Re: [PATCH] lkdtm: Fix race when crashpoint is hit multiple timesbefore checking count

From: Ankita Garg
Date: Wed Jun 15 2011 - 08:58:40 EST


Hi,

On Tue, Jun 14, 2011 at 04:40:27PM -0700, Josh Hunt wrote:
> We observed the crash point count going negative in cases where the crash point
> is hit multiple times before the check of "count == 0" is done. Because of this
> we never call lkdtm_do_action(). This patch just adds a spinlock to protect
> count.
>

Good catch! The fix Looks fine.

> Reported-by: Tapan Dhimant <tdhimant@xxxxxxxxxx>
> Signed-off-by: Josh Hunt <johunt@xxxxxxxxxx>

Acked-by: Ankita Garg <ankita@xxxxxxxxxx>

> ---
> drivers/misc/lkdtm.c | 8 ++++++++
> 1 files changed, 8 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/misc/lkdtm.c b/drivers/misc/lkdtm.c
> index 81d7fa4..150cd70 100644
> --- a/drivers/misc/lkdtm.c
> +++ b/drivers/misc/lkdtm.c
> @@ -120,6 +120,7 @@ static int recur_count = REC_NUM_DEFAULT;
> static enum cname cpoint = CN_INVALID;
> static enum ctype cptype = CT_NONE;
> static int count = DEFAULT_COUNT;
> +static DEFINE_SPINLOCK(count_lock);
>
> module_param(recur_count, int, 0644);
> MODULE_PARM_DESC(recur_count, " Recursion level for the stack overflow test, "\
> @@ -230,11 +231,14 @@ static const char *cp_name_to_str(enum cname name)
> static int lkdtm_parse_commandline(void)
> {
> int i;
> + unsigned long flags;
>
> if (cpoint_count < 1 || recur_count < 1)
> return -EINVAL;
>
> + spin_lock_irqsave(&count_lock, flags);
> count = cpoint_count;
> + spin_unlock_irqrestore(&count_lock, flags);
>
> /* No special parameters */
> if (!cpoint_type && !cpoint_name)
> @@ -349,6 +353,9 @@ static void lkdtm_do_action(enum ctype which)
>
> static void lkdtm_handler(void)
> {
> + unsigned long flags;
> +
> + spin_lock_irqsave(&count_lock, flags);
> count--;
> printk(KERN_INFO "lkdtm: Crash point %s of type %s hit, trigger in %d rounds\n",
> cp_name_to_str(cpoint), cp_type_to_str(cptype), count);
> @@ -357,6 +364,7 @@ static void lkdtm_handler(void)
> lkdtm_do_action(cptype);
> count = cpoint_count;
> }
> + spin_unlock_irqrestore(&count_lock, flags);
> }
>
> static int lkdtm_register_cpoint(enum cname which)

--
Regards,
Ankita Garg (ankita@xxxxxxxxxx)
Linux Technology Center
IBM India Systems & Technology Labs,
Bangalore, India
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/