This is the bug/patch tracker for Rockbox. Click here for more information.
Quick links: Bugs · Patches · Rockbox frontpage
FS#12400 - Quickscreen freezes iAudio X5V on builds newer than rel 3.9.1
Attached to Project:
Rockbox
Opened by Abel (username_already_taken) - Wednesday, 23 November 2011, 01:07 GMT+2
Last edited by Frank Gevaerts (fg) - Sunday, 27 November 2011, 16:37 GMT+2
Opened by Abel (username_already_taken) - Wednesday, 23 November 2011, 01:07 GMT+2
Last edited by Frank Gevaerts (fg) - Sunday, 27 November 2011, 16:37 GMT+2
|
DetailsPresent in builds r31036, 40, 42 and 45, at least.
I initially thought it was somewhat reproducible in a simulator due to the absence of the top/bottom quickscreen defaults, but it seems those were intentionally removed, so the sim is performing correctly. However, my X5V freezes every time I hold REC long enough to call the quickscreen. Sometimes the quickscreen is displayed but cannot be interacted with, others the display freezes immediately. 3.9.1 and earlier releases, that still show the top and bottom defaults, seem to have no issues. Assigning functions to the top and bottom slots before using the quickscreen does not appear to prevent the problem. |
This task depends upon
Closed by Frank Gevaerts (fg)
Sunday, 27 November 2011, 16:37 GMT+2
Reason for closing: Fixed
Additional comments about closing: Fixed in r31075
Sunday, 27 November 2011, 16:37 GMT+2
Reason for closing: Fixed
Additional comments about closing: Fixed in r31075
Seems like it's accessing the disk, as the disk icon appears then it crashes and needs the paperclip to reset.
But it made me think that it *could* be helpful to see your settings - so maybe one (or either) of you could post the config.cfg in the .rockbox folder?
I deleted .rockbox and installed r31045 one more time to repeat the experiment but this time it froze on my very first attempt.
I reinstalled r31045 once more, turned off dircache, rebooted and called the quickscreen: freeze once again. Hope any of this helps.
The current font arguably shouldn't be set to FONT_UI at that point any more, so there might be a bug in the startup code somewhere, but guarding the font_*() functions against invalid ids makes things work.