public inbox for [email protected]
 help / color / mirror / Atom feed
From: Pavel Begunkov <[email protected]>
To: [email protected]
Cc: Jens Axboe <[email protected]>, [email protected]
Subject: [PATCH 6/7] io_uring/rsrc: keep one global dummy_ubuf
Date: Fri, 11 Aug 2023 13:53:46 +0100	[thread overview]
Message-ID: <e4a96dda35ab755914bc43f6781bba0df97ac489.1691757663.git.asml.silence@gmail.com> (raw)
In-Reply-To: <[email protected]>

We set empty registered buffers to dummy_ubuf as an optimisation.
Currently, we allocate the dummy entry for each ring, whenever we can
simply have one global instance.

We're casting out const on assignment, it's fine as we're not going to
change the content of the dummy, the constness gives us an extra layer
of protection if sth ever goes wrong.

Signed-off-by: Pavel Begunkov <[email protected]>
---
 io_uring/io_uring.c |  9 ---------
 io_uring/rsrc.c     | 14 ++++++++++----
 2 files changed, 10 insertions(+), 13 deletions(-)

diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index e57d00939ab9..a7a4d637aee0 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -290,13 +290,6 @@ static __cold struct io_ring_ctx *io_ring_ctx_alloc(struct io_uring_params *p)
 		goto err;
 	if (io_alloc_hash_table(&ctx->cancel_table_locked, hash_bits))
 		goto err;
-
-	ctx->dummy_ubuf = kzalloc(sizeof(*ctx->dummy_ubuf), GFP_KERNEL);
-	if (!ctx->dummy_ubuf)
-		goto err;
-	/* set invalid range, so io_import_fixed() fails meeting it */
-	ctx->dummy_ubuf->ubuf = -1UL;
-
 	if (percpu_ref_init(&ctx->refs, io_ring_ctx_ref_free,
 			    0, GFP_KERNEL))
 		goto err;
@@ -335,7 +328,6 @@ static __cold struct io_ring_ctx *io_ring_ctx_alloc(struct io_uring_params *p)
 	INIT_WQ_LIST(&ctx->submit_state.compl_reqs);
 	return ctx;
 err:
-	kfree(ctx->dummy_ubuf);
 	kfree(ctx->cancel_table.hbs);
 	kfree(ctx->cancel_table_locked.hbs);
 	kfree(ctx->io_bl);
@@ -2897,7 +2889,6 @@ static __cold void io_ring_ctx_free(struct io_ring_ctx *ctx)
 		io_wq_put_hash(ctx->hash_map);
 	kfree(ctx->cancel_table.hbs);
 	kfree(ctx->cancel_table_locked.hbs);
-	kfree(ctx->dummy_ubuf);
 	kfree(ctx->io_bl);
 	xa_destroy(&ctx->io_bl_xa);
 	kfree(ctx);
diff --git a/io_uring/rsrc.c b/io_uring/rsrc.c
index 5e8fdd9b8ca6..d9c853d10587 100644
--- a/io_uring/rsrc.c
+++ b/io_uring/rsrc.c
@@ -33,6 +33,12 @@ static int io_sqe_buffer_register(struct io_ring_ctx *ctx, struct iovec *iov,
 #define IORING_MAX_FIXED_FILES	(1U << 20)
 #define IORING_MAX_REG_BUFFERS	(1U << 14)
 
+static const struct io_mapped_ubuf dummy_ubuf = {
+	/* set invalid range, so io_import_fixed() fails meeting it */
+	.ubuf = -1UL,
+	.ubuf_end = 0,
+};
+
 int __io_account_mem(struct user_struct *user, unsigned long nr_pages)
 {
 	unsigned long page_limit, cur_pages, new_pages;
@@ -132,7 +138,7 @@ static void io_buffer_unmap(struct io_ring_ctx *ctx, struct io_mapped_ubuf **slo
 	struct io_mapped_ubuf *imu = *slot;
 	unsigned int i;
 
-	if (imu != ctx->dummy_ubuf) {
+	if (imu != &dummy_ubuf) {
 		for (i = 0; i < imu->nr_bvecs; i++)
 			unpin_user_page(imu->bvec[i].bv_page);
 		if (imu->acct_pages)
@@ -459,14 +465,14 @@ static int __io_sqe_buffers_update(struct io_ring_ctx *ctx,
 			break;
 
 		i = array_index_nospec(up->offset + done, ctx->nr_user_bufs);
-		if (ctx->user_bufs[i] != ctx->dummy_ubuf) {
+		if (ctx->user_bufs[i] != &dummy_ubuf) {
 			err = io_queue_rsrc_removal(ctx->buf_data, i,
 						    ctx->user_bufs[i]);
 			if (unlikely(err)) {
 				io_buffer_unmap(ctx, &imu);
 				break;
 			}
-			ctx->user_bufs[i] = ctx->dummy_ubuf;
+			ctx->user_bufs[i] = (struct io_mapped_ubuf *)&dummy_ubuf;
 		}
 
 		ctx->user_bufs[i] = imu;
@@ -1077,7 +1083,7 @@ static int io_sqe_buffer_register(struct io_ring_ctx *ctx, struct iovec *iov,
 	int ret, nr_pages, i;
 	struct folio *folio = NULL;
 
-	*pimu = ctx->dummy_ubuf;
+	*pimu = (struct io_mapped_ubuf *)&dummy_ubuf;
 	if (!iov->iov_base)
 		return 0;
 
-- 
2.41.0


  parent reply	other threads:[~2023-08-11 12:55 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-11 12:53 [PATCH 0/7] random fixes and cleanups Pavel Begunkov
2023-08-11 12:53 ` [PATCH 1/7] io_uring/net: don't overflow multishot accept Pavel Begunkov
2023-08-11 12:53 ` [PATCH 2/7] io_uring/net: don't overflow multishot recv Pavel Begunkov
2023-09-14  8:34   ` Jiri Slaby
2023-09-14 13:02     ` Pavel Begunkov
2023-09-14 14:15       ` Jens Axboe
2023-08-11 12:53 ` [PATCH 3/7] io_uring: open code io_fill_cqe_req() Pavel Begunkov
2023-08-11 12:53 ` [PATCH 4/7] io_uring: remove return from io_req_cqe_overflow() Pavel Begunkov
2023-08-11 12:53 ` [PATCH 5/7] io_uring: never overflow io_aux_cqe Pavel Begunkov
2023-08-11 12:53 ` Pavel Begunkov [this message]
2023-08-11 12:53 ` [PATCH 7/7] io_uring: simplify io_run_task_work_sig return Pavel Begunkov
2023-08-11 16:43 ` [PATCH 0/7] random fixes and cleanups 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=e4a96dda35ab755914bc43f6781bba0df97ac489.1691757663.git.asml.silence@gmail.com \
    [email protected] \
    [email protected] \
    [email protected] \
    /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