| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| In the Linux kernel, the following vulnerability has been resolved:
KVM: PPC: Book3S HV: Fix stack handling in idle_kvm_start_guest()
In commit 10d91611f426 ("powerpc/64s: Reimplement book3s idle code in
C") kvm_start_guest() became idle_kvm_start_guest(). The old code
allocated a stack frame on the emergency stack, but didn't use the
frame to store anything, and also didn't store anything in its caller's
frame.
idle_kvm_start_guest() on the other hand is written more like a normal C
function, it creates a frame on entry, and also stores CR/LR into its
callers frame (per the ABI). The problem is that there is no caller
frame on the emergency stack.
The emergency stack for a given CPU is allocated with:
paca_ptrs[i]->emergency_sp = alloc_stack(limit, i) + THREAD_SIZE;
So emergency_sp actually points to the first address above the emergency
stack allocation for a given CPU, we must not store above it without
first decrementing it to create a frame. This is different to the
regular kernel stack, paca->kstack, which is initialised to point at an
initial frame that is ready to use.
idle_kvm_start_guest() stores the backchain, CR and LR all of which
write outside the allocation for the emergency stack. It then creates a
stack frame and saves the non-volatile registers. Unfortunately the
frame it creates is not large enough to fit the non-volatiles, and so
the saving of the non-volatile registers also writes outside the
emergency stack allocation.
The end result is that we corrupt whatever is at 0-24 bytes, and 112-248
bytes above the emergency stack allocation.
In practice this has gone unnoticed because the memory immediately above
the emergency stack happens to be used for other stack allocations,
either another CPUs mc_emergency_sp or an IRQ stack. See the order of
calls to irqstack_early_init() and emergency_stack_init().
The low addresses of another stack are the top of that stack, and so are
only used if that stack is under extreme pressue, which essentially
never happens in practice - and if it did there's a high likelyhood we'd
crash due to that stack overflowing.
Still, we shouldn't be corrupting someone else's stack, and it is purely
luck that we aren't corrupting something else.
To fix it we save CR/LR into the caller's frame using the existing r1 on
entry, we then create a SWITCH_FRAME_SIZE frame (which has space for
pt_regs) on the emergency stack with the backchain pointing to the
existing stack, and then finally we switch to the new frame on the
emergency stack. |
| VisiCut 2.1 allows stack consumption via an XML document with nested set elements, as demonstrated by a java.util.HashMap StackOverflowError when reference='../../../set/set[2]' is used, aka an "insecure deserialization" issue. |
| In the Linux kernel, the following vulnerability has been resolved:
riscv: VMAP_STACK overflow detection thread-safe
commit 31da94c25aea ("riscv: add VMAP_STACK overflow detection") added
support for CONFIG_VMAP_STACK. If overflow is detected, CPU switches to
`shadow_stack` temporarily before switching finally to per-cpu
`overflow_stack`.
If two CPUs/harts are racing and end up in over flowing kernel stack, one
or both will end up corrupting each other state because `shadow_stack` is
not per-cpu. This patch optimizes per-cpu overflow stack switch by
directly picking per-cpu `overflow_stack` and gets rid of `shadow_stack`.
Following are the changes in this patch
- Defines an asm macro to obtain per-cpu symbols in destination
register.
- In entry.S, when overflow is detected, per-cpu overflow stack is
located using per-cpu asm macro. Computing per-cpu symbol requires
a temporary register. x31 is saved away into CSR_SCRATCH
(CSR_SCRATCH is anyways zero since we're in kernel).
Please see Links for additional relevant disccussion and alternative
solution.
Tested by `echo EXHAUST_STACK > /sys/kernel/debug/provoke-crash/DIRECT`
Kernel crash log below
Insufficient stack space to handle exception!/debug/provoke-crash/DIRECT
Task stack: [0xff20000010a98000..0xff20000010a9c000]
Overflow stack: [0xff600001f7d98370..0xff600001f7d99370]
CPU: 1 PID: 205 Comm: bash Not tainted 6.1.0-rc2-00001-g328a1f96f7b9 #34
Hardware name: riscv-virtio,qemu (DT)
epc : __memset+0x60/0xfc
ra : recursive_loop+0x48/0xc6 [lkdtm]
epc : ffffffff808de0e4 ra : ffffffff0163a752 sp : ff20000010a97e80
gp : ffffffff815c0330 tp : ff600000820ea280 t0 : ff20000010a97e88
t1 : 000000000000002e t2 : 3233206874706564 s0 : ff20000010a982b0
s1 : 0000000000000012 a0 : ff20000010a97e88 a1 : 0000000000000000
a2 : 0000000000000400 a3 : ff20000010a98288 a4 : 0000000000000000
a5 : 0000000000000000 a6 : fffffffffffe43f0 a7 : 00007fffffffffff
s2 : ff20000010a97e88 s3 : ffffffff01644680 s4 : ff20000010a9be90
s5 : ff600000842ba6c0 s6 : 00aaaaaac29e42b0 s7 : 00fffffff0aa3684
s8 : 00aaaaaac2978040 s9 : 0000000000000065 s10: 00ffffff8a7cad10
s11: 00ffffff8a76a4e0 t3 : ffffffff815dbaf4 t4 : ffffffff815dbaf4
t5 : ffffffff815dbab8 t6 : ff20000010a9bb48
status: 0000000200000120 badaddr: ff20000010a97e88 cause: 000000000000000f
Kernel panic - not syncing: Kernel stack overflow
CPU: 1 PID: 205 Comm: bash Not tainted 6.1.0-rc2-00001-g328a1f96f7b9 #34
Hardware name: riscv-virtio,qemu (DT)
Call Trace:
[<ffffffff80006754>] dump_backtrace+0x30/0x38
[<ffffffff808de798>] show_stack+0x40/0x4c
[<ffffffff808ea2a8>] dump_stack_lvl+0x44/0x5c
[<ffffffff808ea2d8>] dump_stack+0x18/0x20
[<ffffffff808dec06>] panic+0x126/0x2fe
[<ffffffff800065ea>] walk_stackframe+0x0/0xf0
[<ffffffff0163a752>] recursive_loop+0x48/0xc6 [lkdtm]
SMP: stopping secondary CPUs
---[ end Kernel panic - not syncing: Kernel stack overflow ]--- |
| Connect2id Nimbus JOSE + JWT 10.0.x before 10.0.2 and 9.37.x before 9.37.4 allows a remote attacker to cause a denial of service via a deeply nested JSON object supplied in a JWT claim set, because of uncontrolled recursion. NOTE: this is independent of the Gson 2.11.0 issue because the Connect2id product could have checked the JSON object nesting depth, regardless of what limits (if any) were imposed by Gson. |
| In the Linux kernel, the following vulnerability has been resolved:
misc: fastrpc: fix list iterator in fastrpc_req_mem_unmap_impl
This is another instance of incorrect use of list iterator and
checking it for NULL.
The list iterator value 'map' will *always* be set and non-NULL
by list_for_each_entry(), so it is incorrect to assume that the
iterator value will be NULL if the list is empty (in this case, the
check 'if (!map) {' will always be false and never exit as expected).
To fix the bug, use a new variable 'iter' as the list iterator,
while use the original variable 'map' as a dedicated pointer to
point to the found element.
Without this patch, Kernel crashes with below trace:
Unable to handle kernel access to user memory outside uaccess routines
at virtual address 0000ffff7fb03750
...
Call trace:
fastrpc_map_create+0x70/0x290 [fastrpc]
fastrpc_req_mem_map+0xf0/0x2dc [fastrpc]
fastrpc_device_ioctl+0x138/0xc60 [fastrpc]
__arm64_sys_ioctl+0xa8/0xec
invoke_syscall+0x48/0x114
el0_svc_common.constprop.0+0xd4/0xfc
do_el0_svc+0x28/0x90
el0_svc+0x3c/0x130
el0t_64_sync_handler+0xa4/0x130
el0t_64_sync+0x18c/0x190
Code: 14000016 f94000a5 eb05029f 54000260 (b94018a6)
---[ end trace 0000000000000000 ]--- |
| An issue in phiola/src/afilter/conv.c:115 of phiola v2.0-rc22 allows a remote attacker to cause a denial of service via a crafted .wav file. |
| An issue in the anchors subparser of Showdownjs versions <= 2.1.0 could allow a remote attacker to cause denial of service conditions.
|
| orjson.loads in orjson before 3.9.15 does not limit recursion for deeply nested JSON documents. |
| MongoDB Server may allow upsert operations retried within a transaction to violate unique index constraints, potentially causing an invariant failure and server crash during commit. This issue may be triggered by improper WriteUnitOfWork state management. This issue affects MongoDB Server v6.0 versions prior to 6.0.25, MongoDB Server v7.0 versions prior to 7.0.22 and MongoDB Server v8.0 versions prior to 8.0.12 |
| In Alludo MindManager before 25.0.208 on Windows, attackers could potentially execute code as other local users on the same machine if they could write DLL files to directories within victims' DLL search paths. |
| The express-xss-sanitizer (aka Express XSS Sanitizer) package through 2.0.0 for Node.js has an unbounded recursion depth in sanitize in lib/sanitize.js for a JSON request body. |
| MongoDB Server may be susceptible to stack overflow due to JSON parsing mechanism, where specifically crafted JSON inputs may induce unwarranted levels of recursion, resulting in excessive stack space consumption. Such inputs can lead to a stack overflow that causes the server to crash which could occur pre-authorisation. This issue affects MongoDB Server v7.0 versions prior to 7.0.17 and MongoDB Server v8.0 versions prior to 8.0.5.
The same issue affects MongoDB Server v6.0 versions prior to 6.0.21, but an attacker can only induce denial of service after authenticating. |
| XGrammar is an open-source library for efficient, flexible, and portable structured generation. Prior to version 0.1.21, XGrammar has an infinite recursion issue in the grammar. This issue has been resolved in version 0.1.21. |
| Wasmtime is a runtime for WebAssembly. Prior to versions 24.0.4, 33.0.2, and 34.0.2, a bug in Wasmtime's implementation of the WASIp1 set of import functions can lead to a WebAssembly guest inducing a panic in the host (embedder). The specific bug is triggered by calling `path_open` after calling `fd_renumber` with either two equal argument values or a second argument being equal to a previously-closed file descriptor number value. The corrupt state introduced in `fd_renumber` will lead to the subsequent opening of a file descriptor to panic. This panic cannot introduce memory unsafety or allow WebAssembly to break outside of its sandbox, however. There is no possible heap corruption or memory unsafety from this panic. This bug is in the implementation of Wasmtime's `wasmtime-wasi` crate which provides an implementation of WASIp1. The bug requires a specially crafted call to `fd_renumber` in addition to the ability to open a subsequent file descriptor. Opening a second file descriptor is only possible when a preopened directory was provided to the guest, and this is common amongst embeddings. A panic in the host is considered a denial-of-service vector for WebAssembly embedders and is thus a security issue in Wasmtime. This bug does not affect WASIp2 and embedders using components. In accordance with Wasmtime's release process, patch releases are available as 24.0.4, 33.0.2, and 34.0.2. Users of other release of Wasmtime are recommended to move to a supported release of Wasmtime. Embedders who are using components or are not providing guest access to create more file descriptors (e.g. via a preopened filesystem directory) are not affected by this issue. Otherwise, there is no workaround at this time, and affected embeddings are recommended to update to a patched version which will not cause a panic in the host. |
| Wasmtime is a fast and secure runtime for WebAssembly. Wasmtime's filesystem sandbox implementation on Windows blocks access to special device filenames such as "COM1", "COM2", "LPT0", "LPT1", and so on, however it did not block access to the special device filenames which use superscript digits, such as "COM¹", "COM²", "LPT⁰", "LPT¹", and so on. Untrusted Wasm programs that are given access to any filesystem directory could bypass the sandbox and access devices through those special device filenames with superscript digits, and through them gain access peripheral devices connected to the computer, or network resources mapped to those devices. This can include modems, printers, network printers, and any other device connected to a serial or parallel port, including emulated USB serial ports. Patch releases for Wasmtime have been issued as 24.0.2, 25.0.3, and 26.0.1. Users of Wasmtime 23.0.x and prior are recommended to upgrade to one of these patched versions. There are no known workarounds for this issue. Affected Windows users are recommended to upgrade. |
| Envoy is a cloud-native high-performance edge/middle/service proxy. In affected versions envoy does not properly handle http 1.1 non-101 1xx responses. This can lead to downstream failures in networked devices. This issue has been addressed in versions 1.31.5 and 1.32.3. Users are advised to upgrade. There are no known workarounds for this issue. |
| Envoy is a cloud-native high-performance edge/middle/service proxy. In affected versions `sendOverloadError` is going to assume the active request exists when `envoy.load_shed_points.http1_server_abort_dispatch` is configured. If `active_request` is nullptr, only onMessageBeginImpl() is called. However, the `onMessageBeginImpl` will directly return ok status if the stream is already reset leading to the nullptr reference. The downstream reset can actually happen during the H/2 upstream reset. As a result envoy may crash. This issue has been addressed in releases 1.32.3, 1.31.5, 1.30.9, and 1.29.12. Users are advised to upgrade. Users unable to upgrade may disable `http1_server_abort_dispatch` load shed point and/or use a high threshold. |
| Helm is a package manager for Charts for Kubernetes. A JSON Schema file within a chart can be crafted with a deeply nested chain of references, leading to parser recursion that can exceed the stack size limit and trigger a stack overflow. This issue has been resolved in Helm v3.17.3. |
| KaTeX is a JavaScript library for TeX math rendering on the web. KaTeX users who render untrusted mathematical expressions could encounter malicious input using `\def` or `\newcommand` that causes a near-infinite loop, despite setting `maxExpand` to avoid such loops. KaTeX supports an option named maxExpand which aims to prevent infinitely recursive macros from consuming all available memory and/or triggering a stack overflow error. Unfortunately, support for "Unicode (sub|super)script characters" allows an attacker to bypass this limit. Each sub/superscript group instantiated a separate Parser with its own limit on macro executions, without inheriting the current count of macro executions from its parent. This has been corrected in KaTeX v0.16.10. |
| Envoy is a cloud-native high-performance edge/middle/service proxy. When additional address are not ip addresses, then the Happy Eyeballs sorting algorithm will crash in data plane. This issue has been addressed in releases 1.32.2, 1.31.4, and 1.30.8. Users are advised to upgrade. Users unable to upgrade may disable Happy Eyeballs and/or change the IP configuration. |
