A vulnerability in the Linux kernel's netfilter subsystem related to nf_conntrack_expect was resolved. The issue involved occasional leaking of uninitialized data to userspace via ctnetlink due to missing zeroing at allocation time. The fix involved zeroing the expectation structure at allocation to prevent data leakage. This vulnerability affects the kernel's connection tracking expectations, which are infrequently used. No CVSS score is available for this issue.
A vulnerability in the Linux kernel MMC core was resolved by adding validation for the host-provided max_segs field. The max_segs field is an unsigned short, and if set to an excessively large value by a host driver, it may be truncated to zero. This causes a zero-size allocation request in mmc_alloc_sg(), leading to undefined behavior and a kernel crash due to invalid memory access. The fix adds a validation check to detect zero sg_len and returns an error instead of crashing.
A vulnerability in the Linux kernel's chips-media wave5 driver was fixed by adding range checks to the dec_output_info structure. Without these checks, invalid data could cause the driver to write to invalid memory, leading to a buffer overflow condition. This issue was identified by a static analysis tool (smatch) and corrected to prevent potential memory corruption.
A vulnerability in the Linux kernel's 8250 serial driver related to the recognition of UPIO_AU (type 4) on MIPS Alchemy systems has been resolved. The issue caused warnings during the probe of the 8250 driver due to unsupported UART type 4. The fix involves recognizing UPIO_AU in the set_io_from_upio() function and taking no action since parameters are already set in the platform-specific setup. This fix has been tested on the Alchemy Au1300 platform.
A vulnerability in the Linux kernel's ntfs3 driver related to improper handling of set_blocksize failures has been resolved. The ntfs3 driver uses buffer_heads, which do not handle block sizes larger than PAGE_SIZE correctly. This can cause a kernel BUG_ON failure during mounting due to an invalid offset check in folio_set_bh on the first __bread_gfp call.
A vulnerability in the Linux kernel's drm/gma500 driver for Oaktrail HDMI I2C reads could cause indefinite waiting during I2C transactions due to improper handling of timeouts and interrupts. The issue has been resolved by returning appropriate error codes when timeouts or interrupts occur during the I2C transaction wait, preventing infinite loops.
A vulnerability in the Linux kernel's IMA subsystem was resolved that caused errors during early boot when attempting to change file extended attributes (xattrs) on read-only filesystems. Specifically, on ext4 filesystems, attempts to change xattrs triggered a failing ext4 transaction, leading to kernel warnings and potential boot issues. The fix ensures the system detects the read-only state early and returns an error without starting a failing transaction.
A vulnerability in the Linux kernel's isofs filesystem code was resolved. The issue involved improper handling of set_blocksize failures because isofs uses buffer_heads, which do not handle block sizes larger than PAGE_SIZE correctly. This could cause a kernel BUG_ON failure during mounting due to an offset exceeding the folio size.
A vulnerability in the Linux kernel's mac80211_hwsim component allowed undersized HWSIM_ATTR_TX_INFO attributes to be accepted, leading to out-of-bounds reads. The issue was due to insufficient length validation of the attribute payload before casting and reading it as a struct hwsim_tx_rate pointer. This flaw has been resolved by enforcing exact length checks in the attribute policy to reject malformed inputs.
A vulnerability in the Linux kernel's GICv5 interrupt controller driver caused a delay in priority dropping after interrupt activation, potentially blocking new interrupts and causing a denial of service. The issue was due to the priority drop being performed during interrupt deactivation instead of immediately after acknowledgment. This was fixed by moving the priority drop to occur right after interrupt acknowledgment in the main interrupt handler.
A vulnerability in the Linux kernel arm64 kprobes subsystem allowed improper handling of reentered kprobes during single-stepping, potentially causing incorrect interrupt state handling. The issue was that a kprobe hit in the KPROBE_HIT_SS state was treated as unrecoverable, leading to permanent interrupt disablement in some cases. This has been fixed by allowing reentry in KPROBE_HIT_SS to be handled correctly and preserving the interrupt state across nested kprobes.
A use-after-free vulnerability in the Linux kernel's ksmbd component was resolved by ensuring that stale channels are always removed from all sessions on connection teardown. Previously, channels could remain linked to sessions after a connection's binding flag was cleared, leading to dereferencing freed memory and leaking sessions.
Chatwoot versions before 4.17.0 have an authorization bypass vulnerability that allows authenticated agent bots to access other accounts by using a foreign contact inbox source_id. This enables attackers to create or retrieve conversations and post messages on behalf of other accounts.
Heimdall versions prior to 0.17.17 improperly validate client IP addresses extracted from Forwarded and X-Forwarded-For headers when the trusted_proxies option is enabled. This allows arbitrary, malformed, or obfuscated IP values to be accepted and processed. The Forwarded header parser also incorrectly splits quoted strings, causing parsing errors. This vulnerability has been fixed in version 0.17.17.
Heimdall versions prior to 0.17.17 have an improper input validation vulnerability in proxy mode. The software constructs the Forwarded HTTP header by directly inserting the incoming request's Host header without sanitizing commas or semicolons. This allows attackers to inject additional parameters into the Forwarded header, potentially spoofing IP addresses for upstream services. The issue is patched in version 0.17.17.
CVE-2026-55422 is a medium severity cross-site scripting (XSS) vulnerability in the conda-forge.github.io website. The issue arises because two React components render externally-sourced strings directly as HTML using dangerouslySetInnerHTML without sanitization. These strings include raw output from build logs, patch outputs, and tracebacks generated by the conda-forge autotick bot. No patches are publicly available at the time of publication.
A vulnerability in pghoard, a PostgreSQL backup daemon, prior to version 2.7.1 causes sensitive database connection information, including username and password, to be logged at the debug level when using .pgpass. This issue has been fixed in version 2.7.1.
pghoard is a PostgreSQL backup daemon and restore tooling that stores backup data in cloud object stores. Prior to version 2.7.2, an incomplete path traversal fix in pghoard allows an attacker with write access to the backup object storage (such as S3, GCS, or Azure) to execute arbitrary file writes and directory creation outside the designated PostgreSQL data directory (pgdata). When a database administrator executes a restore operation (via pghoard_restore get-basebackup), a maliciously crafted delta backup manifest with path traversal sequences (e.g., ../../../../etc/cron.d/backdoor) will write arbitrary payloads to the host system. This issue has been patched in version 2.7.2.
CVE-2026-57117 is an OS command injection vulnerability in MervinPraison PraisonAI versions 4.6.10 up to but not including 4.6.59. The vulnerability arises because file operation commands are constructed using raw path arguments without proper neutralization, allowing an attacker to inject arbitrary shell commands. This affects local subprocess execution and Docker container execution environments. The issue has been patched in version 4.6.59.
PraisonAI versions from 4.5.115 up to but not including 4.6.61 have a vulnerability where unauthenticated access to A2U event streaming routes is possible when using the documented CLI entrypoint. This occurs because the CLI wrapper does not enforce API key authentication, allowing exposure of sensitive information without authorization. The issue is fixed in version 4.6.61.