Messages in this thread | | | Date | Fri, 10 May 2002 10:28:55 -0400 (EDT) | From | "Richard B. Johnson" <> | Subject | Re: spin-locks |
| |
On Fri, 10 May 2002, David Woodhouse wrote:
> > root@chaos.analogic.com said: > > Well, here is code that worked on linux 2.2.17. Same CPUs, same > > everything... Just a different version of OS... > > I suspect your code was protected by the BKL in 2.2.17, not by your > 'spinlocks'. > > root@chaos.analogic.com said: > > cli > > lock > > incb (lockf) # Bump lock-value > > Ponder what happens if two CPUs get here at the same time. Lock count is > now two. > > > 1: cmpb $1,(lockf) # See if we own it > > jnz 1b # Nope, spin until we do. > > Now they both spin forever.
Yep. You are correct. Obvious bug that went undetected for over a year.
Cheers, Dick Johnson
Penguin : Linux version 2.4.18 on an i686 machine (797.90 BogoMips).
Windows-2000/Professional isn't.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |