Rockbox

Tasklist

FS#13108 - Random crashes when playing music

Attached to Project: Rockbox
Opened by somedude (freeingdevices) - Thursday, 16 March 2017, 15:24 GMT
Last edited by MichaelGiacomelli (saratoga) - Friday, 17 March 2017, 03:26 GMT
Task Type Bugs
Category Music playback
Status Unconfirmed
Assigned To No-one
Operating System Sansa AMSv2
Severity Critical
Priority Normal
Reported Version Daily build (which?)
Due in Version Undecided
Due Date Undecided
Percent Complete 0%
Votes 0
Private No

Details

For a couple of months my Sansa Fuze v2 has been crashing randomly when playing music. Most of the time I just get a white screen with an error message. The last message I got was :

'undefined instruction at 0006DC8C'

(I think that last number varies at each crash)

This didn't happen with v3.13, but it's been happening with every daily build I've tested IIRC.

It doesn't happen very often so it's hard to reproduce. I listen to music in public transports about twice a day and this happens not more than once a week. It seems to happen more often when the battery is low (less than 10%). It generally happens early in a playback session, when the first or second track is played.

Sometimes I don't get a white screen, but the program freezes and some parts of the interface are surrounded with white frames. I don't know if it's the same bug or if it's unrelated.

Note that I'm always listening to music on a microSD card, and I haven't tried on the main memory.

PS: I chose "Sansa Fuze+" in the OS list because I couldn't find Sansa Fuse v2.
This task depends upon

Comment by somedude (freeingdevices) - Thursday, 16 March 2017, 16:02 GMT
Forgot to say : the version I have currently is 1ba5ef7-170303
Comment by MichaelGiacomelli (saratoga) - Thursday, 16 March 2017, 17:13 GMT
Here is a build of today's rockbox git, but with CPU voltage scaling disabled:

http://web.mit.edu/~mgg6/www/rockbox-fuzev2-novoltagescaling.7z

This is one of the more recent major changes we've made. 3.13 is 4 years old at this point though, so there are a huge number of things that could have been at fault.
Comment by MichaelGiacomelli (saratoga) - Wednesday, 12 April 2017, 23:15 GMT
Did this make any difference?
Comment by Michael Sevakis (MikeS) - Thursday, 13 April 2017, 11:48 GMT
It's also something eefc7c7 may address since previously the combination of file types you had could crash it or not.
Comment by somedude (freeingdevices) - Thursday, 27 April 2017, 09:45 GMT
Sorry for the late answer, somehow the mails sent by the rockbox tracker keep arriving in my Spam/ folder.

I haven't tested the firmware version you've linked, but the bug happened once again this morning and I'm not able to boot the device anymore. I had this device for a year, and I bought it "as new" (it's an old model, but it was still in its box). Maybe it had a faulty battery.

It's the second Sansa Fuze v2 that I buy and it's the second one that doesn't last more than a year of use (the other one stopped booting inexplicably after not using it for a couple of month). I also bought an e200 at one point, who didn't last two years (its screen broke, there again inexplicably, without any shock given to it). I think it's the last time I buy a Sandisk device.
Comment by MichaelGiacomelli (saratoga) - Saturday, 29 April 2017, 16:53 GMT
At this point the battery has probably run dead and the device rebooted, but if not, try holding power for about 30 seconds. You may also need to recharge the battery if you left it hung for a few days.
Comment by somedude (freeingdevices) - Wednesday, 13 September 2017, 15:10 GMT
Hum I thought I had updated this ticket.

So, my Fuze v2 works, I have the new version 3.14 running on it, I thought the crashes wouldn't happen anymore but recently I got new crashes. I don't know if it's the same problem. When it happens again I'll post the error message here.
Comment by somedude (freeingdevices) - Thursday, 12 October 2017, 19:21 GMT
It happened again a few times. The two last times it was weird : the music seemed to jolt a bit and then rockbox crashed. Here the message of the last time :

[first line is blacked out but there seem to be something written under it]
rm_handle fail: 16
pc: 30008E38 sp:300B62E8
bt end
Comment by Michael Sevakis (MikeS) - Thursday, 12 October 2017, 21:17 GMT
Edit: Nevermind! I guess I talked about it already. ok, wtf.

Edit: Nevermind! Bever Bind. I haven't had all my coffee and I reread and noticed you said you _didn't_ test at least the linked version AND that in combination with my initial comment leaves me unsure. I don't you didn't test HEAD, that's all.

That definitely indicates buffer corruption. I don't recall if the release includes eefc7c7, which did address a crashing issue talked abut here: http://forums.rockbox.org/index.php/topic,50793.0.html.

Comment by somedude (freeingdevices) - Thursday, 12 October 2017, 22:45 GMT
No, I didn't test the linked version, when the 3.14 was out I assumed it contained the changes you linked to.

Unrelated: why does google keep putting rockbox tracker e-mails in /Spam ?
Comment by Michael Sevakis (MikeS) - Sunday, 15 October 2017, 03:40 GMT
I asked and I suppose 3.14 does have them. Having a set of files available to me that can cause the issue on demand is essential since I'm not sure what to even consider as a culprit at this point. I have a Fuze v2 and have had no issues. Under Debug->View HW Info, it says mine is a variant 0. What is yours?
Comment by somedude (freeingdevices) - Sunday, 15 October 2017, 10:55 GMT
Variant 0 as well.

The problem is that the crash happens always on different files, and I can't reproduce it.
Comment by Michael Sevakis (MikeS) - Saturday, 21 October 2017, 20:16 GMT
Well, maybe I'll spot something while rummaging through it on some other endeavor.
Comment by somedude (freeingdevices) - Thursday, 07 December 2017, 15:35 GMT
It happened again a couple of times on low battery (below 20%) so I've let it run on a very long album until something happened, and it crashed eventually.

Same as usual :

rm_handle fail: 49
pc: 30008E38 sp:300B62E8
bt end
Comment by Michael Sevakis (MikeS) - Sunday, 10 December 2017, 06:00 GMT
Since I did spot something when doing something else, maybe 6ee3b6f is the magic needed?
Comment by Michael Sevakis (MikeS) - Saturday, 16 December 2017, 03:46 GMT
Make sure to move up to latest build to test. There was a mistake in a recent change that I corrected.

Edit: Continuing to test and report on 3.14 release is pointless and doesn't contribute useful information.

Loading...