Messages in this thread | | | Date | Tue, 19 Jan 2021 11:40:07 +0100 | From | Borislav Petkov <> | Subject | Re: [PATCH v1 06/19] perf/x86/intel/ds: Check insn_get_length() retval |
| |
On Mon, Jan 04, 2021 at 02:19:19PM +0100, Peter Zijlstra wrote: > Do we really still need the !insn.length? That is, it *should* be > impossible to not fail insn_get_length() and still have a 0 length, > seeing how x86 doesn't have 0 length instructions.
I was responding to the "doubly important" thing in the comment scarying me about an infinite loop and thus left the length check in, in case the insn decoder would have a bug and return success but still have insn.length 0.
With the length check the endless loop won't happen but let's be brave here ... :-)
So removed.
-- Regards/Gruss, Boris.
https://people.kernel.org/tglx/notes-about-netiquette
| |