In the Linux kernel, the following vulnerability has been resolved:
can: isotp: fix potential CAN frame reception race in isotp_rcv()
When receiving a CAN frame the current code logic does not consider concurrently receiving processes which do not show up in real world usage.
Ziyang Xuan writes:
The following syz problem is one of the scenarios. so->rx.len is changed by isotprcvff() during isotprcvcf(), so->rx.len equals 0 before allocskb() and equals 4096 after allocskb(). That will trigger skboverpanic() in skb_put().
======================================================= CPU: 1 PID: 19 Comm: ksoftirqd/1 Not tainted 5.16.0-rc8-syzkaller #0 RIP: 0010:skbpanic+0x16c/0x16e net/core/skbuff.c:113 Call Trace: <TASK> skboverpanic net/core/skbuff.c:118 [inline] skbput.cold+0x24/0x24 net/core/skbuff.c:1990 isotprcvcf net/can/isotp.c:570 [inline] isotprcv+0xa38/0x1e30 net/can/isotp.c:668 deliver net/can/afcan.c:574 [inline] canrcvfilter+0x445/0x8d0 net/can/afcan.c:635 canreceive+0x31d/0x580 net/can/afcan.c:665 canrcv+0x120/0x1c0 net/can/afcan.c:696 _netifreceiveskbonecore+0x114/0x180 net/core/dev.c:5465 _netifreceive_skb+0x24/0x1b0 net/core/dev.c:5579
Therefore we make sure the state changes and data structures stay consistent at CAN frame reception time by adding a spinlock in isotprcv(). This fixes the issue reported by syzkaller but does not affect real world operation.