Vulnerabilities

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> PCI: dwc: ep: Prevent changing BAR size/flags in pci_epc_set_bar()<br /> <br /> In commit 4284c88fff0e ("PCI: designware-ep: Allow pci_epc_set_bar() update<br /> inbound map address") set_bar() was modified to support dynamically<br /> changing the backing physical address of a BAR that was already configured.<br /> <br /> This means that set_bar() can be called twice, without ever calling<br /> clear_bar() (as calling clear_bar() would clear the BAR&amp;#39;s PCI address<br /> assigned by the host).<br /> <br /> This can only be done if the new BAR size/flags does not differ from the<br /> existing BAR configuration. Add these missing checks.<br /> <br /> If we allow set_bar() to set e.g. a new BAR size that differs from the<br /> existing BAR size, the new address translation range will be smaller than<br /> the BAR size already determined by the host, which would mean that a read<br /> past the new BAR size would pass the iATU untranslated, which could allow<br /> the host to read memory not belonging to the new struct pci_epf_bar.<br /> <br /> While at it, add comments which clarifies the support for dynamically<br /> changing the physical address of a BAR. (Which was also missing.)

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> KEYS: trusted: dcp: fix improper sg use with CONFIG_VMAP_STACK=y<br /> <br /> With vmalloc stack addresses enabled (CONFIG_VMAP_STACK=y) DCP trusted<br /> keys can crash during en- and decryption of the blob encryption key via<br /> the DCP crypto driver. This is caused by improperly using sg_init_one()<br /> with vmalloc&amp;#39;d stack buffers (plain_key_blob).<br /> <br /> Fix this by always using kmalloc() for buffers we give to the DCP crypto<br /> driver.

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> ocfs2: handle a symlink read error correctly<br /> <br /> Patch series "Convert ocfs2 to use folios".<br /> <br /> Mark did a conversion of ocfs2 to use folios and sent it to me as a<br /> giant patch for review ;-)<br /> <br /> So I&amp;#39;ve redone it as individual patches, and credited Mark for the patches<br /> where his code is substantially the same. It&amp;#39;s not a bad way to do it;<br /> his patch had some bugs and my patches had some bugs. Hopefully all our<br /> bugs were different from each other. And hopefully Mark likes all the<br /> changes I made to his code!<br /> <br /> <br /> This patch (of 23):<br /> <br /> If we can&amp;#39;t read the buffer, be sure to unlock the page before returning.

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> media: uvcvideo: Remove dangling pointers<br /> <br /> When an async control is written, we copy a pointer to the file handle<br /> that started the operation. That pointer will be used when the device is<br /> done. Which could be anytime in the future.<br /> <br /> If the user closes that file descriptor, its structure will be freed,<br /> and there will be one dangling pointer per pending async control, that<br /> the driver will try to use.<br /> <br /> Clean all the dangling pointers during release().<br /> <br /> To avoid adding a performance penalty in the most common case (no async<br /> operation), a counter has been introduced with some logic to make sure<br /> that it is properly handled.

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> tpm: Change to kvalloc() in eventlog/acpi.c<br /> <br /> The following failure was reported on HPE ProLiant D320:<br /> <br /> [ 10.693310][ T1] tpm_tis STM0925:00: 2.0 TPM (device-id 0x3, rev-id 0)<br /> [ 10.848132][ T1] ------------[ cut here ]------------<br /> [ 10.853559][ T1] WARNING: CPU: 59 PID: 1 at mm/page_alloc.c:4727 __alloc_pages_noprof+0x2ca/0x330<br /> [ 10.862827][ T1] Modules linked in:<br /> [ 10.866671][ T1] CPU: 59 UID: 0 PID: 1 Comm: swapper/0 Not tainted 6.12.0-lp155.2.g52785e2-default #1 openSUSE Tumbleweed (unreleased) 588cd98293a7c9eba9013378d807364c088c9375<br /> [ 10.882741][ T1] Hardware name: HPE ProLiant DL320 Gen12/ProLiant DL320 Gen12, BIOS 1.20 10/28/2024<br /> [ 10.892170][ T1] RIP: 0010:__alloc_pages_noprof+0x2ca/0x330<br /> [ 10.898103][ T1] Code: 24 08 e9 4a fe ff ff e8 34 36 fa ff e9 88 fe ff ff 83 fe 0a 0f 86 b3 fd ff ff 80 3d 01 e7 ce 01 00 75 09 c6 05 f8 e6 ce 01 01 0b 45 31 ff e9 e5 fe ff ff f7 c2 00 00 08 00 75 42 89 d9 80 e1<br /> [ 10.917750][ T1] RSP: 0000:ffffb7cf40077980 EFLAGS: 00010246<br /> [ 10.923777][ T1] RAX: 0000000000000000 RBX: 0000000000040cc0 RCX: 0000000000000000<br /> [ 10.931727][ T1] RDX: 0000000000000000 RSI: 000000000000000c RDI: 0000000000040cc0<br /> <br /> The above transcript shows that ACPI pointed a 16 MiB buffer for the log<br /> events because RSI maps to the &amp;#39;order&amp;#39; parameter of __alloc_pages_noprof().<br /> Address the bug by moving from devm_kmalloc() to devm_add_action() and<br /> kvmalloc() and devm_add_action().

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> soc: qcom: socinfo: Avoid out of bounds read of serial number<br /> <br /> On MSM8916 devices, the serial number exposed in sysfs is constant and does<br /> not change across individual devices. It&amp;#39;s always:<br /> <br /> db410c:/sys/devices/soc0$ cat serial_number<br /> 2644893864<br /> <br /> The firmware used on MSM8916 exposes SOCINFO_VERSION(0, 8), which does not<br /> have support for the serial_num field in the socinfo struct. There is an<br /> existing check to avoid exposing the serial number in that case, but it&amp;#39;s<br /> not correct: When checking the item_size returned by SMEM, we need to make<br /> sure the *end* of the serial_num is within bounds, instead of comparing<br /> with the *start* offset. The serial_number currently exposed on MSM8916<br /> devices is just an out of bounds read of whatever comes after the socinfo<br /> struct in SMEM.<br /> <br /> Fix this by changing offsetof() to offsetofend(), so that the size of the<br /> field is also taken into account.

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> Bluetooth: L2CAP: handle NULL sock pointer in l2cap_sock_alloc<br /> <br /> A NULL sock pointer is passed into l2cap_sock_alloc() when it is called<br /> from l2cap_sock_new_connection_cb() and the error handling paths should<br /> also be aware of it.<br /> <br /> Seemingly a more elegant solution would be to swap bt_sock_alloc() and<br /> l2cap_chan_create() calls since they are not interdependent to that moment<br /> but then l2cap_chan_create() adds the soon to be deallocated and still<br /> dummy-initialized channel to the global list accessible by many L2CAP<br /> paths. The channel would be removed from the list in short period of time<br /> but be a bit more straight-forward here and just check for NULL instead of<br /> changing the order of function calls.<br /> <br /> Found by Linux Verification Center (linuxtesting.org) with SVACE static<br /> analysis tool.

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> binfmt_flat: Fix integer overflow bug on 32 bit systems<br /> <br /> Most of these sizes and counts are capped at 256MB so the math doesn&amp;#39;t<br /> result in an integer overflow. The "relocs" count needs to be checked<br /> as well. Otherwise on 32bit systems the calculation of "full_data"<br /> could be wrong.<br /> <br /> full_data = data_len + relocs * sizeof(unsigned long);

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> drm/xe/tracing: Fix a potential TP_printk UAF<br /> <br /> The commit<br /> afd2627f727b ("tracing: Check "%s" dereference via the field and not the TP_printk format")<br /> exposes potential UAFs in the xe_bo_move trace event.<br /> <br /> Fix those by avoiding dereferencing the<br /> xe_mem_type_to_name[] array at TP_printk time.<br /> <br /> Since some code refactoring has taken place, explicit backporting may<br /> be needed for kernels older than 6.10.

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> drm: zynqmp_dp: Fix integer overflow in zynqmp_dp_rate_get()<br /> <br /> This patch fixes a potential integer overflow in the zynqmp_dp_rate_get()<br /> <br /> The issue comes up when the expression<br /> drm_dp_bw_code_to_link_rate(dp-&gt;test.bw_code) * 10000 is evaluated using 32-bit<br /> Now the constant is a compatible 64-bit type.<br /> <br /> Resolves coverity issues: CID 1636340 and CID 1635811

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> drm/msm/gem: prevent integer overflow in msm_ioctl_gem_submit()<br /> <br /> The "submit-&gt;cmd[i].size" and "submit-&gt;cmd[i].offset" variables are u32<br /> values that come from the user via the submit_lookup_cmds() function.<br /> This addition could lead to an integer wrapping bug so use size_add()<br /> to prevent that.<br /> <br /> Patchwork: https://patchwork.freedesktop.org/patch/624696/

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> fs/ntfs3: Mark inode as bad as soon as error detected in mi_enum_attr()<br /> <br /> Extended the `mi_enum_attr()` function interface with an additional<br /> parameter, `struct ntfs_inode *ni`, to allow marking the inode<br /> as bad as soon as an error is detected.