summaryrefslogtreecommitdiff
path: root/fs/aio.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/aio.c')
-rw-r--r--fs/aio.c126
1 files changed, 60 insertions, 66 deletions
diff --git a/fs/aio.c b/fs/aio.c
index 0f4f531c9780..793b7b15ec4b 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -100,7 +100,7 @@ struct kioctx {
unsigned long user_id;
- struct __percpu kioctx_cpu *cpu;
+ struct kioctx_cpu __percpu *cpu;
/*
* For percpu reqs_available, number of slots we move to/from global
@@ -122,7 +122,7 @@ struct kioctx {
unsigned long mmap_base;
unsigned long mmap_size;
- struct page **ring_pages;
+ struct folio **ring_folios;
long nr_pages;
struct rcu_work free_rwork; /* see free_ioctx() */
@@ -160,7 +160,7 @@ struct kioctx {
spinlock_t completion_lock;
} ____cacheline_aligned_in_smp;
- struct page *internal_pages[AIO_RING_PAGES];
+ struct folio *internal_folios[AIO_RING_PAGES];
struct file *aio_ring_file;
unsigned id;
@@ -224,7 +224,7 @@ static unsigned long aio_nr; /* current system wide number of aio requests */
static unsigned long aio_max_nr = 0x10000; /* system wide maximum number of aio requests */
/*----end sysctl variables---*/
#ifdef CONFIG_SYSCTL
-static struct ctl_table aio_sysctls[] = {
+static const struct ctl_table aio_sysctls[] = {
{
.procname = "aio-nr",
.data = &aio_nr,
@@ -334,19 +334,20 @@ static void aio_free_ring(struct kioctx *ctx)
put_aio_ring_file(ctx);
for (i = 0; i < ctx->nr_pages; i++) {
- struct page *page;
- pr_debug("pid(%d) [%d] page->count=%d\n", current->pid, i,
- page_count(ctx->ring_pages[i]));
- page = ctx->ring_pages[i];
- if (!page)
+ struct folio *folio = ctx->ring_folios[i];
+
+ if (!folio)
continue;
- ctx->ring_pages[i] = NULL;
- put_page(page);
+
+ pr_debug("pid(%d) [%d] folio->count=%d\n", current->pid, i,
+ folio_ref_count(folio));
+ ctx->ring_folios[i] = NULL;
+ folio_put(folio);
}
- if (ctx->ring_pages && ctx->ring_pages != ctx->internal_pages) {
- kfree(ctx->ring_pages);
- ctx->ring_pages = NULL;
+ if (ctx->ring_folios && ctx->ring_folios != ctx->internal_folios) {
+ kfree(ctx->ring_folios);
+ ctx->ring_folios = NULL;
}
}
@@ -409,17 +410,7 @@ static int aio_migrate_folio(struct address_space *mapping, struct folio *dst,
struct kioctx *ctx;
unsigned long flags;
pgoff_t idx;
- int rc;
-
- /*
- * We cannot support the _NO_COPY case here, because copy needs to
- * happen under the ctx->completion_lock. That does not work with the
- * migration workflow of MIGRATE_SYNC_NO_COPY.
- */
- if (mode == MIGRATE_SYNC_NO_COPY)
- return -EINVAL;
-
- rc = 0;
+ int rc = 0;
/* mapping->i_private_lock here protects against the kioctx teardown. */
spin_lock(&mapping->i_private_lock);
@@ -441,7 +432,7 @@ static int aio_migrate_folio(struct address_space *mapping, struct folio *dst,
idx = src->index;
if (idx < (pgoff_t)ctx->nr_pages) {
/* Make sure the old folio hasn't already been changed */
- if (ctx->ring_pages[idx] != &src->page)
+ if (ctx->ring_folios[idx] != src)
rc = -EAGAIN;
} else
rc = -EINVAL;
@@ -464,9 +455,10 @@ static int aio_migrate_folio(struct address_space *mapping, struct folio *dst,
* events from being lost.
*/
spin_lock_irqsave(&ctx->completion_lock, flags);
- folio_migrate_copy(dst, src);
- BUG_ON(ctx->ring_pages[idx] != &src->page);
- ctx->ring_pages[idx] = &dst->page;
+ folio_copy(dst, src);
+ folio_migrate_flags(dst, src);
+ BUG_ON(ctx->ring_folios[idx] != src);
+ ctx->ring_folios[idx] = dst;
spin_unlock_irqrestore(&ctx->completion_lock, flags);
/* The old folio is no longer accessible. */
@@ -516,28 +508,30 @@ static int aio_setup_ring(struct kioctx *ctx, unsigned int nr_events)
nr_events = (PAGE_SIZE * nr_pages - sizeof(struct aio_ring))
/ sizeof(struct io_event);
- ctx->ring_pages = ctx->internal_pages;
+ ctx->ring_folios = ctx->internal_folios;
if (nr_pages > AIO_RING_PAGES) {
- ctx->ring_pages = kcalloc(nr_pages, sizeof(struct page *),
- GFP_KERNEL);
- if (!ctx->ring_pages) {
+ ctx->ring_folios = kcalloc(nr_pages, sizeof(struct folio *),
+ GFP_KERNEL);
+ if (!ctx->ring_folios) {
put_aio_ring_file(ctx);
return -ENOMEM;
}
}
for (i = 0; i < nr_pages; i++) {
- struct page *page;
- page = find_or_create_page(file->f_mapping,
- i, GFP_USER | __GFP_ZERO);
- if (!page)
+ struct folio *folio;
+
+ folio = __filemap_get_folio(file->f_mapping, i,
+ FGP_LOCK | FGP_ACCESSED | FGP_CREAT,
+ GFP_USER | __GFP_ZERO);
+ if (IS_ERR(folio))
break;
- pr_debug("pid(%d) page[%d]->count=%d\n",
- current->pid, i, page_count(page));
- SetPageUptodate(page);
- unlock_page(page);
- ctx->ring_pages[i] = page;
+ pr_debug("pid(%d) [%d] folio->count=%d\n", current->pid, i,
+ folio_ref_count(folio));
+ folio_end_read(folio, true);
+
+ ctx->ring_folios[i] = folio;
}
ctx->nr_pages = i;
@@ -570,7 +564,7 @@ static int aio_setup_ring(struct kioctx *ctx, unsigned int nr_events)
ctx->user_id = ctx->mmap_base;
ctx->nr_events = nr_events; /* trusted copy */
- ring = page_address(ctx->ring_pages[0]);
+ ring = folio_address(ctx->ring_folios[0]);
ring->nr = nr_events; /* user copy */
ring->id = ~0U;
ring->head = ring->tail = 0;
@@ -578,7 +572,7 @@ static int aio_setup_ring(struct kioctx *ctx, unsigned int nr_events)
ring->compat_features = AIO_RING_COMPAT_FEATURES;
ring->incompat_features = AIO_RING_INCOMPAT_FEATURES;
ring->header_length = sizeof(struct aio_ring);
- flush_dcache_page(ctx->ring_pages[0]);
+ flush_dcache_folio(ctx->ring_folios[0]);
return 0;
}
@@ -689,9 +683,9 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm)
/* While kioctx setup is in progress,
* we are protected from page migration
- * changes ring_pages by ->ring_lock.
+ * changes ring_folios by ->ring_lock.
*/
- ring = page_address(ctx->ring_pages[0]);
+ ring = folio_address(ctx->ring_folios[0]);
ring->id = ctx->id;
return 0;
}
@@ -1033,7 +1027,7 @@ static void user_refill_reqs_available(struct kioctx *ctx)
* against ctx->completed_events below will make sure we do the
* safe/right thing.
*/
- ring = page_address(ctx->ring_pages[0]);
+ ring = folio_address(ctx->ring_folios[0]);
head = ring->head;
refill_reqs_available(ctx, head, ctx->tail);
@@ -1145,12 +1139,12 @@ static void aio_complete(struct aio_kiocb *iocb)
if (++tail >= ctx->nr_events)
tail = 0;
- ev_page = page_address(ctx->ring_pages[pos / AIO_EVENTS_PER_PAGE]);
+ ev_page = folio_address(ctx->ring_folios[pos / AIO_EVENTS_PER_PAGE]);
event = ev_page + pos % AIO_EVENTS_PER_PAGE;
*event = iocb->ki_res;
- flush_dcache_page(ctx->ring_pages[pos / AIO_EVENTS_PER_PAGE]);
+ flush_dcache_folio(ctx->ring_folios[pos / AIO_EVENTS_PER_PAGE]);
pr_debug("%p[%u]: %p: %p %Lx %Lx %Lx\n", ctx, tail, iocb,
(void __user *)(unsigned long)iocb->ki_res.obj,
@@ -1163,10 +1157,10 @@ static void aio_complete(struct aio_kiocb *iocb)
ctx->tail = tail;
- ring = page_address(ctx->ring_pages[0]);
+ ring = folio_address(ctx->ring_folios[0]);
head = ring->head;
ring->tail = tail;
- flush_dcache_page(ctx->ring_pages[0]);
+ flush_dcache_folio(ctx->ring_folios[0]);
ctx->completed_events++;
if (ctx->completed_events > 1)
@@ -1238,8 +1232,8 @@ static long aio_read_events_ring(struct kioctx *ctx,
sched_annotate_sleep();
mutex_lock(&ctx->ring_lock);
- /* Access to ->ring_pages here is protected by ctx->ring_lock. */
- ring = page_address(ctx->ring_pages[0]);
+ /* Access to ->ring_folios here is protected by ctx->ring_lock. */
+ ring = folio_address(ctx->ring_folios[0]);
head = ring->head;
tail = ring->tail;
@@ -1260,20 +1254,20 @@ static long aio_read_events_ring(struct kioctx *ctx,
while (ret < nr) {
long avail;
struct io_event *ev;
- struct page *page;
+ struct folio *folio;
avail = (head <= tail ? tail : ctx->nr_events) - head;
if (head == tail)
break;
pos = head + AIO_EVENTS_OFFSET;
- page = ctx->ring_pages[pos / AIO_EVENTS_PER_PAGE];
+ folio = ctx->ring_folios[pos / AIO_EVENTS_PER_PAGE];
pos %= AIO_EVENTS_PER_PAGE;
avail = min(avail, nr - ret);
avail = min_t(long, avail, AIO_EVENTS_PER_PAGE - pos);
- ev = page_address(page);
+ ev = folio_address(folio);
copy_ret = copy_to_user(event + ret, ev + pos,
sizeof(*ev) * avail);
@@ -1287,9 +1281,9 @@ static long aio_read_events_ring(struct kioctx *ctx,
head %= ctx->nr_events;
}
- ring = page_address(ctx->ring_pages[0]);
+ ring = folio_address(ctx->ring_folios[0]);
ring->head = head;
- flush_dcache_page(ctx->ring_pages[0]);
+ flush_dcache_folio(ctx->ring_folios[0]);
pr_debug("%li h%u t%u\n", ret, head, tail);
out:
@@ -1341,7 +1335,7 @@ static long read_events(struct kioctx *ctx, long min_nr, long nr,
if (until == 0 || ret < 0 || ret >= min_nr)
return ret;
- hrtimer_init_sleeper_on_stack(&t, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+ hrtimer_setup_sleeper_on_stack(&t, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
if (until != KTIME_MAX) {
hrtimer_set_expires_range_ns(&t.timer, until, current->timer_slack_ns);
hrtimer_sleeper_start_expires(&t, HRTIMER_MODE_REL);
@@ -1513,10 +1507,11 @@ static void aio_complete_rw(struct kiocb *kiocb, long res)
iocb_put(iocb);
}
-static int aio_prep_rw(struct kiocb *req, const struct iocb *iocb)
+static int aio_prep_rw(struct kiocb *req, const struct iocb *iocb, int rw_type)
{
int ret;
+ req->ki_write_stream = 0;
req->ki_complete = aio_complete_rw;
req->private = NULL;
req->ki_pos = iocb->aio_offset;
@@ -1539,7 +1534,7 @@ static int aio_prep_rw(struct kiocb *req, const struct iocb *iocb)
} else
req->ki_ioprio = get_current_ioprio();
- ret = kiocb_set_rw_flags(req, iocb->aio_rw_flags);
+ ret = kiocb_set_rw_flags(req, iocb->aio_rw_flags, rw_type);
if (unlikely(ret))
return ret;
@@ -1591,7 +1586,7 @@ static int aio_read(struct kiocb *req, const struct iocb *iocb,
struct file *file;
int ret;
- ret = aio_prep_rw(req, iocb);
+ ret = aio_prep_rw(req, iocb, READ);
if (ret)
return ret;
file = req->ki_filp;
@@ -1605,7 +1600,7 @@ static int aio_read(struct kiocb *req, const struct iocb *iocb,
return ret;
ret = rw_verify_area(READ, file, &req->ki_pos, iov_iter_count(&iter));
if (!ret)
- aio_rw_done(req, call_read_iter(file, req, &iter));
+ aio_rw_done(req, file->f_op->read_iter(req, &iter));
kfree(iovec);
return ret;
}
@@ -1618,7 +1613,7 @@ static int aio_write(struct kiocb *req, const struct iocb *iocb,
struct file *file;
int ret;
- ret = aio_prep_rw(req, iocb);
+ ret = aio_prep_rw(req, iocb, WRITE);
if (ret)
return ret;
file = req->ki_filp;
@@ -1636,7 +1631,7 @@ static int aio_write(struct kiocb *req, const struct iocb *iocb,
if (S_ISREG(file_inode(file)->i_mode))
kiocb_start_write(req);
req->ki_flags |= IOCB_WRITE;
- aio_rw_done(req, call_write_iter(file, req, &iter));
+ aio_rw_done(req, file->f_op->write_iter(req, &iter));
}
kfree(iovec);
return ret;
@@ -2197,7 +2192,6 @@ SYSCALL_DEFINE3(io_cancel, aio_context_t, ctx_id, struct iocb __user *, iocb,
return -EINVAL;
spin_lock_irq(&ctx->ctx_lock);
- /* TODO: use a hash or array, this sucks. */
list_for_each_entry(kiocb, &ctx->active_reqs, ki_list) {
if (kiocb->ki_res.obj == obj) {
ret = kiocb->ki_cancel(&kiocb->rw);
generated by cgit (git 2.34.1) at 2025-07-16 06:49:15 +0000