[PATCH] Linux 2.6.x VM86 interrupt emulation fixes - the second round

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

[PATCH] Linux 2.6.x VM86 interrupt emulation fixes - the second round

Pavel Pisa
Patch solves VM86 interrupt emulation deadlock on SMP systems.
The VM86 interrupt emulation has been heavily tested and works
well on UP systems after last update, but it seems to deadlock
when we have used it on SMP/HT boxes now.
It seems, that disable_irq() cannot be called from interrupts,
because it waits until disabled interrupt handler finishes
(/kernel/irq/manage.c:synchronize_irq():while(IRQ_INPROGRESS);).
This blocks one CPU after another. Solved by use disable_irq_nosync.
There is the second problem. If IRQ source is fast, it is possible,
that interrupt is sometimes processed and re-enabled by the second
CPU, before it is disabled by the first one, but negative IRQ disable
depths are not allowed. The spinlocking and disabling IRQs over
call to disable_irq_nosync/enable_irq is the only solution found
reliable till now.

Signed-off-by: Michal Sojka <[hidden email]>
Signed-off-by: Pavel Pisa <[hidden email]>

Index: linux-2.6.11.5/arch/i386/kernel/vm86.c
===================================================================
--- linux-2.6.11.5.orig/arch/i386/kernel/vm86.c
+++ linux-2.6.11.5/arch/i386/kernel/vm86.c
@@ -732,12 +732,12 @@ static irqreturn_t irq_handler(int intno
  irqbits |= irq_bit;
  if (vm86_irqs[intno].sig)
  send_sig(vm86_irqs[intno].sig, vm86_irqs[intno].tsk, 1);
- spin_unlock_irqrestore(&irqbits_lock, flags);
  /*
  * IRQ will be re-enabled when user asks for the irq (whether
  * polling or as a result of the signal)
  */
- disable_irq(intno);
+ disable_irq_nosync(intno);
+ spin_unlock_irqrestore(&irqbits_lock, flags);
  return IRQ_HANDLED;
 
 out:
@@ -769,17 +769,20 @@ static inline int get_and_reset_irq(int
 {
  int bit;
  unsigned long flags;
+ int ret = 0;
 
  if (invalid_vm86_irq(irqnumber)) return 0;
  if (vm86_irqs[irqnumber].tsk != current) return 0;
  spin_lock_irqsave(&irqbits_lock, flags);
  bit = irqbits & (1 << irqnumber);
  irqbits &= ~bit;
+ if (bit) {
+ enable_irq(irqnumber);
+ ret = 1;
+        }
+
  spin_unlock_irqrestore(&irqbits_lock, flags);
- if (!bit)
- return 0;
- enable_irq(irqnumber);
- return 1;
+ return ret;
 }
 
 
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [hidden email]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/