[Mesa-dev] [PATCH 2/2] r600g, radeonsi: add a driver query returning GPU load

Marek Olšák maraeo at gmail.com
Mon Apr 27 07:46:28 PDT 2015


From: Marek Olšák <marek.olsak at amd.com>

---
 src/gallium/drivers/radeon/Makefile.am            |   4 +-
 src/gallium/drivers/radeon/Makefile.sources       |   1 +
 src/gallium/drivers/radeon/r600_gpu_load.c        | 141 ++++++++++++++++++++++
 src/gallium/drivers/radeon/r600_pipe_common.c     |   7 +-
 src/gallium/drivers/radeon/r600_pipe_common.h     |  13 ++
 src/gallium/drivers/radeon/r600_query.c           |  11 ++
 src/gallium/winsys/radeon/drm/radeon_drm_winsys.c |  16 +++
 src/gallium/winsys/radeon/drm/radeon_winsys.h     |   3 +
 8 files changed, 194 insertions(+), 2 deletions(-)
 create mode 100644 src/gallium/drivers/radeon/r600_gpu_load.c

diff --git a/src/gallium/drivers/radeon/Makefile.am b/src/gallium/drivers/radeon/Makefile.am
index ef6b999..13d8976 100644
--- a/src/gallium/drivers/radeon/Makefile.am
+++ b/src/gallium/drivers/radeon/Makefile.am
@@ -4,7 +4,9 @@ include $(top_srcdir)/src/gallium/Automake.inc
 
 AM_CFLAGS = \
 	$(GALLIUM_DRIVER_CFLAGS) \
-	$(RADEON_CFLAGS)
+	$(RADEON_CFLAGS) \
+	-Wstrict-overflow=0
+# ^^ disable warnings about overflows (os_time_timeout)
 
 noinst_LTLIBRARIES = libradeon.la
 
diff --git a/src/gallium/drivers/radeon/Makefile.sources b/src/gallium/drivers/radeon/Makefile.sources
index f2b70a1..469f6d1 100644
--- a/src/gallium/drivers/radeon/Makefile.sources
+++ b/src/gallium/drivers/radeon/Makefile.sources
@@ -3,6 +3,7 @@ C_SOURCES := \
 	r600_buffer_common.c \
 	r600_cs.h \
 	r600d_common.h \
+	r600_gpu_load.c \
 	r600_pipe_common.c \
 	r600_pipe_common.h \
 	r600_query.c \
diff --git a/src/gallium/drivers/radeon/r600_gpu_load.c b/src/gallium/drivers/radeon/r600_gpu_load.c
new file mode 100644
index 0000000..a653834
--- /dev/null
+++ b/src/gallium/drivers/radeon/r600_gpu_load.c
@@ -0,0 +1,141 @@
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * 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 (including the next
+ * paragraph) 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.
+ *
+ * Authors: Marek Olšák <maraeo at gmail.com>
+ *
+ */
+
+/* The GPU load is measured as follows.
+ *
+ * There is a thread which samples the GRBM_STATUS register at a certain
+ * frequency and the "busy" or "idle" counter is incremented based on
+ * whether the GUI_ACTIVE bit is set or not.
+ *
+ * Then, the user can sample the counters twice and calculate the average
+ * GPU load between the two samples.
+ */
+
+#include "r600_pipe_common.h"
+#include "os/os_time.h"
+
+/* For good accuracy at 1000 fps or lower. This will be inaccurate for higher
+ * fps (there are too few samples per frame). */
+#define SAMPLES_PER_SEC 10000
+
+#define GRBM_STATUS		0x8010
+#define GUI_ACTIVE(x)		(((x) >> 31) & 0x1)
+
+static bool r600_is_gpu_busy(struct r600_common_screen *rscreen)
+{
+	uint32_t value = 0;
+
+	rscreen->ws->read_registers(rscreen->ws, GRBM_STATUS, 1, &value);
+	return GUI_ACTIVE(value);
+}
+
+static PIPE_THREAD_ROUTINE(r600_gpu_load_thread, param)
+{
+	struct r600_common_screen *rscreen = (struct r600_common_screen*)param;
+	const int period_us = 1000000 / SAMPLES_PER_SEC;
+	int sleep_us = period_us;
+	int64_t cur_time, last_time = os_time_get();
+
+	while (!p_atomic_read(&rscreen->gpu_load_stop_thread)) {
+		if (sleep_us)
+			os_time_sleep(sleep_us);
+
+		/* Make sure we sleep the ideal amount of time to match
+		 * the expected frequency. */
+		cur_time = os_time_get();
+
+		if (os_time_timeout(last_time, last_time + period_us,
+				    cur_time))
+			sleep_us = MAX2(sleep_us - 1, 1);
+		else
+			sleep_us += 1;
+
+		/*printf("Hz: %.1f\n", 1000000.0 / (cur_time - last_time));*/
+		last_time = cur_time;
+
+		/* Update the counters. */
+		if (r600_is_gpu_busy(rscreen))
+			p_atomic_inc(&rscreen->gpu_load_counter_busy);
+		else
+			p_atomic_inc(&rscreen->gpu_load_counter_idle);
+	}
+	p_atomic_dec(&rscreen->gpu_load_stop_thread);
+	return 0;
+}
+
+void r600_gpu_load_kill_thread(struct r600_common_screen *rscreen)
+{
+	if (!rscreen->gpu_load_thread)
+		return;
+
+	p_atomic_inc(&rscreen->gpu_load_stop_thread);
+	pipe_thread_wait(rscreen->gpu_load_thread);
+	rscreen->gpu_load_thread = 0;
+}
+
+static uint64_t r600_gpu_load_read_counter(struct r600_common_screen *rscreen)
+{
+	/* Start the thread if needed. */
+	if (!rscreen->gpu_load_thread) {
+		pipe_mutex_lock(rscreen->gpu_load_mutex);
+		/* Check again inside the mutex. */
+		if (!rscreen->gpu_load_thread)
+			rscreen->gpu_load_thread =
+				pipe_thread_create(r600_gpu_load_thread, rscreen);
+		pipe_mutex_unlock(rscreen->gpu_load_mutex);
+	}
+
+	/* The busy counter is in the lower 32 bits.
+	 * The idle counter is in the upper 32 bits. */
+	return p_atomic_read(&rscreen->gpu_load_counter_busy) |
+	       ((uint64_t)p_atomic_read(&rscreen->gpu_load_counter_idle) << 32);
+}
+
+/**
+ * Just return the counters.
+ */
+uint64_t r600_gpu_load_begin(struct r600_common_screen *rscreen)
+{
+	return r600_gpu_load_read_counter(rscreen);
+}
+
+unsigned r600_gpu_load_end(struct r600_common_screen *rscreen, uint64_t begin)
+{
+	uint64_t end = r600_gpu_load_read_counter(rscreen);
+	unsigned busy = (end & 0xffffffff) - (begin & 0xffffffff);
+	unsigned idle = (end >> 32) - (begin >> 32);
+
+	/* Calculate the GPU load.
+	 *
+	 * If no counters have been incremented, return the current load.
+	 * It's for the case when the load is queried faster than
+	 * the counters are updated.
+	 */
+	if (idle || busy)
+		return busy*100 / (busy + idle);
+	else
+		return r600_is_gpu_busy(rscreen) ? 100 : 0;
+}
diff --git a/src/gallium/drivers/radeon/r600_pipe_common.c b/src/gallium/drivers/radeon/r600_pipe_common.c
index 97eed13..eddb37d 100644
--- a/src/gallium/drivers/radeon/r600_pipe_common.c
+++ b/src/gallium/drivers/radeon/r600_pipe_common.c
@@ -667,7 +667,8 @@ static int r600_get_driver_query_info(struct pipe_screen *screen,
 		{"GTT-usage", R600_QUERY_GTT_USAGE, rscreen->info.gart_size, TRUE},
 		{"temperature", R600_QUERY_GPU_TEMPERATURE, 100, FALSE},
 		{"shader-clock", R600_QUERY_CURRENT_GPU_SCLK, 0, FALSE},
-		{"memory-clock", R600_QUERY_CURRENT_GPU_MCLK, 0, FALSE}
+		{"memory-clock", R600_QUERY_CURRENT_GPU_MCLK, 0, FALSE},
+		{"GPU-load", R600_QUERY_GPU_LOAD, 100, FALSE}
 	};
 	unsigned num_queries;
 
@@ -872,6 +873,7 @@ bool r600_common_screen_init(struct r600_common_screen *rscreen,
 	}
 	util_format_s3tc_init();
 	pipe_mutex_init(rscreen->aux_context_lock);
+	pipe_mutex_init(rscreen->gpu_load_mutex);
 
 	if (rscreen->info.drm_minor >= 28 && (rscreen->debug_flags & DBG_TRACE_CS)) {
 		rscreen->trace_bo = (struct r600_resource*)pipe_buffer_create(&rscreen->b,
@@ -915,6 +917,9 @@ bool r600_common_screen_init(struct r600_common_screen *rscreen,
 
 void r600_destroy_common_screen(struct r600_common_screen *rscreen)
 {
+	r600_gpu_load_kill_thread(rscreen);
+
+	pipe_mutex_destroy(rscreen->gpu_load_mutex);
 	pipe_mutex_destroy(rscreen->aux_context_lock);
 	rscreen->aux_context->destroy(rscreen->aux_context);
 
diff --git a/src/gallium/drivers/radeon/r600_pipe_common.h b/src/gallium/drivers/radeon/r600_pipe_common.h
index c23072c..faa6e0d 100644
--- a/src/gallium/drivers/radeon/r600_pipe_common.h
+++ b/src/gallium/drivers/radeon/r600_pipe_common.h
@@ -58,6 +58,7 @@
 #define R600_QUERY_GPU_TEMPERATURE	(PIPE_QUERY_DRIVER_SPECIFIC + 8)
 #define R600_QUERY_CURRENT_GPU_SCLK	(PIPE_QUERY_DRIVER_SPECIFIC + 9)
 #define R600_QUERY_CURRENT_GPU_MCLK	(PIPE_QUERY_DRIVER_SPECIFIC + 10)
+#define R600_QUERY_GPU_LOAD		(PIPE_QUERY_DRIVER_SPECIFIC + 11)
 
 #define R600_CONTEXT_STREAMOUT_FLUSH		(1u << 0)
 #define R600_CONTEXT_PRIVATE_FLAG		(1u << 1)
@@ -283,6 +284,13 @@ struct r600_common_screen {
 	struct r600_resource		*trace_bo;
 	uint32_t			*trace_ptr;
 	unsigned			cs_count;
+
+	/* GPU load thread. */
+	pipe_mutex			gpu_load_mutex;
+	pipe_thread			gpu_load_thread;
+	unsigned			gpu_load_counter_busy;
+	unsigned			gpu_load_counter_idle;
+	unsigned			gpu_load_stop_thread; /* bool */
 };
 
 /* This encapsulates a state or an operation which can emitted into the GPU
@@ -478,6 +486,11 @@ struct pipe_resource *r600_resource_create_common(struct pipe_screen *screen,
 const char *r600_get_llvm_processor_name(enum radeon_family family);
 void r600_need_dma_space(struct r600_common_context *ctx, unsigned num_dw);
 
+/* r600_gpu_load.c */
+void r600_gpu_load_kill_thread(struct r600_common_screen *rscreen);
+uint64_t r600_gpu_load_begin(struct r600_common_screen *rscreen);
+unsigned r600_gpu_load_end(struct r600_common_screen *rscreen, uint64_t begin);
+
 /* r600_query.c */
 void r600_query_init(struct r600_common_context *rctx);
 void r600_suspend_nontimer_queries(struct r600_common_context *ctx);
diff --git a/src/gallium/drivers/radeon/r600_query.c b/src/gallium/drivers/radeon/r600_query.c
index 758064a..1335087 100644
--- a/src/gallium/drivers/radeon/r600_query.c
+++ b/src/gallium/drivers/radeon/r600_query.c
@@ -89,6 +89,7 @@ static struct r600_resource *r600_new_query_buffer(struct r600_common_context *c
 	case R600_QUERY_GPU_TEMPERATURE:
 	case R600_QUERY_CURRENT_GPU_SCLK:
 	case R600_QUERY_CURRENT_GPU_MCLK:
+	case R600_QUERY_GPU_LOAD:
 		return NULL;
 	}
 
@@ -388,6 +389,7 @@ static struct pipe_query *r600_create_query(struct pipe_context *ctx, unsigned q
 	case R600_QUERY_GPU_TEMPERATURE:
 	case R600_QUERY_CURRENT_GPU_SCLK:
 	case R600_QUERY_CURRENT_GPU_MCLK:
+	case R600_QUERY_GPU_LOAD:
 		skip_allocation = true;
 		break;
 	default:
@@ -459,6 +461,9 @@ static void r600_begin_query(struct pipe_context *ctx, struct pipe_query *query)
 	case R600_QUERY_NUM_BYTES_MOVED:
 		rquery->begin_result = rctx->ws->query_value(rctx->ws, RADEON_NUM_BYTES_MOVED);
 		return;
+	case R600_QUERY_GPU_LOAD:
+		rquery->begin_result = r600_gpu_load_begin(rctx->screen);
+		return;
 	}
 
 	/* Discard the old query buffers. */
@@ -531,6 +536,9 @@ static void r600_end_query(struct pipe_context *ctx, struct pipe_query *query)
 	case R600_QUERY_CURRENT_GPU_MCLK:
 		rquery->end_result = rctx->ws->query_value(rctx->ws, RADEON_CURRENT_MCLK) * 1000000;
 		return;
+	case R600_QUERY_GPU_LOAD:
+		rquery->end_result = r600_gpu_load_end(rctx->screen, rquery->begin_result);
+		return;
 	}
 
 	r600_emit_query_end(rctx, rquery);
@@ -593,6 +601,9 @@ static boolean r600_get_query_buffer_result(struct r600_common_context *ctx,
 	case R600_QUERY_CURRENT_GPU_MCLK:
 		result->u64 = query->end_result - query->begin_result;
 		return TRUE;
+	case R600_QUERY_GPU_LOAD:
+		result->u64 = query->end_result;
+		return TRUE;
 	}
 
 	map = r600_buffer_map_sync_with_rings(ctx, qbuf->buf,
diff --git a/src/gallium/winsys/radeon/drm/radeon_drm_winsys.c b/src/gallium/winsys/radeon/drm/radeon_drm_winsys.c
index b3e3cb2..a6f847f 100644
--- a/src/gallium/winsys/radeon/drm/radeon_drm_winsys.c
+++ b/src/gallium/winsys/radeon/drm/radeon_drm_winsys.c
@@ -582,6 +582,21 @@ static uint64_t radeon_query_value(struct radeon_winsys *rws,
     return 0;
 }
 
+static void radeon_read_registers(struct radeon_winsys *rws,
+                                  unsigned reg_offset,
+                                  unsigned num_registers, uint32_t *out)
+{
+    struct radeon_drm_winsys *ws = (struct radeon_drm_winsys*)rws;
+    unsigned i;
+
+    for (i = 0; i < num_registers; i++) {
+        uint32_t reg = reg_offset + i*4;
+
+        radeon_get_drm_value(ws->fd, RADEON_INFO_READ_REG, "read-reg", &reg);
+        out[i] = reg;
+    }
+}
+
 static unsigned hash_fd(void *key)
 {
     int fd = pointer_to_intptr(key);
@@ -728,6 +743,7 @@ radeon_drm_winsys_create(int fd, radeon_screen_create_t screen_create)
     ws->base.surface_init = radeon_drm_winsys_surface_init;
     ws->base.surface_best = radeon_drm_winsys_surface_best;
     ws->base.query_value = radeon_query_value;
+    ws->base.read_registers = radeon_read_registers;
 
     radeon_bomgr_init_functions(ws);
     radeon_drm_cs_init_functions(ws);
diff --git a/src/gallium/winsys/radeon/drm/radeon_winsys.h b/src/gallium/winsys/radeon/drm/radeon_winsys.h
index 996a201..ee0a904 100644
--- a/src/gallium/winsys/radeon/drm/radeon_winsys.h
+++ b/src/gallium/winsys/radeon/drm/radeon_winsys.h
@@ -589,6 +589,9 @@ struct radeon_winsys {
 
     uint64_t (*query_value)(struct radeon_winsys *ws,
                             enum radeon_value_id value);
+
+    void (*read_registers)(struct radeon_winsys *ws, unsigned reg_offset,
+                           unsigned num_registers, uint32_t *out);
 };
 
 
-- 
2.1.0



More information about the mesa-dev mailing list