From: Pavel Begunkov <asml.silence@gmail.com>
To: Jens Axboe <axboe@kernel.dk>,
io-uring@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: [PATCH 4/5] io_uring: keep all sqe->flags in req->flags
Date: Sun, 12 Apr 2020 02:05:04 +0300 [thread overview]
Message-ID: <92cf6b32f856ea65c99948ff1670874987f20241.1586645520.git.asml.silence@gmail.com> (raw)
In-Reply-To: <cover.1586645520.git.asml.silence@gmail.com>
It's a good idea to not read sqe->flags twice, as it's prone to security
bugs. Instead of passing it around, embeed them in req->flags. It's
already so except for IOSQE_IO_LINK.
1. rename former REQ_F_LINK -> REQ_F_LINK_HEAD
2. introduce and copy REQ_F_LINK, which mimics IO_IOSQE_LINK
And leave req_set_fail_links() using new REQ_F_LINK, because it's more
sensible.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
fs/io_uring.c | 23 +++++++++++++----------
1 file changed, 13 insertions(+), 10 deletions(-)
diff --git a/fs/io_uring.c b/fs/io_uring.c
index 90b806e4022a..9118a0210e0a 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -506,6 +506,7 @@ enum {
REQ_F_FORCE_ASYNC_BIT = IOSQE_ASYNC_BIT,
REQ_F_BUFFER_SELECT_BIT = IOSQE_BUFFER_SELECT_BIT,
+ REQ_F_LINK_HEAD_BIT,
REQ_F_LINK_NEXT_BIT,
REQ_F_FAIL_LINK_BIT,
REQ_F_INFLIGHT_BIT,
@@ -541,6 +542,8 @@ enum {
/* IOSQE_BUFFER_SELECT */
REQ_F_BUFFER_SELECT = BIT(REQ_F_BUFFER_SELECT_BIT),
+ /* head of a link */
+ REQ_F_LINK_HEAD = BIT(REQ_F_LINK_HEAD_BIT),
/* already grabbed next link */
REQ_F_LINK_NEXT = BIT(REQ_F_LINK_NEXT_BIT),
/* fail rest of links */
@@ -1435,7 +1438,7 @@ static bool io_link_cancel_timeout(struct io_kiocb *req)
if (ret != -1) {
io_cqring_fill_event(req, -ECANCELED);
io_commit_cqring(ctx);
- req->flags &= ~REQ_F_LINK;
+ req->flags &= ~REQ_F_LINK_HEAD;
io_put_req(req);
return true;
}
@@ -1471,7 +1474,7 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr)
list_del_init(&req->link_list);
if (!list_empty(&nxt->link_list))
- nxt->flags |= REQ_F_LINK;
+ nxt->flags |= REQ_F_LINK_HEAD;
*nxtptr = nxt;
break;
}
@@ -1482,7 +1485,7 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr)
}
/*
- * Called if REQ_F_LINK is set, and we fail the head request
+ * Called if REQ_F_LINK_HEAD is set, and we fail the head request
*/
static void io_fail_links(struct io_kiocb *req)
{
@@ -1515,7 +1518,7 @@ static void io_fail_links(struct io_kiocb *req)
static void io_req_find_next(struct io_kiocb *req, struct io_kiocb **nxt)
{
- if (likely(!(req->flags & REQ_F_LINK)))
+ if (likely(!(req->flags & REQ_F_LINK_HEAD)))
return;
/*
@@ -1667,7 +1670,7 @@ static inline unsigned int io_sqring_entries(struct io_ring_ctx *ctx)
static inline bool io_req_multi_free(struct req_batch *rb, struct io_kiocb *req)
{
- if ((req->flags & REQ_F_LINK) || io_is_fallback_req(req))
+ if ((req->flags & REQ_F_LINK_HEAD) || io_is_fallback_req(req))
return false;
if (!(req->flags & REQ_F_FIXED_FILE) || req->io)
@@ -2560,7 +2563,7 @@ static int io_read(struct io_kiocb *req, bool force_nonblock)
req->result = 0;
io_size = ret;
- if (req->flags & REQ_F_LINK)
+ if (req->flags & REQ_F_LINK_HEAD)
req->result = io_size;
/*
@@ -2651,7 +2654,7 @@ static int io_write(struct io_kiocb *req, bool force_nonblock)
req->result = 0;
io_size = ret;
- if (req->flags & REQ_F_LINK)
+ if (req->flags & REQ_F_LINK_HEAD)
req->result = io_size;
/*
@@ -5474,7 +5477,7 @@ static struct io_kiocb *io_prep_linked_timeout(struct io_kiocb *req)
{
struct io_kiocb *nxt;
- if (!(req->flags & REQ_F_LINK))
+ if (!(req->flags & REQ_F_LINK_HEAD))
return NULL;
/* for polled retry, if flag is set, we already went through here */
if (req->flags & REQ_F_POLLED)
@@ -5634,7 +5637,7 @@ static int io_submit_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe,
/* same numerical values with corresponding REQ_F_*, safe to copy */
req->flags |= sqe_flags & (IOSQE_IO_DRAIN | IOSQE_IO_HARDLINK |
IOSQE_ASYNC | IOSQE_FIXED_FILE |
- IOSQE_BUFFER_SELECT);
+ IOSQE_BUFFER_SELECT | IOSQE_IO_LINK);
fd = READ_ONCE(sqe->fd);
ret = io_req_set_file(state, req, fd, sqe_flags);
@@ -5685,7 +5688,7 @@ static int io_submit_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe,
req->ctx->drain_next = 0;
}
if (sqe_flags & (IOSQE_IO_LINK|IOSQE_IO_HARDLINK)) {
- req->flags |= REQ_F_LINK;
+ req->flags |= REQ_F_LINK_HEAD;
INIT_LIST_HEAD(&req->link_list);
if (io_alloc_async_ctx(req))
--
2.24.0
next prev parent reply other threads:[~2020-04-11 23:43 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-04-11 23:05 [PATCH 0/5] submission path refactoring pt.2 Pavel Begunkov
2020-04-11 23:05 ` [PATCH 1/5] io_uring: remove obsolete @mm_fault Pavel Begunkov
2020-04-11 23:05 ` [PATCH 2/5] io_uring: track mm through current->mm Pavel Begunkov
2020-04-11 23:05 ` [PATCH 3/5] io_uring: DRY early submission req fail code Pavel Begunkov
2020-04-11 23:05 ` Pavel Begunkov [this message]
2020-04-11 23:05 ` [PATCH 5/5] io_uring: move all request init code in one place Pavel Begunkov
2020-04-12 15:23 ` [PATCH 0/5] submission path refactoring pt.2 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=92cf6b32f856ea65c99948ff1670874987f20241.1586645520.git.asml.silence@gmail.com \
--to=asml.silence@gmail.com \
--cc=axboe@kernel.dk \
--cc=io-uring@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
/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).