Hi Salvatore, Salvatore Bonaccorso <car...@debian.org> (2024-04-10): > On Tue, Apr 09, 2024 at 03:33:09PM +0200, Diederik de Haas wrote: > > Hi Cyril, > > > > On Tuesday, 9 April 2024 01:06:43 CEST Cyril Brulebois wrote: > > > Upgrading from linux-image-6.1.0-18-amd64 to linux-image-6.1.0-19-amd64 > > > leads to losing some SMART information, at least as queried by munin (in > > > Debian 12) when it comes to sensors. > > > > Does the problem go away if you revert the following commits on top of -19? > > > > db6338f45971b4285ea368432a84033690eaf53c > > scsi: core: Move scsi_host_busy() out of host lock for waking up EH handler > > > > 1ebd75cefaac6fd74729a7d3157f6eaa59960ae2 > > scsi: core: Move scsi_host_busy() out of host lock if it is for per-command > > > > cf33e6ca12d814e1be2263cb76960d0019d7fb94 > > scsi: core: Add struct for args to execution functions
Preparing that test right now, thanks Diederik.
> Or if that does not find the culprit, would you be able to bisect the
> upstrema changes beweeen 6.1.76 and 6.1.82?
>
> There would be for instance the following ata related change:
>
> 4b085736e44d ("ata: libata-core: Do not try to set sleeping devices to
> standby")
>
> If you can test it with other kernels, does the same happens on
> 6.7.7-1 and 6.7.9-2?
I'm not really keen on playing kernel ping-pong on this particular
machine (which is important in my infrastructure), but I've verified
that adding a SATA disk to an existing VM running Debian 12 on a
QEMU/libvirt Debian 12 host gives me similar results with -18 and -19
kernels (some data in the former case, no data at all in the latter
one).
I think I'd rather stay with 6.1.y kernels if at all possible, to avoid
having to figure out other changes that might be possibly required to
cope with newer kernels.
Cheers,
--
Cyril Brulebois (k...@debian.org) <https://debamax.com/>
D-I release manager -- Release team member -- Freelance Consultant
signature.asc
Description: PGP signature
