Re: STI, HLT in acpi_cpu_idle_c1

On Saturday 10 July 2004 08:11 am, Scott Long wrote:
> John Baldwin wrote:
> > On Thursday 08 July 2004 08:41 pm, Gerrit Nagelhout wrote:
> >>John Baldwin wrote:
> >>>I think it does handle the interrupt, but that the EOI is
> >>>somehow lost or
> >>>ignored,or that somehow we don't send an EOI in some edge case.
> >>
> >>I switched the 2 CPUs on the system, and the problem did not
> >>track the CPU, so it is unlikely to be a hardware problem.
> >>I've been looking through the interrupt code, and one thing
> >>seems very suspicious, and I am hoping that someone can shed
> >>some light on it ... ie whether it's legal or not.
> >>In apic_vector.s, some of the interrupt handlers (hardclock,
> >>cpuast, and a few others) call doreti.  Under certain conditions,
> >>this can call ast().  The ast function can call PROC_LOCK,
> >>which is a sleep mutex.
> >>What would happen in this case?  How could the interrupt
> >>possibly block?  At this point, the interrupt handler has
> >>already called eoi, but not iret.  Would the processor except
> >>other interrupts, or get stuck?
> >>Thanks,
> >
> > Once the EOI is sent to the APIC, the processor will except other
> > interrupts. However, most of the interrupt code also runs with interrupts
> > disabled.  The exception is in ast(), meaning that another interrupt can
> > come in and nest during an ast(), but ast() is not called during nested
> > returns, so it won't nest infinitely.  iret doesn't do any sort of EOI
> > magic, all it does is popf followed by ret.  The popf usually enables
> > interrupts when it restores the interrupt flag's state.  It might still
> > be a hardware problem in the chipset, esp. since it always happens on a
> > logical CPU.  The fact that no other boxes in the field besides the two
> > identical boxes you have have demonstrated this problem makes me very
> > suspicious with respect to the hardware.
>
> Well, the real question is what happens in when the interrupt context
> has to contest on the proc lock in ast().  What goes to sleep, and how
> it wake up?  Does the interrupt context have a real process/thread
> associated with it, or does it pretend to be curthread, or what?  Since
> the whole point of ithreads was to allow sleep locks in interrupt
> handlers, isn't this a major violation?

The interrupt context is basically just curthread.  It's the same stack and 
all.  However, since you've already EOI'd before you call ast(), you've 
released the PIC so it is free to interrupt again if need be.  When you block 
on the proc lock in ast() you are blocking curthread just like usual.  The 
reason fast interrupt handlers (which != ast()) can't use a sleep lock is 
that they would block whatever thread they interrupted.  ast() however 
performs actions on the current thread, so it's ok for curthread to block in 
ast().

-- 
John Baldwin <jhb_at_FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve"  =  http://www.FreeBSD.org