| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| In the Linux kernel, the following vulnerability has been resolved:
io_uring: free the task context after a grace period
A task other than the owner can free a request, io_ring_exit_work()
running the local task_work of a DEFER_TASKRUN ring for example, or
io_do_iopoll() reaping completions. io_put_task() then takes the else
branch:
percpu_counter_sub(&tctx->inflight, 1);
if (unlikely(atomic_read(&tctx->in_cancel)))
wake_up(&tctx->wait);
put_task_struct(tctx->task);
The first line is what lets the owner leave io_uring_cancel_generic(),
and on exec it frees the tctx right after. Nothing holds that off for
the remaining three lines, the task reference only covers a tctx freed
along with the task_struct. It used to work out because
io_uring_del_tctx_node() took the uring_lock that the freeing side holds
across io_put_task(), until that moved to tctx_lock.
Free the tctx with kfree_rcu(), and enter an RCU read section in
io_put_task() before the count is dropped. |
| In the Linux kernel, the following vulnerability has been resolved:
io_uring/sqpoll: don't wait for the thread to exit
io_sq_thread_stop() wakes the SQPOLL thread and waits for it to exit.
The thread belongs to the thread group that created the ring, so a
SIGSTOP or the cgroup freezer holds it in get_signal(), and the wake up
doesn't change that. io_ring_exit_work() then sleeps in that wait until
someone continues or kills the process. iou_exit has 64 slots, so 64
closed SQPOLL rings of a stopped process keep every other ring from
being freed, for everyone.
Don't wait. Set the stop bit, wake the thread, and let it exit when it
gets to run again. The sqd can now outlive its last ring, so give it a
second refcount with one reference for the rings and one for the thread,
and free it with the last one. The thread can't reach a freed ring,
rings leave ctx_list before dropping their reference, and the io-wq hash
is refcounted. |
| In the Linux kernel, the following vulnerability has been resolved:
io_uring: defer foreign io-wq enqueues to task_work
An IOPOLL ring without DEFER_TASKRUN can be reaped by a task other than
the one that submitted the I/O. If a polled request is marked
REQ_F_REISSUE, io_free_batch_list() calls io_queue_iowq() with the
original tctx. Ordinary task exit can tear down tctx->io_wq after
draining only the tracked requests. The live request pins the task, but
not its io-wq, so the foreign task can race io_wq_put_and_exit() and
enqueue through a freed or NULL queue.
Forward foreign userspace enqueues to the owning task before accessing
tctx->io_wq. Use a dedicated task_work callback so reissues still go
straight to io-wq, and fail them if task_work is running in cancellation
mode. Preserve the existing PF_KTHREAD cancellation path. This also
covers the other callers of io_queue_iowq() without changing the
worker-owned enqueue path. |
| In the Linux kernel, the following vulnerability has been resolved:
io_uring/cancel: drop sqd->lock while SQPOLL waits
An exiting SQPOLL thread calls io_uring_cancel_generic() with sqd->lock
held, and keeps it across cond_resched() and the sleep until nothing is
in flight anymore. A zero-copy send notification can't be canceled and
stays pending as long as the peer doesn't read, so the lock is held for
as long as a user cares to. io_ring_exit_work() needs it to park the
thread, and so do IORING_REGISTER_IOWQ_AFF and IOWQ_MAX_WORKERS. With
64 such rings sharing the thread, every iou_exit slot sleeps on the lock
and no ring gets freed for anyone.
Drop the lock around cond_resched() and the sleep, like the main loop of
the thread does. Each walk of ctx_list is still under one hold, and
nothing is kept across the unlock. The thread drops its io-wq, frees its
tctx and clears sqd->thread under one hold after the loop, so whoever
holds the lock and finds sqd->thread set can still use them.
[axboe: adapted to the schedule_timeout() based loop] |
| A buffer overflow vulnerability exists in the web management service of the Ethernet Remote I/Os products. An unauthenticated remote attacker could exploit insufficient bounds checking of specially crafted HTTP requests, potentially causing the web management service to crash and resulting in a denial-of-service condition. |
| A buffer overflow vulnerability exists in the SNMP service of the Ethernet Remote I/Os products. An unauthenticated remote attacker could exploit improper handling of specially crafted SNMP requests, potentially leading to a denial-of-service condition on the affected device. |
| Loop with Unreachable Exit Condition ('Infinite Loop') vulnerability in benoitc hackney allows Excessive Allocation.
The Alt-Svc response header parser in src/hackney_altsvc.erl does not guarantee forward progress. When parse_token/2 receives a non-token, non-whitespace, non-comma byte (e.g. !, @, =, ;), it returns the input unchanged. skip_comma/1 also returns the buffer unchanged when the first byte is not a comma. parse_entries/2 then recurses with identical data, creating a tight infinite tail-recursive loop that pins a scheduler at 100% CPU. The calling process never returns.
The entry point parse_and_cache/3 is called synchronously in the connection process on every HTTP response. A single-byte Alt-Svc: ! response header is sufficient to trigger the hang; the header is fully controlled by any HTTP origin the client connects to.
This issue affects hackney: from 2.0.0-beta.1 before 4.0.1. |
| A flaw was found in Vim's netrw plugin. A crafted filename containing quote characters and expression fragments can break out of the quoted context during mark/unmark operations, allowing arbitrary Vimscript execution. This can be leveraged to run shell commands with the privileges of the user running Vim. |
| Unexpected Status Code or Return Value vulnerability in ninenines gun (gun_http module) allows a malicious HTTP server to force the client into raw protocol mode via an unsolicited 101 Switching Protocols response.
In gun_http:handle_inform/8, when a 101 Switching Protocols response is received over HTTP/1.1, the function verifies only that the Upgrade header is syntactically valid and that the stream reference is a plain reference(). It does not check whether the client ever sent an Upgrade or Connection: upgrade header on the corresponding request. Because this check is absent, any 101 response (solicited or not) causes gun to dispatch a gun_upgrade message to the caller and transition the entire connection to raw protocol mode.
A malicious or compromised HTTP server can send an unsolicited 101 response to any HTTP/1.1 request, causing the gun client to abandon HTTP framing for that connection. Once in raw mode, gun_raw applies no flow control (flow=infinity) and re-arms socket active mode after every received packet, so the server can flood the client with arbitrary bytes. These are forwarded as unbounded gun_data messages to the owner process, exhausting its mailbox and BEAM memory, ultimately crashing the VM.
This issue affects gun: from 2.0.0-pre.1 before 2.4.0. |
| A flaw was found in cifs-utils. This flaw allows an unprivileged local user to achieve privilege escalation. When a privileged service mounts a Common Internet File System (CIFS) share using a user-controlled credentials file, input values are not properly sanitized before being assembled into mount options. A local attacker can exploit this vulnerability to inject unauthorized mount options, allowing them to bypass permission checks and execute files with administrative (root) privileges. |
| Uncontrolled Resource Consumption vulnerability in ninenines gun (gun_http module) allows a malicious server to exhaust client memory via unbounded HTTP/1.1 response buffering.
In gun_http:handle/5, three clauses accumulate incoming TCP data into the connection's buffer field using binary concatenation with no upper-bound check: the head clause appends data until the \r\n\r\n header terminator is found; the body_chunked clause appends data whenever cow_http_te:stream_chunked/2 returns a more result indicating an incomplete chunk boundary; and the body_trailer clause appends data until the trailing \r\n\r\n is found. In each case, when the expected terminator never arrives, the enlarged binary is stored back into state and the process waits for more data, with no configurable or hard-coded ceiling on buffer size.
A malicious or compromised server can exploit this by sending a partial response that never completes. For example, a response may begin with HTTP/1.1 200 OK\r\nX-Pad: followed by an unbounded stream of arbitrary bytes, never sending the header terminator. The gun connection process will continuously append the incoming data to its buffer, causing unbounded heap growth. Because BEAM imposes no per-process heap limit by default, a single malicious connection can exhaust all available memory on the node, causing a node-wide out-of-memory crash.
This issue affects gun: from 1.0.0-pre.1 before 2.4.0. |
| Origin Validation Error vulnerability in ninenines gun (gun_http2 module) allows cross-origin cookie injection via unvalidated HTTP/2 PUSH_PROMISE authority.
In gun_http2:push_promise_frame/7, the :authority pseudo-header from an incoming PUSH_PROMISE frame is stored verbatim into the promised stream record without checking that it matches the connection's origin. When gun_http2:headers_frame/9 later processes the response headers for the promised stream, it calls gun_cookies:set_cookie_header/7 with the unvalidated server-supplied authority before any status branching and before user code can act. This violates RFC 7540 §10.6 / RFC 9113 §8.4, which require receivers to treat as a protocol error any push for a resource the server is not authoritative for.
A malicious or compromised HTTP/2 server can plant cookies scoped to arbitrary third-party domains into the client's shared cookie store. This enables session fixation attacks against those domains and, if the planted cookie overrides a legitimate session token, may result in account takeover. No user interaction beyond making a normal HTTP/2 request to the attacker-controlled server is required.
This issue affects gun: from 2.0.0-pre.2 before 2.4.0. |
| A flaw was found in Red Hat Ansible Automation Platform's automation-
controller. Four debug views that trigger the internal task, dependency, and
workflow schedulers are configured to allow any user (including unauthenticated
clients) and are routed in production builds because their URL include is not
gated on the debug setting. An unauthenticated remote attacker can repeatedly
invoke these endpoints to acquire the cluster-wide scheduler advisory lock;
because the legitimate scheduler acquires the same lock without waiting, the
attacker causes real scheduler runs to be skipped, stalling job dispatch for
all tenants, while also consuming controller web workers. The debug root view
additionally discloses the list of debug endpoints to unauthenticated callers. |
| Allocation of Resources Without Limits or Throttling vulnerability in phenixdigital phoenix_storybook allows unauthenticated denial of service via atom table exhaustion. LiveView event handlers and request parameters of the storybook convert client-supplied strings to atoms. Atoms are never garbage-collected, so a client that sends enough unique strings fills the atom table, and the BEAM node terminates.
PhoenixStorybook.ExtraAssignsHelpers.handle_set_variation_assign/3 and handle_toggle_variation_assign/3 call String.to_atom/1 on parameter keys, attribute names, variation ids, and values of attributes declared as :atom or :boolean. PhoenixStorybook.Story.Playground.handle_event/3 and PhoenixStorybook.StoryLive do the same for tab names, theme names, and playground field names. Releases published as phx_live_storybook have the same calls in the PhxLiveStorybook namespace, at first in PhxLiveStorybook.EntryLive.
This issue affects phoenix_storybook: from 0.5.0 before 1.1.0; phx_live_storybook: from 0.2.0 onward. |
| A flaw has been found in H3C ICG2000 Unified Communication Gateway 2023. This issue affects some unknown processing of the file /wnm/check.j of the component Web Management Module. Executing a manipulation can lead to missing authentication. The attack may be performed from remote. The exploit has been published and may be used. Upgrading to version R9141PXX is capable of addressing this issue. It is recommended to upgrade the affected component. The vendor confirms: "We have confirmed this issue is a known issue and has been fixed in R9141PXX versions." |
| Code Injection vulnerability in phenixdigital phoenix_storybook allows unauthenticated remote code execution via HEEx template injection in the component playground. A client can send attribute values over the playground WebSocket. A value that closes the attribute quote and adds a HEEx expression is evaluated as Elixir code on the server.
PhoenixStorybook.Story.PlaygroundPreviewLive.handle_event/3 accepts attribute names and values from the client and stores them through PhoenixStorybook.ExtraAssignsHelpers.handle_set_variation_assign/3. PhoenixStorybook.Rendering.ComponentRenderer.attributes_markup/1 interpolates string values into a HEEx template without escaping. The template is compiled with EEx.compile_string/2 and evaluated. Releases published as phx_live_storybook have the same code in the PhxLiveStorybook namespace, where the first client entry point is PhxLiveStorybook.Entry.Playground.handle_event/3.
This issue affects phoenix_storybook: from 0.5.0 before 1.1.0; phx_live_storybook: from 0.3.0 onward. |
| In the Linux kernel, the following vulnerability has been resolved:
drm/vc4: Drop the BO reservations outside of job_lock
vc4_queue_submit() calls drm_exec_fini() with job_lock held and interrupts
disabled. Besides unlocking the reservations, drm_exec_fini() frees the
object array with kvfree(), which asks for preemptible task context or a
non-NMI interrupt, and may reach might_sleep() through vfree().
Considering that an array that exceeds a page can take the vmalloc
fallback, call drm_exec_fini() once job_lock is dropped. The reservations
are then held across the job kick, which is harmless as nothing in the
completion path takes a BO reservation.
Note that vc4_attach_fences() has to stay inside the critical section, as
it walks exec, which the job done worker is free to release as soon as
job_lock is dropped. |
| In the Linux kernel, the following vulnerability has been resolved:
drm/loongson: Fix double free on BO initialization failure
If ttm_bo_init_validate() fails, it drops the buffer object's reference
and cleans it up through the supplied destroy callback. lsdc_bo_destroy()
releases the GEM object and frees the enclosing lsdc_bo, so freeing it
again in lsdc_bo_create() causes a double free on a synchronous failure
path.
Let TTM own the cleanup after initialization has started and return the
error directly. Keep the explicit free on drm_gem_object_init() failure,
which occurs before ownership is passed to TTM.
Tested on LS7A2000 by making drm_vma_offset_add() return -ENOSPC for one
selected dumb-buffer creation. The error reached userspace, the destroy
callback ran once, and no handle was published or tracked BO retained.
Normal buffer creation, zeroing, mapping, readback and release passed.
AI assistance was used for the ownership analysis, fix, fault-injection
tools, build and test execution. |
| In the Linux kernel, the following vulnerability has been resolved:
mm/huge_memory: fix list corruption when deferred_split_scan() races with deferred_split_folio()
Since the deferred split queue became a list_lru, the isolate callback
moves the folios it took a reference on to an on-stack list, linked
through folio->_deferred_list, and deferred_split_scan() takes them off
that list with list_del_init() without holding the list_lru lock.
deferred_split_folio() does not know about the on-stack list. It only
holds the list_lru lock when it looks at the same list_head, and it can
run for such a folio: a THP is queued when it is mapped, as a candidate
for the underused check, and is queued "again" when a part of it gets
unmapped:
CPU0 deferred_split_scan() CPU1 deferred_split_folio()
list_del_init(&folio->_deferred_list)
next->prev = prev
prev->next = next
entry->next = entry
list_lru_lock_irqsave()
folio_set_partially_mapped()
__list_lru_add()
list_empty(item) -> true
list_add_tail(item, &l->list)
item->next = &l->list
item->prev = tail
entry->prev = entry
The folio now sits on the list_lru with ->prev pointing to itself. The
next removal leaves the neighbours pointing to it, and they still do so
after the folio has been freed. With CONFIG_DEBUG_LIST:
list_del corruption. prev->next should be ffd4000000270090, but was
ffd4000000278090. (prev=ffd4000000270090)
Fix it by grabbing the lock around list_del_init() so
deferred_split_folio() cannot race with it. |
| In the Linux kernel, the following vulnerability has been resolved:
vhost-vdpa: fix pinned page leak in vhost_vdpa_pa_map() error path
When pin_user_pages() or vhost_vdpa_map() fails, vhost_vdpa_pa_map()
is supposed to unpin the pages pinned after the last successfully
mapped chunk, i.e. the range [map_pfn, last_pfn]. The unpin is guarded
by nchunks, but nchunks is reset to 0 every time a chunk is mapped from
inside the current page list, so a failure after any such boundary
leaves that pending range pinned forever. With physically scattered
memory nearly every page list contains a boundary, so the leak happens
on practically every failed IOTLB update.
The pending range is non-empty exactly when a page has been pinned,
which last_pfn already tells us, so drop nchunks and test last_pfn. |