Report of a nonworking P5100/SB800 TCO WatchDog after 1f182aca2300 ("Watchdog: sp5100_tco: Refactor MMIO base address initialization") changes

Report of a nonworking P5100/SB800 TCO WatchDog after 1f182aca2300 ("Watchdog: sp5100_tco: Refactor MMIO base address initialization") changes

[Salvatore Bonaccorso]

Hi

A user in Debian reported quite a long back an issue that after the
update to the 5.18 version but present still up to recent kernel
versions his SP5100/SB800 TCO WatchDog would not work anymore.

The report is at: https://bugs.debian.org/1054231

> Hello. WatchDog broken since version 5.17. Kernel write an error to syslog.
> Tested versions: 5.10, 5.15, 5.16, 5.17, 5.19, 6.1, 6.5.
> 
> From syslog:
> KERNEL 5.16 (and lower)
> Oct 19 12:24:12 debian-gateway kernel: Linux version 5.16.0-6-amd64 (debian-kernel@lists.debian.org) (gcc-11 (Debian 11.2.0-19) 11.2.0, GNU ld (GNU Binutils for Debian) 2.38) #1 SMP PREEMPT Debian 5.16.18-1 (2022-03-29)
> ...
> Oct 19 12:24:13 debian-gateway kernel: sp5100_tco: SP5100/SB800 TCO WatchDog Timer Driver
> 
> KERNEL 5.17 (and higher)
> Oct 18 20:36:00 debian-gateway kernel: Linux version 5.17.0-3-amd64 (debian-kernel@lists.debian.org) (gcc-11 (Debian 11.3.0-3) 11.3.0, GNU ld (GNU Binutils for Debian) 2.38) #1 SMP PREEMPT Debian 5.17.11-1 (2022-05-26)
> ...
> Oct 18 20:36:01 debian-gateway kernel: sp5100_tco: SP5100/SB800 TCO WatchDog Timer Driver
> Oct 18 20:36:01 debian-gateway kernel: sp5100-tco sp5100-tco: Failed to reserve MMIO or alternate MMIO region
> Oct 18 20:36:01 debian-gateway kernel: sp5100-tco: probe of sp5100-tco failed with error -16

I had a look only recently at open bugs which were stalled in Debian
and had a look at upstrema changes, and I'm not sure if this was a
kernel problem at all an the change is introduced around


5559598742fb ("drm/amd/display: Pass non-null to dcn20_validate_apply_pipe_split_flags")
(https://git.kernel.org/linus/1f182aca230086d4a4469c0f9136a6ea762d6385)
which is 5.18-rc1 and got backported to 5.17.10 and 5.15.42.

The reporter confirmed that the problem does not show up with the old
packaged image we can get from the snapshots archive 5.7.6-1 but
happends reconfirmed with 6.7.11-1 and say there are no MMIO related
settings to adjust in BIOS.

Does the above ring some bell?

Regards,
Salvatore

Re: Report of a nonworking P5100/SB800 TCO WatchDog after 1f182aca2300 ("Watchdog: sp5100_tco: Refactor MMIO base address initialization") changes

[Salvatore Bonaccorso]

On Fri, Feb 28, 2025 at 06:45:15AM +0100, Salvatore Bonaccorso wrote:
> Hi
> 
> A user in Debian reported quite a long back an issue that after the
> update to the 5.18 version but present still up to recent kernel
> versions his SP5100/SB800 TCO WatchDog would not work anymore.
> 
> The report is at: https://bugs.debian.org/1054231
> 
> > Hello. WatchDog broken since version 5.17. Kernel write an error to syslog.
> > Tested versions: 5.10, 5.15, 5.16, 5.17, 5.19, 6.1, 6.5.
> > 
> > From syslog:
> > KERNEL 5.16 (and lower)
> > Oct 19 12:24:12 debian-gateway kernel: Linux version 5.16.0-6-amd64 (debian-kernel@lists.debian.org) (gcc-11 (Debian 11.2.0-19) 11.2.0, GNU ld (GNU Binutils for Debian) 2.38) #1 SMP PREEMPT Debian 5.16.18-1 (2022-03-29)
> > ...
> > Oct 19 12:24:13 debian-gateway kernel: sp5100_tco: SP5100/SB800 TCO WatchDog Timer Driver
> > 
> > KERNEL 5.17 (and higher)
> > Oct 18 20:36:00 debian-gateway kernel: Linux version 5.17.0-3-amd64 (debian-kernel@lists.debian.org) (gcc-11 (Debian 11.3.0-3) 11.3.0, GNU ld (GNU Binutils for Debian) 2.38) #1 SMP PREEMPT Debian 5.17.11-1 (2022-05-26)
> > ...
> > Oct 18 20:36:01 debian-gateway kernel: sp5100_tco: SP5100/SB800 TCO WatchDog Timer Driver
> > Oct 18 20:36:01 debian-gateway kernel: sp5100-tco sp5100-tco: Failed to reserve MMIO or alternate MMIO region
> > Oct 18 20:36:01 debian-gateway kernel: sp5100-tco: probe of sp5100-tco failed with error -16
> 
> I had a look only recently at open bugs which were stalled in Debian
> and had a look at upstrema changes, and I'm not sure if this was a
> kernel problem at all an the change is introduced around
> 
> 
> 5559598742fb ("drm/amd/display: Pass non-null to dcn20_validate_apply_pipe_split_flags")
> (https://git.kernel.org/linus/1f182aca230086d4a4469c0f9136a6ea762d6385)
> which is 5.18-rc1 and got backported to 5.17.10 and 5.15.42.

And me failing twice in a row, the above should clearly be
1f182aca2300 ("Watchdog: sp5100_tco: Refactor MMIO base address
initialization").

> The reporter confirmed that the problem does not show up with the old
> packaged image we can get from the snapshots archive 5.7.6-1 but
> happends reconfirmed with 6.7.11-1 and say there are no MMIO related
> settings to adjust in BIOS.
> 
> Does the above ring some bell?
> 
> Regards,
> Salvatore

Regards,
Salvatore