| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| llama.cpp b11003 before b12556 contains an integer overflow in the hrm_text loader load_arch_hparams() that allows attackers to bypass the block count check using crafted GGUF metadata. Attackers can supply a GGUF model whose layer factors wrap modulo 2^32, causing load_arch_tensors() to write tensor pointers past model.layers, corrupting heap memory and potentially enabling code execution. |
| A vulnerability was found in ByteDance Lark Desktop 8.0.0.14. This vulnerability affects the function FUN_18001b770/FUN_18001ba40 in the library backend.dll. Performing a manipulation of the argument day results in signed to unsigned conversion error. The attack needs to be approached locally. The exploit has been made public and could be used. The vendor was contacted early about this disclosure but did not respond in any way. |
| TLS protocol dissector memory leak in 4.6.0 to 4.6.9 and 4.4.0 to 4.4.19 allows denial of service |
| Issue summary: Calling EVP_EncodeBlock() with an input length close to INT_MAX
may overflow the calculation of the encoded output size and write past the end
of the caller-supplied output buffer.
Impact summary: No path by which an attacker can trigger this has been
identified. An application that passes an input length near INT_MAX to
EVP_EncodeBlock() could suffer a heap buffer overflow, crashing or corrupting
its own memory, but it must itself supply such a length.
CWE: CWE-190: Integer Overflow or Wraparound
Description: EVP_EncodeBlock() base64-encodes a buffer of a caller-supplied
length into an output buffer that the caller allocates. The caller is expected
to reserve 4*((len+2)/3)+1 bytes. For input lengths close to INT_MAX, the
arithmetic that computes the encoded length overflows a signed integer, so the
function writes more data than the documented output size. The result is a
heap buffer overflow in the calling process.
Reaching this requires an application to pass an input length near INT_MAX to
EVP_EncodeBlock(). OpenSSL's TLS, QUIC and CMS implementations never encode
buffers of that size, and the command line tools cannot practically be made to
do so: the limits on reading a proxy password from a file and on bignum sizes
prevent it. An application that accepts an attacker-controlled length and
forwards it to EVP_EncodeBlock() without its own bound check is exposed.
The severity is Low because no attacker-controlled path to the overflow is
known. The caller must already hold a buffer of about two gigabytes before the
defect can be reached.
FIPS impact: no
The affected code is outside the FIPS module boundary. |
| In Eclipse Mosquitto versions 1.6 to 2.0.22 and 2.1.0 to 2.1.2, an MQTT v5.0 client connecting with a will-delay-interval that subsequently disconnects with a shorter interval causes all delayed wills to be suppressed. |
| Go-Guerrilla SMTP Daemon is a lightweight SMTP server written in Go. In 1.6.7 and earlier, according to the repository advisory's structured affected range, server.go defines the seven-byte cmdXCLIENT token, but the XCLIENT handler slices input[8:] after matching only that seven-byte verb, so a bare command causes an out-of-bounds panic in handleClient and terminates the daemon because that connection goroutine does not recover the panic. The issue is reachable by an unauthenticated client when the non-default xclient_on setting is enabled; no HELO command is required. The repository fix is present in version 1.6.7-1. |
| Dell PowerStore contains an integer underflow vulnerability. An attacker could exploit this vulnerability, potentially leading to denial of service of a cryptographic service. |
| In Eclipse Mosquitto versions 2.1 through 2.1.2, a listener configured with `protocol websockets` mishandles empty WebSocket data frames. `net__read_ws()` in `lib/net_ws.c` returns the number of framing bytes consumed (the 4 masking-key bytes) for a data frame with a zero-length payload, without writing any bytes to the caller’s buffer. The MQTT packet reader in `packet__read_single()` (`lib/packet_mosq.c`) trusts that return value as a payload byte count. A remote client can therefore advance the read position over an uninitialised heap allocation, causing the broker to process and forward uninitialised heap memory (information disclosure, including heap and libc pointers), and can underflow the unsigned `to_process` counter so that subsequent attacker-controlled frame data is written past the end of the allocation (heap out-of-bounds write). The out-of-bounds write is reachable from the CONNECT packet before any authentication check, so an unauthenticated client can corrupt the heap and crash the broker even when `allow_anonymous false` is set. Combining the disclosure and the write allows remote code execution as the broker’s user by any client able to establish a session. |
| Numeric truncation error in Windows Bluetooth Service allows an authorized attacker to elevate privileges locally. |
| Heap-based buffer overflow in Windows Local Security Authority Subsystem Service (LSASS) allows an authorized attacker to elevate privileges locally. |
| Kafka protocol dissector infinite loop in 4.6.0 to 4.6.9 and 4.4.0 to 4.4.19 allows denial of service |
| HTSlib is a library for reading and writing bioinformatics file formats. Prior to 1.21.2, 1.22.3, 1.23.2, and 1.24, the BCF reader can experience an integer overflow in a record-length check that verifies whether a record contains enough bytes for the expected number and size of stored items. A crafted BCF record can bypass this check, causing reads beyond the stored data and, in some cases, writes beyond the heap buffer when the reader sets the least significant bit of accessed bytes. Opening such a file can crash the process, corrupt data or heap structures, and may permit arbitrary code execution. This issue is fixed in versions 1.21.2, 1.22.3, 1.23.2, and 1.24. |
| Integer overflow in Autofill in Google Chrome on on Android prior to 158.0.8127.36 allowed a remote attacker leveraging social engineering to obtain cross-origin data via a crafted HTML page. (Chromium security severity: Medium) |
| OFFIS DCMTK 3.5.4 through 3.7.1 contains an integer overflow in DJCodecEncoder::encodeTrueLossless size checks that allows attackers to trigger a heap out-of-bounds read. Attackers can supply a crafted DICOM file with large Columns, Rows and SamplesPerPixel values that, when compressed with true-lossless JPEG, crashes the application. |
| In the Linux kernel, the following vulnerability has been resolved:
9p: don't decrement a server-owned zero nlink
A 9P2000.L server can report nlink=0 for a regular file that still
exists. On a cache=loose or cache=fscache mount, lookup stores that
value with set_nlink(). A later successful unlink calls
v9fs_dec_count(), which still calls drop_nlink() for a non-directory.
drop_nlink() warns, and the unsigned link count wraps to 4294967295.
fstat() on a descriptor opened before the unlink shows 0 before and
4294967295 after.
The warning on the unpatched kernel, trimmed:
WARNING: fs/inode.c:408 at drop_nlink+0xac/0xd0, CPU#1: trigger/188
CPU: 1 UID: 0 PID: 188 Comm: trigger Tainted: G N 7.3.0-rc4-00537-ga3ff15db6820 #1 PREEMPT(lazy)
RIP: 0010:drop_nlink+0xac/0xd0
Call Trace:
<TASK>
v9fs_remove+0x529/0x750
vfs_unlink+0x2f0/0xbd0
filename_unlinkat+0x340/0x6c0
__x64_sys_unlink+0x46/0x70
do_syscall_64+0xe8/0x4f0
entry_SYSCALL_64_after_hwframe+0x77/0x7f
</TASK>
---[ end trace 0000000000000000 ]---
Directories whose cached nlink is at most 2 already skip this
decrement, because the exported filesystem may not maintain nlink
(commit ac89b2ef9b559). cache=none skips it as well (commit
574aa0b479947). cache=mmap does not take this path: it sets neither
CACHE_META nor CACHE_LOOSE. A regular file on cache=loose or
cache=fscache still does. Its count came from GETATTR and may
already be zero, so skip the decrement when i_nlink is already 0.
Verified on 7.3-rc4-00537-ga3ff15db6820 with only this change. A
server that advertises nlink=0 and accepts the unlink no longer
warns, and the open descriptor stays at nlink 0. An honest server
on the same mount still goes from 1 to 0. A rejected unlink leaves
the count unchanged. cache=none is unchanged. With panic_on_warn
the unpatched kernel panics on the warning; the patched kernel does
not. |
| In the Linux kernel, the following vulnerability has been resolved:
mm/gup: check for overflow when converting nr_pages to bytes in GUP-fast
The overflow check in gup_fast_fallback() only covers the addition to
start. The preceding nr_pages << PAGE_SHIFT can already have truncated
the length on 32bit.
Users such as RapidIO's rio_dma_transfer() can derive such page counts
from a 64-bit userspace length.
Let's just check_shl_overflow() to reject any such overflow with
-EOVERFLOW early. |
| In the Linux kernel, the following vulnerability has been resolved:
SUNRPC: Copy the deferred RPC Call from the head buffer
svc_defer() snapshots the RPC Call for replay once a pending cache
upcall resolves. It copies the Call starting (rq_arg.len -
head[0].iov_len) bytes before head[0].iov_base, a back-up that once
covered a transport header. No transport has had one since commit
983084b2672c ("SUNRPC: Remove svc_rqst::rq_xprt_hlen"), so the two
lengths are meant to be equal.
An RPCSEC_GSS Call breaks that assumption. Unwrapping the integrity
service truncates rq_arg.len by the length of the trailing checksum,
and unwrapping the privacy service leaves rq_arg.len short by an
amount the client sets with trailing bytes after the wrap token.
Neither adjusts head[0].iov_len. Both lengths are size_t, so the
subtraction underflows and the copy starts past head[0].iov_base.
Every deferred integrity Call replays a shifted snapshot.
A deferred replay skips GSS verification, so a shifted snapshot runs
as an authenticated Call. With the privacy service the offset is
client-chosen. A forged Call placed in that window, naming another
user's GSS context handle, runs under that identity.
Copy the Call from head[0].iov_base. Refuse to defer a Call whose
length runs past the head buffer, so the copy cannot read out of
bounds. |
| In the Linux kernel, the following vulnerability has been resolved:
xfs: don't subtract excess usage twice when computing agf_btreeblks
LOLLM noticed that the free space btree repair code calculates
agf_btreeblks from the number of blocks we wrote to rebuild the free
space btrees (minus the roots). That's correct. However, it observes
that we subtract the "excess" reservation, which is not correct. Fix
this accounting problem. |
| In the Linux kernel, the following vulnerability has been resolved:
mm/page-writeback: avoid 32-bit overflow in dirty threshold shares
The ratios used by __wb_calc_thresh() are scaled to parts per million, but
the threshold and ratio operands are unsigned long. On 32-bit kernels,
their multiplication can wrap before the result is assigned to u64.
For example, with a threshold of 25600 pages and the default maximum ratio
of 1000000, wb_max_thresh becomes 4125 pages instead of 25600. The shared
threshold and minimum share calculations can overflow in the same way,
producing incorrect per-writeback dirty limits.
Use a u64 threshold for the three ratio multiplications and div_u64() to
retain the existing division order and rounding. |
| In the Linux kernel, the following vulnerability has been resolved:
mm: fix address wraparound when scanning large folio PTEs
numa_rebuild_large_mapping() and do_swap_page() can be tricked into
generating an address wraparound, crashing the kernel. All that's required
is to map a tail page of a large folio close enough to the start/end of
the address space to e.g., make
unsigned long addr_start = addr - (nr << PAGE_SHIFT);
to wrap around.
Let's fix it by providing a better helper, folio_scan_range(), that
takes care of staying within VMA and page table boundaries while
handling possible wraparounds.
There are some nasty details, like not being able to use pmd_addr_end()
in this context. Add some helpful comments for these cases. |