| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Per-task restrictions apply to all rings created by a task. Once
installed, they should not be dropped across exec.
For a task that has used io_uring, the exec cancellation path calls
__io_uring_free(). This frees both the task context and the per-task
restriction, so a ring created after exec is unrestricted.
Split task context cleanup into io_uring_free_tctx(), and use it from
the exec cancellation path. Keep __io_uring_free() for final task
cleanup, where both the context and restriction are released.
Fixes: ed82f35b926b ("io_uring: allow registration of per-task restrictions")
Cc: stable@vger.kernel.org # 7.1+
Signed-off-by: Kyumin Lee <fyonglkm@gmail.com>
Link: https://patch.msgid.link/20260730192734.459247-1-fyonglkm@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
import_zcrx() attaches an existing ifq to another ring. It never calls
zcrx_set_ring_ctx() and so never takes the ->master_ctx reference, but
its error path still passes @ctx to zcrx_unregister(), which clears
->master_ctx and drops its percpu_ref whenever ifq->master_ctx == ctx.
That condition is reachable. A ring that registers an ifq with a
non-zero event type_mask gets ->master_ctx pointed at itself, and
nothing stops it from exporting that ifq with ZCRX_CTRL_EXPORT and
importing the resulting fd back into the same ring. Failing the import
after the refcount bumps -- an argument page mapped PROT_READ makes the
copy_to_user() in import_zcrx() return -EFAULT -- then clears the
->master_ctx owned by the original registration, which is still live.
Refcounts stay balanced and nothing is freed early, so there is no
splat. The ring silently stops receiving ZCRX_EVENT_ALLOC_FAIL and
ZCRX_EVENT_COPY: zcrx_send_notif() returns early on a NULL
->master_ctx, and ->master_ctx is only ever set on a freshly allocated
ifq, so it cannot be restored without tearing the ring down.
Pass NULL instead, matching zcrx_box_release() and the zcrx_export()
error path. io_register_zcrx() only gets away with passing @ctx
because zcrx_set_ring_ctx() runs after its last goto err.
Fixes: 00d91481279f ("io_uring/zcrx: share an ifq between rings")
Signed-off-by: Woraphat Khiaodaeng <worapat.kd2@gmail.com>
Link: https://patch.msgid.link/20260730162741.1125-1-worapat.kd2@gmail.com
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
[axboe: add pavel edit]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Turns out that my naive assumption that surely nobody is using io_uring
to manage nested epoll instances was wrong - libuv is in fact doing that
very thing. Since we have users in the wild using that feature, we
cannot remove it from upstream. Revert the commit that introduced that
limitation.
This reverts commit cfa1539b24aff18ecb71c6334e7270f810d145bb.
Link: https://lore.kernel.org/io-uring/605939add4dfb674bd7ab39ef8dbd3fd22e754e4.camel@xry111.site/
Reported-by: Xi Ruoyao <xry111@xry111.site>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_ring_buffers_peek() builds an iovec array from provided buffers, and
that in turn can be handed off to a lower level provider. Be prudent and
cap the total size to MAX_RW_COUNT, which is the Linux default for how
much IO do to in a single call.
No bugs here, but it's a good preventative measure to avoid truncation
issues.
Reviewed-by: Gabriel Krisman Bertazi <krisman@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Commit:
6a8afb9fff64 ("io_uring/net: allow multishot receive per-invocation cap")
changed how io_mshot_prep_retry() set sr->len, and added the same
initialization in io_mshot_prep_retry(). But it neglected to touch the
send path, which may also uses the mshot retry path. Ensure that
sr->mshot_len always gets initialized correctly.
Fixes: 6a8afb9fff64 ("io_uring/net: allow multishot receive per-invocation cap")
Cc: stable@vger.kernel.org
Reported-by: Sung Keum <kambodi127@gmail.com>
Reviewed-by: Gabriel Krisman Bertazi <krisman@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux
Pull io_uring fixes from Jens Axboe:
- Fix a missing ERESTARTSYS conversion in the read paths, which got
messed up back when some code consolidation was done for read
multishot support
- zcrx UAPI rename, dropping the abbreviated "notif" naming in favor of
"event" for consistency and to be less ambiguous for users. This was
added for 7.2, so let's rename it while we still can. No functional
or code changes, just a strict rename
* tag 'io_uring-7.2-20260724' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:
io_uring/zcrx: rename notif to event
io_uring/zcrx: rename ZCRX_NOTIF_NO_BUFFERS
io_uring/zcrx: drop "notif" from stats struct names
io_uring/rw: fix missing ERESTARTSYS conversion in read paths
|
|
"Notification" is too long and the abbreviated version is used in
several places, which is inconsistent and more ambiguous for users.
Rename it to event, which is easier to keep consistent. To keep the
change small, only change uapi/ + do necessary fix ups, and the rest of
internals can be adjusted in the next release.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://patch.msgid.link/f95ca6717da3c8d3649a1a7f0d883a563f545052.1784726895.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
ZCRX_NOTIF_NO_BUFFERS tells when page pool fails to allocate memory from
zcrx. "No buffers" could be more confusing, rename it to
ZCRX_NOTIF_ALLOC_FAIL.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://patch.msgid.link/29bd4fc069bc89691868beba0627ffbe570c2722.1784726895.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Keep zcrx statistics generic and don't stick "notif" to its uapi
definitions. Stats dosn't need to be bound to notification details,
it makes it cleaner and more readable.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://patch.msgid.link/6a39676b6f71b67d3f89c6ebab7a3739873834a3.1784726895.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Both read and write may receive internal restart error codes from
the filesystem layer and should be converted to -EINTR. However,
when multishot read support was added, the error code normalization
was lost for both io_read() and io_read_mshot().
Extract the conversion into io_fixup_restart_res() and apply it
in all three locations: io_rw_done(), io_read(), and io_read_mshot().
Fixes: a08d195b586a ("io_uring/rw: split io_read() into a helper")
Cc: stable@vger.kernel.org
Signed-off-by: Yitang Yang <yi1tang.yang@gmail.com>
Link: https://patch.msgid.link/20260722124551.130563-1-yi1tang.yang@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux
Pull io_uring fixes from Jens Axboe:
- Fix a use-after-free in the bpf-ops struct_ops path, where the same
io_uring_bpf_ops map could be registered more than once.
- Fix the deferred iovec free for the provided-buffer grow path, which
could leave the caller with a dangling iovec and result in repeated
frees. Follow-up to the earlier fix in this series.
- Zero-check the unused addr3/pad2 SQE fields for unlinkat
* tag 'io_uring-7.2-20260717' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:
io_uring/bpf-ops: reject re-registration of an already-bound ops
io_uring/fs: check unused sqe fields for unlinkat
io_uring/kbuf: free the replaced iovec after a successful grow
|
|
io_install_bpf() only rejects a second registration on the ctx side
(ctx->bpf_ops) and sets the per-map back-pointer ops->priv
unconditionally. The struct_ops link path never advances a map past
BPF_STRUCT_OPS_STATE_READY, so the same io_uring_bpf_ops map can be
registered more than once, and bpf_io_reg() re-resolves the target ring
via fget(ops->ring_fd) on every call. A caller can therefore point the
same ring_fd at a different io_ring_ctx between two BPF_LINK_CREATE
calls.
The second registration passes the ctx->bpf_ops check (the new ctx has
none) and overwrites ops->priv, orphaning the first ctx. Teardown
(io_eject_bpf()/bpf_io_unreg()) only reaches a ctx through ops->priv, so
the orphaned ctx is never torn down: its ctx->loop_step keeps pointing
into the struct_ops trampoline, which is freed once the map is gone. A
later io_uring_enter() on the orphaned ring then calls the dangling
ctx->loop_step from io_run_loop() -- a use-after-free of freed
executable memory, reachable by a task with CAP_BPF + CAP_PERFMON.
Reject registration when ops->priv is already set, as hid_bpf_reg()
does for its struct_ops.
Cc: stable@vger.kernel.org
Fixes: 98f37634b12b ("io_uring/bpf-ops: implement bpf ops registration")
Signed-off-by: Woraphat Khiaodaeng <worapat.kd2@gmail.com>
Reviewed-by: Gabriel Krisman Bertazi <krisman@suse.de>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://patch.msgid.link/20260717154537.129736-1-worapat.kd2@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Zero check unused SQE fields addr3 and pad2 for unlinkat. They're
not needed now, but could be used sometime in the future.
Signed-off-by: Yi Xie <xieyi@kylinos.cn>
Reviewed-by: Gabriel Krisman Bertazi <krisman@suse.de>
Link: https://patch.msgid.link/20260714030306.64820-1-xieyi@kylinos.cn
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
The provided-buffer validation fix deferred freeing a cached iovec
until validation completed. However, the deferred free uses arg->iovs.
After a grow, that points to the newly allocated array. Without a grow,
it points to the cached array that remains in use.
This leaves the caller with a dangling iovec in both cases and can
result in repeated frees. Only free org_iovs when arg->iovs actually
replaced it.
Fixes: cd053d788c3f ("io_uring: fix dangling iovec after provided-buffer bundle grow failure")
Assisted-by: Codex:gpt-5.3-codex-spark
Signed-off-by: Jaeyeong Lee <iostreampy@proton.me>
Link: https://patch.msgid.link/20260712142612.188695595-iostreampy@proton.me
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux
Pull io_uring fixes from Jens Axboe:
- Restore full RCU read section in io_req_local_work_add(), which was
mistakenly dropped with the DEFER_TASKRUN rework in this merge
window. Revert the commit that grabbed the RCU read lock in
io_ctx_mark_taskrun(), as that's no longer required with the previous
fix.
- Fix a dangling iovec after a provided-buffer bundle grow failure,
also an issue introduced in this merge window.
- Reject IORING_CQE_F_32 flag pass-through in MSG_RING to rings that
weren't setup with CQE32 or CQE_MIXED.
- Return -EINVAL rather than -ENOMEM from get_unmapped_area() when mmap
validation fails, matching io_uring_mmap().
* tag 'io_uring-7.2-20260710' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:
Revert "io_uring: grab RCU read lock marking task run"
io_uring: restore RCU read section in io_req_local_work_add()
io_uring: fix dangling iovec after provided-buffer bundle grow failure
io_uring/uring_cmd: fix uring_cmd.c comments
io_uring/msg_ring: reject CQE32 flag pass-through to normal rings
io_uring/memmap: return -EINVAL from get_unmapped_area() on bad mmap
|
|
This reverts commit ed64f5c546b3d5e3a4840f6c055448ce90edf56c.
Since commit:
648790e09527 ("io_uring: restore RCU read section in io_req_local_work_add()")
io_ctx_mark_taskrun() is only ever called with the RCU read lock
already held, like previously. Hence's there's no need for this commit
anymore, which grabbed the RCU read lock inside io_ctx_mark_taskrun().
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
The task-work refactor that moved io_req_local_work_add() out of
io_uring.c into the new io_uring/tw.c dropped the whole-body
guard(rcu)() that used to cover the function body.
For DEFER_TASKRUN rings the ring teardown still relies on that RCU read
section pairing with its grace period:
/* pairs with RCU read section in io_req_local_work_add() */
if (ctx->flags & IORING_SETUP_DEFER_TASKRUN)
synchronize_rcu();
io_ring_ctx_free(ctx);
io_req_local_work_add() keeps dereferencing ctx after mpscq_push() has
published the request to the work list (ctx->cq_wait_nr, and
ctx->submitter_task in the final wake_up_state()), without holding a ctx
reference across that window. The RCU read section was the only thing
guaranteeing an in-flight adder had finished touching ctx before
io_ring_ctx_free() ran; synchronize_rcu() only waits for readers that
are actually inside an RCU read-side critical section. With the guard
gone the grace period no longer pairs with anything on the add side, so
ctx can be freed and reused while io_req_local_work_add() is still using
it.
Fixes: d46ab2c98aba ("io_uring: switch local task_work to a mpscq")
Signed-off-by: Woraphat Khiaodaeng <worapat.kd2@gmail.com>
Link: https://patch.msgid.link/20260709035100.2269-1-worapat.kd2@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
When growing a provided-buffer bundle, the old cached iovec is freed
before the new buffers have all been validated. If validation fails, the
request still points at the freed iovec, which can be freed again during
completion cleanup.
Fix this by deferring the free of the old cached iovec until validation
has succeeded. On failure, free the newly allocated iovec and leave the
request pointing at the original one.
Fixes: 46800585ae04 ("io_uring/kbuf: validate ring provided buffer addresses with access_ok()")
Signed-off-by: Hao-Yu Yang <naup96721@gmail.com>
Link: https://patch.msgid.link/20260706183304.919275-1-naup96721@gmail.com
Suggested-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Fix "concelable" -> "cancelable" in the comment above
io_uring_cmd_mark_cancelable(), and fix the memory ordering comment
in __io_uring_cmd_done() to reference io_do_iopoll() and
->iopoll_completed.
Signed-off-by: Yang Xiuwei <yangxiuwei@kylinos.cn>
Link: https://patch.msgid.link/20260702082937.3707134-3-yangxiuwei@kylinos.cn
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
IORING_OP_MSG_RING with IORING_MSG_RING_FLAGS_PASS allows a sender to
pass completion flags through sqe->file_index. If the sender sets
IORING_CQE_F_32 in file_index, the target-side completion path treats it
as a 32b CQE and writes big_cqe[0] and big_cqe[1] into the CQ ring
regardless of whether the target ring was created with
IORING_SETUP_CQE32 or IORING_SETUP_CQE_MIXED.
On a normal 16b CQE ring, this writes 16 extra bytes (two u64 big_cqe
fields) into the next CQE slot in the ring buffer. As the receiving ring
doesn't understand 32b CQEs, this is incorrect and they should be
rejected.
Fixes: cbeb47a7b5f0 ("io_uring/msg_ring: Pass custom flags to the cqe")
Signed-off-by: Melbin K Mathew <mlbnkm1@gmail.com>
Link: https://patch.msgid.link/20260701081145.196730-1-mlbnkm1@gmail.com
[axboe: edit commit message]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
get_unmapped_area() returns -ENOMEM when io_uring_validate_mmap_request()
fails, but validation errors are -EINVAL. Propagate that errno to
userspace, like io_uring_mmap() already does.
Signed-off-by: Yi Xie <xieyi@kylinos.cn>
Link: https://patch.msgid.link/20260630091206.126206-1-xieyi@kylinos.cn
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux
Pull io_uring fixes from Jens Axboe:
- Fix a file reference leak in the nop opcode when used with
IOSQE_FIXED_FILE
- Preserve the SQ array entries when resizing the ring via the register
path
- Preserve the partial result for an iopoll request rather than
overwriting it
- Don't audit log IORING_OP_RECV_ZC
- Bound io_pin_pages() by the page array byte size in the memmap path
- Follow-up cleanup to the task_work mpscq conversion, getting rid of
the now-unnecessary tw_pending tracking for the !DEFER_TASKRUN path
- Switch a system_unbound_wq user over to system_dfl_wq
* tag 'io_uring-7.2-20260625' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:
io_uring/memmap: bound io_pin_pages() by page array byte size
io_uring: Use system_dfl_wq instead of system_unbound_wq
io_uring/register: preserve SQ array entries on resize
io_uring, audit: don't log IORING_OP_RECV_ZC
io_uring: get rid of tw_pending for !DEFER task work
io_uring/rw: preserve partial result for iopoll
io_uring/nop: fix file reference leak with IOSQE_FIXED_FILE
|
|
io_pin_pages() checks that nr_pages does not exceed INT_MAX, then
allocates a struct page * array of nr_pages entries. kvmalloc() limits
allocations to INT_MAX bytes, but the check counts pages, not bytes.
On 64-bit each entry is 8 bytes, so the array hits the INT_MAX byte
limit at INT_MAX / sizeof(struct page *) pages, well before the page
count check fires.
Since commit b4e41050b212 ("io_uring/rsrc: raise registered buffer 1GB
limit") raised the per-buffer cap to 1TB, a buffer near that cap maps
~2^28 pages, making the array allocation exceed INT_MAX bytes. This
passes the page count check, reaches kvmalloc(), and triggers the
WARN_ON_ONCE() for oversized allocations in __kvmalloc_node_noprof().
Check nr_pages against INT_MAX / sizeof(struct page *) so the buffer is
rejected with -EOVERFLOW before the allocation is attempted.
Reported-by: syzbot+f99b00a963915b6b52c6@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=f99b00a963915b6b52c6
Fixes: b4e41050b212 ("io_uring/rsrc: raise registered buffer 1GB limit")
Tested-by: syzbot+f99b00a963915b6b52c6@syzkaller.appspotmail.com
Signed-off-by: Deepanshu Kartikey <kartikey406@gmail.com>
Reviewed-by: Gabriel Krisman Bertazi <krisman@suse.de>
Link: https://patch.msgid.link/20260621012933.50571-1-kartikey406@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux
Pull io_uring epoll update from Jens Axboe:
"As discussed a few months ago, this pull request gets rid of allowing
nested epoll notification contexts via io_uring.
Nested contexts have been a source of issues on the epoll side, and
there should not be a need to support them from io_uring. The epoll
io_uring side exists mainly to facilitate a gradual migration from a
notification based epoll setup to an io_uring ditto"
* tag 'for-7.2/io_uring-epoll-20260616' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:
io_uring/epoll: disallow adding an epoll file to an epoll context
io_uring/epoll: switch to using do_epoll_ctl_file() interface
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next
Pull networking updates from Jakub Kicinski:
"Core & protocols:
- Work on removing rtnl_lock protection throughout the stack
continues. In this chapter:
- don't use rtnl_lock for IPv6 multicast routing configuration
- don't take rtnl_lock in ethtool for modern drivers
- prepare Qdisc dump callbacks for rtnl_lock removal
- Support dumping just ifindex + name of all interfaces, under RCU.
It's a common operation for Netlink CLI tools (when translating
names to ifindexes) and previously required full rtnl_lock.
- Support dumping qdiscs and page pools for a specific netdev. Even
tho user space wants a dump of all netdevs, most of the time, the
OOO programming model results in repeating the dump for each
netdev. Which, in absence of a cache, leads to a O(n^2) behavior.
- Flush nexthops once on multi-nexthop removal (e.g. when device goes
down), another O(n^2) -> O(n) improvement.
- Rehash locally generated traffic to a different nexthop on
retransmit timeout.
- Honor oif when choosing nexthop for locally generated IPv6 traffic.
- Convert TCP Auth Option to crypto library, and drop non-RFC algos.
- Increase subflow limits in MPTCP to 64 and endpoint limit to 256.
- Support MPTCP signaling of IPv6 address + port (ADD_ADDR). We need
to selectively skip reporting of the standard TCP Timestamp option,
because they won't fit into the header space together (12 + 30 >
40).
- Support using bridge neighbor suppression, Duplicate Address
Detection, Gratuitous ARP and unsolicited NA forwarding - in EVPN
deployments, e.g. VXLAN fabrics (IPv4 and IPv6).
- Improve link state reporting for upper netdevs (e.g. macvlan) over
tunnel devices (again, mostly for EVPN deployments).
- Support binding GENEVE tunnels to a local address.
- Speed up UDP tunnel destruction (remove one synchronize_rcu()).
- Support exponential field encoding in multicast (IGMPv3 and MLDv2).
- Support attaching PSP crypto offload to containers (veth, netkit).
- Add a new IPSec Netlink message XFRM_MSG_MIGRATE_STATE that allows
migrating individual IPsec SAs independently of their policies.
The existing XFRM_MSG_MIGRATE is tightly coupled to policy+SA
migration, lacks SPI for unique SA identification, and cannot
express reqid changes or migrate Transport mode selectors.
The new interface identifies the SA via SPI and mark, supports
reqid changes, address family changes, encap removal, and uses an
atomic create+install flow under x->lock to prevent SN/IV reuse
during AEAD SA migration.
- Implement GRO/GSO support for PPPoE.
- Convert sockopt callbacks in a number of protocols to iov_iter.
Cross-tree stuff:
- Remove support for Crypto TFM cloning (unblocked after the TCP Auth
Option rework). This feature regressed performance for all crypto
API users, since it changed crypto transformation objects into
reference-counted objects.
- Add FCrypt-PCBC implementation to rxrpc and remove it from the
global crypto API as obsolete and insecure.
Wireless:
- Major rework of station bandwidth handling, fixing issues with
lower capability than AP.
- Cleanups for EMLSR spec issues (drafts differed).
- More Neighbor Awareness Networking (Wi-Fi Aware) work (multicast,
schedule improvements, multi-station etc.)
- Some Ultra High Reliability (UHR) / IEEE 802.11bn (D1.4) work
(e.g. non-primary channel access, UHR DBE support).
- Fine Timing Measurement ranging (i.e. distance measurement) APIs.
Netfilter:
- Use per-rule hash initval in nf_conncount. This avoids unnecessary
lock contention with short keys (e.g. conntrack zones) in different
namespaces.
- Various safety improvements, both in packet parsing and object
lifetimes. Notably add refcounts to conntrack timeout policy.
Deletions:
- Remove TLS + sockmap integration. TLS wants to pin user pages to
avoid a copy, and sockmap wants to write to the input stream. More
work on this integration is clearly needed, and we can't find any
users (original author admitted that they never deployed it).
- Remove support for TLS offload with TCP Offload Engine (the far
more common opportunistic offload is retained). The locking looks
unfixable (driver sleeps under TCP spin locks) and people from the
vendor that added this are AWOL.
- Remove more ATM code, trying to leave behind only what PPPoATM
needs, AAL5 and br2684 with permanent circuits.
- Remove AppleTalk. Let it join hamradio in our out of tree protocol
graveyard, I mean, repository.
- Disable 32-bit x_tables compatibility (32bit binaries on 64bit
kernel) interface in user namespaces. To be deleted completely,
soon.
- Remove 5/10 MHz support from cfg80211/mac80211.
Drivers:
- Software:
- Support DEVMEM/DMABUF Tx over NETMEM_TX_NO_DMA devices (netkit)
- bonding: add knob to strictly follow 802.3ad for link state
- New drivers:
- Alibaba Elastic Ethernet Adaptor (cloud vNIC).
- NXP NETC switch within i.MX94.
- DPLL:
- Add operational state to pins (implement in zl3073x).
- Add generic DPLL type, for daisy-chaining DPLLs (implement in ice).
- Ethernet high-speed NICs:
- Huawei (hinic3):
- enhance tc flow offload support with queue selection,
tunnels
- nVidia/Mellanox:
- avoid over-copying payload to the skb's linear part (up to
60% win for LRO on slow CPUs like ARM64 V2)
- expose more per-queue stats over the standard API
- support additional, unprivileged PFs in the DPU
configuration
- support Socket Direct (multi-PF) with switchdev offloads
- add a pool / frag allocator for DMA mapped buffers for
control objects, save memory on systems with 64kB page size
- take advantage of the ability to dynamically change RSS
table size, even when table is configured by the user
- increase the max RSS table size for even traffic
distribution
- Ethernet NICs:
- Marvell/Aquantia:
- AQC113 PTP support
- Realtek USB (r8152):
- support 10Gbit Link Speeds and Energy-Efficient Ethernet
(EEE)
- support firmware loaded (for RTL8157/RTL8159)
- support for the RTL8159
- Intel (ixgbe):
- support Energy-Efficient Ethernet (EEE) on E610 devices
- Ethernet switches:
- Airoha:
- support multiple netdevs on a single GDM block / port
- Marvell (mv88e6xxx):
- support SERDES of mv88e6321
- Microchip (ksz8/9):
- rework the driver callbacks to remove one indirection layer
- Motorcomm (yt921x):
- support port rate policing
- support TBF qdisc offload
- support ACL/flower offload
- nVidia/Mellanox:
- expose per-PG rx_discards
- Realtek:
- rtl8365mb: bridge offloading and VLAN support
- Ethernet PHYs:
- Airoha:
- support Airoha AN8801R Gigabit PHYs.
- Micrel:
- implement 3 low-loss cable tunables
- Realtek:
- support MDI swapping for RTL8226-CG
- support MDIO for RTL931x
- Qualcomm:
- at803x: Rx and Tx clock management for IPQ5018 PHY
- Motorcomm:
- support YT8522 100M RMII PHY
- set drive strength in YT8531s RGMII
- TI:
- dp83822: add optional external PHY clock
- Bluetooth:
- hci_sync: add support for HCI_LE_Set_Host_Feature [v2]
- SMP: use AES-CMAC library API
- Intel:
- support Product level reset
- support smart trigger dump
- Mediatek:
- add event filter to filter specific event
- Realtek:
- fix RTL8761B/BU broken LE extended scan
- WiFi:
- Broadcom (b43):
- new support for a 11n device
- MediaTek (mt76):
- support mt7927
- mt792x: broken usb transport detection
- mt7921: regulatory improvements
- Qualcomm (ath9k):
- GPIO interface improvements
- Qualcomm (ath12k):
- WDS support
- replace dynamic memory allocation in WMI Rx path
- thermal throttling/cooling device support
- 6 GHz incumbent interference detection
- channel 177 in 5 GHz
- Realtek (rt89):
- RTL8922AU support
- USB 3 mode switch for performance
- better monitor radiotap support
- RTL8922DE preparations"
* tag 'net-next-7.2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1778 commits)
ipv4: fib_rule: Move fib4_rules_exit() to ->exit().
net: serialize netif_running() check in enqueue_to_backlog()
net: skmsg: preserve sg.copy across SG transforms
appletalk: move the protocol out of tree
appletalk: stop storing per-interface state in struct net_device
selftests/bpf: test that TLS crypto is rejected on a sockmap socket
selftests/bpf: drop the unused kTLS program from test_sockmap
selftests/bpf: remove sockmap + ktls tests
tls: remove dead sockmap (psock) handling from the SW path
tls: reject the combination of TLS and sockmap
atm: remove orphaned uAPI for deleted drivers, protocols and SVCs
atm: remove unused ATM PHY operations
atm: remove the unused pre_send and send_bh device operations
atm: remove the unused change_qos device operation
atm: remove SVC socket support and the signaling daemon interface
atm: remove the local ATM (NSAP) address registry
atm: remove dead SONET PHY ioctls
atm: remove the unused send_oam / push_oam callbacks
atm: remove AAL3/4 transport support
net: dsa: sja1105: fix lastused timestamp in flower stats
...
|
|
Commit de7341ffe49e ("io_uring: switch normal task_work to a mpscq")
added a use of system_unbound_wq, which is deprecated in favor of
system_dfl_wq added by commit 128ea9f6ccfb ("workqueue: Add
system_percpu_wq and system_dfl_wq"). An upcoming warning in the
workqueue tree flags this with:
workqueue: work func io_tctx_fallback_work enqueued on deprecated workqueue. Use system_{percpu|dfl}_wq instead.
Switch to system_dfl_wq to clear up the warning.
Fixes: de7341ffe49e ("io_uring: switch normal task_work to a mpscq")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Link: https://patch.msgid.link/20260616-io_uring-fix-wq-warning-v1-1-cfc9d934eedb@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Ring resizing copies pending SQEs from the old SQE array into the new
one so submissions queued before the resize can still be consumed
afterwards.
That copy currently walks the SQ head/tail range directly. This is only
correct when there is no SQ array indirection. With a regular SQ array,
each pending SQ entry contains an index into the SQE array. After resize,
ctx->sq_array is repointed at the newly allocated array, so pending
entries lose their old logical-to-physical mapping and may submit the
wrong SQE.
Remember the old and new SQ arrays while migrating pending SQ entries. For
each pending entry, copy the SQE selected by the old array into the new
destination slot and rebuild the new array entry to point at the copied
SQE. Keep invalid user-provided entries invalid so the normal submission
path still drops them after resize.
Fixes: 79cfe9e59c2a1 ("io_uring/register: add IORING_REGISTER_RESIZE_RINGS")
Signed-off-by: guzebing <guzebing1612@gmail.com>
Link: https://patch.msgid.link/20260608133316.3656440-1-guzebing1612@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
IORING_OP_RECV_ZC is a read operation. Audit only tracks file/socket
creation, not subsequent reads. Set audit_skip to align with
audit-userspace uringop_table.h.
Fixes: 11ed914bbf94 ("io_uring/zcrx: add io_recvzc request")
Suggested-by: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Ricardo Robaina <rrobaina@redhat.com>
Acked-by: Paul Moore <paul@paul-moore.com>
Link: https://patch.msgid.link/20260616123632.3209545-1-rrobaina@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
The normal task_work path used a tw_pending bit to ensure the callback
was only added once: the mpscq drains incrementally, so a single
tctx_task_work() run can take the queue through empty -> non-empty
several times, and each transition would otherwise re-add the already
pending callback_head. This corrupts the task_work list, and is what
tw_pending protects again.
This can go away, if we stop running the task_work as soon as the queue
empties.
Suggested-by: Caleb Sander Mateos <csander@purestorage.com>
Reviewed-by: Caleb Sander Mateos <csander@purestorage.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
A partial read will store the completed byte count in io->bytes_done.
The regular completion path applies io_fixup_rw_res() so that, when the
following operation reaches EOF, the number of bytes already read is
returned.
The iopoll completion path does not apply this fixup to the return value
and can return zero instead.
Use the fixup result when updating the CQE, and the raw result for the
reissue check.
Cc: stable@vger.kernel.org
Fixes: 4d9cb92ca41d ("io_uring/rw: fix short rw error handling")
Signed-off-by: Michael Wigham <michael@wigham.net>
Link: https://patch.msgid.link/20260613225240.34032-1-michael@wigham.net
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
NOP file-acquisition support choses between a fixed (registered) file and
a normal fget()'d file based on its own IORING_NOP_FIXED_FILE flag in
sqe->nop_flags. However, a request's REQ_F_FIXED_FILE is set
independently from the generic IOSQE_FIXED_FILE sqe flag during request
init, before the issue handler runs.
If a NOP is submitted with IOSQE_FIXED_FILE set (so REQ_F_FIXED_FILE is
set) but without IORING_NOP_FIXED_FILE, io_nop() takes the normal path
and grabs a real reference via io_file_get_normal(). On completion,
io_put_file() only drops the reference when REQ_F_FIXED_FILE is clear,
so the fget()'d file is never released and leaks:
BUG: memory leak
unreferenced object 0xffff88800f42c240 (size 176):
kmem_cache_alloc_noprof+0x358/0x440
alloc_empty_file+0x57/0x180
path_openat+0x44/0x1e50
do_file_open+0x121/0x200
do_sys_openat2+0xa7/0x150
__x64_sys_openat+0x82/0xf0
Decide between fixed and normal file acquisition from REQ_F_FIXED_FILE,
the same way io_assign_file() does for every other opcode, and fold
IORING_NOP_FIXED_FILE into REQ_F_FIXED_FILE at prep time.
Cc: stable@vger.kernel.org
Fixes: a85f31052bce ("io_uring/nop: add support for testing registered files and buffers")
Reported-by: syzbot+2cd473471e77bda12b0e@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?id=879092631b98f73a28ea405adacfa5bb34a14a25
Signed-off-by: Vasileios Almpanis <vasilisalmpanis@gmail.com>
Link: https://patch.msgid.link/20260615144619.482749-1-vasilisalmpanis@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux
Pull block updates from Jens Axboe:
- NVMe pull request via Keith:
- Per-controller admin and IO timeout sysfs attributes, and
letting the block layer set request timeouts (Maurizio,
Maximilian)
- Multipath passthrough iostats, and PCI P2PDMA enablement for
multipath devices (Keith, Kiran)
- A new diag sysfs attribute group exporting per-controller
counters (retries, multipath failover, error counters, requeue
and failure counts, reset and reconnect events) (Nilay)
- FDP configuration validation and bounds check fixes (liuxixin)
- Various nvmet fixes, including a pre-auth out-of-bounds read in
the Discovery Get Log Page handler, auth payload bounds
validation, and tcp error-path leak fixes (Bryam, Tianchu,
Geliang)
- nvme-tcp lockdep and workqueue fixes (Shin'ichiro, Kuniyuki,
Eric)
- Assorted other fixes and cleanups (John, Yao, Chao, Mateusz,
Achkinazi, Wentao)
- MD pull request via Yu Kuai:
- raid1/raid10 fixes for a deadlock in the read error recovery
path, error-path detection and bio accounting with cloned bios,
and an nr_pending leak in the REQ_ATOMIC bad-block error path
(Abd-Alrhman)
- PCI P2PDMA propagation from member devices to the RAID device
(Kiran)
- dm-raid bio requeue fix, and various smaller fixes and cleanups
(Benjamin, Chen, Li, Thorsten)
- Enable Clang lock context analysis for the block layer, with the
accompanying annotations across queue limits, the blk_holder_ops
callbacks, crypto, cgroup, iocost, kyber and mq-deadline (Bart)
- Block status code infrastructure work: a tagged status table, a
str_to_blk_op() helper, a bio_endio_status() helper, and on top of
that a new configurable block-layer error injection facility
(Christoph)
- DRBD netlink rework, replacing the genl_magic machinery with explicit
netlink serialization and moving the DRBD UAPI headers to
include/uapi/linux/ (Christoph Böhmwalder)
- bvec improvements: a bvec_folio() helper and making the bvec_iter
helpers proper inline functions (Willy, Christoph)
- ublk cleanups and a canceling-flag fix for the disk-not-allocated
case (Caleb, Ming)
- Partition handling fixes: bound the AIX pp_count scan, fix an of_node
refcount leak, and replace __get_free_page() with kmalloc() (Bryam,
Wentao, Mike)
- Convert numa_node to int in blk_mq_hw_ctx and ->init_request, and add
WQ_PERCPU to the block workqueue users (Mateusz, Marco)
- Block statistics and tracing: propagate in-flight to the whole disk
on partition IO, export passthrough stats, and a new
block_rq_tag_wait tracepoint (Tang, Keith, Aaron)
- A round of removals, unexports and cleanups across bio, direct-io and
the bvec helpers (Christoph)
- Various driver fixes (mtip32xx use-after-free, rbd snap_count
validation and strscpy conversion, nbd socket lockdep reclassify,
virtio-blk zone report clamp, floppy) and a batch of MAINTAINERS
email/list updates (Coly, Li, Yu, Christoph Böhmwalder)
- Other little fixes and cleanups all over
* tag 'for-7.2/block-20260615' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux: (117 commits)
MAINTAINERS: Update Coly Li's email address
block: check bio split for unaligned bvec
nbd: Reclassify sockets to avoid lockdep circular dependency
block: add configurable error injection
block: add a str_to_blk_op helper
block: add a "tag" for block status codes
block: add a macro to initialize the status table
floppy: Drop unused pnp driver data
block: propagate in_flight to whole disk on partition I/O
virtio-blk: clamp zone report to the report buffer capacity
block: optimize I/O merge hot path with unlikely() hints
drivers/block/rbd: Use strscpy() to copy strings into arrays
partitions: aix: bound the pp_count scan to the ppe array
block: Enable lock context analysis
block/mq-deadline: Make the lock context annotations compatible with Clang
block/Kyber: Make the lock context annotations compatible with Clang
block/blk-mq-debugfs: Improve lock context annotations
block/blk-iocost: Inline iocg_lock() and iocg_unlock()
block/blk-iocost: Split ioc_rqos_throttle()
block/crypto: Annotate the crypto functions
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux
Pull io_uring updates from Jens Axboe:
- Rework the task_work infrastructure.
Both the local (DEFER_TASKRUN) and the normal (tctx) task_work lists
were llist based, which is LIFO ordered, and hence each run had to do
an O(n) list reversal pass first to restore queue order.
Additionally, to cap the amount of task_work run, each method needed
a retry list as well.
Add a lockless MPCS FIFO queue (based on Dmitry Vyukov's intrusive
MPSC algorithm) and switch both task_work lists to it. It performs
better than llists and we can then also ditch the retry lists as well
as entries are popped one-at-the-time.
On top of those changes, run the tctx fallback task_work directly and
remove the now-unused per-ctx fallback machinery entirely.
- zcrx user notifications.
Add a mechanism for zcrx to communicate conditions back to userspace
via a dedicated CQE, with the initial users being notification on
running out of buffers and on a frag copy fallback, plus
shared-memory notification statistics.
Alongside that, a series of zcrx reliability and cleanup fixes: more
reliable scrubbing, poisoning pointers on unregistration, dropping an
extra ifq close, adding a ctx back-pointer, reordering fd allocation
in the export path, and killing a dead 'sock' member.
- Allow using io_uring registered buffers for plain SEND and RECV, not
just for the zero-copy send path.
This enables targets like ublk's NBD backend to push/pull IO data
directly to/from a registered buffer over a plain send/recv on a TCP
socket.
- Registered buffer improvements: account huge pages correctly, bump
the io_mapped_ubuf length field to size_t, and raise the previous 1GB
registered buffer size limit.
- Restrict the ctx access exposed to io_uring BPF struct_ops programs
by handing them an opaque type rather than the full io_ring_ctx, and
add a separate MAINTAINERS entry for the bpf-ops code.
- Allow opcode filtering on IORING_OP_CONNECT.
- Validate ring-provided buffer addresses with access_ok(), and align
the legacy buffer add limit with MAX_BIDS_PER_BGID.
- Various other cleanups and minor fixes, including avoiding msghdr
async data on connect/bind, dropping async_size for OP_LISTEN, making
the POLL_FIRST receive side checks consistent, re-checking
IO_WQ_BIT_EXIT for each linked work item, and using
trace_call__##name() at guarded tracepoint call sites.
* tag 'for-7.2/io_uring-20260615' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux: (31 commits)
io_uring/bpf-ops: add a separate maintainer entry
io_uring/net: make POLL_FIRST receive side checks consistent
io_uring: remove the per-ctx fallback task_work machinery
io_uring: run the tctx task_work fallback directly
io_uring: switch normal task_work to a mpscq
io_uring: switch local task_work to a mpscq
io_uring/mpscq: add lockless multi-producer, single-consumer FIFO queue
io_uring: grab RCU read lock marking task run
io_uring/zcrx: kill dead 'sock' member in struct io_zcrx_args
io_uring/kbuf: validate ring provided buffer addresses with access_ok()
io_uring/net: support registered buffer for plain send and recv
io_uring/nop: Drop a wrong comment in struct io_nop
io_uring/net: Remove async_size for OP_LISTEN
io_uring/net: Avoid msghdr on op_connect/op_bind async data
io_uring/bpf-ops: restrict ctx access to BPF
io_uring/io-wq: re-check IO_WQ_BIT_EXIT for each linked work item
io_uring/kbuf: align legacy buffer add limit with MAX_BIDS_PER_BGID
io_uring/zcrx: add shared-memory notification statistics
io_uring/zcrx: notify user on frag copy fallback
io_uring/zcrx: notify user when out of buffers
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Pull crypto updates from Herbert Xu:
"API:
- Drop support for off-CPU cryptography in af_alg
- Document that af_alg is *always* slower
- Document the deprecation of af_alg
- Remove zero-copy support from skcipher and aead in af_alg
- Cap AEAD AD length to 0x80000000 in af_alg
- Free default RNG on module exit
Algorithms:
- Fix vli multiplication carry overflow in ecc
- Drop unused cipher_null crypto_alg
- Remove unused variants of drbg
- Use lib/crypto in drbg
- Use memcpy_from/to_sglist in authencesn
- Allow authenc(hmac(sha{256,384}),cts(cbc(aes))) in FIPS mode
- Disallow RSA PKCS#1 SHA-1 sig algs in FIPS mode
- Filter out async aead implementations at alloc in krb5
- Fix non-parallel fallback by rstoring callback in pcrypt
- Validate poly1305 template argument in chacha20poly1305
Drivers:
- Add sysfs PCI reset support to qat
- Add KPT support for GEN6 devices to qat
- Remove unused character device and ioctls from qat
- Add support for hw access via SMCC to mtk
- Remove prng support from crypto4xx
- Remove prng support from hisi-trng
- Remove prng support from sun4i-ss
- Remove prng support from xilinx-trng
- Remove loongson-rng
- Remove exynos-rng
Others:
- Remove support for AIO on sockets"
* tag 'v7.2-p1' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (196 commits)
crypto: tegra - fix refcount leak in tegra_se_host1x_submit()
crypto: rng - Free default RNG on module exit
crypto: testmgr - allow authenc(hmac(sha{256,384}),cts(cbc(aes))) in FIPS mode
hwrng: jh7110 - fix refcount leak in starfive_trng_read()
crypto: atmel-ecc - drop dead code in atmel_ecdh_max_size
crypto: cavium/cpt - fix DMA cleanup using wrong loop index
crypto: marvell/octeontx - fix DMA cleanup using wrong loop index
MAINTAINERS: make myself the maintainer of the Qualcomm QCE driver
crypto: amcc - convert irq_of_parse_and_map to platform_get_irq
crypto: sun4i-ss - Remove insecure and unused rng_alg
hwrng: xilinx - Move xilinx-rng into drivers/char/hw_random/
crypto: xilinx-trng - Replace crypto_drbg_ctr_df() with HMAC-SHA512
crypto: xilinx-trng - Fix return value of xtrng_hwrng_trng_read()
crypto: xilinx-trng - Remove crypto_rng interface
crypto: exynos-rng - Remove exynos-rng driver
hwrng: hisi-trng - Move hisi-trng into drivers/char/hw_random/
crypto: hisi-trng - Remove crypto_rng interface
crypto: loongson - Remove broken and unused loongson-rng
crypto: crypto4xx - Remove insecure and unused rng_alg
crypto: qat - validate RSA CRT component lengths
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/vbabka/slab
Pull slab updates from Vlastimil Babka:
- Support for "allocation tokens" (currently available in Clang 22+)
for smarter partitioning of kmalloc caches based on the allocated
object type, which can be enabled instead of the "random"
per-caller-address-hash partitioning.
It should be able to deterministically separate types containing a
pointer from those that do not (Marco Elver)
- Improvements and simplification of the kmem_cache_alloc_bulk() and
mempool_alloc_bulk() API. This includes adaptation of callers
(Christoph Hellwig)
- Performance improvements and cleanups related mostly to sheaves
refill (Hao Li, Shengming Hu, Vlastimil Babka)
- Several fixups for the slabinfo tool (Xuewen Wang)
* tag 'slab-for-7.2' of git://git.kernel.org/pub/scm/linux/kernel/git/vbabka/slab:
mm/slab: do not limit zeroing to orig_size when only red zoning is enabled
mm/slub: preserve original size in _kmalloc_nolock_noprof retry path
mm: simplify the mempool_alloc_bulk API
mm/slab: improve kmem_cache_alloc_bulk
mm/slub: detach and reattach partial slabs in batch
mm/slub: introduce helpers for node partial slab state
mm/slub: use empty sheaf helpers for oversized sheaves
tools/mm/slabinfo: remove redundant slab->partial assignment
tools/mm/slabinfo: remove dead assignment in get_obj_and_str()
tools/mm/slabinfo: Fix trace disable logic inversion
MAINTAINERS: add slab-related scripts and tools to SLAB ALLOCATOR
mm/slub: fix typo in sheaves comment
mm, slab: simplify returning slab in __refill_objects_node()
mm, slab: add an optimistic __slab_try_return_freelist()
slab: fix kernel-docs for mm-api
slab: improve KMALLOC_PARTITION_RANDOM randomness
slab: support for compiler-assisted type-based slab cache partitioning
mm/slub: defer freelist construction until after bulk allocation from a new slab
|
|
This adds observability for the io_uring zcrx rx-buf-len configuration.
Signed-off-by: Dragos Tatulea <dtatulea@nvidia.com>
Reviewed-by: Yael Chemla <ychemla@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Acked-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://patch.msgid.link/20260612211709.1456966-3-dtatulea@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
gitolite.kernel.org:pub/scm/linux/kernel/git/tip/tip
Pull locking updates from Ingo Molnar:
"Futex updates:
- Optimize futex hash bucket access patterns (Peter Zijlstra)
- Large series to address the robust futex unlock race for real, by
Thomas Gleixner:
"The robust futex unlock mechanism is racy in respect to the
clearing of the robust_list_head::list_op_pending pointer because
unlock and clearing the pointer are not atomic.
The race window is between the unlock and clearing the pending op
pointer. If the task is forced to exit in this window, exit will
access a potentially invalid pending op pointer when cleaning up
the robust list.
That happens if another task manages to unmap the object
containing the lock before the cleanup, which results in an UAF.
In the worst case this UAF can lead to memory corruption when
unrelated content has been mapped to the same address by the time
the access happens.
User space can't solve this problem without help from the kernel.
This series provides the kernel side infrastructure to help it
along:
1) Combined unlock, pointer clearing, wake-up for the
contended case
2) VDSO based unlock and pointer clearing helpers with a
fix-up function in the kernel when user space was interrupted
within the critical section.
... with help by André Almeida:
- Add a note about robust list race condition (André Almeida)
- Add self-tests for robust release operations (André Almeida)
Context analysis updates:
- Implement context analysis for 'struct rt_mutex'. (Bart Van Assche)
- Bump required Clang version to 23 (Marco Elver)
Guard infrastructure updates:
- Series to remove NULL check from unconditional guards (Dmitry
Ilvokhin)
Lockdep updates:
- Restore self-test migrate_disable() and sched_rt_mutex state on
PREEMPT_RT (Karl Mehltretter)
Membarriers updates:
- Use per-CPU mutexes for targeted commands (Aniket Gattani)
- Modernize membarrier_global_expedited with cleanup guards (Aniket
Gattani)
- Add rseq stress test for CFS throttle interactions (Aniket Gattani)
percpu-rwsems updates:
- Extract __percpu_up_read() to optimize inlining overhead (Dmitry
Ilvokhin)
Seqlocks updates:
- Allow UBSAN_ALIGNMENT to fail optimizing (Heiko Carstens)
Lock tracing:
- Add contended_release tracepoint to sleepable locks such as
mutexes, percpu-rwsems, rtmutexes, rwsems and semaphores (Dmitry
Ilvokhin)
MAINTAINERS updates:
- MAINTAINERS: Add RUST [SYNC] entry (Boqun Feng)
Misc updates and fixes by Randy Dunlap, YE WEI-HONG, Fabricio Parra,
Dmitry Ilvokhin and Peter Zijlstra"
* tag 'locking-core-2026-06-14' of gitolite.kernel.org:pub/scm/linux/kernel/git/tip/tip: (36 commits)
locking: Add contended_release tracepoint to sleepable locks
locking/percpu-rwsem: Extract __percpu_up_read()
tracing/lock: Remove unnecessary linux/sched.h include
futex: Optimize futex hash bucket access patterns
rust: sync: completion: Mark inline complete_all and wait_for_completion
MAINTAINERS: Add RUST [SYNC] entry
cleanup: Specify nonnull argument index
selftests: futex: Add tests for robust release operations
Documentation: futex: Add a note about robust list race condition
x86/vdso: Implement __vdso_futex_robust_try_unlock()
x86/vdso: Prepare for robust futex unlock support
futex: Provide infrastructure to plug the non contended robust futex unlock race
futex: Add robust futex unlock IP range
futex: Add support for unlocking robust futexes
futex: Cleanup UAPI defines
x86: Select ARCH_MEMORY_ORDER_TSO
uaccess: Provide unsafe_atomic_store_release_user()
futex: Provide UABI defines for robust list entry modifiers
futex: Move futex related mm_struct data into a struct
futex: Make futex_mm_init() void
...
|
|
io_recv() and io_recvzc() are the odd ones out, as they checks for
whether POLL_FIRST should be honored before checking if the file is a
socket. It doesn't really matter, but might as well make it consistent
across all receive and send types.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
With the tctx fallback running its entries directly, the per-ctx
fallback work has a single user left: moving local (DEFER_TASKRUN)
task_work entries out of a ring that is going away. Both of its call
sites are process context and don't hold ->uring_lock, the same
conditions the deferred fallback work itself ran under - so run the
entries in cancel mode right there instead, and rename the helper to
io_cancel_local_task_work() to match what it now does.
With that, ->fallback_llist, ->fallback_work, io_fallback_req_func()
and __io_fallback_tw() can all go away, along with the fallback work
flushing in the ring exit and cancel paths. Requests that get
orphaned by an exiting task now run via the tctx fallback work, which
the ring exit side implicitly waits on through the ctx refs those
requests hold.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
The fallback work drains the tctx queue only to redistribute the entries
into the per-ctx fallback lists, bouncing them through a second
(per-ctx) work item before they finally run. That made sense when the
producer side did the draining and could be in any context, but the
fallback work is a regular process context kworker: it can just run the
entries itself. Reuse the normal run loop - if run from the fallback
kernel thread, ts.cancel will get set, and the work terminated.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Like the local task_work list, the normal (tctx) task_work list is an
llist, and hence needs the O(n) llist_reverse_order() pass before
running entries in queue order. On top of that, capped runs - sqpoll
processing IORING_TW_CAP_ENTRIES_VALUE entries at a time - need the
claimed-but-unprocessed leftovers carried in a separate retry_list,
as they can't be pushed back to the shared list.
Switch tctx->task_list to a mpscq, like what was done for the
DEFER_TASKRUN paths as well.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
The local (DEFER_TASKRUN) task_work list is an llist, which is LIFO
ordered, and hence __io_run_local_work() has to restore the right
running order with an O(n) llist_reverse_order() pass first. On top of
that, a batch that gets capped by max_events needs the leftover entries
parked on a separate ->retry_llist, as they can't be pushed back to the
shared list.
Switch it to the FIFO mpscq. Adds are wait-free instead of a cmpxchg
retry loop, entries are popped in queue order with no reversal pass,
capping a run simply leaves the remainder on the queue, and
->retry_llist goes away entirely. The consumer cursor, ->work_head,
lives with the rest of the ->uring_lock protected state rather than
next to the queue, so that popping entries doesn't dirty the producer
side cacheline.
For low amounts of task_work, this ends up being a bit more efficient
than the existing scheme. As an example of that, doing multishot
receives for 8 clients has the following task_work overhead:
1.02% sock-test [kernel.kallsyms] [k] io_req_local_work_add
0.88% sock-test [kernel.kallsyms] [k] __io_run_local_work_loop
0.60% sock-test [kernel.kallsyms] [k] llist_reverse_order
0.14% sock-test [kernel.kallsyms] [k] __io_run_local_work
2.64% at ~46Gb/sec
and after this change:
1.08% sock-test [kernel.kallsyms] [k] io_req_local_work_add
1.03% sock-test [kernel.kallsyms] [k] __io_run_local_work
2.11% at ~53Gb/sec
which has less overhead even though that test run was faster. For a case
of having 1024 clients on a single ring:
2.22% sock-test [kernel.kallsyms] [k] llist_reverse_order
0.84% sock-test [kernel.kallsyms] [k] __io_run_local_work_loop
0.42% sock-test [kernel.kallsyms] [k] io_req_local_work_add
0.02% sock-test [kernel.kallsyms] [k] __io_run_local_work
3.50% at ~24Gb/sec
we start to see the llist reversing taking a considerable amount of
time, and the total add+run task_work overhead is around 3.5%. After
the change:
0.90% sock-test [kernel.kallsyms] [k] __io_run_local_work
0.42% sock-test [kernel.kallsyms] [k] io_req_local_work_add
1.32% at ~26Gb/sec
most of that overhead is gone, and performance is better as well.
Caleb Sander Mateos <csander@purestorage.com> reports that it improves
the performance of a ublk 4kb workload by 4% [1], while testing v1 of
this patchset.
[1] https://lore.kernel.org/io-uring/CADUfDZr-MMYBaP-e+y9+xuRhuiunO2sBTUCmwZyd7AgT8sVtiQ@mail.gmail.com/
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Local task_work is currently using llists for managing the work,
but that's a LIFO type of list. This means that running this task_work
needs to reverse the list first, to ensure fairness in running the
queued items.
Add a lockless FIFO queued, based on Dmitry Vyukov's intrusive MPSC
node-based queue algorithm, modified with an externally held consumer
cursor and conditional stub reinsertion. See comments in the header.
Producers are wait-free: a push is a single xchg() on the queue tail,
which serializes concurrent producers and defines the FIFO order, plus
a store linking the node to its predecessor. There are no cmpxchg retry
loops, and pushing is safe from any context, including hardirq.
The cost of linked list FIFO ordering is that a push publishes the node
in two steps - the xchg() makes it visible as the new tail before the
subsequent store links it into the chain that is reachable from the
head. A consumer hitting that window gets a NULL from mpscq_pop() while
mpscq_empty() reports false, and must retry later rather than treat the
queue as empty. The window is two instructions wide, but a producer can
get preempted inside it, so the consumer must not busy wait on it.
The consumer side supports a single consumer at a time, with callers
providing their own serialization. A stub node, which also defines the
empty state (tail == stub), allows the consumer to detach the final
node without racing against producer link stores: that node is only
handed out once the stub has been cmpxchg'ed back in as the tail. This
also guarantees that the previous tail returned by mpscq_push() cannot
get freed before that push has linked it, making it always valid for
comparisons.
The consumer cursor is deliberately not part of the queue struct - the
caller owns it and passes it to mpscq_pop(). This is done to separate
the consumer and producers cacheline. The cursor is written for every
popped entry, and keeping it on the same cacheline as ->tail would have
the consumer invalidating the line that producers need for every push.
Keeping it external lets the caller place it with its own consumer side
data instead.
Reviewed-by: Caleb Sander Mateos <csander@purestorage.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Not required right now, as io_req_local_work_add() already calls this
helper with the RCU read lock held. But in preparation for that not
being the case, grab it locally.
Reviewed-by: Caleb Sander Mateos <csander@purestorage.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux
Pull io_uring fixes from Jens Axboe:
- Tweak for an off-by-one in the CQ ring accounting for the min wait
support.
- Don't truncate end buffer length for a bundle, as the transfer might
not happen. It's not required in the first place, as the completion
side handles this condition already.
* tag 'io_uring-7.1-20260611' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:
io_uring/wait: fix min_timeout behavior
io_uring/kbuf: don't truncate end buffer for bundles
|
|
This member is only ever assigned, never read. Kill it.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Commit:
809b997a5ce9 ("x86-64/arm64/powerpc: clean up and rename __copy_from_user_flushcache")
sanitized that any provided copy helper should separately validate
destination and source addresses, but we should also ensure that
anything that is retrieved from a buffer is validated upfront. For ring
provided buffers, always include an access_ok() when grabbing a new
buffer.
Fixes: c7fb19428d67 ("io_uring: add support for ring mapped supplied buffers")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
So far IORING_RECVSEND_FIXED_BUF is only honoured on the SEND_ZC path,
even though the import wiring is already present for plain send and
completely absent for recv. Targets such as ublk's NBD backend want to
push/pull I/O data directly to/from an io_uring registered buffer over a
plain send/recv on a TCP socket.
Wire IORING_RECVSEND_FIXED_BUF into the plain IORING_OP_SEND and
IORING_OP_RECV paths:
- Accept the flag in SENDMSG_FLAGS / RECVMSG_FLAGS and, at prep time,
restrict it to the non-vectorized IORING_OP_SEND / IORING_OP_RECV
opcodes. It is mutually exclusive with buffer select, bundles and
(for recv) multishot, and records sqe->buf_index.
- For recv, set REQ_F_IMPORT_BUFFER in setup so the registered buffer
is imported lazily at issue time, mirroring the send path.
- In io_send()/io_recv(), import the registered buffer via
io_import_reg_buf() (ITER_SOURCE for send, ITER_DEST for recv) and
clear REQ_F_IMPORT_BUFFER. The resulting bvec iter persists in
async_data, so MSG_WAITALL partial send/recv retries resume at the
right offset.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Link: https://patch.msgid.link/20260608142511.659240-2-ming.lei@redhat.com
[axboe: combine flags checks]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
The wakeup condition if a min timeout is present and has expired is that
at least _one_ CQE was posted. Thus set the cq_tail target to
->cq_min_tail + 1. Without this commit a spurious wakeup can result in a
premature wakeup because io_should_wake() will return true even if _no_
CQE was posted at all.
Cc: Tip ten Brink <tip@tenbrinkmeijs.com>
Fixes: e15cb2200b93 ("io_uring: fix min_wait wakeups for SQPOLL")
Cc: stable@vger.kernel.org
Signed-off-by: Christian A. Ehrhardt <lk@c--e.de>
Link: https://patch.msgid.link/20260606201120.1441447-1-lk@c--e.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
If buffers have been peeked for a bundle receive, the kernel will
truncate the end buffer, if the available length is shorter than the
buffer itself. This is unnecessary, as applications iterating bundle
receives must always use the minimum size of the buffer length and the
remaining number of bytes in the bundle. The examples in liburing do
that as well, eg examples/proxy.c.
If the kernel does truncate this buffer AND the current transfer fails,
then the buffer will be left with a smaller size than what is otherwise
available.
Just remove the buffer truncation, as it's not necessary in the first
place.
Link: https://lore.kernel.org/io-uring/CAAEr8jbY60noGj1fw_k91UJRBkyiRVoS6=nLhZ7Svwidjn4CAA@mail.gmail.com/
Reported-by: Federico Brasili <federico.brasili@gmail.com>
Cc: stable@vger.kernel.org
Fixes: 35c8711c8fc4 ("io_uring/kbuf: add helpers for getting/peeking multiple buffers")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
