Messages in this thread | | | Subject | Re: [PATCH RFC V2 3/5] jump_label: if a key has already been initialized, don't nop it out | From | Steven Rostedt <> | Date | Tue, 04 Oct 2011 14:05:51 -0400 |
| |
On Tue, 2011-10-04 at 13:53 -0400, Jason Baron wrote: > On Tue, Oct 04, 2011 at 09:30:01AM -0700, H. Peter Anvin wrote: > > On 10/04/2011 07:10 AM, Jason Baron wrote: > > > > > > 1) The jmp +0, is a 'safe' no-op that I know is going to initially > > > boot for all x86. I'm not sure if there is a 5-byte nop that works on > > > all x86 variants - but by using jmp +0, we make it much easier to debug > > > cases where we may be using broken no-ops. > > > > > > > There are *plenty*. jmp+0 is about as pessimal as you can get. > > > > The current recommendation when you don't know the CPU you're running at is: > > > > 3E 8D 74 26 00 (GENERIC_NOP5_ATOMIC) > > > > ... on 32 bits and ... > > > > 0F 1F 44 00 00 (P6_NOP5_ATOMIC) > > > > ... on 64 bits. > > > > -hpa > > > > We're currently patching the code at run-time (boot and module load > time), with the 'ideal' no-op anyway, so the initial no-op doesn't > really matter much (other than to save patching if the initial and ideal > match).
Out of correctness, we should still update this to use the proper "default" nops, as mcount already does.
-- Steve
| |