linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jens Axboe <axboe@kernel.dk>
To: linux-fsdevel@vger.kernel.org, linux-block@vger.kernel.org
Cc: viro@ZenIV.linux.org.uk, Jens Axboe <axboe@kernel.dk>
Subject: [PATCH 6/7] io_uring: retry bulk slab allocs as single allocs
Date: Fri, 15 Mar 2019 08:54:41 -0600	[thread overview]
Message-ID: <20190315145442.21127-7-axboe@kernel.dk> (raw)
In-Reply-To: <20190315145442.21127-1-axboe@kernel.dk>

I've seen cases where bulk alloc fails, since the bulk alloc API
is all-or-nothing - either we get the number we ask for, or it
returns 0 as number of entries.

If we fail a batch bulk alloc, retry a "normal" kmem_cache_alloc()
and just use that instead of failing with -EAGAIN.

While in there, ensure we use GFP_KERNEL. That was an oversight in
the original code, when we switched away from GFP_ATOMIC.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 fs/io_uring.c | 22 ++++++++++++++++------
 1 file changed, 16 insertions(+), 6 deletions(-)

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 443ca5615554..5be6e4f99a9e 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -399,13 +399,14 @@ static void io_ring_drop_ctx_refs(struct io_ring_ctx *ctx, unsigned refs)
 static struct io_kiocb *io_get_req(struct io_ring_ctx *ctx,
 				   struct io_submit_state *state)
 {
+	gfp_t gfp = GFP_KERNEL | __GFP_NOWARN;
 	struct io_kiocb *req;
 
 	if (!percpu_ref_tryget(&ctx->refs))
 		return NULL;
 
 	if (!state) {
-		req = kmem_cache_alloc(req_cachep, __GFP_NOWARN);
+		req = kmem_cache_alloc(req_cachep, gfp);
 		if (unlikely(!req))
 			goto out;
 	} else if (!state->free_reqs) {
@@ -413,13 +414,22 @@ static struct io_kiocb *io_get_req(struct io_ring_ctx *ctx,
 		int ret;
 
 		sz = min_t(size_t, state->ios_left, ARRAY_SIZE(state->reqs));
-		ret = kmem_cache_alloc_bulk(req_cachep, __GFP_NOWARN, sz,
-						state->reqs);
-		if (unlikely(ret <= 0))
-			goto out;
+		ret = kmem_cache_alloc_bulk(req_cachep, gfp, sz, state->reqs);
+
+		/*
+		 * Bulk alloc is all-or-nothing. If we fail to get a batch,
+		 * retry single alloc to be on the safe side.
+		 */
+		if (ret <= 0) {
+			req = kmem_cache_alloc(req_cachep, gfp);
+			if (unlikely(!req))
+				goto out;
+			ret = 1;
+		} else {
+			req = state->reqs[0];
+		}
 		state->free_reqs = ret - 1;
 		state->cur_req = 1;
-		req = state->reqs[0];
 	} else {
 		req = state->reqs[state->cur_req];
 		state->free_reqs--;
-- 
2.17.1


  parent reply	other threads:[~2019-03-15 14:55 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-03-15 14:54 [PATCHSET v2 0/7] io_uring fixes Jens Axboe
2019-03-15 14:54 ` [PATCH 1/7] io_uring: use regular request ref counts Jens Axboe
2019-03-15 14:54 ` [PATCH 2/7] io_uring: make io_read/write return an integer Jens Axboe
2019-03-15 14:54 ` [PATCH 3/7] io_uring: add prepped flag Jens Axboe
2019-03-15 14:54 ` [PATCH 4/7] io_uring: fix fget/fput handling Jens Axboe
2019-03-15 14:54 ` [PATCH 5/7] io_uring: fix poll races Jens Axboe
2019-03-15 14:54 ` Jens Axboe [this message]
2019-03-15 14:54 ` [PATCH 7/7] io_uring: mark me as the maintainer Jens Axboe

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20190315145442.21127-7-axboe@kernel.dk \
    --to=axboe@kernel.dk \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=viro@ZenIV.linux.org.uk \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).