CVE-2026-98234
Low risk· EPSS 7%Exploitation Probability (EPSS)
Low risk7th percentile - higher than 7% of all known CVEs
Summary
In the Linux kernel net/sched subsystem, hhf_change() stores TCA_HHF_HH_FLOWS_LIMIT with no upper bound. A huge hh_flows_limit lets each new heavy-hitter flow pass the hh_flows_current_cnt check in alloc_new_hh() and forces a fixed-size kzalloc(GFP_ATOMIC) per flow under spoofed traffic, causing unbounded memory growth.
Risk Assessment
An attacker with CAP_NET_ADMIN in a user namespace can exhaust kernel memory, causing a denial of service (DoS) or system crash.
Recommendation
Update the Linux kernel to a version containing the fix that bounds the attribute to 2*HH_FLOWS_CNT via NLA_POLICY_MAX() and sets the default before hhf_change().
Related vulnerabilities
- CVE-2023-22793Critical
A vulnerability allowing an unauthenticated attacker to perform remote code execution (RCE) on servers running the Veeam Virtualization Extensions service of Veeam Management Pack for Microsoft System Center.
- CVE-2026-61707Critical
In SOCFortress CoPilot prior to 0.1.76, case-report templates uploaded by the lowest-privilege customer_user role are later executed by non-sandboxed Jinja2 engines when an admin or analyst generates a report. Lack of template content validation and cross-tenant scoping allows an attacker to inject malicious Jinja expressions that are evaluated as Python on the backend.
- CVE-2026-61645Critical
Docat has an arbitrary file upload vulnerability in its Upload API because the API does not validate the uploaded file path or type. An attacker can overwrite sensitive configuration files or place malicious files on the Docat web server.
- CVE-2026-103093Critical
ServiceNow remediated a sandbox escape vulnerability in the ServiceNow AI Platform. An authenticated user with the admin role could bypass Rhino sandbox restrictions and execute arbitrary code on the underlying host operating system.
- CVE-2026-118769Critical
In the Botan C++ cryptography library, the C API function botan_block_cipher_block_size converted a size_t block size to int without a bounds check. In an application that accepts external algorithm parameter strings and uses FFI raw ECB mode, an attacker can select the variable-block-size Lion cipher, causing the reported size to be truncated, insufficient memory allocation, and later memory corruption. The issue is fixed in version 3.13.0.
- CVE-2026-118762Critical
In the Monkeytype repository, the .github/workflows/fix-formatting.yml workflow runs on pull_request_target with repository write permissions, checks out a fork pull request, installs dependencies from the fork's package.json, and invokes the fork-controlled oxfmt binary. Because actions/checkout persists the elevated GITHUB_TOKEN in Git configuration until after untrusted code executes, an external contributor can run arbitrary code, read or exfiltrate the token, and push repository content or modify pull requests and issues. The issue affects the repository CI workflow, not the Monkeytype application or user data, and is fixed in version 26.43.0.
- CVE-2026-107058Critical
An issue in Shelter (versionCode 445) allows attackers to register their own auth_key and take over another user's binder service API and preference state.
- CVE-2026-93680Critical
IBM Langflow OSS versions 1.0.0 through 1.12.8 contain an improper authentication flaw that could allow a remote attacker to bypass security restrictions.
- CVE-2026-90868Critical
Progress Software WS_FTP Server versions prior to 2024.0.7 and 2025.0.4 on Windows contain an OS command injection vulnerability that could allow arbitrary command execution under certain configurations.
- CVE-2026-90866Critical
Progress Software WS_FTP Server versions prior to 2024.0.7 and 2025.0.4 on Windows contain a path traversal vulnerability that could allow a delegated Host Administrator to access locations outside their authorized scope, potentially enabling remote code execution.
Original NVD description (English source)
In the Linux kernel, the following vulnerability has been resolved: net/sched: hhf: cap hh_flows_limit at change time hhf_change() stores TCA_HHF_HH_FLOWS_LIMIT with no upper bound. A huge hh_flows_limit lets each new heavy-hitter flow pass the hh_flows_current_cnt check in alloc_new_hh() and forces a fixed-size kzalloc(GFP_ATOMIC) per flow under spoofed traffic, for unbounded memory growth. Bound the attribute with NLA_POLICY_MAX() at 2*HH_FLOWS_CNT (the hhf_init() default) and report the rejected value via extack. The deprecated nested parse is kept: legacy tc does not set NLA_F_NESTED on TCA_OPTIONS. Configs relying on hh_limit above the default were relying on unbounded, unsafe behaviour and are not supported going forward. hhf_init() also ran hhf_change() before setting the default hh_flows_limit, so a user-supplied hh_limit at add time was clobbered back to 2048. Set the default before hhf_change() so the configured value sticks. This is a follow-up to commit eb56a495f59b ("net/sched: hhf: clamp quantum in change and init paths"), which bounded the quantum of the same qdisc; the hh_flows_limit bound is the remaining unbounded knob of that series' scope. Conditions to recreate the bug: CAP_NET_ADMIN in a user namespace; tc qdisc change dev X root hhf hh_limit 4294967295 succeeds and the value is echoed by tc qdisc show, unbounding heavy-hitter flow allocations; also tc qdisc add dev X root hhf hh_limit 500 stores 2048 instead of 500.
Vulnerability data from NVD (NIST) · CISA KEV · EPSS

