONS, + NVE4_HW_SM_QUERY_LOCAL_LD, + NVE4_HW_SM_QUERY_LOCAL_LD_TRANSACTIONS, + NVE4_HW_SM_QUERY_LOCAL_ST, + NVE4_HW_SM_QUERY_LOCAL_ST_TRANSACTIONS, + NVE4_HW_SM_QUERY_PROF_TRIGGER_0, + NVE4_HW_SM_QUERY_PROF_TRIGGER_1, + NVE4_HW_SM_QUERY_PROF_TRIGGER_2, + NVE4_HW_SM_QUERY_PROF_TRIGGER_3, + NVE4_HW_SM_QUERY_PROF_TRIGGER_4, + NVE4_HW_SM_QUERY_PROF_TRIGGER_5, + NVE4_HW_SM_QUERY_PROF_TRIGGER_6, + NVE4_HW_SM_QUERY_PROF_TRIGGER_7, + NVE4_HW_SM_QUERY_SHARED_LD, + NVE4_HW_SM_QUERY_SHARED_LD_REPLAY, + NVE4_HW_SM_QUERY_SHARED_ST, + NVE4_HW_SM_QUERY_SHARED_ST_REPLAY, + NVE4_HW_SM_QUERY_SM_CTA_LAUNCHED, + NVE4_HW_SM_QUERY_THREADS_LAUNCHED, + NVE4_HW_SM_QUERY_UNCACHED_GLD_TRANSACTIONS, + NVE4_HW_SM_QUERY_WARPS_LAUNCHED, + NVE4_HW_SM_QUERY_METRIC_IPC, + NVE4_HW_SM_QUERY_METRIC_IPAC, + NVE4_HW_SM_QUERY_METRIC_IPEC, + NVE4_HW_SM_QUERY_METRIC_MP_OCCUPANCY, + NVE4_HW_SM_QUERY_METRIC_MP_EFFICIENCY, + NVE4_HW_SM_QUERY_METRIC_INST_REPLAY_OHEAD, + NVE4_HW_SM_QUERY_COUNT +}; + +#define NVC0_HW_SM_QUERY(i) (PIPE_QUERY_DRIVER_SPECIFIC + 2048 + (i)) +#define NVC0_HW_SM_QUERY_LAST NVC0_HW_SM_QUERY(NVC0_HW_SM_QUERY_COUNT - 1) +enum nvc0_pm_queries +{ + NVC0_HW_SM_QUERY_ACTIVE_CYCLES = 0, + NVC0_HW_SM_QUERY_ACTIVE_WARPS, + NVC0_HW_SM_QUERY_ATOM_COUNT, + NVC0_HW_SM_QUERY_BRANCH, + NVC0_HW_SM_QUERY_DIVERGENT_BRANCH, + NVC0_HW_SM_QUERY_GLD_REQUEST, + NVC0_HW_SM_QUERY_GRED_COUNT, + NVC0_HW_SM_QUERY_GST_REQUEST, + NVC0_HW_SM_QUERY_INST_EXECUTED, + NVC0_HW_SM_QUERY_INST_ISSUED1_0, + NVC0_HW_SM_QUERY_INST_ISSUED1_1, + NVC0_HW_SM_QUERY_INST_ISSUED2_0, + NVC0_HW_SM_QUERY_INST_ISSUED2_1, + NVC0_HW_SM_QUERY_LOCAL_LD, + NVC0_HW_SM_QUERY_LOCAL_ST, + NVC0_HW_SM_QUERY_PROF_TRIGGER_0, + NVC0_HW_SM_QUERY_PROF_TRIGGER_1, + NVC0_HW_SM_QUERY_PROF_TRIGGER_2, + NVC0_HW_SM_QUERY_PROF_TRIGGER_3, + NVC0_HW_SM_QUERY_PROF_TRIGGER_4, + NVC0_HW_SM_QUERY_PROF_TRIGGER_5, + NVC0_HW_SM_QUERY_PROF_TRIGGER_6, + NVC0_HW_SM_QUERY_PROF_TRIGGER_7, + NVC0_HW_SM_QUERY_SHARED_LD, + NVC0_HW_SM_QUERY_SHARED_ST, + NVC0_HW_SM_QUERY_THREADS_LAUNCHED, + NVC0_HW_SM_QUERY_TH_INST_EXECUTED_0, + NVC0_HW_SM_QUERY_TH_INST_EXECUTED_1, + NVC0_HW_SM_QUERY_TH_INST_EXECUTED_2, + NVC0_HW_SM_QUERY_TH_INST_EXECUTED_3, + NVC0_HW_SM_QUERY_WARPS_LAUNCHED, + NVC0_HW_SM_QUERY_COUNT +}; + +struct nvc0_query * +nvc0_hw_create_query(struct nvc0_context *, unsigned, unsigned); +void +nvc0_hw_query_pushbuf_submit(struct nouveau_pushbuf *, struct nvc0_query *, + unsigned); +void +nvc0_hw_query_fifo_wait(struct nouveau_pushbuf *, struct nvc0_query *); + +#endif diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_shader_state.c b/src/gallium/drivers/nouveau/nvc0/nvc0_shader_state.c index 2fade982b83..af837fc4a33 100644 --- a/src/gallium/drivers/nouveau/nvc0/nvc0_shader_state.c +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_shader_state.c @@ -26,6 +26,7 @@ #include "util/u_inlines.h" #include "nvc0/nvc0_context.h" +#include "nvc0/nvc0_query_hw.h" static inline void nvc0_program_update_context_state(struct nvc0_context *nvc0, @@ -272,14 +273,14 @@ nvc0_tfb_validate(struct nvc0_context *nvc0) continue; if (!targ->clean) - nvc0_query_fifo_wait(push, nvc0_query(targ->pq)); + nvc0_hw_query_fifo_wait(push, nvc0_query(targ->pq)); BEGIN_NVC0(push, NVC0_3D(TFB_BUFFER_ENABLE(b)), 5); PUSH_DATA (push, 1); PUSH_DATAh(push, buf->address + targ->pipe.buffer_offset); PUSH_DATA (push, buf->address + targ->pipe.buffer_offset); PUSH_DATA (push, targ->pipe.buffer_size); if (!targ->clean) { - nvc0_query_pushbuf_submit(push, nvc0_query(targ->pq), 0x4); + nvc0_hw_query_pushbuf_submit(push, nvc0_query(targ->pq), 0x4); } else { PUSH_DATA(push, 0); /* TFB_BUFFER_OFFSET */ targ->clean = false; diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_state.c b/src/gallium/drivers/nouveau/nvc0/nvc0_state.c index 269c75b03a6..742bef39247 100644 --- a/src/gallium/drivers/nouveau/nvc0/nvc0_state.c +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_state.c @@ -29,6 +29,7 @@ #include "nvc0/nvc0_stateobj.h" #include "nvc0/nvc0_context.h" +#include "nvc0/nvc0_query_hw.h" #include "nvc0/nvc0_3d.xml.h" #include "nv50/nv50_texture.xml.h" @@ -1070,7 +1071,7 @@ nvc0_so_target_create(struct pipe_context *pipe, if (!targ) return NULL; - targ->pq = pipe->create_query(pipe, NVC0_QUERY_TFB_BUFFER_OFFSET, 0); + targ->pq = pipe->create_query(pipe, NVC0_HW_QUERY_TFB_BUFFER_OFFSET, 0); if (!targ->pq) { FREE(targ); return NULL; diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_vbo.c b/src/gallium/drivers/nouveau/nvc0/nvc0_vbo.c index 9be25cfe66e..c464904d6d4 100644 --- a/src/gallium/drivers/nouveau/nvc0/nvc0_vbo.c +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_vbo.c @@ -29,6 +29,7 @@ #include "translate/translate.h" #include "nvc0/nvc0_context.h" +#include "nvc0/nvc0_query_hw.h" #include "nvc0/nvc0_resource.h" #include "nvc0/nvc0_3d.xml.h" @@ -775,7 +776,7 @@ nvc0_draw_stream_output(struct nvc0_context *nvc0, res->status &= ~NOUVEAU_BUFFER_STATUS_GPU_WRITING; PUSH_SPACE(push, 2); IMMED_NVC0(push, NVC0_3D(SERIALIZE), 0); - nvc0_query_fifo_wait(push, nvc0_query(so->pq)); + nvc0_hw_query_fifo_wait(push, nvc0_query(so->pq)); if (nvc0->screen->eng3d->oclass < GM107_3D_CLASS) IMMED_NVC0(push, NVC0_3D(VERTEX_ARRAY_FLUSH), 0); @@ -791,7 +792,7 @@ nvc0_draw_stream_output(struct nvc0_context *nvc0, BEGIN_NVC0(push, NVC0_3D(DRAW_TFB_STRIDE), 1); PUSH_DATA (push, so->stride); BEGIN_NVC0(push, NVC0_3D(DRAW_TFB_BYTES), 1); - nvc0_query_pushbuf_submit(push, nvc0_query(so->pq), 0x4); + nvc0_hw_query_pushbuf_submit(push, nvc0_query(so->pq), 0x4); IMMED_NVC0(push, NVC0_3D(VERTEX_END_GL), 0); mode |= NVC0_3D_VERTEX_BEGIN_GL_INSTANCE_NEXT;