fio: replace malloc+memset with calloc

[fio.git] / t / io_uring.c

diff --git a/t/io_uring.c b/t/io_uring.c
index 9d580b5af11b529af3233298547f8ed051ae88d7..6b0efef85cb745fecc71cefff3f52d7eada4477a 100644 (file)
--- a/t/io_uring.c
+++ b/t/io_uring.c
@@ -156,7 +156,7 @@ static float plist[] = { 1.0, 5.0, 10.0, 20.0, 30.0, 40.0, 50.0, 60.0, 70.0,
 static int plist_len = 17;
 
 #ifndef IORING_REGISTER_MAP_BUFFERS
-#define IORING_REGISTER_MAP_BUFFERS    22
+#define IORING_REGISTER_MAP_BUFFERS    26
 struct io_uring_map_buffers {
        __s32   fd;
        __u32   buf_start;
@@ -487,11 +487,10 @@ 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));
+       p = calloc(1, 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;
@@ -530,8 +529,11 @@ static unsigned long long get_offset(struct submitter *s, struct file *f)
        long r;
 
        if (random_io) {
+               unsigned long long block;
+
                r = __rand64(&s->rand_state);
-               offset = (r % (f->max_blocks - 1)) * bs;
+               block = r % f->max_blocks;
+               offset = block * (unsigned long long) bs;
        } else {
                offset = f->cur_off;
                f->cur_off += bs;
@@ -542,16 +544,10 @@ static unsigned long long get_offset(struct submitter *s, struct file *f)
        return offset;
 }
 
-static void init_io(struct submitter *s, unsigned index)
+static struct file *init_new_io(struct submitter *s)
 {
-       struct io_uring_sqe *sqe = &s->sqes[index];
        struct file *f;
 
-       if (do_nop) {
-               sqe->opcode = IORING_OP_NOP;
-               return;
-       }
-
        if (s->nr_files == 1) {
                f = &s->files[0];
        } else {
@@ -563,7 +559,22 @@ static void init_io(struct submitter *s, unsigned index)
                        f = &s->files[s->cur_file];
                }
        }
+
        f->pending_ios++;
+       return f;
+}
+
+static void init_io(struct submitter *s, unsigned index)
+{
+       struct io_uring_sqe *sqe = &s->sqes[index];
+       struct file *f;
+
+       if (do_nop) {
+               sqe->opcode = IORING_OP_NOP;
+               return;
+       }
+
+       f = init_new_io(s);
 
        if (register_files) {
                sqe->flags = IOSQE_FIXED_FILE;
@@ -603,30 +614,10 @@ static void init_io_pt(struct submitter *s, unsigned index)
        struct nvme_uring_cmd *cmd;
        unsigned long long slba;
        unsigned long long nlb;
-       long r;
 
-       if (s->nr_files == 1) {
-               f = &s->files[0];
-       } else {
-               f = &s->files[s->cur_file];
-               if (f->pending_ios >= file_depth(s)) {
-                       s->cur_file++;
-                       if (s->cur_file == s->nr_files)
-                               s->cur_file = 0;
-                       f = &s->files[s->cur_file];
-               }
-       }
-       f->pending_ios++;
+       f = init_new_io(s);
 
-       if (random_io) {
-               r = __rand64(&s->rand_state);
-               offset = (r % (f->max_blocks - 1)) * bs;
-       } else {
-               offset = f->cur_off;
-               f->cur_off += bs;
-               if (f->cur_off + bs > f->max_size)
-                       f->cur_off = 0;
-       }
+       offset = get_offset(s, f);
 
        if (register_files) {
                sqe->fd = f->fixed_fd;
@@ -650,6 +641,10 @@ static void init_io_pt(struct submitter *s, unsigned index)
        cmd->cdw12 = nlb;
        cmd->addr = (unsigned long) s->iovecs[index].iov_base;
        cmd->data_len = bs;
+       if (fixedbufs) {
+               sqe->uring_cmd_flags = IORING_URING_CMD_FIXED;
+               sqe->buf_index = index;
+       }
        cmd->nsid = f->nsid;
        cmd->opcode = 2;
 }
@@ -657,8 +652,12 @@ static void init_io_pt(struct submitter *s, unsigned index)
 static int prep_more_ios_uring(struct submitter *s, int max_ios)
 {
        struct io_sq_ring *ring = &s->sq_ring;
-       unsigned index, tail, next_tail, prepped = 0;
-       unsigned int head = atomic_load_acquire(ring->head);
+       unsigned head, index, tail, next_tail, prepped = 0;
+
+       if (sq_thread_poll)
+               head = atomic_load_acquire(ring->head);
+       else
+               head = *ring->head;
 
        next_tail = tail = *ring->tail;
        do {
@@ -704,7 +703,7 @@ static int get_file_size(struct file *f)
                                        bs, lbs);
                        return -1;
                }
-               f->max_blocks = nlba / bs;
+               f->max_blocks = nlba;
                f->max_size = nlba;
                f->lba_shift = ilog2(lbs);
                return 0;
@@ -737,7 +736,6 @@ static int reap_events_uring(struct submitter *s)
        do {
                struct file *f;
 
-               read_barrier();
                if (head == atomic_load_acquire(ring->tail))
                        break;
                cqe = &ring->cqes[head & cq_ring_mask];
@@ -792,7 +790,6 @@ static int reap_events_uring_pt(struct submitter *s)
        do {
                struct file *f;
 
-               read_barrier();
                if (head == atomic_load_acquire(ring->tail))
                        break;
                index = head & cq_ring_mask;
@@ -856,7 +853,10 @@ static int detect_node(struct submitter *s, const char *name)
        char str[128];
        int ret, fd, node;
 
-       sprintf(str, "/sys/block/%s/device/numa_node", base);
+       if (pt)
+               sprintf(str, "/sys/class/nvme-generic/%s/device/numa_node", base);
+       else
+               sprintf(str, "/sys/block/%s/device/numa_node", base);
        fd = open(str, O_RDONLY);
        if (fd < 0)
                return -1;
@@ -1112,18 +1112,7 @@ static int prep_more_ios_aio(struct submitter *s, int max_ios, struct iocb *iocb
        while (index < max_ios) {
                struct iocb *iocb = &iocbs[index];
 
-               if (s->nr_files == 1) {
-                       f = &s->files[0];
-               } else {
-                       f = &s->files[s->cur_file];
-                       if (f->pending_ios >= file_depth(s)) {
-                               s->cur_file++;
-                               if (s->cur_file == s->nr_files)
-                                       s->cur_file = 0;
-                               f = &s->files[s->cur_file];
-                       }
-               }
-               f->pending_ios++;
+               f = init_new_io(s);
 
                io_prep_pread(iocb, f->real_fd, s->iovecs[index].iov_base,
                                s->iovecs[index].iov_len, get_offset(s, f));
@@ -1410,18 +1399,7 @@ static void *submitter_sync_fn(void *data)
                uint64_t offset;
                struct file *f;
 
-               if (s->nr_files == 1) {
-                       f = &s->files[0];
-               } else {
-                       f = &s->files[s->cur_file];
-                       if (f->pending_ios >= file_depth(s)) {
-                               s->cur_file++;
-                               if (s->cur_file == s->nr_files)
-                                       s->cur_file = 0;
-                               f = &s->files[s->cur_file];
-                       }
-               }
-               f->pending_ios++;
+               f = init_new_io(s);
 
 #ifdef ARCH_HAVE_CPU_CLOCK
                if (stats)