Release Beta Test: Upcoming (HF5) Firmware Image BETA 3

[EvilDragon]

Aaaand here we go with a new HF5-Betatesting image for you guys


This time we added some new stuff, so please test them and report back


It's still not in PND-form, so flashing this puts your Pandora back to the state you got it. Be sure to backup anything you put on the internal storage if you need it.


Here are the changes that happened from Beta 2 to Beta 3:

* omap3-sgx-modules: backport spinlock fix from omap3630 (thanks to notaz)
* mount.sh: Small optimization (changed rm to rmdir)
* New: Added CPU-Settings script
* pnd_run.sh: Remove empty AUFS-Directories after unmounting (thanks to sebt3)
* MiniMenu: Various bugfixes and minor enhancements (i.e. About screen, etc.)
* MiniMenu: Directory browser is now available in config panel. Will currently only run files that look like they're executable or pnd's.
* MiniMenu: added 'subcats as folders' option and support, set ...

As you can see, most changes have happened with MiniMenu!


It now shows subcategories as folders inside the tabs (can be switched off, if you don't like it) and it has a directory browser included (disabled as default).


It follows strict Freedesktop conventions, any PNDs that are using non-standard categories will appear in OTHER.


There are quite a few PNDs out there with wrong categories that need to be fixed.


Please test MiniMenu extensively and report back!


Another thing that needs to be tested are the fresh and new CPU-Settings (can be found within the settings menu).


Here you can change settings like the allowed maximum OPP, disable the CPU overclock warning, set maximum values allowed for the CPU, set a default speed on bootup, etc.


Please also test it and report back.


In case you haven't checked the changes of HF5 Beta 1 or 2 yet, here is the link to the the Beta 1 thread and here to the Beta 2 thread


How to flash the image?


Easy: Download the zipped image and extract the full contents onto the root of your SD Card.


Put the SD Card into slot 1 (the left one), switch on the Pandora while holding the R-Shoulder button and select "boot from SD1:1"


The flashing process should start. After it's finished, reboot the unit.


The first boot will take a while (up to 10 minutes). Patiently wait and you'll be greeted with the First Boot Wizard.


EDIT: New version available here.

[/CODE]

 

[Wrath Of Khan]

Thanks for all that work ed.I dont have my pandora so cant test this right now.Will use it when my pandys back from repairs.

 

[lordbobjones]

Wow. You are cranking out the updates. Downloading now to try.

 

[meandu229]

Flashed but when I run pidgin this is the output


./bin/pidgin: error while loading shared libraries: libXss.so.1: cannot open shared object file: No such file or directory


Any ideas?

 

[Coldbird]

Just a suggestion from my end, not a bug report. You guys should include a easy configurator, like the CPU one, for creating swapfiles on SD cards...


We all know swap is bad and should be avoided, but with hungry apps like Firefox it would be a big help on making this device more versatile.


One could still leave the default setting at "no swap", but having a simple configurator there to allow setting one, instead of having to manually create the file and tell the OS to use, would make this device way more useful to the average endusers.


Just my 2 cents.

 

[EvilDragon]

Flashed but when I run pidgin this is the output


./bin/pidgin: error while loading shared libraries: libXss.so.1: cannot open shared object file: No such file or directory


Any ideas?

Are you using the latest version?

 

[lordbobjones]

Just flashed it a few minutes ago and am trying things out right now. So far looks great. No issues.

 

[tarwin]

Here is a bug report regarding that CPU-Settings Menu:


I pressed cancel at the dialog where i could have change the cpu setting (after the big warning message).


This was followed by an "The default clockspeed was set to ." message.


This message appers now every time i try to access "set default mhz", the slide to change it doesn't show up anymore.


Thanks for your work! =)


-T

 

[the_marshal]

I wanted to post the same thing... there is a bug in the CPU-Settings. Same windows (OPP slider), only I did click ok (after moving the slider, I reselected OPP3 though). Same message and same symptom.


Also just a small idea... maybe it would be nice to group the Pandora specific settings in a folder inside the settings menu. There are quite a lot of XFCE settings and people might miss the Pandora ones.

 

[tarwin]

I have run into another bug:


i copyed a big file (psx iso) from a sdcard in slot two to slot one. works fine. Then ive copyed the same file from the same cards but switched slots: the copy crashes after about two seconds with io error on slot one. dmesg says:


[ 1654.175659] mmci-omap mmci-omap.0: MMC IRQ 0x208000 : ERRI DCRC


[ 1654.175720] mmcblk0: error -84 transferring data


[ 1654.180419] end_request: I/O error, dev mmcblk0, sector 13861400


I this error is repeatable. slot one to two gives an error, slot two to one works. I ve tried it with different target cards, problem persists.

 

[Zenbot]

ED -- Some bugs for your consideration. I'm no coder, so I'll try and describe what I did as best I can in hopes that you can figure out what went wrong.


This first major bug happened in MiniMenu after freshly flashing the newest beta:


1. Hid some PNDs from showing up in the tabs. Saw that TV-Out and the user management app failed to hide.


2. Hid some tabs.


3. Set CPU on leaving to 800 MHz. Saved/exited config.


4. Launching any app causes machine to freeze.


5. Upon subsequent hard resets, loading MiniMenu took noticeably longer.


6. Reflashed again. Apps now load properly.


7. Set CPU speed on leaving to 800 MHz, and now apps won't load again. So I'd guess the issue is here. In beta 1, this worked fine. [EDIT: No it didn't. Just tried it and it causes apps to freeze in beta 1, too.]


8. Disabled "set CPU speed on leaving" and now apps load properly.


In general it seemed that navigating apps in MiniMenu using the D-pad is not as responsive (slight lag between hitting key and seeing movement on the screen).


This next bug is about the battery: Since the first beta I've been noticing oddness in the way the battery power shows: After charging for a while, sometimes it shows as 0 or 1%; at one point it showed at 47% and then upon reset, it dropped to about 27%. I also left it plugged in all day yesterday and still it was showing only 34% or so. Before trying the betas there was no issue.


Third, I've noticed on occasion even just setting the Pandora down causes it to restart. Perhaps just a battery seating issue?


And finally, as of beta 1 the nubs would occasionally stop working. Resetting the nub settings (which I never changed in the first place?) fixed the issue. I haven't encountered this issue yet in beta 3.


Any questions, let me know and I'll do my best to help!

 

[EvilDragon]

I wanted to post the same thing... there is a bug in the CPU-Settings. Same windows (OPP slider), only I did click ok (after moving the slider, I reselected OPP3 though). Same message and same symptom.


Also just a small idea... maybe it would be nice to group the Pandora specific settings in a folder inside the settings menu. There are quite a lot of XFCE settings and people might miss the Pandora ones.

Ah, thanks. I'll fix that one.


Yeah, have thought about the Pandora-Settings menu myself

I have run into another bug:


i copyed a big file (psx iso) from a sdcard in slot two to slot one. works fine. Then ive copyed the same file from the same cards but switched slots: the copy crashes after about two seconds with io error on slot one. dmesg says:


[ 1654.175659] mmci-omap mmci-omap.0: MMC IRQ 0x208000 : ERRI DCRC


[ 1654.175720] mmcblk0: error -84 transferring data


[ 1654.180419] end_request: I/O error, dev mmcblk0, sector 13861400


I this error is repeatable. slot one to two gives an error, slot two to one works. I ve tried it with different target cards, problem persists.

That's an old one - since the very first release.


Seems to be a bug in the MMC driver, needs some serious bughunting

ED -- Some bugs for your consideration. I'm no coder, so I'll try and describe what I did as best I can in hopes that you can figure out what went wrong.


This first major bug happened in MiniMenu after freshly flashing the newest beta:


1. Hid some PNDs from showing up in the tabs. Saw that TV-Out and the user management app failed to hide.


2. Hid some tabs.


3. Set CPU on leaving to 800 MHz. Saved/exited config.


4. Launching any app causes machine to freeze.


5. Upon subsequent hard resets, loading MiniMenu took noticeably longer.


6. Reflashed again. Apps now load properly.


7. Set CPU speed on leaving to 800 MHz, and now apps won't load again. So I'd guess the issue is here. In beta 1, this worked fine. [EDIT: No it didn't. Just tried it and it causes apps to freeze in beta 1, too.]


8. Disabled "set CPU speed on leaving" and now apps load properly.

Those are all MiniMenu bugs, so that's one thing for skeezix


What happens if you clock to 800MHz and then run the apps?


Can your machine do 800MHz?

This next bug is about the battery: Since the first beta I've been noticing oddness in the way the battery power shows: After charging for a while, sometimes it shows as 0 or 1%; at one point it showed at 47% and then upon reset, it dropped to about 27%. I also left it plugged in all day yesterday and still it was showing only 34% or so. Before trying the betas there was no issue.


Third, I've noticed on occasion even just setting the Pandora down causes it to restart. Perhaps just a battery seating issue?

This is all typical for a not properly aligned battery, no software bug


Try to put a piece of paper between the battery and the lid (best is the area where the contacts are), that should definately help if the battery is a bit too loose there.

And finally, as of beta 1 the nubs would occasionally stop working. Resetting the nub settings (which I never changed in the first place?) fixed the issue. I haven't encountered this issue yet in beta 3.

Hmm, nothing has changed there.


Are you sure they didn't stop after running MUPEN (which sets the nub to joystick mode but doesn't seem to set them back)?

Any questions, let me know and I'll do my best to help!

Any of those reports are a good help, thanks

 

[Link]

Yesterday while going from Mini-Menu to XFCE, I toggled the WiFi on, browsed the forums, toggled Wi-Fi off, my left nub started slowing down on response in one direction. Then I went into setting to try to reset the nubs and I got an error that the .PND wouldn't run for the Nubs. I powered the unit off, everything came back fine, nub worked, I could run the .PND to reset nubs too. Just wanted to mention this as it seemed odd.

 

[tarwin]

I have run into another bug:


i copyed a big file (psx iso) from a sdcard in slot two to slot one. works fine. Then ive copyed the same file from the same cards but switched slots: the copy crashes after about two seconds with io error on slot one. dmesg says:


[ 1654.175659] mmci-omap mmci-omap.0: MMC IRQ 0x208000 : ERRI DCRC


[ 1654.175720] mmcblk0: error -84 transferring data


[ 1654.180419] end_request: I/O error, dev mmcblk0, sector 13861400


I this error is repeatable. slot one to two gives an error, slot two to one works. I ve tried it with different target cards, problem persists.

That's an old one - since the very first release.


Seems to be a bug in the MMC driver, needs some serious bughunting

Damn. =/


Any idea if it's always like this, that you can't copy stuff from port 1 to 2?


Anyway I can help?


-T

 

[urjaman]

I wanted to post the same thing... there is a bug in the CPU-Settings. Same windows (OPP slider), only I did click ok (after moving the slider, I reselected OPP3 though). That's an old one - since the very first release.


Seems to be a bug in the MMC driver, needs some serious bughunting

Needs more bughunting to really fix, but see my comment at http://bugs.openpandora.org/index.php?do=details&task_id=152 - I think including this workaround would be much better than the alternative of not including ...

 

[Zenbot]

...


Those are all MiniMenu bugs, so that's one thing for skeezix


What happens if you clock to 800MHz and then run the apps?


Can your machine do 800MHz?


...


This is all typical for a not properly aligned battery, no software bug


Try to put a piece of paper between the battery and the lid (best is the area where the contacts are), that should definately help if the battery is a bit too loose there.


...


Hmm, nothing has changed there.


Are you sure they didn't stop after running MUPEN (which sets the nub to joystick mode but doesn't seem to set them back)?

Well sir, I have been able to set apps to 800 MHz in the past. Using the CPU app I haven't had problem with this. But, is there a more optimum speed I should be using? Are some Pandoras not able to go that high? I got mine a few weeks ago.


I'll try the battery thing you suggested. Thanks!


And while I have tried using Mupen64 and encountered the issue with that app, the problem I'm describing re the nubs can happen even when I haven't used Mupen64.


All that being said, I'm pretty impressed with the OS overall. That I can say I finished a few games already is gratifying (not to mention a big relief after all that waiting!).


Cheers mate!

 

[NoeleyC]

I decided to make this update a bit more interesting (and slightly less time consuming) for myself this time, so on my 2 hour drive across Ireland from my parent's in Cork to where I live, I downloaded the entire update via my O2 Hotspot onto my Pandora in a moving car! Download speed clocked in at over 400k for some bit of it, which is imressive for a Pandora that used to get disgruntled at anything over 30kb. Updated my machine when I got there, and tried out the new cpu speed settings, seem to work great and nicely userfriendly. it's nice to eke out a few extra fps in thigns like PSX emulation, didn't really try much N64 though yet. All in all, no problems found by me yet


Not really used Minimenu much at all, but I'll try it out and see

 

[notaz]

I wanted to post the same thing... there is a bug in the CPU-Settings. Same windows (OPP slider), only I did click ok (after moving the slider, I reselected OPP3 though). That's an old one - since the very first release.


Seems to be a bug in the MMC driver, needs some serious bughunting

Needs more bughunting to really fix, but see my comment at http://bugs.openpandora.org/index.php?do=details&task_id=152 - I think including this workaround would be much better than the alternative of not including ...

Hm haven't seen that one, does it help? I'll probably cherry pick this from mainline to retain attribution though, instead of using your version.

 

[urjaman]

Hm haven't seen that one, does it help? I'll probably cherry pick this from mainline to retain attribution though, instead of using your version.

IIRC I had to do some changes to the function (adapt to kernel internal interface changes in time). But maybe you can work it out someway pretty, like do a set of commits retaining attribution, then one commit by you or me (I dont care for attribution that much, but whatever) doing the interface fixing.


Anyhows the randomly found-out workaround codepath is as follows:


1. Try to read big block (2 or more sectors)


2. Get spurious read error due to concurrent write on the other channel. (Why? - resolving this would be the real bug fix)


3. New code tries to find out what sector failed by reading the big block 1 sector at a time (old 2.6.27 code marked lots of sectors as bad at this point)


4. No sector fails (why does this work?, dunno but works), and code returns with success (as the re-reading is done with a flag and a goto to retry point - so it can also normally exit with success.)

 

[skeezix]

Re: CPU speed opn leaving .. if set up, then when mmenu 'exits' (to run an app, say) it'll hit the script to set the speed. If the system wedges after that, then I woudl think it also wedges if you set the clock up using the normal methods. ie: Minimenu isn't doing anything tricky there.


It _could_ be some oddity with X11 and fork though .. there some seem to be an oddity on occasion if I run something direct from mmenu, where X gets weird (noticed that when running non-pnds from the dir browser.. now it runs 'sh -c /path/foo' and no problem. Maybe when mmenu is running the clock setitng script, it should hand it to /bin/sh as well. Doesn't make sense, so I'll have to look into it, but that ought to be a good workaround for now.


I'll do that


jeff


Also, there is a .1-.2 sec delay possible when you hit keys; I never thought anyone would notice, but that was a short term thing; its going away next build, as well.