Date: Wed, 9 Mar 2016 20:34:08 +0200 From: Jukka Ukkonen <jau789@gmail.com> To: Ian Lepore <ian@freebsd.org>, Hans Petter Selasky <hps@selasky.org> Cc: freebsd-arm <freebsd-arm@FreeBSD.org> Subject: Re: Odd hang during boot on RPI2 for 2 days now Message-ID: <56E06CA0.4030107@gmail.com> In-Reply-To: <1457365690.13785.179.camel@freebsd.org> References: <56D99C2E.4020301@gmail.com> <56DD5DF8.9090700@gmail.com> <56DD6476.80603@selasky.org> <56DD9F60.7010509@gmail.com> <1457365690.13785.179.camel@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 03/07/16 17:48, Ian Lepore wrote: > On Mon, 2016-03-07 at 17:33 +0200, Jukka Ukkonen wrote: >> On 03/07/16 13:22, Hans Petter Selasky wrote: >>> On 03/07/16 11:54, Jukka Ukkonen wrote: >>>> >>>> Continuing where I previously left with this problem... >>>> When I set boot_verbose=1 in loader.conf I got a little >>>> more information before the poor RPI2 simply froze. >>>> Now the last shown line was... >>>> >>>> random: harvesting attach, 8 bytes (4 bits) from ukbd0 >>>> >>>> If I disconnect the USB keyboard and then reconnect it, >>>> the kernel prints the normal detach and attach messages, >>>> once again returns back to printing the "random: ..." >>>> line shown above, and freezes again. >>>> >>> >>> Then I think the init process is waiting for something. Would you >>> manage >>> to press CTRL+T in the console and see what is printed? >> >> The ctrl-T trick might work with a keyboard connected via >> the serial line pins. With my USB keyboard and HDMI display >> combo it doesn't do anything at all. >> Surely the system must be waiting for something. Of that I >> am quite certain. It does not panic and it keeps reacting to >> the USB keyboard being disconnected and connected. Obviously >> it is still otherwise fine, at least sort of, but also firmly >> stuck waiting an event which either never happens or which >> goes undetected for some reason. > > Part of what's needed to diagnose this problem is dmesg output from a > failed boot that is more than 1 or 2 lines snipped out of context. It > sounds to me like this has nothing to do with the usb keyboard except > for the random coincidence that it is the last usb device to be > enumerated. It seems like the system is hung waiting for an interrupt, > and there's no reason to think it's necessarily a usb interrupt that's > missing. Getting the dmesg output for the failed boot would require a serial console cable connecting the RPI2 to another computer and script running on the other computer to record the output. Lacking such a cable this is a no-go of course. A new fresh dmesg.boot file will be generated only after a normal successful boot. And after a successful boot I would no longer need the content of the file. > Also, it seems unlikely that ^T will produce any output, because it > sounds like init hasn't even started yet, and the low-level console > used during boot doesn't respond to ^T/SIGINFO. At least the system has already tried to start init. There is a line ... start_init: trying /sbin/init shown some 30 lines before the freeze. If anyone wishes to see a photo of what is shown on the display when the system freezes, let me know. I will send a copy separately. There are some 60-70 lines of text shown in the image. I don't wish to spam the whole mailing list with a large image file, though. --jau
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?56E06CA0.4030107>