* [PATCH] drivers/rtc/rtc-m41t93.c: don't let get_time() reset error state
[not found] <201204231230.q3NCUbbP015327@gatekeeper.vosshq.de>
@ 2012-04-23 20:53 ` Andrew Morton
2012-04-24 4:51 ` Voss, Nikolaus
0 siblings, 1 reply; 2+ messages in thread
From: Andrew Morton @ 2012-04-23 20:53 UTC (permalink / raw)
To: linux-arm-kernel
On Mon, 23 Apr 2012 12:51:23 +0200
Nikolaus Voss <n.voss@weinmann.de> wrote:
> If the rtc reports the time might be invalid due to oscillator
> failure, this flags must not be reset by get_time() as the read
> operation doesn't make the time valid.
>
> Instead, the flag is reset in set_time() when a valid time is
> to be written.
This is rather vague - what is/are "these flags"? From the patch I
think you're referring to (the maddeningly undocumented)
M41T93_FLAG_ST? If so, I'd suggest something like
: If the rtc reports the time might be invalid due to oscillator failure,
: the M41T93_FLAG_ST flag must not be reset by get_time() as the read
: operation doesn't make the time valid.
:
: Instead, the M41T93_FLAG_ST flag is reset in set_time() when a valid time
: is to be written.
Secondly, you provided no description of the user-visible effects of
the bug. Hence I cannot work out which kernel version(s) this patch
should be merged into.
^ permalink raw reply [flat|nested] 2+ messages in thread
* [PATCH] drivers/rtc/rtc-m41t93.c: don't let get_time() reset error state
2012-04-23 20:53 ` [PATCH] drivers/rtc/rtc-m41t93.c: don't let get_time() reset error state Andrew Morton
@ 2012-04-24 4:51 ` Voss, Nikolaus
0 siblings, 0 replies; 2+ messages in thread
From: Voss, Nikolaus @ 2012-04-24 4:51 UTC (permalink / raw)
To: linux-arm-kernel
Andrew Morton wrote on 2012-04-23:
> On Mon, 23 Apr 2012 12:51:23 +0200
> Nikolaus Voss <n.voss@weinmann.de> wrote:
>
>> If the rtc reports the time might be invalid due to oscillator
>> failure, this flags must not be reset by get_time() as the read
>> operation doesn't make the time valid.
>>
>> Instead, the flag is reset in set_time() when a valid time is
>> to be written.
>
> This is rather vague - what is/are "these flags"?
Sorry, that's true. I speak of the M41T93_FLAG_OF
(Oscillator Fail) flag.
[...]
> Secondly, you provided no description of the user-visible effects of
> the bug. Hence I cannot work out which kernel version(s) this patch
> should be merged into.
What the user sees is that the first get_time() read reports
an error (invalid time) and the second reports no error, so the
first read magically heals the rtc time.
The patch applies to all kernel versions with support for m41t93 rtc.
I'll repost the patch with a proper description.
Niko
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2012-04-24 4:51 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <201204231230.q3NCUbbP015327@gatekeeper.vosshq.de>
2012-04-23 20:53 ` [PATCH] drivers/rtc/rtc-m41t93.c: don't let get_time() reset error state Andrew Morton
2012-04-24 4:51 ` Voss, Nikolaus
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).