SuperZaxxon v1.73 - maintenance release

[ingoreis]

Yes had Problems with Busybox too.


An Pandora System Updgrade is not more possible.


Reflashed here to 1.73 with the Fullflash.


This /media/mmcblk0 is little anoying because Thunar itself show the Name 128GBSD and many other Filechooser show /media/mmcblk0/


The Name of the SD Cards do not mount correctly into the /media/ Folder.


Maybe because this Busybox Errors in Backround?


Maybe i can copy/paste this Errors this Evening.


Greeting from Germany

 

[notaz]

OK I can see the problem on my fully reflashed pandora (the upgraded one is not affected, strangely). It's not really related to the "not linking" errors, we had those since day one, what actually happened is that blkid package went MIA for some reason and that broke SD card naming.


Try running the upgrade again, should hopefully fix it (SD names should return on card replug or reboot).

 

[ingoreis]

Thanks i will try that this Evening.

 

[canseco]

It's working fine again, thanks.

 

[ingoreis]

I can prove that.


It work realy wonderfull again..no other Errors found.


Aaand little surprise for tha Pandora Android Users.


The Wlan Connection do not more disconnect after 30 Seconds..wonderfull stable Wlan Connection

 

[notaz]

The Wlan Connection do not more disconnect after 30 Seconds..wonderfull stable Wlan Connection

Hmm I haven't changed anything in that area, so it's just wifi firmware being unpredictable.


BTW here is a full flash with the fix:


http://notaz.gp2x.de/tmp/SuperZaxxon173u1.zip (deleted, use updated image from the first post)


I've tested and it works for me, but I'll wait for anyone else to test it too before moving it to official location and updating the first post.

 

[Linux-SWAT]

I confirm that the wrong SD naming problem is gone with this new fullflash.

 

[notaz]

Thanks for testing, moved the file to official location and updated the links.

 

[FBnil]

Thanks Notaz, will find a time next weekend to reflash and put back my bashrc, tv firmware, etc.

 

[Hooka]

Notaz: any chance to get the newest version of libSDL-ttf pulled from opkg and installed in the firmware? My newest port of gmenu2x requires some of the new hinting code in it. Your call on if it's worth it I'd just rather have it included if I'm going to release the menu instead of having to use LD_Library_Path and include the lib with the menu.

 

[Galaxis]

While we're talking about blkid - after using my Pandora for several years, I notice that automounting is done by /etc/udev/scripts/mount.sh, and there's a blacklist file (/etc/udev/mount.blacklist). This comes in very handy when booting from sd card, and you don't want your root volume to be automounted a second time...

When I was juggling cards today it has irked me that I have to use device names for the blacklist, and those are bound to the sd card slot. How about the following patch to mount.sh that allows using UUIDs (I'm not really adept with bash, so I don't know if changing the if condition that way fails on some weird corner cases):

--- /etc/udev/scripts/mount.sh.old  2015-12-25 01:45:46.468753107 +0100
+++ /etc/udev/scripts/mount.sh  2015-12-25 02:12:07.734375570 +0100
@@ -12,11 +12,12 @@
name="`basename "$DEVNAME"`"
name2="`$blkid "$DEVNAME" -c /dev/null -o value -s LABEL | sed 's/ /_/g'`"
name3=$(echo $DEVNAME | sed 's/.*dev.//g')
+name4="`$blkid "$DEVNAME" -c /dev/null -o value -s UUID`"
uid=$(cat /tmp/currentuid)

for line in `grep -v ^# /etc/udev/mount.blacklist`
do
-  if ( echo "$DEVNAME" | grep -q "$line" )
+  if [ "$DEVNAME" = "$line" -o "$name4" = "$line" ]
  then
  logger "udev/mount.sh" "[$DEVNAME] is blacklisted, ignoring"
  exit 0

[edit] Now I just need a way to prevent the excluded volumes from showing up on the desktop...

 

[notaz]

Notaz: any chance to get the newest version of libSDL-ttf pulled from opkg and installed in the firmware?

I could do it but updating things in the past has resulted in many broken PNDs. There was an automatic ABI checker but it seems it went down, so I have no idea how safe it would be to update. So in short if you could somehow show it's safe I can do it.

-  if ( echo "$DEVNAME" | grep -q "$line" )
+  if [ "$DEVNAME" = "$line" -o "$name4" = "$line" ]

Hmm without grep it will change old behavior (having something like /dev/mmcblk in blacklist won't affect all cards), which may break some's setup.

 

[bukkit]

The u-boot fixes require to use the flasher. If you want the u-boot update but don't want to wipe your system, you can prepare the SD card like you would for a reflash, only leave rootfs.img and rootfs.md5 files out. It will then reflash u-boot but leave your rootfs intact (you can then update that using "upgrade OS" tool).

If I do that, I get the new kernel, but not the respective modules, so I end up with a half-system which doesn't get me into the internet (wifi modules missing), so I can't finish the update. Maybe it works if you take out all uImage* from bootf.tgz, but I didn't try. (I fixed it by copying the modules.)

Other than that: Thanks a lot for your work. It's appreciated, though probably not enough.

 

[notaz]

If I do that, I get the new kernel, but not the respective modules, so I end up with a half-system which doesn't get me into the internet (wifi modules missing), so I can't finish the update.

Hmm right, haven't thought about that, updated the instructions to ask to run OS update before updating the bootloader.

 

[dgame]

Thanks for the update.

Full Flashed 1.73u1 now getting SQUASHFS errors with Pale Moon

[37892.820343] SQUASHFS error: xz_dec_run error, data probably corrupt
[37892.820465] SQUASHFS error: squashfs_read_data failed to read block 0x7b3e3a
[37892.820617] SQUASHFS error: Unable to read data cache entry [7b3e3a]
[37892.820678] SQUASHFS error: Unable to read page, block 7b3e3a, size 6290
[37892.820831] palemoon: unhandled page fault (7) at 0x406bfc60, code 0x80000007 (read)
[37892.820861] palemoon: pc=406bfc60 (libnss3.so+7ac60), lr=40658d2c

Reboot fixes it temporarily.
Did not happen with 1.71

Any suggestions?

 

[notaz]

Works fine here, I think your SD card (or at least palemoon PND) is actually corrupt.

 

[Galaxis]

Hmm without grep it will change old behavior (having something like /dev/mmcblk in blacklist won't affect all cards), which may break some's setup.

Right, the grep does a substring search... Tbh I ignored that part since I couldn't figure out how to combine a subshell call in an if clause with a OR condition...

 

[Yoyobuae]

I'm trying to do a full reflash on a 1Ghz Pandora with SuperSaxxon version 1.61. I haven't done full reflash since SuperSaxxon 1.55 if that matters.

I can get the "Boot from SD1:1" option on the menu that shows while holding R button. But when I select it I get this:

reading boot.scr

5158740 bytes read
## Executing script at 82000000
Wrong image format for "source" command
script failed.

And pressing B button just takes me back to the previous menu where "Boot from SD1:1" is.

The files seems to have downloaded fine:

gerald-openpandora:~$ cd /media/PANCARDB/
gerald-openpandora:/media/PANCARDB$ md5sum boot.scr bootf.tgz rootfs.img      
c96550f0b03102e62967b2ad38e46182  boot.scr
f8017682954caed9e272527f721192d9  bootf.tgz
dfa618151de9cc16cbc911667058e4f7  rootfs.img
gerald-openpandora:/media/PANCARDB$ cat rootfs.md5
dfa618151de9cc16cbc911667058e4f7  rootfs.img

I'm using a 64GB SD card formatted FAT32 for this.

 

[ZXDunny]

I'm using a 64GB SD card formatted FAT32 for this.

Yeah, don't.

Use a smaller SD card.

D.

 

[Linux-SWAT]

Or use FAT16.