forbid using open/close/statx with {SQ,IO}POLL
cover forbid using open/close/statx with {SQ,IO}POLL
The first one adds checks {SQPOLL,IOPOLL} for open, close, etc. And 3 others are just cleanups on top.
notes:
- it cures symptoms, but would be great what's with null-deref in [1/4]
- need to look whether epoll and sfr are affected by {SQ,IO}POLL
- it still allows statx with SQPOLL, because it has a bunchg of rules for ignoring dirfd, e.g. when absolute path is specified. Let's handle it separately.
Pavel Begunkov (4):
io_uring: fix open/close/statx with {SQ,IO}POLL
io_uring: do build_open_how() only once
io_uring: deduplicate io_openat{,2}_prep()
io_uring: move send/recv IOPOLL check into prep
fs/io_uring.c | 78 +++++++++++++++++++++------------------------------ 1 file changed, 32 insertions(+), 46 deletions(-)
Signature · 1 line
2.24.0
PATCH 1/4 io_uring: fix open/close/statx with {SQ,IO}POLL
Trying to use them with IORING_SETUP_IOPOLL:
RIP: 0010:io_iopoll_getevents+0x111/0x5a0 Call Trace: ? _raw_spin_unlock_irqrestore+0x24/0x40 ? do_send_sig_info+0x64/0x90 io_iopoll_reap_events.part.0+0x5e/0xa0 io_ring_ctx_wait_and_kill+0x132/0x1c0 io_uring_release+0x20/0x30 __fput+0xcd/0x230 ____fput+0xe/0x10 task_work_run+0x67/0xa0 do_exit+0x353/0xb10 ? handle_mm_fault+0xd4/0x200 ? syscall_trace_enter+0x18c/0x2c0 do_group_exit+0x43/0xa0 __x64_sys_exit_group+0x18/0x20 do_syscall_64+0x60/0x1e0 entry_SYSCALL_64_after_hwframe+0x44/0xa9
Also SQPOLL thread can't know which file table to use with open/close. Disallow all these cases.
--- fs/io_uring.c | 8 ++++++++ 1 file changed, 8 insertions(+)
fs/io_uring.c +8 -0
@@ -2990,6 +2990,8 @@ static int io_openat_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)const char __user *fname;int ret; +if (unlikely (req->ctx->flags & (IORING_SETUP_IOPOLL |IORING_SETUP_SQPOLL ))) +return -EINVAL ;if (sqe->ioprio || sqe->buf_index)return -EINVAL ;if (req->flags &REQ_F_FIXED_FILE )
@@ -3023,6 +3025,8 @@ static int io_openat2_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)size_t len;int ret; +if (unlikely (req->ctx->flags & (IORING_SETUP_IOPOLL |IORING_SETUP_SQPOLL ))) +return -EINVAL ;if (sqe->ioprio || sqe->buf_index)return -EINVAL ;if (req->flags &REQ_F_FIXED_FILE )
@@ -3373,6 +3377,8 @@ static int io_fadvise(struct io_kiocb *req, bool force_nonblock)static int io_statx_prep (struct io_kiocb *req,const struct io_uring_sqe *sqe) { +if (unlikely (req->ctx->flags &IORING_SETUP_IOPOLL )) +return -EINVAL ;if (sqe->ioprio || sqe->buf_index)return -EINVAL ;if (req->flags &REQ_F_FIXED_FILE )
@@ -3417,6 +3423,8 @@ static int io_close_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)*/ req->work.flags |=IO_WQ_WORK_NO_CANCEL ; +if (unlikely (req->ctx->flags & (IORING_SETUP_IOPOLL |IORING_SETUP_SQPOLL ))) +return -EINVAL ;if (sqe->ioprio || sqe->off || sqe->addr || sqe->len || sqe->rw_flags || sqe->buf_index)return -EINVAL ;
Signature · 1 line
2.24.0
Re: [PATCH 1/4] io_uring: fix open/close/statx with {SQ,IO}POLL
Quoted 11 lines · Pavel Begunkov
Trying to use them with
IORING_SETUP_IOPOLL:RIP: 0010:io_iopoll_getevents+0x111/0x5a0 Call Trace: ? _raw_spin_unlock_irqrestore+0x24/0x40 ? do_send_sig_info+0x64/0x90 io_iopoll_reap_events.part.0+0x5e/0xa0 io_ring_ctx_wait_and_kill+0x132/0x1c0 io_uring_release+0x20/0x30 __fput+0xcd/0x230 ____fput+0xe/0x10
Quoted 8 lines
task_work_run+0x67/0xa0 do_exit+0x353/0xb10 ? handle_mm_fault+0xd4/0x200 ? syscall_trace_enter+0x18c/0x2c0 do_group_exit+0x43/0xa0 __x64_sys_exit_group+0x18/0x20 do_syscall_64+0x60/0x1e0 entry_SYSCALL_64_after_hwframe+0x44/0xa9
io_do_iopoll()
{
...
ret = kiocb->*ki_filp*->f_op->iopoll(kiocb, spin);
}Hmm, I'll double check later that only read*/write* can be done with IOPOLL, and send a follow-up patch if necessary.
Quoted 50 lines
Also SQPOLL thread can't know which file table to use with open/close. Disallow all these cases.
Signed-off-by Pavel Begunkov <[email protected]>--- fs/io_uring.c | 8 ++++++++ 1 file changed, 8 insertions(+)fs/io_uring.c +8
-0@@ -2990,6 +2990,8 @@ static int io_openat_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)const char __user *fname;int ret; +if (unlikely (req->ctx->flags & (IORING_SETUP_IOPOLL |IORING_SETUP_SQPOLL ))) +return -EINVAL ;if (sqe->ioprio || sqe->buf_index)return -EINVAL ;if (req->flags &REQ_F_FIXED_FILE )@@ -3023,6 +3025,8 @@ static int io_openat2_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)size_t len;int ret; +if (unlikely (req->ctx->flags & (IORING_SETUP_IOPOLL |IORING_SETUP_SQPOLL ))) +return -EINVAL ;if (sqe->ioprio || sqe->buf_index)return -EINVAL ;if (req->flags &REQ_F_FIXED_FILE )@@ -3373,6 +3377,8 @@ static int io_fadvise(struct io_kiocb *req, bool force_nonblock)static int io_statx_prep (struct io_kiocb *req,const struct io_uring_sqe *sqe) { +if (unlikely (req->ctx->flags &IORING_SETUP_IOPOLL )) +return -EINVAL ;if (sqe->ioprio || sqe->buf_index)return -EINVAL ;if (req->flags &REQ_F_FIXED_FILE )@@ -3417,6 +3423,8 @@ static int io_close_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)*/ req->work.flags |=IO_WQ_WORK_NO_CANCEL ; +if (unlikely (req->ctx->flags & (IORING_SETUP_IOPOLL |IORING_SETUP_SQPOLL ))) +return -EINVAL ;if (sqe->ioprio || sqe->off || sqe->addr || sqe->len || sqe->rw_flags || sqe->buf_index)return -EINVAL ;
Signature · 1 line
Pavel Begunkov
PATCH 2/4 io_uring: do build_open_how() only once
In reply to Pavel Begunkovbuild_open_how() is just adjusting open_flags/mode. Do it once during prep. It looks better than storing raw values for the future.
--- fs/io_uring.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-)
fs/io_uring.c +7 -6
@@ -2988,6 +2988,7 @@ static int io_fallocate(struct io_kiocb *req, bool force_nonblock)static int io_openat_prep (struct io_kiocb *req,const struct io_uring_sqe *sqe) {const char __user *fname; +u64 flags, mode;int ret;if (unlikely (req->ctx->flags & (IORING_SETUP_IOPOLL |IORING_SETUP_SQPOLL )))
@@ -2999,13 +3000,14 @@ static int io_openat_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)if (req->flags &REQ_F_NEED_CLEANUP )return 0 ; - req->open.dfd =READ_ONCE (sqe->fd); - req->open.how.mode =READ_ONCE (sqe->len); - fname =u64_to_user_ptr (READ_ONCE (sqe->addr)); - req->open.how.flags =READ_ONCE (sqe->open_flags); + mode =READ_ONCE (sqe->len); + flags =READ_ONCE (sqe->open_flags);if (force_o_largefile ()) - req->open.how.flags |=O_LARGEFILE ; + flags |=O_LARGEFILE ; + req->open.how =build_open_how (flags, mode); + req->open.dfd =READ_ONCE (sqe->fd); + fname =u64_to_user_ptr (READ_ONCE (sqe->addr)); req->open.filename =getname (fname);if (IS_ERR (req->open.filename)) { ret =PTR_ERR (req->open.filename);
@@ -3099,7 +3101,6 @@ static int io_openat2(struct io_kiocb *req, bool force_nonblock)static int io_openat (struct io_kiocb *req,bool force_nonblock) { - req->open.how =build_open_how (req->open.how.flags, req->open.how.mode);return io_openat2 (req, force_nonblock); }
Signature · 1 line
2.24.0
PATCH 3/4 io_uring: deduplicate io_openat{,2}_prep()
In reply to Pavel Begunkovio_openat_prep() and io_openat2_prep() are identical except for how struct open_how is built. Deduplicate it with a helper.
--- fs/io_uring.c | 55 ++++++++++++++++++--------------------------------- 1 file changed, 19 insertions(+), 36 deletions(-)
fs/io_uring.c +19 -36
@@ -2985,26 +2985,21 @@ static int io_fallocate(struct io_kiocb *req, bool force_nonblock)return 0 ; } -static int io_openat_prep (struct io_kiocb *req,const struct io_uring_sqe *sqe) +static int __io_openat_prep (struct io_kiocb *req,const struct io_uring_sqe *sqe) {const char __user *fname; -u64 flags, mode;int ret;if (unlikely (req->ctx->flags & (IORING_SETUP_IOPOLL |IORING_SETUP_SQPOLL )))return -EINVAL ; -if (sqe->ioprio || sqe->buf_index) +if (unlikely (sqe->ioprio || sqe->buf_index))return -EINVAL ; -if (req->flags &REQ_F_FIXED_FILE ) +if (unlikely (req->flags &REQ_F_FIXED_FILE ))return -EBADF ; -if (req->flags &REQ_F_NEED_CLEANUP ) -return 0 ; - mode =READ_ONCE (sqe->len); - flags =READ_ONCE (sqe->open_flags); -if (force_o_largefile ()) - flags |=O_LARGEFILE ; - req->open.how =build_open_how (flags, mode); +/* open.how should be already initialised */ +if (!(req->open.how.flags &O_PATH ) &&force_o_largefile ()) + req->open.how.flags |=O_LARGEFILE ; req->open.dfd =READ_ONCE (sqe->fd); fname =u64_to_user_ptr (READ_ONCE (sqe->addr));
@@ -3014,33 +3009,33 @@ static int io_openat_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) req->open.filename =NULL ;return ret; } - req->open.nofile =rlimit (RLIMIT_NOFILE ); req->flags |=REQ_F_NEED_CLEANUP ;return 0 ; } +static int io_openat_prep (struct io_kiocb *req,const struct io_uring_sqe *sqe) +{ +u64 flags, mode; + +if (req->flags &REQ_F_NEED_CLEANUP ) +return 0 ; + mode =READ_ONCE (sqe->len); + flags =READ_ONCE (sqe->open_flags); + req->open.how =build_open_how (flags, mode); +return __io_openat_prep (req, sqe); +} +static int io_openat2_prep (struct io_kiocb *req,const struct io_uring_sqe *sqe) {struct open_how __user *how; -const char __user *fname;size_t len;int ret; -if (unlikely (req->ctx->flags & (IORING_SETUP_IOPOLL |IORING_SETUP_SQPOLL ))) -return -EINVAL ; -if (sqe->ioprio || sqe->buf_index) -return -EINVAL ; -if (req->flags &REQ_F_FIXED_FILE ) -return -EBADF ;if (req->flags &REQ_F_NEED_CLEANUP )return 0 ; - - req->open.dfd =READ_ONCE (sqe->fd); - fname =u64_to_user_ptr (READ_ONCE (sqe->addr)); how =u64_to_user_ptr (READ_ONCE (sqe->addr2)); len =READ_ONCE (sqe->len); -if (len <OPEN_HOW_SIZE_VER0 )return -EINVAL ;
@@ -3049,19 +3044,7 @@ static int io_openat2_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)if (ret)return ret; -if (!(req->open.how.flags &O_PATH ) &&force_o_largefile ()) - req->open.how.flags |=O_LARGEFILE ; - - req->open.filename =getname (fname); -if (IS_ERR (req->open.filename)) { - ret =PTR_ERR (req->open.filename); - req->open.filename =NULL ; -return ret; - } - - req->open.nofile =rlimit (RLIMIT_NOFILE ); - req->flags |=REQ_F_NEED_CLEANUP ; -return 0 ; +return __io_openat_prep (req, sqe); }static int io_openat2 (struct io_kiocb *req,bool force_nonblock)
Signature · 1 line
2.24.0
PATCH 4/4 io_uring: move send/recv IOPOLL check into prep
In reply to Pavel BegunkovFail recv/send in case of IORING_SETUP_IOPOLL earlier during prep, so it'd be done only once. Removes duplication as well
--- fs/io_uring.c | 18 ++++++------------ 1 file changed, 6 insertions(+), 12 deletions(-)
fs/io_uring.c +6 -12
@@ -3541,6 +3541,9 @@ static int io_sendmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)struct io_async_ctx *io = req->io;int ret; +if (unlikely (req->ctx->flags &IORING_SETUP_IOPOLL )) +return -EINVAL ; + sr->msg_flags =READ_ONCE (sqe->msg_flags); sr->msg =u64_to_user_ptr (READ_ONCE (sqe->addr)); sr->len =READ_ONCE (sqe->len);
@@ -3570,9 +3573,6 @@ static int io_sendmsg(struct io_kiocb *req, bool force_nonblock)struct socket *sock;int ret; -if (unlikely (req->ctx->flags &IORING_SETUP_IOPOLL )) -return -EINVAL ; - sock =sock_from_file (req->file, &ret);if (sock) {struct io_async_ctx io;
@@ -3626,9 +3626,6 @@ static int io_send(struct io_kiocb *req, bool force_nonblock)struct socket *sock;int ret; -if (unlikely (req->ctx->flags &IORING_SETUP_IOPOLL )) -return -EINVAL ; - sock =sock_from_file (req->file, &ret);if (sock) {struct io_sr_msg *sr = &req->sr_msg;
@@ -3781,6 +3778,9 @@ static int io_recvmsg_prep(struct io_kiocb *req,struct io_async_ctx *io = req->io;int ret; +if (unlikely (req->ctx->flags &IORING_SETUP_IOPOLL )) +return -EINVAL ; + sr->msg_flags =READ_ONCE (sqe->msg_flags); sr->msg =u64_to_user_ptr (READ_ONCE (sqe->addr)); sr->len =READ_ONCE (sqe->len);
@@ -3809,9 +3809,6 @@ static int io_recvmsg(struct io_kiocb *req, bool force_nonblock)struct socket *sock;int ret, cflags =0 ; -if (unlikely (req->ctx->flags &IORING_SETUP_IOPOLL )) -return -EINVAL ; - sock =sock_from_file (req->file, &ret);if (sock) {struct io_buffer *kbuf;
@@ -3873,9 +3870,6 @@ static int io_recv(struct io_kiocb *req, bool force_nonblock)struct socket *sock;int ret, cflags =0 ; -if (unlikely (req->ctx->flags &IORING_SETUP_IOPOLL )) -return -EINVAL ; - sock =sock_from_file (req->file, &ret);if (sock) {struct io_sr_msg *sr = &req->sr_msg;
Signature · 1 line
2.24.0