CVE-2026-23243: RDMA/umad: Reject negative data_len in ib_umad_write
In the Linux kernel, the following vulnerability has been resolved:
RDMA/umad: Reject negative datalen in ibumadwrite
ibumadwrite computes datalen from user-controlled count and the MAD header sizes. With a mismatched user MAD header size and RMPP header length, datalen can become negative and reach ibcreatesendmad(). This can make the padding calculation exceed the segment size and trigger an out-of-bounds memset in allocsendrmpplist().
Add an explicit check to reject negative datalen before creating the send buffer.
KASAN splat: [ 211.363464] BUG: KASAN: slab-out-of-bounds in ibcreatesendmad+0xa01/0x11b0 [ 211.364077] Write of size 220 at addr ffff88800c3fa1f8 by task spraythread/102 [ 211.365867] ibcreatesendmad+0xa01/0x11b0 [ 211.365887] ibumadwrite+0x853/0x1c80
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Configuration
Add an explicit check in ib_umad_write to reject negative data_len before creating the MAD/RMPP structures (before calling ib_create_send_mad), so that alloc_send_rmpp_list() cannot compute an out-of-bounds memset due to padding calculations exceeding the segment size.
Linux kernel RDMA/umad (ib_umad_write, alloc_send_rmpp_list, ib_create_send_mad) data_len = reject negative values (check for data_len < 0 and return error)
Event History
Frequently Asked Questions
What is the severity of CVE-2026-23243?
CVE-2026-23243 has a medium severity rating as it involves improper input validation that could lead to unwanted behavior.
How do I fix CVE-2026-23243?
To fix CVE-2026-23243, upgrade your Linux kernel to the patched version provided by your distribution.
What software versions are affected by CVE-2026-23243?
CVE-2026-23243 affects various versions of the Linux kernel that utilize the RDMA/umad subsystem.
What are the potential impacts of CVE-2026-23243?
The potential impacts of CVE-2026-23243 include denial of service or system instability resulting from improper handling of user-controlled input.
Is there a known exploit for CVE-2026-23243?
As of now, there are no publicly known exploits specifically targeting CVE-2026-23243.