| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| vLLM is a library for LLM inference and serving. From 0.12.0 to before 0.24.0, sending a pure prompt embeds payload in a /v1/completions request with a model using M-RoPE causes EngineCore to fail an assertion and fatally crash, shutting down the entire server application. Any remote user who is authorized to make a /v1/completions request can make such a request and induce a crash. This issue is fixed in version 0.24.0. |
| In the Linux kernel, the following vulnerability has been resolved:
ext4: replace BUG_ON with proper error handling in ext4_read_inline_folio
Replace BUG_ON() with proper error handling when inline data size
exceeds PAGE_SIZE. This prevents kernel panic and allows the system to
continue running while properly reporting the filesystem corruption.
The error is logged via ext4_error_inode(), the buffer head is released
to prevent memory leak, and -EFSCORRUPTED is returned to indicate
filesystem corruption. |
| In the Linux kernel, the following vulnerability has been resolved:
signal: clear JOBCTL_PENDING_MASK for caller in zap_other_threads()
When a multi-threaded process receives a stop signal (e.g., SIGSTOP),
do_signal_stop() sets JOBCTL_STOP_PENDING and JOBCTL_STOP_CONSUME on all
threads and sets signal->group_stop_count to the number of threads. If
one of the threads concurrently calls execve(), de_thread() invokes
zap_other_threads() to kill all other threads. zap_other_threads()
aborts the pending group stop by resetting signal->group_stop_count to 0
and clears the JOBCTL_PENDING_MASK for all other threads. However, it
fails to clear the job control flags for the calling thread.
When execve() completes, the calling thread returns to user mode and
checks for pending signals. Seeing the stale JOBCTL_STOP_PENDING flag,
it calls do_signal_stop(), which invokes task_participate_group_stop().
Since JOBCTL_STOP_CONSUME is still set, it attempts to decrement the
already-zero signal->group_stop_count, triggering a warning:
sig->group_stop_count == 0
WARNING: CPU: 1 PID: 6475 at kernel/signal.c:373
task_participate_group_stop+0x215/0x2d0
Call Trace:
<TASK>
do_signal_stop+0x3be/0x5c0 kernel/signal.c:2619
get_signal+0xa8c/0x1330 kernel/signal.c:2884
arch_do_signal_or_restart+0xbc/0x840 arch/x86/kernel/signal.c:337
exit_to_user_mode_loop+0x8c/0x4d0 kernel/entry/common.c:98
do_syscall_64+0x33e/0xf80 arch/x86/entry/syscall_64.c:100
entry_SYSCALL_64_after_hwframe+0x77/0x7f
</TASK>
Fix this race condition by clearing the JOBCTL_PENDING_MASK for the
calling thread in zap_other_threads(), ensuring it does not retain any
stale job control state after the thread group is destroyed. This aligns
with other functions that tear down a thread group and abort group
stops, such as zap_process() and complete_signal(), which correctly
clear these flags for all threads including the current one. |
| In the Linux kernel, the following vulnerability has been resolved:
libceph: handle rbtree insertion error in decode_choose_args()
A message of type CEPH_MSG_OSD_MAP contains an OSD map that itself
contains a CRUSH map. The received CRUSH map may optionally contain
choose_args that get decoded in decode_choose_args(). In this function,
num_choose_arg_maps is read from the message, and a corresponding number
of crush_choose_arg_maps gets decoded afterwards. Each
crush_choose_arg_map has a choose_args_index, which serves as the key
when inserting it into the choose_args rbtree of the decoded crush_map.
If a (potentially corrupted) message contains two crush_choose_arg_maps
with the same index, the assertion in insert_choose_arg_map() triggers a
kernel BUG when trying to insert the second crush_choose_arg_map.
This patch fixes the issue by switching to the non-asserting rbtree
insertion function and rejecting the message if the insertion fails.
[ idryomov: changelog ] |
| In the Linux kernel, the following vulnerability has been resolved:
accel/ethosu: reject NPU_OP_RESIZE commands from userspace
NPU_OP_RESIZE is a U85-only command that the driver does not yet
implement. The existing WARN_ON(1) placeholder fires unconditionally
whenever userspace submits this command via DRM_IOCTL_ETHOSU_GEM_CREATE,
causing unbounded kernel log spam.
If panic_on_warn is set the kernel panics, giving any unprivileged user
with access to the DRM device a trivial denial-of-service primitive.
Replace the WARN_ON(1) with an explicit -EINVAL return so the ioctl
rejects the command before it reaches hardware. |
| In EmberZNet v9.0.2 and earlier, malformed Color Control messages can lead to asserts that terminate the process. These messages must come from a device that has already joined the network. Only devices supporting the Color Control cluster may be impacted. |
| In EmberZNet v9.0.2 and earlier, malformed Color Control messages can lead to asserts that terminate the process. These messages must come from a device that has already joined the network. Only devices supporting the Color Control cluster may be impacted. |
| CWE-617 Reachable Assertion vulnerability exists that could allow an authenticated attacker to trigger a denial-of-service condition, impacting system availability when a specially crafted request is sent to a vulnerable network-exposed service. |
| NSD from version 4.13.0 has a heap use-after-free bug in logging errors on TLS connections, causing a crash of the server process, which can be triggered trivially by sending a DNS query over a DoT connection, and closing the connection without reading the response. |
| vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.22.0, an assert-based security check in vLLM's activation function loading allows any unauthenticated attacker to achieve arbitrary code execution on the server by publishing a malicious HuggingFace model, when vLLM runs in Python optimized mode (python -O or PYTHONOPTIMIZE=1). This vulnerability is fixed in 0.22.0. |
| In the Linux kernel, the following vulnerability has been resolved:
libceph: replace overzealous BUG_ON in osdmap_apply_incremental()
If the osdmap is (maliciously) corrupted such that the incremental
osdmap epoch is different from what is expected, there is no need to
BUG. Instead, just declare the incremental osdmap to be invalid. |
| The $_internalConvertBucketIndexStats stage used PauseExecution as a way to signal "skip this document" when an index stats conversion failed. But PauseExecution is not a general purpose skip mechanism, but rather a TeeBuffer-internal signal used solely by $facet to coordinate its sub-pipelines. When this stage is placed before $facet in a pipeline, TeeBuffer receives the unexpected PauseExecution from upstream and hits a hard invariant assertion, crashing mongod. |
| An authenticated user can cause a MongoDB server to crash or return incorrect results by creating documents that interfere with internal metadata processing during query execution. This stems from insufficient separation between user-controlled document fields and internal metadata in certain execution paths. |
| Adding fromRouter:true and runtimeConstants.userRoles could cause aggregations to crash mongodb server. |
| In the Linux kernel, the following vulnerability has been resolved:
drm/amdgpu/sdma4: replace BUG_ON with WARN_ON in fence emission
sdma_v4_0_ring_emit_fence() contains two BUG_ON(addr & 0x3) assertions
that verify fence writeback addresses are dword-aligned. These
assertions can be reached from unprivileged userspace via crafted
DRM_IOCTL_AMDGPU_CS submissions, causing a fatal kernel panic in a
scheduler worker thread.
Replace both BUG_ON() calls with WARN_ON() to log the condition without
crashing the kernel. A misaligned fence address at this point indicates
a driver bug, but crashing the kernel is never the correct response when
the assertion is reachable from userspace.
The CS IOCTL path is the correct place to filter invalid submissions;
the ring emission callback is too late to do anything about it.
(cherry picked from commit b90250bd933afd1ba94d86d6b13821997b22b18e) |
| Nimiq is a Rust implementation of the Nimiq Proof-of-Stake protocol based on the Albatross consensus algorithm. Prior to version 1.5.0, a remote peer can crash any full node by sending a RequestBatchSet message containing the genesis block's hash. The handler calls get_epoch_chunks which iterates backwards through macro blocks using Policy::macro_block_before. When it reaches the genesis block number, macro_block_before panics with "No macro blocks before genesis block". This issue has been patched in version 1.5.0. |
| A vulnerability has been found in some Dahua products could allow an authenticated remote attacker to send a specially crafted packet, triggering an exception that causes the system to reboot unexpectedly, resulting in a denial of service. |
| A vulnerability has been found in some Dahua products could
allow an unauthenticated remote attacker to send a specially crafted packet,
triggering an exception that causes the system to reboot unexpectedly,
resulting in a denial of service. |
| This issue can occur when running an aggregation pipeline that uses the internal $exchange stage configured with key-range partitioning and order-preserving delivery. If a single key range produces enough documents to fill its exchange buffer (that is, many results are routed to the same consumer), the server reaches the code path where a full per-consumer buffer is detected but the internal "high watermark" for that key range is not updated as intended. |
| When using $changestreams and $_requestReshardingResumeToken with the exchange option the server hits an invariant which causes the server to crash. There are no special privileges needed. The user must be logged in to issue the statement. |
