AZL-115091: CVE-2026-99554 affecting package kernel 6.6.161.1-1
In the Linux kernel, the following vulnerability has been resolved:
net: usb: pegasus: do not advance rxskb->data on dropped frames
For the 0x8513 chip, readbulkcallback() skips the 2-byte receive header by advancing pegasus->rxskb->data in place. This is done before the pktlen > PEGASUSMTU check, and when that check drops the frame the good path resubmits the same skb with the advanced rxskb->data as the transfer buffer, still PEGASUSMTU bytes long.
Every frame with an oversized length field therefore moves the receive buffer two more bytes towards the end of the skb data area. A device that keeps sending such frames makes the host controller write received data over skbsharedinfo and past the end of the allocation, and the corrupted skbsharedinfo is used when the skb is eventually freed. With an emulated device sending frames filled with 0x5a on a KASAN kernel:
BUG: KASAN: slab-out-of-bounds in dummytimer+0xade/0x3f00 [dummyhcd] Write of size 1536 at addr ffff888105821202 by task swapper/1/0 ... Oops: general protection fault, probably for non-canonical address 0xeb4b474b4b4b4b4c: 0000 [#1] PREEMPT SMP KASAN NOPTI KASAN: maybe wild-memory-access in range [0x5a5a5a5a5a5a5a60-0x5a5a5a5a5a5a5a67] RIP: 0010:skbreleasedata+0x36f/0x800
Only skip the header with skbreserve() once the frame is going to be passed up, so the receive buffer always starts at the beginning of the skb data area. Unlike the old pointer adjustment, skbreserve() also moves skb->tail, so the two header bytes now count against the buffer size: lower the length limit for the 0x8513 accordingly, otherwise a pktlen of PEGASUSMTU - 1 or PEGASUSMTU would trip skboverpanic().
Found by code review with LLM assistance and reproduced with an emulated USB device (raw-gadget + dummyhcd).