This is the mail archive of the
systemtap@sourceware.org
mailing list for the systemtap project.
Re: [PATCH]Fix the order of atomic operations in restore_previous_kprobes on ia64
- From: Shaohua Li <shaohua dot li at intel dot com>
- To: Masami Hiramatsu <mhiramat at redhat dot com>
- Cc: "Luck, Tony" <tony dot luck at intel dot com>, ia64 <linux-ia64 at vger dot kernel dot org>, LKML <linux-kernel at vger dot kernel dot org>, systemtap-ml <systemtap at sources dot redhat dot com>
- Date: Tue, 22 Jan 2008 09:07:46 +0800
- Subject: Re: [PATCH]Fix the order of atomic operations in restore_previous_kprobes on ia64
- References: <4790D1DE.8030700@redhat.com>
On Sat, 2008-01-19 at 00:20 +0800, Masami Hiramatsu wrote:
> From: Masami Hiramatsu <mhiramat@redhat.com>
>
> Fix the order of atomic operations to prevent overwriting
> prev_kprobe[0].
> To pop values from stack, we must decrement stack index right AFTER
> reading values.
>
> Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com>
> ---
> Details of this issue was reported to
> http://sources.redhat.com/bugzilla/show_bug.cgi?id=2071
>
> arch/ia64/kernel/kprobes.c | 7 ++++---
> 1 file changed, 4 insertions(+), 3 deletions(-)
>
> Index: 2.6.24-rc8/arch/ia64/kernel/kprobes.c
> ===================================================================
> --- 2.6.24-rc8.orig/arch/ia64/kernel/kprobes.c 2008-01-17
> 21:14:01.000000000 -0500
> +++ 2.6.24-rc8/arch/ia64/kernel/kprobes.c 2008-01-17
> 21:14:01.000000000 -0500
> @@ -381,9 +381,10 @@
> static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk
> *kcb)
> {
> unsigned int i;
> - i = atomic_sub_return(1, &kcb->prev_kprobe_index);
> - __get_cpu_var(current_kprobe) = kcb->prev_kprobe[i].kp;
> - kcb->kprobe_status = kcb->prev_kprobe[i].status;
> + i = atomic_read(&kcb->prev_kprobe_index);
> + __get_cpu_var(current_kprobe) = kcb->prev_kprobe[i-1].kp;
> + kcb->kprobe_status = kcb->prev_kprobe[i-1].status;
> + atomic_sub(1, &kcb->prev_kprobe_index);
> }
>
> static void __kprobes set_current_kprobe(struct kprobe *p,
Acked. Thanks for the patch.
Thanks,
Shaohua