PT-2026-64657 · Linux · Linux
CVE-2026-64436
·
Published
2026-07-25
·
Updated
2026-07-25
None
No severity ratings or metrics are available. When they are, we'll update the corresponding info on the page.
In the Linux kernel, the following vulnerability has been resolved:
net: af key: initialize alg key len for IPComp states
pfkey msg2xfrm state() handles the IPComp (SADB X SATYPE IPCOMP) case by
allocating x->calg and copying only the algorithm name:
x->calg = kmalloc obj(*x->calg);
if (!x->calg) {
err = -ENOMEM;
goto out;
}
strcpy(x->calg->alg name, a->name);
x->props.calgo = sa->sadb sa encrypt;Unlike the authentication (x->aalg) and encryption (x->ealg) branches of
the same function, the compression branch never initializes
calg->alg key len. IPComp carries no key and the allocation only
reserves sizeof(struct xfrm algo) (i.e. no room for a key), so the field
is left containing uninitialized slab data.
calg->alg key len is later used as a length by xfrm algo clone() when an
IPComp state is cloned during XFRM MSG MIGRATE:
xfrm state migrate()
xfrm state clone and setup()
x->calg = xfrm algo clone(orig->calg);
kmemdup(orig, xfrm alg len(orig));where xfrm alg len() returns sizeof(*alg) + (alg key len + 7) / 8. With
a non-zero garbage alg key len, kmemdup() reads past the end of the
68-byte calg object. Adding an IPComp SA via PF KEY and then migrating
it triggers (net-next, KASAN, init on alloc=0):
BUG: KASAN: slab-out-of-bounds in kmemdup noprof+0x44/0x60
Read of size 4164 at addr ff11000025a74980 by task diag2/9287
CPU: 3 UID: 0 PID: 9287 Comm: diag2 7.1.0-rc6-g903db046d557 #1
Call Trace:
dump stack lvl+0x10e/0x1f0
print report+0xf7/0x600
kasan report+0xe4/0x120
kasan check range+0x105/0x1b0
asan memcpy+0x23/0x60
kmemdup noprof+0x44/0x60
xfrm state migrate+0x70a/0x1da0
xfrm migrate+0x753/0x18a0
xfrm do migrate+0xb47/0xf10
xfrm user rcv msg+0x411/0xb50
netlink rcv skb+0x158/0x420
xfrm netlink rcv+0x71/0x90
netlink unicast+0x584/0x850
netlink sendmsg+0x8b0/0xdc0
sys sendmsg+0x9f7/0xb90
sys sendmsg+0x134/0x1d0
sys sendmsg+0x16d/0x220
do syscall 64+0x116/0x7d0
entry SYSCALL 64 after hwframe+0x77/0x7f
Allocated by task 9287:
kasan save stack+0x33/0x60
kasan save track+0x14/0x30
kasan kmalloc+0xaa/0xb0
pfkey add+0x2652/0x2ea0
pfkey process+0x6d0/0x830
pfkey sendmsg+0x42c/0x850
sys sendto+0x461/0x4b0
x64 sys sendto+0xe0/0x1c0
do syscall 64+0x116/0x7d0
entry SYSCALL 64 after hwframe+0x77/0x7f
The buggy address belongs to the object at ff11000025a74980
which belongs to the cache kmalloc-96 of size 96
The buggy address is located 0 bytes inside of
allocated 68-byte region [ff11000025a74980, ff11000025a749c4)
Depending on the uninitialized value the same field can instead request
an oversized kmemdup() allocation and make the migration clone fail.
The XFRM netlink path is not affected: verify one alg() rejects an
XFRMA ALG COMP attribute shorter than xfrm alg len(), so a calg added via
XFRM MSG NEWSA is always self-consistent.
Initialize calg->alg key len to 0, matching the aalg/ealg branches.
Found an issue in the description? Have something to add? Feel free to write us 👾
Related Identifiers
Affected Products
Linux