Date: Wed, 2 Oct 2024 08:59:09 -0400 From: mike tancsa <mike@sentex.net> To: Chris6 via freebsd-hardware <freebsd-hardware@freebsd.org> Subject: Re: watchdog timer programming (progress) Message-ID: <24fba042-3c77-4dc5-9660-c73a102e33ca@sentex.net> In-Reply-To: <86r08ydfb5.fsf@cthulhu.stephaner.labo.int> References: <3065debc-8d4f-4487-abbb-c9408810cea6@sentex.net> <86plotbk5b.fsf@cthulhu.stephaner.labo.int> <9008b389-ab06-401d-9a95-84f849ca602a@sentex.net> <86plosdv48.fsf@cthulhu.stephaner.labo.int> <78e9461c-b93d-403f-b3a1-3568548b9283@sentex.net> <86h6a1egcs.fsf@cthulhu.stephaner.labo.int> <c4343727-606c-409a-8618-9da732fc3059@sentex.net> <868qvddwph.fsf@cthulhu.stephaner.labo.int> <2d850ccc-2e90-4a1a-927c-045d4750d570@sentex.net> <864j5xehes.fsf@cthulhu.stephaner.labo.int> <be51c81f-11c7-4f90-9df7-9635ec1df2c4@sentex.net> <86zfnocpb8.fsf@cthulhu.stephaner.labo.int> <8b730043-a759-4bb4-b7ee-323a317ce6d2@sentex.net> <1b346afb-d6ed-4f00-8dcf-5cdd389d210b@sentex.net> <82dc6dbf-8aa7-45ef-8fe9-08dc54973c2c@sentex.net> <86r08ydfb5.fsf@cthulhu.stephaner.labo.int>
index | next in thread | previous in thread | raw e-mail
<trimming a bunch>
On 10/2/2024 5:13 AM, Stephane Rochoy wrote:
>>
>> it seems the soft timeout value action is never overridden for some
>> reason.
>>
>> This kinda feels like a bug / pr ?
>
> Well, honestly I'm puzzled:
> - in one hand, watchdog.c don't seems to use wd_softtimeout_act
> - and on the other hand hardclock seems to directly call
> watchdog_fire which just kdb_enter or panic.
>
> Note that wd_timeout_cb seems to be about both pretimeout and
> timeout handling.
>
I was able to get things to work the way I want by setting the
pre-action timeout. But, I need to set things 'right' the first time I
call it. If I do
0{p9999}# watchdogd --pretimeout-action panic --softtimeout-action panic
-t 10
0{p9999}# killall -9 watchdogd
0{p9999}# KDB: stack backtrace:
#0 0xffffffff80b7fefd at kdb_backtrace+0x5d
#1 0xffffffff80abec93 at hardclock+0x103
#2 0xffffffff80abfe8b at handleevents+0xab
#3 0xffffffff80ac0b7c at timercb+0x24c
#4 0xffffffff810d0ebb at lapic_handle_timer+0xab
#5 0xffffffff80fd8a71 at Xtimerint+0xb1
#6 0xffffffff804b3685 at acpi_cpu_idle+0x2c5
#7 0xffffffff80fc48f6 at cpu_idle_acpi+0x46
#8 0xffffffff80fc49ad at cpu_idle+0x9d
#9 0xffffffff80b67bb6 at sched_idletd+0x576
#10 0xffffffff80aecf7f at fork_exit+0x7f
#11 0xffffffff80fd7dae at fork_trampoline+0xe
0{p9999}#
0{p9999}# watchdogd --pretimeout-action panic --softtimeout
--softtimeout-action panic -t 10
watchdogd: setting WDIOC_SETSOFT 1: Invalid argument
watchdogd: patting the dog: Invalid argument
71{p9999}#
But if I reboot the box and make sure nothing is set, and start the daemon
watchdogd --pretimeout-action panic --softtimeout --softtimeout-action
panic -t 10
it works
0{p9999}# watchdogd --pretimeout-action panic --softtimeout
--softtimeout-action panic -t 10
0{p9999}# killall -9 watchdogd
0{p9999}# panic: watchdog soft-timeout, WD_SOFT_PANIC set
cpuid = 0
time = 1727873819
KDB: stack backtrace:
#0 0xffffffff80b7fefd at kdb_backtrace+0x5d
#1 0xffffffff80b32bd1 at vpanic+0x131
#2 0xffffffff80b32a93 at panic+0x43
#3 0xffffffff809827bb at wd_timeout_cb+0x6b
#4 0xffffffff80b50b0c at softclock_call_cc+0x12c
#5 0xffffffff80b52355 at softclock_thread+0xe5
#6 0xffffffff80aecf7f at fork_exit+0x7f
#7 0xffffffff80fd7dae at fork_trampoline+0xe
Timeout initializing vt_vga
Uptime: 50s
Automatic reboot in 15 seconds - press a key on the console to abort
I think some of the dead ends I ran into was due to this reason. My
stock image has watchdogd_enable="YES" and having that start up would
set something that would then lead to dead ends. But, if I JUST start up
watchdogd --pretimeout-action panic --softtimeout --softtimeout-action
panic -t 10
it works. I wonder if warrants a PR for the docs at least. Anyways,
thanks again for helping me work through all this!
---Mike
help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?24fba042-3c77-4dc5-9660-c73a102e33ca>