CVE-2026-80575
Description détaillée
In the Linux kernel, the following vulnerability has been resolved: Input: cs40l50-vibra - validate custom data from user space cs40l50_add() copies the custom data of an FF_PERIODIC/FF_CUSTOM effect straight from the ff_effect the user passed to EVIOCSFF, without requiring it to hold anything: work_data.custom_data = memdup_array_user(periodic->custom_data, periodic->custom_len, sizeof(s16)); work_data.custom_len = periodic->custom_len; The driver then reads two words out of that buffer: custom_data[0] as the waveform bank in cs40l50_effect_bank_set(), and custom_data[1] as the index within the bank in cs40l50_effect_index_set(). Neither read is covered by a length check, and custom_len is fully user controlled: - custom_len == 0 makes memdup_array_user() call memdup_user() with a length of zero, which returns ZERO_SIZE_PTR rather than an error, so custom_data[0] dereferences it. - custom_len == 1 allocates two bytes. A bank of ROM or RAM keeps effect->type out of the OWT case, and custom_data[1] is then read one word past the allocation. The bank value itself is also mishandled. It is masked with CS40L50_CUSTOM_DATA_MASK (0xffff) but stored in an s16, so a custom_data[0] of 0x8000 or above wraps to a negative value that passes the "bank_type >= CS40L50_WVFRM_BANK_NUM" test. cs40l50_effect_index_set() indexes vib->dsp.banks[] with it before the switch statement's default case gets a chance to reject it: base_index = vib->dsp.banks[effect->type].base_index; max_index = vib->dsp.banks[effect->type].max_index; Require the two words the driver reads to be present, and hold the masked bank in a u32 so the existing upper-bound test covers the whole range. The da7280 haptic driver already range checks custom_len this way.
Références et Patchs
Dernières Vulnérabilités
CVE-2026-76784
Multiple TP-Link Kasa smart home devices contain insufficient cryptographic protections in the local device communication protocol. An adjacent network attacker may intercept, replay or forge locally exchanged control messages, potentially resulting in unauthorized device control. Successful exploitation could allow an attacker to manipulate the operational state of an affected device, resulting in unauthorized state changes, disruption of normal device functionality or a denial-of-service condition.
CVE-2026-58474
whichllm before 0.5.16 contains a code injection vulnerability in the run and snippet commands that allows a remote attacker who controls a HuggingFace repository to achieve arbitrary code execution by crafting a malicious GGUF filename containing double quotes or other special characters. The script generation function in cli.py interpolates HuggingFace-derived values, including GGUF variant filenames from the Hub API siblings rfilename field, directly into Python source code without escaping, allowing the crafted filename to break out of the generated string literal and execute injected code on the user's machine before any model download occurs.
CVE-2026-54256
Winter CMS is a content management system built on the Laravel PHP framework. In versions up to and including 1.2.12, the backend FileUpload form widget trusted an attacker-controlled file_id POST parameter when resolving the attachment it operates on, allowing an authenticated backend user to read and modify attachment records belonging to other users or records. The widget's getFileRecord() lookup resolved the posted id against the global system_files table without verifying that the file belonged to the widget's own relation, parent record, or deferred-binding session. Because all attachments share a single File model and table and attachment ids are sequential integers that are easily enumerated, a user reaching any form with a fileupload field, including the built-in My Account avatar field that requires no specific permission, could target arbitrary attachments to modify their title and description via onSaveAttachmentConfig and change their sort order via onSortAttachments, which passed posted ids straight to an unscoped update. CSRF tokens remain enforced, so exploitation requires a valid authenticated backend session with any level of access. This issue is fixed in version 1.2.13.