CVE Vulnerability Catalog
Translated CVE descriptions from NVD NIST - in English
Browse vulnerabilities by packageCISA KEV catalog updated: (v2026.09.18)
Weekly CVE digest
One email a week with newly published vulnerabilities worth knowing about. No account needed.
This digest covers new vulnerabilities in general, not your servers. If you want to know which of them actually run in your infrastructure, that is what Secvalis does: it scans your machines and reports only what concerns them.
A problem with interrupt context was found in the smc91x driver for the Linux kernel with PREEMPT_RT. The smc_special_trylock() function disabled IRQs even on PREEMPT_RT, but smc_special_unlock() did not restore them correctly, causing a workqueue error.
In the Linux kernel, a vulnerability in the seqiv module was found where req->iv is dereferenced after crypto_aead_encrypt returns. Since the underlying request may be freed by an asynchronous completion, this leads to use-after-free.
In the Linux kernel DRM subsystem for Intel GPUs (i915), a vulnerability was found due to uninitialized eb.vma array in i915_gem_do_execbuffer. Missing initialization can lead to NULL pointer dereference during error handling in eb_lookup_vmas(), causing system crash.
In the Linux kernel, a bug was found in the ERSPAN code for GRE tunnels where the options_len field is not initialized before referencing the flexible array member options in the ip_tunnel_info structure. This causes a kernel panic when compiled with GCC 15+ and FORTIFY_SOURCE enabled.
A vulnerability in the Linux kernel's mac80211 subsystem allows accepting Beacon frames sent to a non-broadcast address. According to IEEE 802.11, Beacon frames must be sent to the broadcast address only, and unicast Beacons can be used in CSA attacks to move stations to a rogue channel. The issue is critical when Beacon protection (BIGTK) is enabled, as the current logic fails to drop unicast frames with the protection bit set, allowing bypass of protection.
A deadlock was discovered in the Linux kernel's MPTCP implementation during fallback. The issue occurs when the packet scheduler attempts reinjection after receiving MP_FAIL before the infinite map is transmitted, causing a double-lock attempt on fallback_lock.
In the Linux kernel's ext4 filesystem, a vulnerability was found in the parse_apply_sb_mount_options() function due to improper string copying. Using strscpy_pad() on a non-null-terminated string can cause an out-of-bounds read and trigger a kernel warning.
In the Linux kernel's SUNRPC subsystem, a NULL pointer dereference vulnerability was found in the gss_read_proxy_verf function. A zero-length GSS token results in empty memory pages, and accessing page_address(in_token->pages[0]) causes a NULL dereference even when the copy length is 0.
In the Linux kernel, the HNS3 network driver lacks VLAN ID validation before use. Receiving a configuration mailbox from a VF with an invalid VLAN ID can cause out-of-bounds memory access.
W jądrze Linuxa zidentyfikowano podatność, która została naprawiona. Problem dotyczy funkcji defer_free(), która nieprawidłowo obsługuje tagi pamięci po zwolnieniu obiektu, co prowadzi do raportów o użyciu pamięci po zwolnieniu (use-after-free).
In the Linux kernel for MIPS architecture, the ftrace mechanism has a memory corruption vulnerability when the kernel is located beyond 32 bits. The UASM_i_LA_mostly macro can generate more than 2 instructions, but ftrace code assumes at most 2, causing a buffer overflow and corruption of variables in the __read_mostly section.
In the Crazy Bubble Tea mobile application, an authenticated attacker can obtain personal information about other users by enumerating a `loyaltyGuestId` parameter. The server does not verify the permissions required to obtain the data. This issue was fixed in version 915 (Android) and 7.4.1 (iOS).
Lack of authorization of the InputManager D-Bus interface in InputPlumber before v0.63.0 allows a local attacker to invoke unauthorized operations. This can lead to denial-of-service, information leak, or privilege escalation within the context of the currently active user session.
In Polkit versions before v0.69.0, authentication is disabled by default and a race condition in the authorization check can lead to the same issues as in CVE-2025-66005.
A vulnerability in the Google Gemini connector configuration allows an attacker to disclose arbitrary files via a crafted credentials JSON payload. This requires authenticated access with privileges to create or modify connectors (Alerts & Connectors: All). The server processes the configuration without proper validation, enabling arbitrary network requests and file reads.
In certain Arm CPUs, executing a CPP RCTX instruction on one Processing Element (PE) may inhibit TLB invalidation when a TLBI is issued to that PE, either by the same PE or another PE in the shareability domain. As a result, the PE may retain stale TLB entries that should have been invalidated by the TLBI.
Wtyczka Shipping Rate By Cities dla WordPressa jest podatna na atak typu SQL Injection poprzez parametr 'city' we wszystkich wersjach do 2.0.0 włącznie, z powodu niewystarczającego zabezpieczenia dostarczonego przez użytkownika oraz braku odpowiedniego przygotowania istniejącego zapytania SQL. Umożliwia to nieautoryzowanym atakującym dodawanie dodatkowych zapytań SQL do już istniejących zapytań.
Wtyczka AJS Footnotes dla WordPressa jest podatna na przechowywane ataki Cross-Site Scripting (XSS) poprzez parametry 'note_list_class' i 'popup_display_effect_in' we wszystkich wersjach do 1.0 włącznie. Problem wynika z braku autoryzacji i weryfikacji nonce przy zapisywaniu ustawień oraz niewystarczającej sanitizacji wejścia i ucieczki wyjścia.
Wtyczka Name Directory dla WordPressa jest podatna na przechowywane ataki Cross-Site Scripting (XSS) poprzez parametry 'name_directory_name' i 'name_directory_description' we wszystkich wersjach do 1.30.3 włącznie, z powodu niewystarczającej sanitizacji wejścia i ucieczki wyjścia. Umożliwia to nieautoryzowanym atakującym wstrzykiwanie dowolnych skryptów webowych na stronach, które będą wykonywane przy każdym dostępie użytkownika do zainfekowanej strony.
Wtyczka GeekyBot — Generowanie treści AI bez podpowiedzi, chatbot i generowanie leadów dla WordPressa jest podatna na przechowywane ataki Cross-Site Scripting (XSS) poprzez pole wiadomości czatu we wszystkich wersjach do 1.1.8 włącznie, z powodu niewystarczającej sanitizacji wejścia i ucieczki wyjścia. Umożliwia to nieautoryzowanym atakującym wstrzykiwanie dowolnych skryptów webowych na stronach, które będą wykonywane, gdy administrator uzyska dostęp do strony historii czatu.

