Samedi 24 octobre 2026

Apache Software Foundation Apache Camel · Apache camel

Improper Authentication vulnerability in Apache Camel Platform HTTP Main component. This issue affects Apache Camel: from 4.8.0 before 4.22.0. The camel-main embedded HTTP server can protect its endpoints with JWT authentication, configured through authenticationEnabled together with the JWT keystore properties. JWTAuthenticationConfigurer.buildJwtOptions returned null when neither jwtIssuer nor jwtAudience was configured, and the caller then skipped the JWTAuthOptions.setJWTOptions call entirely, so the Vert.x JWTAuth instance was built from the keystore alone. The result was that inbound tokens were checked only for signature and expiry: the iss and aud claims were not validated at all. Nothing signalled this - the server started normally and reported no warning - so a deployment configured the documented way silently enforced less than the operator believed it had enabled, and the component documentation itself presented signature and expiry checking as the default with issuer and audience as an optional extra. Both the application server and the management server were affected, because the omission was in each of the two configureAuthentication paths. Any unexpired token signed by any key the configured keystore trusts was therefore accepted, regardless of which issuer minted it or which audience it was intended for. How far that reaches depends on the trust set of the keystore: where the signing key belongs to a shared or multi-tenant identity provider, a token legitimately issued for an entirely different audience is accepted, while a keystore holding a dedicated signer narrows it to reuse of tokens minted for other services within the same trust domain. The jwtIssuer and jwtAudience options did not exist before 4.21.0, so on earlier releases there was no supported way to have these claims enforced at all. Users are recommended to upgrade to version 4.22.0, which fixes the issue. From 4.22.0 the server refuses to start when a JWT keystore is configured but neither jwtIssuer nor jwtAudience is set, naming the properties involved, and a deployment that genuinely wants signature and expiry validation only must say so explicitly with the new jwtAllowMissingIssuerAndAudience option, which defaults to false. This behaviour is fixed only on 4.22.0. The 4.14.9 and 4.18.4 releases do not change the default: they add the jwtIssuer and jwtAudience options so that operators on those maintenance lines can enforce the claims by configuration, and an installation that upgrades to 4.14.9 or 4.18.4 without also setting at least one of those two properties is still accepting any unexpired token signed by a trusted key. Users on 4.14.x or 4.18.x should therefore upgrade to 4.14.9 or 4.18.4 and then set jwtIssuer, jwtAudience, or both. Releases from 4.8.0 up to and including 4.21.x offer no way to enforce these claims and should be moved to a version that does. Independently of version, restrict the JWT keystore to the smallest possible trust set - ideally a signer dedicated to this service rather than a shared identity-provider key - and where a gateway already validates issuer and audience in front of the server, ensure it cannot be bypassed. Notes: The JIRA ticket:  https://issues.apache.org/jira/browse/CAMEL-24281  refers to the various commits that resolved the issue, and has more details. The fail-closed guard could not be backported. The jwtIssuer and jwtAudience options were themselves only introduced in 4.21.0 by CAMEL-23525, so on camel-4.18.x and camel-4.14.x there was nothing an operator could set to satisfy the requirement and the guard would have broken every JWT deployment on those branches with no remedy available.

Linux

In the Linux kernel, the following vulnerability has been resolved: xfs: fix exchange-range-to-eof file size exchange Norbert Szetei posted a patch containing an incomplete description of a bug in XFS_IOC_EXCHANGE_RANGE's TO_EOF flag. The bug tricks exchange-range into modifying a file so that it has shared blocks starting beyond EOF, which is never allowed because files never have written data beyond EOF, and you can only share written data blocks. This is key to all the incorrect behavior that follows. Eventually I got from Norbert a description of what's going wrong: > Fair. Here it is with physical block numbers, 4k blocks. peer is the file > that gets rewritten and file1 is the one that ends up with the flag > cleared. > > Start, nothing shared: > > peer size 4096 -> [100] > file1 size 12288 -> [200] [201] [202] > donor1 size 4096 -> [300] > donor2 size 4096 -> [400] > > 1. FICLONERANGE peer's block into file1's middle slot: > > peer size 4096 -> [100] > file1 size 12288 -> [200] [100] [202] reflink flag set > ^^^^^ both files own block 100 > > 2. XFS_IOC_EXCHANGE_RANGE file1 against donor1, file1_offset 0, > file2_offset 8192, length 0, TO_EOF. One block is exchanged, and the sizes > are exchanged with it: > > file1 size 4096 -> [200] [100] [300] reflink flag set > ^^^^^^^^^^^ still owned, now past the size > donor1 size 12288 -> [202] > > file1 says it is one block long while it still owns three. Nothing was > unmapped. file2_offset is block 2, so this exchange cannot clear a flag. > > 3. XFS_IOC_EXCHANGE_RANGE file1 against donor2, both offsets 0, length 4096. > By i_disk_size this is two one-block files exchanging everything, so > xmi_can_exchange_reflink_flags() agrees and the flag is cleared: > > file1 size 4096 -> [400] [100] [300] reflink flag CLEARED > donor2 size 4096 -> [200] > > Block 100 is still shared with peer. > > 4. pwrite(file1, 4096, 4096), which is file1's second slot, block 100. > xfs_is_cow_inode(file1) is false, so no CoW: > > peer size 4096 -> [100] same block, new contents > > Block 100 is the only one that matters here. PoC below, it reads a file > called peer in the current directory. The filesystem needs reflink and > exchange-range, which mkfs.xfs 7.x turns on by default. So yes, TO_EOF is screwing up the file size even if it's actually exchanging the data block correctly. The following is the intended behavior of XFS_EXCHANGE_RANGE_TO_EOF, as described by its manpage: "Ignore the length parameter. All bytes in file1_fd from file1_offset to EOF are moved to file2_fd, and file2's size is set to (file2_offset+(file1_length-file1_offset)). Meanwhile, all bytes in file2 from file2_offset to EOF are moved to file1 and file1's size is set to (file1_offset+(file2_length-file2_offset))." In other words, if we only exchange a portion of the file data, then we should only exchange the portion of the file size that relates to the exchanged data. Unfortunately, the code swaps the ondisk file size and the incore file sizes, which is incorrect if the caller passes in a nonzero offset. Having confused the file sizes, the PoC uses a second exchange-range call on what looks like a non-reflinked single-block file. Because the file size is set incorrectly, exchange-range thinks it's exchanging the full contents of two files and clears the reflink flag on the broken file. That enables an extending write of the broken file to rewrite the shared block that's just past EOF. This has become known colloquially as refluxfs. Once the file sizes are set correctly, the second exchange-range no longer thinks that it's doing a full-contents swap, so it won't clear the reflink flag on either of its file arguments. Norbert followed up to an earlier version of this patch with a second complaint about us not handling the i_disk_size state correctly: > The offsets are bounded against the incore i_size but subtracted from > ---truncated---

apptainer · Lfprojects apptainer · Sylabs singularity · Redhat enterprise linux

Apptainer is an open source container platform for Linux. There is an ext4 use-after-free flaw that is exploitable through versions of Apptainer < 1.1.0 and installations that include apptainer-suid < 1.1.8 on older operating systems where that CVE has not been patched. That includes Red Hat Enterprise Linux 7, Debian 10 buster (unless the linux-5.10 package is installed), Ubuntu 18.04 bionic and Ubuntu 20.04 focal. Use-after-free flaws in the kernel can be used to attack the kernel for denial of service and potentially for privilege escalation. Apptainer 1.1.8 includes a patch that by default disables mounting of extfs filesystem types in setuid-root mode, while continuing to allow mounting of extfs filesystems in non-setuid "rootless" mode using fuse2fs. Some workarounds are possible. Either do not install apptainer-suid (for versions 1.1.0 through 1.1.7) or set `allow setuid = no` in apptainer.conf. This requires having unprivileged user namespaces enabled and except for apptainer 1.1.x versions will disallow mounting of sif files, extfs files, and squashfs files in addition to other, less significant impacts. (Encrypted sif files are also not supported unprivileged in apptainer 1.1.x.). Alternatively, use the `limit containers` options in apptainer.conf/singularity.conf to limit sif files to trusted users, groups, and/or paths, and set `allow container extfs = no` to disallow mounting of extfs overlay files. The latter option by itself does not disallow mounting of extfs overlay partitions inside SIF files, so that's why the former options are also needed.