linux-hwmon.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Re: nct6775: current code detects too few fan sensors for some NCT6793D chips
       [not found] <20190127215959.4431B320112@apps1.cs.toronto.edu>
@ 2019-01-28  0:21 ` Guenter Roeck
  0 siblings, 0 replies; only message in thread
From: Guenter Roeck @ 2019-01-28  0:21 UTC (permalink / raw)
  To: Chris Siebenmann, linux-hwmon

On 1/27/19 1:59 PM, Chris Siebenmann wrote:
>   I have an Asus PRIME Z370-A motherboard, on which I run Fedora. In
> Fedora's 4.19.15 kernel, the nct6775 driver detected all six fan sensors
> on my motherboard.  In their recently released 4.20.3 kernel, it only
> detects five, dropping the sixth fan sensor (which happens to be the
> second chassis fan on this motherboard).
> 
>   I filed a Fedora bug report, and after discussions with the Fedora
> maintainer for this area we think we have narrowed the likely issue down
> to this change:
> 
> hwmon: (nct6775) Separate fan/pwm configuration detection for NCT6793D
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=2d99925a15b639026b67bd96419df6f9d760b212
> 
> If I'm reading the before and after code right, the code before the change
> did:
>      fan6pin = !dsw_en && (cr2d & BIT(1));
>      fan6pin |= creb & BIT(3);
> 
> And the code after dropped the '!dsw_en' part and became just:
>      fan6pin = creb & BIT(3);
> 

Sigh. Too much complexity in those chips. Yes, the first line got lost,
but it applies for NCT6793D. Sorry for that. I'll write a patch.

> (I don't know if it's possible for me to probe things in some way to
> see if dsw_en would be set, or even dump the full raw state of these
> registers.)
> 
>   The full Fedora bug report is here, with various things like dmesg
> output and so on from the two kernels on my hardware:
> 
> 	https://bugzilla.redhat.com/show_bug.cgi?id=1669564
> 
> (Hopefully emailing you directly instead of the linux-hwmon mailing
> list is the right choice. I looked at the mailing list archive and
> it seemed to be for development, with lots of patches, instead of
> issue reports.)
> 

e-mailing me, or any other maintainer, directly without copying the
respective mailing list is never the correct choice. It is ok
to copy me, but it is never ok to not let the rest of the community
know that there may be a problem.

Guenter

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2019-01-28  0:21 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <20190127215959.4431B320112@apps1.cs.toronto.edu>
2019-01-28  0:21 ` nct6775: current code detects too few fan sensors for some NCT6793D chips Guenter Roeck

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).