2013-06-06 17:46:58

by cbrand

Subject: Re-send of patch to broader audience

I originally sent this patch 4 Apr [1] and 17 Apr [2] to the rtc
mailing list and maintainer, but saw no responses. I think it's
pretty clear that the original code is wrong.

Hopefully third time's the charm...

Chris

[1] https://groups.google.com/forum/?fromgroups=#!topic/rtc-linux/GrsRzUH1ORc
[2] https://groups.google.com/forum/?fromgroups=#!topic/rtc-linux/R3U764kTnhU

2013-06-06 17:45:43

by cbrand

Subject: [PATCH] RTC: Return -EBUSY, not -EACCES when device is busy

From: Chris Brand <[email protected]>

If rtc->irq_task is non-NULL and task is NULL, they always
differ, which means that in rtc_irq_set_state() and
rtc_irq_set_freq(), whenever err is set to -EBUSY it will
then immediately be set to -EACCES, misleading the caller
as to the underlying problem.
---
drivers/rtc/interface.c | 8 ++++----
1 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/rtc/interface.c b/drivers/rtc/interface.c
index 42bd57d..40d4032 100644
--- a/drivers/rtc/interface.c
+++ b/drivers/rtc/interface.c
@@ -698,9 +698,9 @@ retry:
spin_lock_irqsave(&rtc->irq_task_lock, flags);
if (rtc->irq_task != NULL && task == NULL)
err = -EBUSY;
- if (rtc->irq_task != task)
+ else if (rtc->irq_task != task)
err = -EACCES;
- if (!err) {
+ else {
if (rtc_update_hrtimer(rtc, enabled) < 0) {
spin_unlock_irqrestore(&rtc->irq_task_lock, flags);
cpu_relax();
@@ -734,9 +734,9 @@ retry:
spin_lock_irqsave(&rtc->irq_task_lock, flags);
if (rtc->irq_task != NULL && task == NULL)
err = -EBUSY;
- if (rtc->irq_task != task)
+ else if (rtc->irq_task != task)
err = -EACCES;
- if (!err) {
+ else {
rtc->irq_freq = freq;
if (rtc->pie_enabled && rtc_update_hrtimer(rtc, 1) < 0) {
spin_unlock_irqrestore(&rtc->irq_task_lock, flags);
--
1.7.5.4

2013-06-06 18:00:14

by Alessandro Zummo

Subject: Re: Re-send of patch to broader audience

On Thu, 6 Jun 2013 10:46:08 -0700
[email protected] wrote:

> I originally sent this patch 4 Apr [1] and 17 Apr [2] to the rtc
> mailing list and maintainer, but saw no responses. I think it's
> pretty clear that the original code is wrong.

Hi, your patch seems reasonable to me.

Acked-by: Alessandro Zummo <[email protected]>

--

Best regards,

Alessandro Zummo,
Tower Technologies - Torino, Italy

http://www.towertech.it