X-Git-Url: https://git.kernel.dk/?a=blobdiff_plain;f=t%2Fio_uring.c;h=538cc7d4e44c99b0dfbde4502780334be7d89b6a;hb=543196617ce45b6a04fb039a3b9c6d06c9b58309;hp=363cba3eba86c9c2c80098edaa9c9791a93145a9;hpb=d9c50de7c1d95e5c173ac6d7f5b3f0d63131f8b4;p=fio.git

diff --git a/t/io_uring.c b/t/io_uring.c
index 363cba3e..538cc7d4 100644
--- a/t/io_uring.c
+++ b/t/io_uring.c
@@ -46,7 +46,6 @@ struct io_cq_ring {
 #define DEPTH			128
 #define BATCH_SUBMIT		32
 #define BATCH_COMPLETE		32
-
 #define BS			4096
 
 #define MAX_FDS			16
@@ -63,7 +62,7 @@ struct file {
 struct submitter {
 	pthread_t thread;
 	int ring_fd;
-	struct drand48_data rand;
+	int index;
 	struct io_sq_ring sq_ring;
 	struct io_uring_sqe *sqes;
 	struct io_cq_ring cq_ring;
@@ -71,7 +70,6 @@ struct submitter {
 	unsigned long reaps;
 	unsigned long done;
 	unsigned long calls;
-	unsigned long cachehit, cachemiss;
 	volatile int finish;
 
 	__s32 *fds;
@@ -88,6 +86,7 @@ static volatile int finish;
 static int depth = DEPTH;
 static int batch_submit = BATCH_SUBMIT;
 static int batch_complete = BATCH_COMPLETE;
+static int bs = BS;
 static int polled = 1;		/* use IO polling */
 static int fixedbufs = 1;	/* use fixed user buffers */
 static int register_files = 1;	/* use fixed files */
@@ -95,13 +94,16 @@ static int buffered = 0;	/* use buffered IO, not O_DIRECT */
 static int sq_thread_poll = 0;	/* use kernel submission/poller thread */
 static int sq_thread_cpu = -1;	/* pin above thread to this CPU */
 static int do_nop = 0;		/* no-op SQ ring commands */
+static int nthreads = 1;
+
+static int vectored = 1;
 
 static int io_uring_register_buffers(struct submitter *s)
 {
 	if (do_nop)
 		return 0;
 
-	return syscall(__NR_sys_io_uring_register, s->ring_fd,
+	return syscall(__NR_io_uring_register, s->ring_fd,
 			IORING_REGISTER_BUFFERS, s->iovecs, depth);
 }
 
@@ -118,26 +120,51 @@ static int io_uring_register_files(struct submitter *s)
 		s->files[i].fixed_fd = i;
 	}
 
-	return syscall(__NR_sys_io_uring_register, s->ring_fd,
+	return syscall(__NR_io_uring_register, s->ring_fd,
 			IORING_REGISTER_FILES, s->fds, s->nr_files);
 }
 
 static int io_uring_setup(unsigned entries, struct io_uring_params *p)
 {
-	return syscall(__NR_sys_io_uring_setup, entries, p);
+	return syscall(__NR_io_uring_setup, entries, p);
+}
+
+static void io_uring_probe(int fd)
+{
+	struct io_uring_probe *p;
+	int ret;
+
+	p = malloc(sizeof(*p) + 256 * sizeof(struct io_uring_probe_op));
+	if (!p)
+		return;
+
+	memset(p, 0, sizeof(*p) + 256 * sizeof(struct io_uring_probe_op));
+	ret = syscall(__NR_io_uring_register, fd, IORING_REGISTER_PROBE, p, 256);
+	if (ret < 0)
+		goto out;
+
+	if (IORING_OP_READ > p->ops_len)
+		goto out;
+
+	if ((p->ops[IORING_OP_READ].flags & IO_URING_OP_SUPPORTED))
+		vectored = 0;
+out:
+	free(p);
 }
 
 static int io_uring_enter(struct submitter *s, unsigned int to_submit,
 			  unsigned int min_complete, unsigned int flags)
 {
-	return syscall(__NR_sys_io_uring_enter, s->ring_fd, to_submit,
-			min_complete, flags, NULL, 0);
+	return syscall(__NR_io_uring_enter, s->ring_fd, to_submit, min_complete,
+			flags, NULL, 0);
 }
 
+#ifndef CONFIG_HAVE_GETTID
 static int gettid(void)
 {
 	return syscall(__NR_gettid);
 }
+#endif
 
 static unsigned file_depth(struct submitter *s)
 {
@@ -169,8 +196,8 @@ static void init_io(struct submitter *s, unsigned index)
 	}
 	f->pending_ios++;
 
-	lrand48_r(&s->rand, &r);
-	offset = (r % (f->max_blocks - 1)) * BS;
+	r = lrand48();
+	offset = (r % (f->max_blocks - 1)) * bs;
 
 	if (register_files) {
 		sqe->flags = IOSQE_FIXED_FILE;
@@ -182,8 +209,13 @@ static void init_io(struct submitter *s, unsigned index)
 	if (fixedbufs) {
 		sqe->opcode = IORING_OP_READ_FIXED;
 		sqe->addr = (unsigned long) s->iovecs[index].iov_base;
-		sqe->len = BS;
+		sqe->len = bs;
 		sqe->buf_index = index;
+	} else if (!vectored) {
+		sqe->opcode = IORING_OP_READ;
+		sqe->addr = (unsigned long) s->iovecs[index].iov_base;
+		sqe->len = bs;
+		sqe->buf_index = 0;
 	} else {
 		sqe->opcode = IORING_OP_READV;
 		sqe->addr = (unsigned long) &s->iovecs[index];
@@ -203,8 +235,7 @@ static int prep_more_ios(struct submitter *s, int max_ios)
 	next_tail = tail = *ring->tail;
 	do {
 		next_tail++;
-		read_barrier();
-		if (next_tail == *ring->head)
+		if (next_tail == atomic_load_acquire(ring->head))
 			break;
 
 		index = tail & sq_ring_mask;
@@ -214,12 +245,8 @@ static int prep_more_ios(struct submitter *s, int max_ios)
 		tail = next_tail;
 	} while (prepped < max_ios);
 
-	if (*ring->tail != tail) {
-		/* order tail store with writes to sqes above */
-		write_barrier();
-		*ring->tail = tail;
-		write_barrier();
-	}
+	if (prepped)
+		atomic_store_release(ring->tail, tail);
 	return prepped;
 }
 
@@ -235,10 +262,10 @@ static int get_file_size(struct file *f)
 		if (ioctl(f->real_fd, BLKGETSIZE64, &bytes) != 0)
 			return -1;
 
-		f->max_blocks = bytes / BS;
+		f->max_blocks = bytes / bs;
 		return 0;
 	} else if (S_ISREG(st.st_mode)) {
-		f->max_blocks = st.st_size / BS;
+		f->max_blocks = st.st_size / bs;
 		return 0;
 	}
 
@@ -256,30 +283,27 @@ static int reap_events(struct submitter *s)
 		struct file *f;
 
 		read_barrier();
-		if (head == *ring->tail)
+		if (head == atomic_load_acquire(ring->tail))
 			break;
 		cqe = &ring->cqes[head & cq_ring_mask];
 		if (!do_nop) {
 			f = (struct file *) (uintptr_t) cqe->user_data;
 			f->pending_ios--;
-			if (cqe->res != BS) {
+			if (cqe->res != bs) {
 				printf("io: unexpected ret=%d\n", cqe->res);
 				if (polled && cqe->res == -EOPNOTSUPP)
-					printf("Your filesystem doesn't support poll\n");
+					printf("Your filesystem/driver/kernel doesn't support polled IO\n");
 				return -1;
 			}
 		}
-		if (cqe->flags & IOCQE_FLAG_CACHEHIT)
-			s->cachehit++;
-		else
-			s->cachemiss++;
 		reaped++;
 		head++;
 	} while (1);
 
-	s->inflight -= reaped;
-	*ring->head = head;
-	write_barrier();
+	if (reaped) {
+		s->inflight -= reaped;
+		atomic_store_release(ring->head, head);
+	}
 	return reaped;
 }
 
@@ -291,11 +315,12 @@ static void *submitter_fn(void *data)
 
 	printf("submitter=%d\n", gettid());
 
-	srand48_r(pthread_self(), &s->rand);
+	srand48(pthread_self());
 
 	prepped = 0;
 	do {
 		int to_wait, to_submit, this_reap, to_prep;
+		unsigned ring_flags = 0;
 
 		if (!prepped && s->inflight < depth) {
 			to_prep = min(depth - s->inflight, batch_submit);
@@ -314,15 +339,20 @@ submit:
 		 * Only need to call io_uring_enter if we're not using SQ thread
 		 * poll, or if IORING_SQ_NEED_WAKEUP is set.
 		 */
-		if (!sq_thread_poll || (*ring->flags & IORING_SQ_NEED_WAKEUP)) {
+		if (sq_thread_poll)
+			ring_flags = atomic_load_acquire(ring->flags);
+		if (!sq_thread_poll || ring_flags & IORING_SQ_NEED_WAKEUP) {
 			unsigned flags = 0;
 
 			if (to_wait)
 				flags = IORING_ENTER_GETEVENTS;
-			if ((*ring->flags & IORING_SQ_NEED_WAKEUP))
+			if (ring_flags & IORING_SQ_NEED_WAKEUP)
 				flags |= IORING_ENTER_SQ_WAKEUP;
 			ret = io_uring_enter(s, to_submit, to_wait, flags);
 			s->calls++;
+		} else {
+			/* for SQPOLL, we submitted it all effectively */
+			ret = to_submit;
 		}
 
 		/*
@@ -376,10 +406,25 @@ submit:
 	return NULL;
 }
 
+static struct submitter *get_submitter(int offset)
+{
+	void *ret;
+
+	ret = submitter;
+	if (offset)
+		ret += offset * (sizeof(*submitter) + depth * sizeof(struct iovec));
+	return ret;
+}
+
 static void sig_int(int sig)
 {
+	int j;
+
 	printf("Exiting on signal %d\n", sig);
-	submitter->finish = 1;
+	for (j = 0; j < nthreads; j++) {
+		struct submitter *s = get_submitter(j);
+		s->finish = 1;
+	}
 	finish = 1;
 }
 
@@ -420,6 +465,8 @@ static int setup_ring(struct submitter *s)
 	}
 	s->ring_fd = fd;
 
+	io_uring_probe(fd);
+
 	if (fixedbufs) {
 		ret = io_uring_register_buffers(s);
 		if (ret < 0) {
@@ -468,19 +515,22 @@ static int setup_ring(struct submitter *s)
 
 static void file_depths(char *buf)
 {
-	struct submitter *s = submitter;
 	char *p;
-	int i;
+	int i, j;
 
 	buf[0] = '\0';
 	p = buf;
-	for (i = 0; i < s->nr_files; i++) {
-		struct file *f = &s->files[i];
+	for (j = 0; j < nthreads; j++) {
+		struct submitter *s = get_submitter(j);
 
-		if (i + 1 == s->nr_files)
-			p += sprintf(p, "%d", f->pending_ios);
-		else
-			p += sprintf(p, "%d, ", f->pending_ios);
+		for (i = 0; i < s->nr_files; i++) {
+			struct file *f = &s->files[i];
+
+			if (i + 1 == s->nr_files)
+				p += sprintf(p, "%d", f->pending_ios);
+			else
+				p += sprintf(p, "%d, ", f->pending_ios);
+		}
 	}
 }
 
@@ -489,16 +539,18 @@ static void usage(char *argv)
 	printf("%s [options] -- [filenames]\n"
 		" -d <int> : IO Depth, default %d\n"
 		" -s <int> : Batch submit, default %d\n"
-		" -c <int> : Batch complete, default %d\n",
-		argv, DEPTH, BATCH_SUBMIT, BATCH_COMPLETE);
+		" -c <int> : Batch complete, default %d\n"
+		" -b <int> : Block size, default %d\n"
+		" -p <bool> : Polled IO, default %d\n",
+		argv, DEPTH, BATCH_SUBMIT, BATCH_COMPLETE, BS, polled);
 	exit(0);
 }
 
 int main(int argc, char *argv[])
 {
 	struct submitter *s;
-	unsigned long done, calls, reap, cache_hit, cache_miss;
-	int err, i, flags, fd, opt;
+	unsigned long done, calls, reap;
+	int err, i, j, flags, fd, opt;
 	char *fdepths;
 	void *ret;
 
@@ -507,7 +559,7 @@ int main(int argc, char *argv[])
 		return 1;
 	}
 
-	while ((opt = getopt(argc, argv, "d:s:c:h?")) != -1) {
+	while ((opt = getopt(argc, argv, "d:s:c:b:p:B:F:n:h?")) != -1) {
 		switch (opt) {
 		case 'd':
 			depth = atoi(optarg);
@@ -518,6 +570,21 @@ int main(int argc, char *argv[])
 		case 'c':
 			batch_complete = atoi(optarg);
 			break;
+		case 'b':
+			bs = atoi(optarg);
+			break;
+		case 'p':
+			polled = !!atoi(optarg);
+			break;
+		case 'B':
+			fixedbufs = !!atoi(optarg);
+			break;
+		case 'F':
+			register_files = !!atoi(optarg);
+			break;
+		case 'n':
+			nthreads = atoi(optarg);
+			break;
 		case 'h':
 		case '?':
 		default:
@@ -526,18 +593,25 @@ int main(int argc, char *argv[])
 		}
 	}
 
-	submitter = malloc(sizeof(*submitter) + depth * sizeof(struct iovec));
-	memset(submitter, 0, sizeof(*submitter) + depth * sizeof(struct iovec));
-	s = submitter;
+	submitter = calloc(nthreads, sizeof(*submitter) +
+				depth * sizeof(struct iovec));
+	for (j = 0; j < nthreads; j++) {
+		s = get_submitter(j);
+		s->index = j;
+		s->done = s->calls = s->reaps = 0;
+	}
 
 	flags = O_RDONLY | O_NOATIME;
 	if (!buffered)
 		flags |= O_DIRECT;
 
+	j = 0;
 	i = optind;
+	printf("i %d, argc %d\n", i, argc);
 	while (!do_nop && i < argc) {
 		struct file *f;
 
+		s = get_submitter(j);
 		if (s->nr_files == MAX_FDS) {
 			printf("Max number of files (%d) reached\n", MAX_FDS);
 			break;
@@ -560,9 +634,11 @@ int main(int argc, char *argv[])
 		}
 		f->max_blocks--;
 
-		printf("Added file %s\n", argv[i]);
+		printf("Added file %s (submitter %d)\n", argv[i], s->index);
 		s->nr_files++;
 		i++;
+		if (++j >= nthreads)
+			j = 0;
 	}
 
 	if (fixedbufs) {
@@ -578,51 +654,51 @@ int main(int argc, char *argv[])
 
 	arm_sig_int();
 
-	for (i = 0; i < depth; i++) {
-		void *buf;
+	for (j = 0; j < nthreads; j++) {
+		s = get_submitter(j);
+		for (i = 0; i < depth; i++) {
+			void *buf;
 
-		if (posix_memalign(&buf, BS, BS)) {
-			printf("failed alloc\n");
-			return 1;
+			if (posix_memalign(&buf, bs, bs)) {
+				printf("failed alloc\n");
+				return 1;
+			}
+			s->iovecs[i].iov_base = buf;
+			s->iovecs[i].iov_len = bs;
 		}
-		s->iovecs[i].iov_base = buf;
-		s->iovecs[i].iov_len = BS;
 	}
 
-	err = setup_ring(s);
-	if (err) {
-		printf("ring setup failed: %s, %d\n", strerror(errno), err);
-		return 1;
+	for (j = 0; j < nthreads; j++) {
+		s = get_submitter(j);
+
+		err = setup_ring(s);
+		if (err) {
+			printf("ring setup failed: %s, %d\n", strerror(errno), err);
+			return 1;
+		}
 	}
-	printf("polled=%d, fixedbufs=%d, buffered=%d", polled, fixedbufs, buffered);
+	s = get_submitter(0);
+	printf("polled=%d, fixedbufs=%d, register_files=%d, buffered=%d", polled, fixedbufs, register_files, buffered);
 	printf(" QD=%d, sq_ring=%d, cq_ring=%d\n", depth, *s->sq_ring.ring_entries, *s->cq_ring.ring_entries);
 
-	pthread_create(&s->thread, NULL, submitter_fn, s);
+	for (j = 0; j < nthreads; j++) {
+		s = get_submitter(j);
+		pthread_create(&s->thread, NULL, submitter_fn, s);
+	}
 
-	fdepths = malloc(8 * s->nr_files);
-	cache_hit = cache_miss = reap = calls = done = 0;
+	fdepths = malloc(8 * s->nr_files * nthreads);
+	reap = calls = done = 0;
 	do {
 		unsigned long this_done = 0;
 		unsigned long this_reap = 0;
 		unsigned long this_call = 0;
-		unsigned long this_cache_hit = 0;
-		unsigned long this_cache_miss = 0;
 		unsigned long rpc = 0, ipc = 0;
-		double hit = 0.0;
 
 		sleep(1);
-		this_done += s->done;
-		this_call += s->calls;
-		this_reap += s->reaps;
-		this_cache_hit += s->cachehit;
-		this_cache_miss += s->cachemiss;
-		if (this_cache_hit && this_cache_miss) {
-			unsigned long hits, total;
-
-			hits = this_cache_hit - cache_hit;
-			total = hits + this_cache_miss - cache_miss;
-			hit = (double) hits / (double) total;
-			hit *= 100.0;
+		for (j = 0; j < nthreads; j++) {
+			this_done += s->done;
+			this_call += s->calls;
+			this_reap += s->reaps;
 		}
 		if (this_call - calls) {
 			rpc = (this_done - done) / (this_call - calls);
@@ -630,18 +706,20 @@ int main(int argc, char *argv[])
 		} else
 			rpc = ipc = -1;
 		file_depths(fdepths);
-		printf("IOPS=%lu, IOS/call=%ld/%ld, inflight=%u (%s), Cachehit=%0.2f%%\n",
+		printf("IOPS=%lu, IOS/call=%ld/%ld, inflight=%u (%s)\n",
 				this_done - done, rpc, ipc, s->inflight,
-				fdepths, hit);
+				fdepths);
 		done = this_done;
 		calls = this_call;
 		reap = this_reap;
-		cache_hit = s->cachehit;
-		cache_miss = s->cachemiss;
 	} while (!finish);
 
-	pthread_join(s->thread, &ret);
-	close(s->ring_fd);
+	for (j = 0; j < nthreads; j++) {
+		s = get_submitter(j);
+		pthread_join(s->thread, &ret);
+		close(s->ring_fd);
+	}
 	free(fdepths);
+	free(submitter);
 	return 0;
 }