CVE-2026-80876

In the Linux kernel, the following vulnerability has been resolved:

ring-buffer: Fix event length with forced 8-byte alignment

When RB FORCE 8BYTE ALIGNMENT is true, rb calculate event length() reserves the space of event->array[0] for placing the data length and rb update event() stores the data length in event->array[0] accordingly. As a result the whole event length will add extra 4 bytes for sizeof(event.array[0]) unconditionally.

But ring buffer event length() only subtracts the sizeof(event->array[0]) for events larger than RB MAX SMALL DATA + sizeof(event->array[0]). As a result, small events on architectures with RB FORCE 8BYTE ALIGNMENT=true report a data length that is 4 bytes larger than expected.

To fix it, add the RB FORCE 8BYTE ALIGNMENT as a condition to subtract the size of that length field whenever RB FORCE 8BYTE ALIGNMENT is true.

This issue is observed in a riscv64 kernel with CONFIG HAVE 64BIT ALIGNED ACCESS set to y, when we run ftrace selftest trace marker raw.tc, we get the weird log: for cases where the id is 1..100, the number of data field is 8N, but once id exceeds 100, the number of data field becomes 8N+4:

...

...

After applying this change, the number of data field keeps being 8*N+4 consistently.