From b948f9d5d3057b01188e36664e7c7604d1c8ecb5 Mon Sep 17 00:00:00 2001 From: Junxi Qian Date: Sun, 29 Mar 2026 23:39:09 +0800 Subject: io_uring/net: fix slab-out-of-bounds read in io_bundle_nbufs() sqe->len is __u32 but gets stored into sr->len which is int. When userspace passes sqe->len values exceeding INT_MAX (e.g. 0xFFFFFFFF), sr->len overflows to a negative value. This negative value propagates through the bundle recv/send path: 1. io_recv(): sel.val = sr->len (ssize_t gets -1) 2. io_recv_buf_select(): arg.max_len = sel->val (size_t gets 0xFFFFFFFFFFFFFFFF) 3. io_ring_buffers_peek(): buf->len is not clamped because max_len is astronomically large 4. iov[].iov_len = 0xFFFFFFFF flows into io_bundle_nbufs() 5. io_bundle_nbufs(): min_t(int, 0xFFFFFFFF, ret) yields -1, causing ret to increase instead of decrease, creating an infinite loop that reads past the allocated iov[] array This results in a slab-out-of-bounds read in io_bundle_nbufs() from the kmalloc-64 slab, as nbufs increments past the allocated iovec entries. BUG: KASAN: slab-out-of-bounds in io_bundle_nbufs+0x128/0x160 Read of size 8 at addr ffff888100ae05c8 by task exp/145 Call Trace: io_bundle_nbufs+0x128/0x160 io_recv_finish+0x117/0xe20 io_recv+0x2db/0x1160 Fix this by rejecting negative sr->len values early in both io_sendmsg_prep() and io_recvmsg_prep(). Since sqe->len is __u32, any value > INT_MAX indicates overflow and is not a valid length. Fixes: a05d1f625c7a ("io_uring/net: support bundles for send") Cc: stable@vger.kernel.org Signed-off-by: Junxi Qian Link: https://patch.msgid.link/20260329153909.279046-1-qjx1298677004@gmail.com Signed-off-by: Jens Axboe --- io_uring/net.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/io_uring/net.c b/io_uring/net.c index d27adbe3f20b..8885d944130a 100644 --- a/io_uring/net.c +++ b/io_uring/net.c @@ -421,6 +421,8 @@ int io_sendmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) sr->done_io = 0; sr->len = READ_ONCE(sqe->len); + if (unlikely(sr->len < 0)) + return -EINVAL; sr->flags = READ_ONCE(sqe->ioprio); if (sr->flags & ~SENDMSG_FLAGS) return -EINVAL; @@ -791,6 +793,8 @@ int io_recvmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) sr->umsg = u64_to_user_ptr(READ_ONCE(sqe->addr)); sr->len = READ_ONCE(sqe->len); + if (unlikely(sr->len < 0)) + return -EINVAL; sr->flags = READ_ONCE(sqe->ioprio); if (sr->flags & ~RECVMSG_FLAGS) return -EINVAL; -- cgit v1.2.3 From 111a12b422a8cfa93deabaef26fec48237163214 Mon Sep 17 00:00:00 2001 From: Qi Tang Date: Mon, 30 Mar 2026 00:49:36 +0800 Subject: io_uring/rsrc: reject zero-length fixed buffer import validate_fixed_range() admits buf_addr at the exact end of the registered region when len is zero, because the check uses strict greater-than (buf_end > imu->ubuf + imu->len). io_import_fixed() then computes offset == imu->len, which causes the bvec skip logic to advance past the last bio_vec entry and read bv_offset from out-of-bounds slab memory. Return early from io_import_fixed() when len is zero. A zero-length import has no data to transfer and should not walk the bvec array at all. BUG: KASAN: slab-out-of-bounds in io_import_reg_buf+0x697/0x7f0 Read of size 4 at addr ffff888002bcc254 by task poc/103 Call Trace: io_import_reg_buf+0x697/0x7f0 io_write_fixed+0xd9/0x250 __io_issue_sqe+0xad/0x710 io_issue_sqe+0x7d/0x1100 io_submit_sqes+0x86a/0x23c0 __do_sys_io_uring_enter+0xa98/0x1590 Allocated by task 103: The buggy address is located 12 bytes to the right of allocated 584-byte region [ffff888002bcc000, ffff888002bcc248) Fixes: 8622b20f23ed ("io_uring: add validate_fixed_range() for validate fixed buffer") Signed-off-by: Qi Tang Link: https://patch.msgid.link/20260329164936.240871-1-tpluszz77@gmail.com Signed-off-by: Jens Axboe --- io_uring/rsrc.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/io_uring/rsrc.c b/io_uring/rsrc.c index 842e231c8a7c..3f82ed7dd11e 100644 --- a/io_uring/rsrc.c +++ b/io_uring/rsrc.c @@ -1062,6 +1062,10 @@ static int io_import_fixed(int ddir, struct iov_iter *iter, return ret; if (!(imu->dir & (1 << ddir))) return -EFAULT; + if (unlikely(!len)) { + iov_iter_bvec(iter, ddir, NULL, 0, 0); + return 0; + } offset = buf_addr - imu->ubuf; -- cgit v1.2.3 From 61a11cf4812726aceaee17c96432e1c08f6ed6cb Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 31 Mar 2026 07:07:47 -0600 Subject: io_uring: protect remaining lockless ctx->rings accesses with RCU Commit 96189080265e addressed one case of ctx->rings being potentially accessed while a resize is happening on the ring, but there are still a few others that need handling. Add a helper for retrieving the rings associated with an io_uring context, and add some sanity checking to that to catch bad uses. ->rings_rcu is always valid, as long as it's used within RCU read lock. Any use of ->rings_rcu or ->rings inside either ->uring_lock or ->completion_lock is sane as well. Do the minimum fix for the current kernel, but set it up such that this basic infra can be extended for later kernels to make this harder to mess up in the future. Thanks to Junxi Qian for finding and debugging this issue. Cc: stable@vger.kernel.org Fixes: 79cfe9e59c2a ("io_uring/register: add IORING_REGISTER_RESIZE_RINGS") Reviewed-by: Junxi Qian Tested-by: Junxi Qian Link: https://lore.kernel.org/io-uring/20260330172348.89416-1-qjx1298677004@gmail.com/ Signed-off-by: Jens Axboe --- io_uring/io_uring.c | 7 +++++-- io_uring/io_uring.h | 34 +++++++++++++++++++++++++++++----- io_uring/wait.c | 50 +++++++++++++++++++++++++++++++------------------- io_uring/wait.h | 7 +++++-- 4 files changed, 70 insertions(+), 28 deletions(-) diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c index 20ec8fdafcae..48f2f627319d 100644 --- a/io_uring/io_uring.c +++ b/io_uring/io_uring.c @@ -2015,7 +2015,7 @@ int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr) if (ctx->flags & IORING_SETUP_SQ_REWIND) entries = ctx->sq_entries; else - entries = io_sqring_entries(ctx); + entries = __io_sqring_entries(ctx); entries = min(nr, entries); if (unlikely(!entries)) @@ -2250,7 +2250,9 @@ static __poll_t io_uring_poll(struct file *file, poll_table *wait) */ poll_wait(file, &ctx->poll_wq, wait); - if (!io_sqring_full(ctx)) + rcu_read_lock(); + + if (!__io_sqring_full(ctx)) mask |= EPOLLOUT | EPOLLWRNORM; /* @@ -2270,6 +2272,7 @@ static __poll_t io_uring_poll(struct file *file, poll_table *wait) if (__io_cqring_events_user(ctx) || io_has_work(ctx)) mask |= EPOLLIN | EPOLLRDNORM; + rcu_read_unlock(); return mask; } diff --git a/io_uring/io_uring.h b/io_uring/io_uring.h index 0fa844faf287..ee24bc5d77b3 100644 --- a/io_uring/io_uring.h +++ b/io_uring/io_uring.h @@ -142,16 +142,28 @@ struct io_wait_queue { #endif }; +static inline struct io_rings *io_get_rings(struct io_ring_ctx *ctx) +{ + return rcu_dereference_check(ctx->rings_rcu, + lockdep_is_held(&ctx->uring_lock) || + lockdep_is_held(&ctx->completion_lock)); +} + static inline bool io_should_wake(struct io_wait_queue *iowq) { struct io_ring_ctx *ctx = iowq->ctx; - int dist = READ_ONCE(ctx->rings->cq.tail) - (int) iowq->cq_tail; + struct io_rings *rings; + int dist; + + guard(rcu)(); + rings = io_get_rings(ctx); /* * Wake up if we have enough events, or if a timeout occurred since we * started waiting. For timeouts, we always want to return to userspace, * regardless of event count. */ + dist = READ_ONCE(rings->cq.tail) - (int) iowq->cq_tail; return dist >= 0 || atomic_read(&ctx->cq_timeouts) != iowq->nr_timeouts; } @@ -431,9 +443,9 @@ static inline void io_cqring_wake(struct io_ring_ctx *ctx) __io_wq_wake(&ctx->cq_wait); } -static inline bool io_sqring_full(struct io_ring_ctx *ctx) +static inline bool __io_sqring_full(struct io_ring_ctx *ctx) { - struct io_rings *r = ctx->rings; + struct io_rings *r = io_get_rings(ctx); /* * SQPOLL must use the actual sqring head, as using the cached_sq_head @@ -445,9 +457,15 @@ static inline bool io_sqring_full(struct io_ring_ctx *ctx) return READ_ONCE(r->sq.tail) - READ_ONCE(r->sq.head) == ctx->sq_entries; } -static inline unsigned int io_sqring_entries(struct io_ring_ctx *ctx) +static inline bool io_sqring_full(struct io_ring_ctx *ctx) { - struct io_rings *rings = ctx->rings; + guard(rcu)(); + return __io_sqring_full(ctx); +} + +static inline unsigned int __io_sqring_entries(struct io_ring_ctx *ctx) +{ + struct io_rings *rings = io_get_rings(ctx); unsigned int entries; /* make sure SQ entry isn't read before tail */ @@ -455,6 +473,12 @@ static inline unsigned int io_sqring_entries(struct io_ring_ctx *ctx) return min(entries, ctx->sq_entries); } +static inline unsigned int io_sqring_entries(struct io_ring_ctx *ctx) +{ + guard(rcu)(); + return __io_sqring_entries(ctx); +} + /* * Don't complete immediately but use deferred completion infrastructure. * Protected by ->uring_lock and can only be used either with diff --git a/io_uring/wait.c b/io_uring/wait.c index 0581cadf20ee..91df86ce0d18 100644 --- a/io_uring/wait.c +++ b/io_uring/wait.c @@ -79,12 +79,15 @@ static enum hrtimer_restart io_cqring_min_timer_wakeup(struct hrtimer *timer) if (io_has_work(ctx)) goto out_wake; /* got events since we started waiting, min timeout is done */ - if (iowq->cq_min_tail != READ_ONCE(ctx->rings->cq.tail)) - goto out_wake; - /* if we have any events and min timeout expired, we're done */ - if (io_cqring_events(ctx)) - goto out_wake; + scoped_guard(rcu) { + struct io_rings *rings = io_get_rings(ctx); + if (iowq->cq_min_tail != READ_ONCE(rings->cq.tail)) + goto out_wake; + /* if we have any events and min timeout expired, we're done */ + if (io_cqring_events(ctx)) + goto out_wake; + } /* * If using deferred task_work running and application is waiting on * more than one request, ensure we reset it now where we are switching @@ -186,9 +189,9 @@ int io_cqring_wait(struct io_ring_ctx *ctx, int min_events, u32 flags, struct ext_arg *ext_arg) { struct io_wait_queue iowq; - struct io_rings *rings = ctx->rings; + struct io_rings *rings; ktime_t start_time; - int ret; + int ret, nr_wait; min_events = min_t(int, min_events, ctx->cq_entries); @@ -201,15 +204,23 @@ int io_cqring_wait(struct io_ring_ctx *ctx, int min_events, u32 flags, if (unlikely(test_bit(IO_CHECK_CQ_OVERFLOW_BIT, &ctx->check_cq))) io_cqring_do_overflow_flush(ctx); - if (__io_cqring_events_user(ctx) >= min_events) + + rcu_read_lock(); + rings = io_get_rings(ctx); + if (__io_cqring_events_user(ctx) >= min_events) { + rcu_read_unlock(); return 0; + } init_waitqueue_func_entry(&iowq.wq, io_wake_function); iowq.wq.private = current; INIT_LIST_HEAD(&iowq.wq.entry); iowq.ctx = ctx; - iowq.cq_tail = READ_ONCE(ctx->rings->cq.head) + min_events; - iowq.cq_min_tail = READ_ONCE(ctx->rings->cq.tail); + iowq.cq_tail = READ_ONCE(rings->cq.head) + min_events; + iowq.cq_min_tail = READ_ONCE(rings->cq.tail); + nr_wait = (int) iowq.cq_tail - READ_ONCE(rings->cq.tail); + rcu_read_unlock(); + rings = NULL; iowq.nr_timeouts = atomic_read(&ctx->cq_timeouts); iowq.hit_timeout = 0; iowq.min_timeout = ext_arg->min_time; @@ -240,14 +251,6 @@ int io_cqring_wait(struct io_ring_ctx *ctx, int min_events, u32 flags, trace_io_uring_cqring_wait(ctx, min_events); do { unsigned long check_cq; - int nr_wait; - - /* if min timeout has been hit, don't reset wait count */ - if (!iowq.hit_timeout) - nr_wait = (int) iowq.cq_tail - - READ_ONCE(ctx->rings->cq.tail); - else - nr_wait = 1; if (ctx->flags & IORING_SETUP_DEFER_TASKRUN) { atomic_set(&ctx->cq_wait_nr, nr_wait); @@ -298,11 +301,20 @@ int io_cqring_wait(struct io_ring_ctx *ctx, int min_events, u32 flags, break; } cond_resched(); + + /* if min timeout has been hit, don't reset wait count */ + if (!iowq.hit_timeout) + scoped_guard(rcu) + nr_wait = (int) iowq.cq_tail - + READ_ONCE(io_get_rings(ctx)->cq.tail); + else + nr_wait = 1; } while (1); if (!(ctx->flags & IORING_SETUP_DEFER_TASKRUN)) finish_wait(&ctx->cq_wait, &iowq.wq); restore_saved_sigmask_unless(ret == -EINTR); - return READ_ONCE(rings->cq.head) == READ_ONCE(rings->cq.tail) ? ret : 0; + guard(rcu)(); + return READ_ONCE(io_get_rings(ctx)->cq.head) == READ_ONCE(io_get_rings(ctx)->cq.tail) ? ret : 0; } diff --git a/io_uring/wait.h b/io_uring/wait.h index 5e236f74e1af..3a145fcfd3dd 100644 --- a/io_uring/wait.h +++ b/io_uring/wait.h @@ -28,12 +28,15 @@ void io_cqring_do_overflow_flush(struct io_ring_ctx *ctx); static inline unsigned int __io_cqring_events(struct io_ring_ctx *ctx) { - return ctx->cached_cq_tail - READ_ONCE(ctx->rings->cq.head); + struct io_rings *rings = io_get_rings(ctx); + return ctx->cached_cq_tail - READ_ONCE(rings->cq.head); } static inline unsigned int __io_cqring_events_user(struct io_ring_ctx *ctx) { - return READ_ONCE(ctx->rings->cq.tail) - READ_ONCE(ctx->rings->cq.head); + struct io_rings *rings = io_get_rings(ctx); + + return READ_ONCE(rings->cq.tail) - READ_ONCE(rings->cq.head); } /* -- cgit v1.2.3 From aa35dd6bdd033dea8aa3e20cbbbe10e06b2d044f Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 31 Mar 2026 08:16:58 -0600 Subject: io_uring/bpf_filters: retain COW'ed settings on parse failures If io_parse_restrictions() fails, it ends up clearing any restrictions currently set. The intent is only to clear whatever it already applied, but it ends up clearing everything, including whatever settings may have been applied in a copy-on-write fashion already. Ensure that those are retained. Link: https://lore.kernel.org/io-uring/CAK8a0jzF-zaO5ZmdOrmfuxrhXuKg5m5+RDuO7tNvtj=kUYbW7Q@mail.gmail.com/ Reported-by: antonius Fixes: ed82f35b926b ("io_uring: allow registration of per-task restrictions") Signed-off-by: Jens Axboe --- io_uring/register.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/io_uring/register.c b/io_uring/register.c index 5f3eb018fb32..837324bf0223 100644 --- a/io_uring/register.c +++ b/io_uring/register.c @@ -178,9 +178,17 @@ static __cold int io_register_restrictions(struct io_ring_ctx *ctx, return -EBUSY; ret = io_parse_restrictions(arg, nr_args, &ctx->restrictions); - /* Reset all restrictions if an error happened */ + /* + * Reset all restrictions if an error happened, but retain any COW'ed + * settings. + */ if (ret < 0) { + struct io_bpf_filters *bpf = ctx->restrictions.bpf_filters; + bool cowed = ctx->restrictions.bpf_filters_cow; + memset(&ctx->restrictions, 0, sizeof(ctx->restrictions)); + ctx->restrictions.bpf_filters = bpf; + ctx->restrictions.bpf_filters_cow = cowed; return ret; } if (ctx->restrictions.op_registered) -- cgit v1.2.3