[Xen-changelog] Fix calling point for do_softirq, which should only be checked

[Xen patchbot -unstable <patchbot-unstable@xxxxxxxxxxxxxxxxxxx>]

# HG changeset patch
# User fred@xxxxxxxxxxxxxxxxxxxxx
# Node ID 9be7fe98a5565bf8f6297332e0367cd6904f52af
# Parent  bd77de43ed4ad21f0bf96afc58f1dce0695df12f
Fix calling point for do_softirq, which should only be checked
at the very point back to guest domain. Or else unexpected domain
switch may happen in nested interrupt in Xen.

Signed-off-by Kevin Tian <kevin.tian@xxxxxxxxx>

diff -r bd77de43ed4a -r 9be7fe98a556 xen/arch/ia64/patch/linux-2.6.11/irq_ia64.c
--- a/xen/arch/ia64/patch/linux-2.6.11/irq_ia64.c       Tue Aug  2 10:13:02 2005
+++ b/xen/arch/ia64/patch/linux-2.6.11/irq_ia64.c       Tue Aug  2 10:47:41 2005
@@ -20,11 +20,19 @@
                        __do_IRQ(local_vector_to_irq(vector), regs);
  
                        /*
-@@ -167,6 +173,95 @@
+@@ -167,6 +173,103 @@
        irq_exit();
  }
  
 +#ifdef  CONFIG_VTI
++#define vmx_irq_enter()               \
++      add_preempt_count(HARDIRQ_OFFSET);
++
++/* Now softirq will be checked when leaving hypervisor, or else
++ * scheduler irq will be executed too early.
++ */
++#define vmx_irq_exit(void)    \
++      sub_preempt_count(HARDIRQ_OFFSET);
 +/*
 + * That's where the IVT branches when we get an external
 + * interrupt. This branches to the correct hardware IRQ handler via
@@ -72,7 +80,7 @@
 +       * 16 (without this, it would be ~240, which could easily lead
 +       * to kernel stack overflows).
 +       */
-+      irq_enter();
++      vmx_irq_enter();
 +      saved_tpr = ia64_getreg(_IA64_REG_CR_TPR);
 +      ia64_srlz_d();
 +      while (vector != IA64_SPURIOUS_INT_VECTOR) {
@@ -106,7 +114,7 @@
 +       * handler needs to be able to wait for further keyboard interrupts, 
which can't
 +       * come through until ia64_eoi() has been done.
 +       */
-+      irq_exit();
++      vmx_irq_exit();
 +      if ( wake_dom0 && current != dom0 ) 
 +              domain_wake(dom0->vcpu[0]);
 +}

_______________________________________________
Xen-changelog mailing list
Xen-changelog@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-changelog