Port-arm archive

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

Re: Kernel crash on RPI 2 + 3, netbsd-current from May 28, earmv7hf

Added an assertion, received this panic:

Jun  6 06:21:05 armv7 syslogd[314]: restart
Jun  6 06:21:05 armv7 /netbsd: panic: kernel diagnostic assertion "task != NULL" failed: file "../../../../dev/sdmmc/ld_sdmmc.c", line 205 
Jun  6 06:21:05 armv7 /netbsd: cpu0: Begin traceback...
Jun  6 06:21:05 armv7 /netbsd: 0xbac69ecc: netbsd:db_panic+0xc
Jun  6 06:21:05 armv7 /netbsd: 0xbac69ee4: netbsd:vpanic+0x1b4
Jun  6 06:21:05 armv7 /netbsd: 0xbac69efc: netbsd:__udivmoddi4
Jun  6 06:21:05 armv7 /netbsd: 0xbac69f34: netbsd:ld_sdmmc_start+0xb8
Jun  6 06:21:05 armv7 /netbsd: 0xbac69f54: netbsd:ld_diskstart+0x70
Jun  6 06:21:05 armv7 /netbsd: 0xbac69f84: netbsd:dk_start+0xc4
Jun  6 06:21:05 armv7 /netbsd: 0xbac69fac: netbsd:sdmmc_task_thread+0x74
Jun  6 06:21:05 armv7 /netbsd: cpu0: End traceback...
Jun  6 06:21:05 armv7 /netbsd: 
Jun  6 06:21:05 armv7 /netbsd: dump to dev 92,1 not possible
Jun  6 06:21:05 armv7 /netbsd: rebooting...

On 06 Jun 2017, at 08:06, Emilian Bold <emilian.bold%gmail.com@localhost> wrote:

If you have power issues you should see constantly the little rainbow square in the top-right corner of your display.


On Tue, Jun 6, 2017 at 8:42 AM, Benjamin Lorenz <mail%benjaminlorenz.com@localhost> wrote:
Apologies for the bad screenshot. Indeed it is almost in best focus, but my little HDMI monitor screws it up, fighting with the resolution. Was not able yet to reconfigure for a lower one.

Sure I can transcribe into a PR. Was seeing the exact same crash multiple times, with both RPI2 and RPI3 devices, so my intuition tells me this is not a power supply issue. But nevertheless a good point. Might try with another one and see how this changes things. During boot, I see the kernel hanging from time to time, always in the USB detection area. Maybe another indication for bad power supply? Or have you faced this yourself?


On 06 Jun 2017, at 07:03, Paul Goyette <paul%whooppee.com@localhost> wrote:

> This looks like we have some passed an invalid pointer to ld_diskstart() which in turn passed that pointer to ld_sdmmc_start().  But it's not clear how that can happen.
> It would be great if you (Benjamin) could transcribe the backtrace (the screen shot is badly out of focus and nearly impossible to read!), and include it in a PR (use the send-pr utility, or submit via the web at http://www.netbsd.org/cgi-bin/sendpr.cgi?gndb=netbsd )
> On Mon, 5 Jun 2017, Michael Cheponis wrote:
>> Is your unit adequately heat-sinked, adequate air flow, and (especially)
>> adequate power supply?  The power supply is crucial.   Also, no
>> over-clocking, right?
>> On Mon, Jun 5, 2017 at 5:06 AM, Benjamin Lorenz <mail%benjaminlorenz.com@localhost>
>> wrote:
>>> On 05 Jun 2017, at 10:13, Benjamin Lorenz <mail%benjaminlorenz.com@localhost> wrote:
>>> Hi, getting crashes like from screenshot below, after quite some uptime
>>> and compilations. Forgot to display backtrace output, after sync the system
>>> unfortunately froze completely.
>>> Got the same crash again, this time after just a couple of minutes of
>>> uptime, on an RPI 2 (last crash was on an RPI 3), and with 8.99.1 earmv7hf.
>>> Did a backtrace, here you go. See screenshot.
>>> Regards,
>>> Benjamin
>> !DSPAM:5936362b170541839418673!
> +------------------+--------------------------+----------------------------+
> | Paul Goyette     | PGP Key fingerprint:     | E-mail addresses:          |
> | (Retired)        | FA29 0E3B 35AF E8AE 6651 | paul at whooppee dot com   |
> | Kernel Developer | 0786 F758 55DE 53BA 7731 | pgoyette at netbsd dot org |
> +------------------+--------------------------+----------------------------+

Home | Main Index | Thread Index | Old Index