Vulnerabilities

Vim is a greatly improved version of the good old UNIX editor Vi. Vim allows to redirect screen messages using the `:redir` ex command to register, variables and files. It also allows to show the contents of registers using the `:registers` or `:display` ex command. When redirecting the output of `:display` to a register, Vim will free the register content before storing the new content in the register. Now when redirecting the `:display` command to a register that is being displayed, Vim will free the content while shortly afterwards trying to access it, which leads to a use-after-free. Vim pre 9.1.1115 checks in the ex_display() function, that it does not try to redirect to a register while displaying this register at the same time. However this check is not complete, and so Vim does not check the `+` and `*` registers (which typically donate the X11/clipboard registers, and when a clipboard connection is not possible will fall back to use register 0 instead. In Patch 9.1.1115 Vim will therefore skip outputting to register zero when trying to redirect to the clipboard registers `*` or `+`. Users are advised to upgrade. There are no known workarounds for this vulnerability.

A vulnerability was found in OpenSSH when the VerifyHostKeyDNS option is enabled. A machine-in-the-middle attack can be performed by a malicious machine impersonating a legit server. This issue occurs due to how OpenSSH mishandles error codes in specific conditions when verifying the host key. For an attack to be considered successful, the attacker needs to manage to exhaust the client's memory resource first, turning the attack complexity high.

SPID.AspNetCore.Authentication is an AspNetCore Remote Authenticator for SPID. Authentication using Spid and CIE is based on the SAML2 standard which provides two entities: Identity Provider (IDP): the system that authenticates users and provides identity information (SAML affirmation) to the Service Provider, in essence, is responsible for the management of the credentials and identity of users; Service Provider (SP): the system that provides a service to the user and relies on the Identity Provider to authenticate the user, receives SAML assertions from the IdP to grant access to resources. The validation logic of the signature is central as it ensures that you cannot create a SAML response with arbitrary assertions and then impersonate other users. There is no guarantee that the first signature refers to the root object, it follows that if an attacker injects an item signed as the first element, all other signatures will not be verified. The only requirement is to have an XML element legitimately signed by the IdP, a condition that is easily met using the IdP's public metadata. An attacker could create an arbitrary SAML response that would be accepted by SPs using vulnerable SDKs, allowing him to impersonate any Spid and/or CIE user. This vulnerability has been addressed in version 3.4.0 and all users are advised to upgrade. There are no known workarounds for this vulnerability.

CIE.AspNetCore.Authentication is an AspNetCore Remote Authenticator for CIE 3.0. Authentication using Spid and CIE is based on the SAML2 standard which provides two entities: 1. Identity Provider (IDP): the system that authenticates users and provides identity information (SAML affirmation) to the Service Provider, in essence, is responsible for the management of the credentials and identity of users; 2. Service Provider (SP): the system that provides a service to the user and relies on the Identity Provider to authenticate the user, receives SAML assertions from the IdP to grant access to resources. The library cie-aspnetcore refers to the second entity, the SP, and implements the validation logic of SAML assertions within SAML responses. In affected versions there is no guarantee that the first signature refers to the root object, it follows that if an attacker injects an item signed as the first element, all other signatures will not be verified. The only requirement is to have an XML element legitimately signed by the IdP, a condition that is easily met using the IdP's public metadata. An attacker could create an arbitrary SAML response that would be accepted by SPs using vulnerable SDKs, allowing him to impersonate any Spid and/or CIE user. This issue has been addressed in version 2.1.0 and all users are advised to upgrade. There are no known workarounds for this vulnerability.

The ZOO-Project is an open source processing platform, released under MIT/X11 Licence. A vulnerability in ZOO-Project's WPS (Web Processing Service) implementation allows unauthorized access to files outside the intended directory through path traversal. Specifically, the Gdal_Translate service, when processing VRT (Virtual Format) files, does not properly validate file paths referenced in the VRTRasterBand element, allowing attackers to read arbitrary files on the system. The vulnerability exists because the service doesn't properly sanitize the SourceFilename parameter in VRT files, allowing relative path traversal sequences (../). When combined with VRT's raw data handling capabilities, this allows reading arbitrary files as raw binary data and converting them to TIFF format, effectively exposing their contents. This vulnerability is particularly severe because it allows attackers to read sensitive system files, potentially exposing configuration data, credentials, or other confidential information stored on the server. An unauthenticated attacker can read arbitrary files from the system through path traversal, potentially accessing sensitive information such as configuration files, credentials, or other confidential data stored on the server. The vulnerability requires no authentication and can be exploited remotely through the WPS service. This issue has been addressed in commit `5f155a8` and all users are advised to upgrade. There are no known workarounds for this vulnerability.

Meshtastic is an open source mesh networking solution. In affected firmware versions crafted packets over MQTT are able to appear as a DM in client to a node even though they were not decoded with PKC. This issue has been addressed in version 2.5.19 and all users are advised to upgrade. There are no known workarounds for this vulnerability.

Server-Side Access Control Bypass vulnerability in WombatDialer before 25.02 could allow unauthorized users to potentially call certain services without the necessary access level. This issue is limited to services used by the client (not the general-use JSON services) and requires reverse engineering of the proprietary serialization protocol, making it difficult to exploit.

Incorrect cookie session handling in WombatDialer before 25.02 results in the full session identity being written to system logs and could be used by a malicious attacker to impersonate an existing user session.

A flaw was found in grub2. A specially crafted JPEG file can cause the JPEG parser of grub2 to incorrectly check the bounds of its internal buffers, resulting in an out-of-bounds write. The possibility of overwriting sensitive information to bypass secure boot protections is not discarded.

Duende.AccessTokenManagement is a set of .NET libraries that manage OAuth and OpenId Connect access tokens. Duende.AccessTokenManagement contains a race condition when requesting access tokens using the client credentials flow. Concurrent requests to obtain an access token using differing protocol parameters can return access tokens obtained with the wrong scope, resource indicator, or other protocol parameters. Such usage is somewhat atypical, and only a small percentage of users are likely to be affected. Duende.AccessTokenManagement can request access tokens using the client credentials flow in several ways. In basic usage, the client credentials flow is configured once and the parameters do not vary. In more advanced situations, requests with varying protocol parameters may be made by calling specific overloads of these methods: `HttpContext.GetClientAccessTokenAsync()` and `IClientCredentialsTokenManagementService.GetAccessTokenAsync()`. There are overloads of both of these methods that accept a `TokenRequestParameters` object that customizes token request parameters. However, concurrent requests with varying `TokenRequestParameters` will result in the same token for all concurrent calls. Most users can simply update the NuGet package to the latest version. Customizations of the `IClientCredentialsTokenCache` that derive from the default implementation (`DistributedClientCredentialsTokenCache`) will require a small code change, as its constructor was changed to add a dependency on the `ITokenRequestSynchronization` service. The synchronization service will need to be injected into the derived class and passed to the base constructor. The impact of this vulnerability depends on how Duende.AccessTokenManagement is used and on the security architecture of the solution. Most users will not be vulnerable to this issue. More advanced users may run into this issue by calling the methods specified above with customized token request parameters. The impact of obtaining an access token with different than intended protocol parameters will vary depending on application logic, security architecture, and the authorization policy of the resource servers.

Webkul QloApps v1.6.1 exposes authentication tokens in URLs during redirection. When users access the admin panel or other protected areas, the application appends sensitive authentication tokens directly to the URL.

smartbanner.js is a customizable smart app banner for iOS and Android. Prior to version 1.14.1, clicking on smartbanner `View` link and navigating to 3rd party page leaves `window.opener` exposed. It may allow hostile third parties to abuse `window.opener`, e.g. by redirection or injection on the original page with smartbanner. `rel="noopener"` is automatically populated to links as of `v1.14.1` which is a recommended upgrade to resolve the vulnerability. Some workarounds are available for those who cannot upgrade. Ensure `View` link is only taking users to App Store or Google Play Store where security is guarded by respective app store security teams. If `View` link is going to a third party page, limit smartbanner.js to be used on iOS that decreases the scope of the vulnerability since as of Safari 12.1, `rel="noopener"` is imposed on all `target="_blank"` links. Version 1.14.1 of smartbanner.js contains a fix for the issue.