Re: [10.99.12] Panic (softints stuck)

BERTRAND Joël <[email protected]> Tue, 18 Feb 2025 17:37:41 +0100

Newsgroups	gmane.os.netbsd.devel.kernel
Message-ID	<[email protected]>

Michael van Elst a écrit :
> [email protected] (=?UTF-8?Q?BERTRAND_Jo=c3=abl?=) writes:
> 
>> [ 148425,692985] dumping to dev 18,1 (offset=3D253015, size=3D4162677):
>> [ 148425,692985] dump [   1.0000000] Copyright (c) 1996, 1997, 1998,
> 
> 
> My guess is that dump on a raid partition is broken.
> 
> raid checks the partition to be of type FS_SWAP and
> redirects the request to the FS_RAID partition of the component.
> 
> The component driver does the same check and fails.
> 
> Please try:
> 
> Index: sys/dev/dksubr.c
> ===================================================================
> RCS file: /cvsroot/src/sys/dev/dksubr.c,v
> retrieving revision 1.117
> diff -p -u -r1.117 dksubr.c
> --- sys/dev/dksubr.c    6 Feb 2025 20:11:46 -0000       1.117
> +++ sys/dev/dksubr.c    18 Feb 2025 09:58:56 -0000
> @@ -831,7 +831,7 @@ dk_dump(struct dk_softc *dksc, dev_t dev
>         } else {
>                 int nsects, sectoff;
>  
> -               if (p->p_fstype != FS_SWAP) {
> +               if (p->p_fstype != FS_SWAP && p->p_fstype != FS_RAID) {
>                         DPRINTF(DKDB_DUMP, ("%s: bad fstype %d\n", __func__,
>                             p->p_fstype));
>                         return ENXIO;
> 

	Your patch fixes this issue :

Checking for core dump...
savecore: reboot after panic:
[ 454.5104887] panic: crashme plain old panic
savecore: system went down at Tue Feb 18 17:28:28 2025
savecore: writing compressed core to /var/crash/netbsd.6.core.gz

	Thanks,

	JB

signature.asc (application/pgp-signature, 228 B)