SanDisk Sansa c200v2, m200v4, clipv1, clipv2, clip+, and fuzev2

[funman]

I had a crashed while waiting for TRAN state yesterday (didn't think to write down the state in which it was).

I'll try highering the delay from 1 sec to 2 secs

EDIT: crashed again in state DATA

[Xanikseo]

Just like to report that the scroll-wheel randomly stopped working on r26086. Cause is not known, but this has never happened before... A reset fixed the problem.

[Jennifur]

Just like to report that the scroll-wheel randomly stopped working on r26086. Cause is not known, but this has never happened before... A reset fixed the problem.

This happened to me quite a few times. I thought it was just me messing around with the scrollwheel code. I'm trying r26092-100516 right now to see if that fixes it.

EDIT:
r26092-100516 has the same problem. I had it paused while it was on the charger for about 20 min while I was doing something else. The scroll-wheel stopped working. The previous version I used that worked just fine was: r25938-100514 with and without me messing around with the scroll-wheel code.

[funman]

Perhaps the problem is caused by r26060 ? Can you check if reverting the commit helps?

[bor_ka]

I believe I've detected a bug in the r25299. Sansa Fuze v1, 8GB. May be I can trace it, but I need help.

In short - player hangs in database population. More detailed description and steps to reproduce are in the FlySpray http://www.rockbox.org/tracker/task/11267. BTW, look at my last comment, the initial description was not accurate (!).

I can build custom fw, and may be debug it - but need a help, since I've never worked with the external hardware. Linux/BSD/kernel/C though are Ok .

[csavery]

Regarding scrollwheel freezing. I get this in r26096 after some minutes of no button activity while playing.
I reverted to 26059 and 26060 and both didn't freeze for me.
I'm on r26060 now and letting it go longer without button activity to see if freezing occurs.

[saratoga]

I believe I've detected a bug in the r25299. Sansa Fuze v1, 8GB. May be I can trace it, but I need help.

In short - player hangs in database population. More detailed description and steps to reproduce are in the FlySpray http://www.rockbox.org/tracker/task/11267. BTW, look at my last comment, the initial description was not accurate (!).

I can build custom fw, and may be debug it - but need a help, since I've never worked with the external hardware. Linux/BSD/kernel/C though are Ok .

I see you're using some odd compiler version as well as EABI.  Does this still occur if you use the normal compiler (4.0.3)?

[nomak]

Hello! Thanks for GREAT work! I use rockbox first time. Maybe I do smth wrong...But i can't clear dynamic playlist and to do other delete operations(delete folders, files doesn't work) rockbox r26043-100515 Fuze V2

Now evrything OK. I pressed >|| instead of central button to confirm delete operations.
- Sometimes can't exit from rockpaint.
- once in brickmania in 24(or higher level)  ball was disappeared in right bottom corner , so I must to play again from 1st level =)

[Xanikseo]

Try the most recent build here.
I'm so pleased that AMS are willing to give the datasheet for rockbox development . http://svn.rockbox.org/viewvc.cgi?view=rev&revision=26116

[funman]

I'm so pleased that AMS are willing to give the datasheet for rockbox development . http://svn.rockbox.org/viewvc.cgi?view=rev&revision=26116

Looking at the commit log I think I must precise I got the datasheet for the AS3543 (audio front end and power management unit), not the "AS3525v2" which is only the name we gave to this SoC, not knowing what it is exactly (well it's a newer version of as3525, or at least compatible).

kudos to them !

[bilditup1]

The 'auto-reboot to OF on USB' feature from 26116 is really cool. Kudos to funman et al for working on this, and to AMS for being considerate. I wonder though, funman - did you ask them for other datasheets as well, or was this the most important one? What's the upshot of getting these sheets from an audio perspective?

[funman]

Well I only asked for this one (yesterday), I wouldn't know what else I could ask for.

As for audio, I don't know, audio works pretty well already.

It seems even line out works, but we need to detect dock plugging (tenfoot is working on it)

[csavery]

funman: I'm still seeing scrollwheel lock up in r26114. I have an idea what it is but I'm not familiar with the code so I'm pretty much just poking in the dark and just putting this out there as a question.

In button-fuzev2.c there is the change as of r26060,

long v = (time - last_wheel_post);
    if (v < WHEEL_LOOP_INTERVAL) /* avoid too frequent updates */
        return ;

I see v is defined as long and not unsigned long. According to my calcs with timer freq of 1.5 MHz this will wrap into a negative number in about 23 minutes. This is round about the time taken for the scroll wheel to lock up. So I had the idea that if the scroll wheel doesn't move for 23 minutes and no event is posted then the time delta would be negative and cause this function to return without processing the event for another 23 minutes until it wraps positive again. Is this possible?

In my tests it takes approx this much time to freeze and only with no wheel activity. Unfortunately due to how long you have to wait it's rather slow to test.

[ssorgatem]

I've observed the scrollwheel freeze too, after leaving the fuze some minutes inactive (they may well hace been more than 23 minutes)

[funman]

csavery: it looks possible, i have seen the wheel stop working and work again when testing charging.

I didn't measure the time elapsed but it might have been multiples of 23 minutes.

Instead of counting the number of timer ticks I think we should just count rockbox ticks separately because they overflow in much longer time (248 days with HZ=100).

If you have a patch can you post it on flyspray?

thanks!


If you use a current build with Clip+ it will try to reboot to OF on USB plug, but with current mkamsboot it will fail and power off the Clip+
So you should:

• Not plug USB
• Keep center button pressed when plugging USB
• Reinstall the bootloader (the last bootloader will work) with a mkamsboot build from SVN or wait for a new mkamsboot release.

I'm not sure if we will make a mkamsboot release soon because there is some good progress on USB support in rockbox on all the models (AMSv1 & v2), perhaps we should just wait for the drivers to work.

What is the opinion of the other devs on this?