[PATCH] kprobes: fix bug when probed on task and isr functions

This patch fixes a race condition where in system used to hang or sometime
crash within minutes when kprobes are inserted on ISR routine and a task
routine.

The fix has been stress tested on i386, ia64, pp64 and on x86_64.  To
reproduce the problem insert kprobes on schedule() and do_IRQ() functions
and you should see hang or system crash.

Signed-off-by: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
Signed-off-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Acked-by: Prasanna S Panchamukhi <prasanna@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

1 files changed, 22 insertions, 0 deletions

diff --git a/kernel/kprobes.c b/kernel/kprobes.c
index 3b7653f2e7ae..f3ea492ab44d 100644
--- a/kernel/kprobes.c
+++ b/kernel/kprobes.c
@@ -155,14 +155,36 @@ void __kprobes free_insn_slot(kprobe_opcode_t *slot)
 /* Locks kprobe: irqs must be disabled */
 void __kprobes lock_kprobes(void)
 {
+        unsigned long flags = 0;
+
+        /* Avoiding local interrupts to happen right after we take the kprobe_lock
+         * and before we get a chance to update kprobe_cpu, this to prevent
+         * deadlock when we have a kprobe on ISR routine and a kprobe on task
+         * routine
+         */
+        local_irq_save(flags);
+
         spin_lock(&kprobe_lock);
         kprobe_cpu = smp_processor_id();
+
+        local_irq_restore(flags);
 }
 
 void __kprobes unlock_kprobes(void)
 {
+        unsigned long flags = 0;
+
+        /* Avoiding local interrupts to happen right after we update
+         * kprobe_cpu and before we get a a chance to release kprobe_lock,
+         * this to prevent deadlock when we have a kprobe on ISR routine and
+         * a kprobe on task routine
+         */
+        local_irq_save(flags);
+
         kprobe_cpu = NR_CPUS;
         spin_unlock(&kprobe_lock);
+
+        local_irq_restore(flags);
 }
 
 /* You have to be holding the kprobe_lock */