i915/gem_exec_schedule: Fix measurement of ring size to use right engine

gem_ring_measure_inflight() uses the legacy ring idx, whereas
gem_exec_schedule is using the new dynamic ctx->engine[] indices.
Mismatch and hanging tests ensue.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>

6 files changed, 123 insertions, 7 deletions

diff --git a/lib/i915/gem_submission.c b/lib/i915/gem_submission.c
index 6bea6aa3d..e5deb6cea 100644
--- a/lib/i915/gem_submission.c
+++ b/lib/i915/gem_submission.c
@@ -23,8 +23,10 @@
 
 #include <errno.h>
 #include <fcntl.h>
+#include <signal.h>
 #include <stdbool.h>
 #include <sys/ioctl.h>
+#include <sys/time.h>
 
 #include <i915_drm.h>
 
@@ -278,3 +280,116 @@ bool gem_engine_has_mutable_submission(int i915, unsigned int engine)
 	return gem_class_has_mutable_submission(i915,
 						gem_execbuf_flags_to_engine_class(engine));
 }
+
+static int __execbuf(int i915, struct drm_i915_gem_execbuffer2 *execbuf)
+{
+	int err;
+
+	err = 0;
+	if (ioctl(i915, DRM_IOCTL_I915_GEM_EXECBUFFER2, execbuf)) {
+		err = -errno;
+		igt_assume(err);
+	}
+
+	errno = 0;
+	return err;
+}
+
+static void alarm_handler(int sig)
+{
+}
+
+static unsigned int
+__measure_ringsize(int i915, unsigned int engine)
+{
+	struct sigaction old_sa, sa = { .sa_handler = alarm_handler };
+	struct drm_i915_gem_exec_object2 obj[2];
+	struct drm_i915_gem_execbuffer2 execbuf;
+	const uint32_t bbe = MI_BATCH_BUFFER_END;
+	unsigned int last[2]= { -1, -1 }, count;
+	struct itimerval itv;
+	IGT_CORK_HANDLE(cork);
+
+	memset(obj, 0, sizeof(obj));
+	obj[1].handle = gem_create(i915, 4096);
+	gem_write(i915, obj[1].handle, 0, &bbe, sizeof(bbe));
+
+	memset(&execbuf, 0, sizeof(execbuf));
+	execbuf.buffers_ptr = to_user_pointer(&obj[1]);
+	execbuf.buffer_count = 1;
+	execbuf.flags = engine;
+	gem_execbuf(i915, &execbuf);
+
+	obj[0].handle = igt_cork_plug(&cork, i915);
+
+	execbuf.buffers_ptr = to_user_pointer(obj);
+	execbuf.buffer_count = 2;
+
+	sigaction(SIGALRM, &sa, &old_sa);
+	itv.it_interval.tv_sec = 0;
+	itv.it_interval.tv_usec = 1000;
+	itv.it_value.tv_sec = 0;
+	itv.it_value.tv_usec = 10000;
+	setitimer(ITIMER_REAL, &itv, NULL);
+
+	count = 0;
+	do {
+		int err = __execbuf(i915, &execbuf);
+
+		if (err == 0) {
+			count++;
+			continue;
+		}
+
+		if (err == -EWOULDBLOCK)
+			break;
+
+		if (last[1] == count)
+			break;
+
+		/* sleep until the next timer interrupt (woken on signal) */
+		pause();
+		last[1] = last[0];
+		last[0] = count;
+	} while (1);
+	igt_assert(count > 2);
+
+	memset(&itv, 0, sizeof(itv));
+	setitimer(ITIMER_REAL, &itv, NULL);
+	sigaction(SIGALRM, &old_sa, NULL);
+
+	igt_cork_unplug(&cork);
+	gem_close(i915, obj[0].handle);
+	gem_close(i915, obj[1].handle);
+
+	/* Be conservative, expect relocations, in case we must wrap later */
+	return count / 2 - 1;
+}
+
+unsigned int gem_submission_measure(int i915, unsigned int engine)
+{
+	unsigned int size;
+	bool nonblock;
+
+	nonblock = fcntl(i915, F_GETFD) & O_NONBLOCK;
+	if (!nonblock)
+		fcntl(i915, F_SETFD, fcntl(i915, F_GETFD) | O_NONBLOCK);
+
+	if (engine == ALL_ENGINES) {
+		struct intel_execution_engine2 *e;
+
+		size = -1;
+		__for_each_physical_engine(i915, e) {
+			unsigned int this =  __measure_ringsize(i915, e->flags);
+			if (this < size)
+				size = this;
+		}
+	} else {
+		size =  __measure_ringsize(i915, engine);
+	}
+
+	if (!nonblock)
+		fcntl(i915, F_SETFD, fcntl(i915, F_GETFD) & ~O_NONBLOCK);
+
+	return size;
+}
diff --git a/lib/i915/gem_submission.h b/lib/i915/gem_submission.h
index b030ed9fc..773e7b516 100644
--- a/lib/i915/gem_submission.h
+++ b/lib/i915/gem_submission.h
@@ -46,6 +46,8 @@ static inline bool gem_has_cmdparser(int i915, uint32_t engine)
 bool gem_has_blitter(int i915);
 void gem_require_blitter(int i915);
 
+unsigned int gem_submission_measure(int i915, unsigned int engine);
+
 void gem_test_engine(int fd, unsigned int engine);
 
 #endif /* GEM_SUBMISSION_H */
diff --git a/tests/i915/gem_busy.c b/tests/i915/gem_busy.c
index a3cff23fc..ee6a5ca0e 100644
--- a/tests/i915/gem_busy.c
+++ b/tests/i915/gem_busy.c
@@ -232,7 +232,7 @@ static void xchg_u32(void *array, unsigned i, unsigned j)
 static void close_race(int fd)
 {
 	const unsigned int ncpus = sysconf(_SC_NPROCESSORS_ONLN);
-	const unsigned int nhandles = gem_measure_ring_inflight(fd, ALL_ENGINES, 0) / 2;
+	const unsigned int nhandles = gem_submission_measure(fd, ALL_ENGINES);
 	const struct intel_execution_engine2 *e;
 	unsigned int engines[16], nengine;
 	unsigned long *control;
diff --git a/tests/i915/gem_exec_fence.c b/tests/i915/gem_exec_fence.c
index 4140bff24..ad22f9f57 100644
--- a/tests/i915/gem_exec_fence.c
+++ b/tests/i915/gem_exec_fence.c
@@ -1756,7 +1756,7 @@ igt_main
 		long ring_size = 0;
 
 		igt_fixture {
-			ring_size = gem_measure_ring_inflight(i915, ALL_ENGINES, 0) - 1;
+			ring_size = gem_submission_measure(i915, ALL_ENGINES);
 			igt_info("Ring size: %ld batches\n", ring_size);
 			igt_require(ring_size);
 
diff --git a/tests/i915/gem_exec_schedule.c b/tests/i915/gem_exec_schedule.c
index c551e23ea..c5390493b 100644
--- a/tests/i915/gem_exec_schedule.c
+++ b/tests/i915/gem_exec_schedule.c
@@ -1651,7 +1651,7 @@ static void deep(int fd, unsigned ring)
 		ctx[n] = gem_context_clone_with_engines(fd, 0);
 	}
 
-	nreq = gem_measure_ring_inflight(fd, ring, 0) / (4 * XS) * MAX_CONTEXTS;
+	nreq = gem_submission_measure(fd, ring) / (4 * XS) * MAX_CONTEXTS;
 	if (nreq > max_req)
 		nreq = max_req;
 	igt_info("Using %d requests (prio range %d)\n", nreq, max_req);
@@ -1796,9 +1796,8 @@ static int __execbuf(int fd, struct drm_i915_gem_execbuffer2 *execbuf)
 
 static void wide(int fd, unsigned ring)
 {
+	const unsigned int ring_size = gem_submission_measure(fd, ring);
 	struct timespec tv = {};
-	unsigned int ring_size = gem_measure_ring_inflight(fd, ring, MEASURE_RING_NEW_CTX);
-
 	IGT_CORK_FENCE(cork);
 	uint32_t result;
 	uint32_t result_read[MAX_CONTEXTS];
@@ -1842,12 +1841,12 @@ static void wide(int fd, unsigned ring)
 
 static void reorder_wide(int fd, unsigned ring)
 {
+	const unsigned int ring_size = gem_submission_measure(fd, ring);
 	const int gen = intel_gen(intel_get_drm_devid(fd));
 	struct drm_i915_gem_relocation_entry reloc;
 	struct drm_i915_gem_exec_object2 obj[2];
 	struct drm_i915_gem_execbuffer2 execbuf;
 	struct timespec tv = {};
-	unsigned int ring_size = gem_measure_ring_inflight(fd, ring, MEASURE_RING_NEW_CTX);
 	IGT_CORK_FENCE(cork);
 	uint32_t result, target;
 	uint32_t result_read[1024];
diff --git a/tests/perf_pmu.c b/tests/perf_pmu.c
index e80f730cf..f4db3c10a 100644
--- a/tests/perf_pmu.c
+++ b/tests/perf_pmu.c
@@ -1287,7 +1287,7 @@ static void cpu_hotplug(int gem_fd)
 
 static int target_num_interrupts(int i915)
 {
-	return min(gem_measure_ring_inflight(i915, I915_EXEC_DEFAULT, 0), 30);
+	return min(gem_submission_measure(i915, I915_EXEC_DEFAULT), 30);
 }
 
 static void