Re: new problem?

On Sunday, April 09, 2017 06:08:04 PM Jeffrey Bouquet wrote:
> i386_set_ldt: start=-1 num=1 descs=0x38449fac

This message is a harmless debug printf.  Somehow you compiled
sys/i386/i386/sys_machdep.c with 'DEBUG' defined to enable it
though.  (The message is not enabled by default.)
 
> Tons of those, eight at a time, newly spamming /var/log/messages, maybe after a
> fsck_ffs -y the partitions after a crash, that fixed it,
> 
> as in
> df 
> df output, then eight of the above line with the hex values varying only
> ls
> ls output, then eight of the above line with the hex values varying only
> ......................................
> 2nd problem
> persistent LOR in  [several]
> 
> nvidia_os:c:662, 824
>  
> [dmesg only, above, ]
> 
> vfs_mount.c:1277
> ffs_softdep.c:1908
> vfs_subr.c:2600,  2150, 
> ffs_vnops.c:280
> vfs_bio.c:3500
> ufs_dirhash.c:281
> vfs_syscalls.c:3364
> 
> [some above also in dmesg, most in /var/log/messages...

There are several known LORs in the VFS/UFS code that are either false
positives or ones that have effectively been harmless.

-- 
John Baldwin