tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: [10.0_STABLE] Hard lock



Taylor R Campbell a écrit :
>> Date: Fri, 18 Oct 2024 17:52:24 +0200
>> From: BERTRAND Jo�l <joel.bertrand%systella.fr@localhost>
>>
>> 	I have tested. Same hard lock with iscsi patches from -current.
> 
> Exactly what date from current?

	I have patched my -10 tree with following patches :

(1/2) https://mail-index.netbsd.org/source-changes/2023/12/28/msg149090.html
Use correct status value SCSI_BUSY (0x08) instead of XS_BUSY (7)...

(2/2) https://mail-index.netbsd.org/source-changes/2024/08/24/msg153012.html
Avoid race in timeout handling.
Don't try to wake up CCB without connection (which led to a NULL pointer
deref).

> In particular, you're using altq, right?  Can you confirm with

	Yes.

> ident /netbsd | grep -F cprng_fast.c

legendre:[/] > ident /netbsd | grep -F cprng_fast.c
     $NetBSD: cprng_fast.c,v 1.18.4.1 2023/08/11 14:35:25 martin Exp $

> (or similarly on whatever kerneel you've booted) whether you're running
> with this change?
> 
> https://mail-index.netbsd.org/source-changes/2024/10/15/msg153958.html

	Is this patch in -10 tree ?

> When the system locks up while running a current kernel, do you get a
> heartbeat panic after 15sec?

	I haven't tested -current kernel yet. I will try.

> Do you have a dump device configured?

	Yes, raid0b, but no crash dump is written. System is just freezed.

> When the system locks up, is
> there any indication of disk activity?  (That might be the system
> dumping to disk, which would be useful for diagnostics later -- the
> amount of time it takes depends on the size of RAM and the latency of
> disk I/O.)
> 
> When you said `DIAGNOSTIC doesn't return usable information' in your
> original message, what does that mean?  When the system locks up, can
> you reset it (note: reset, not power off and back on), and share dmesg
> output, which (on reset) should include the previous boot?

	DIAGNOSTIC kernel only panics (when it tries to access to iscsi
targets). And after each dead lock or panic, I have to mannualy done
fsck as soft updates are not enough to keep ffs2 in a good state (one
hour to run fsck -fpP in single user mode).

	Regards,

	JB

Attachment: signature.asc
Description: OpenPGP digital signature



Home | Main Index | Thread Index | Old Index