Search results

Last person who complained about these types of errors had a corrupted SD card.

Probably newer gcc is more strict about duplicate symbols (I'm still on 4.9), I guess you'll need to disable rtl8712 module in .config.

You need to load the right kernel modules from kernel/drivers/gpu/pvr/ , there are many versions of them there. See http://pandorawiki.org/SGX_drivers , all versions of .so files, pvrsrvctl and kernel modules must match. Also see /etc/init.d/pvr-init script from SZ.

No, everything related to SGX is closed source. There is no pvrsrvinit for armhf, they removed it in 4.08.00.01/1.9.19.2139099 and replaced it with pvrsrvctl, which you have to run as "pvrsrvctl --start --no-module". TI have only released armhf blobs for 4.08 or newer.

I don't think partitioning will help, it's the longer init time that large cards sometimes take that is the problem, older bootloader just isn't waiting long enough. You need to find smaller/older cards, those usually work.

sudo opkg update sudo opkg upgrade sudo opkg install fuse-exfat

Mouse is just relative movement, it can't be mapped to touchscreen easily (absolute input). In some rare cases it can be done (uae4all?), but it does that by hacking the AmigaOS, from what I understand. I guess the best result could be obtained by statically recompiling the Windows version, but...

Any reason to suffer with the OS on SD card? Setups like that are never tested, so you are on your own when using them. If I were you I'd just reflash the NAND to latest and live happily ever after. If you are concerned about NAND wear, the driver has good wear leveling, and there have been no...

It's like 10-100us per second, or something like that.

Are you sure it was different on other SZ versions? Unfortunately this is not true, there is a small drift and always was, I've only minimized it as much as I could. The drift even differs between different pandoras, I'm not even sure why.

That symlink looks indeed broken though, but judging by the date it has been like that for years...

I've noticed that if you reboot a PC, at the moment the PC resets pandora's OTG sometimes gets knocked off and won't work until cable replug. To get around that you have to plug the pandora in at the time window between the actual PC reset (when other USB device LEDs go out for a short time) and...

Thanks to this easy reproducer I've found the bug - the new power saving thing (off mode) was losing the gamma table. This is perfectly normal as the off mode loses all the state, and the driver already handles it properly by restoring all the registers, but for gamma table it's a bit different...

Should actually still boot but with missing drivers (no wifi, touchscreen, PNDs, etc), from which it's possible to recover by downloading http://www.openpandora.org/feeds/unstable/omap3-pandora/omap3-pandora-kernel2_3.2-pandora-r33.5_omap3-pandora.ipk and doing "opkg install" on it. Hmm haven't...

Finally rebuilt the image and updated the first post link to the new one to avoid such surprises for others. Yeah 3.2 changes OPPs in a completely different way by using the dedicated i2c bus to PMIC for setting voltage instead of the "normal" one, so maybe it affects things. Could also be that...

Yes I've already fixed it in opkg upgrade, but the flashable image is unfortunately broken (you can run OS upgrade after reflasing to fix it though). I can rebuild the image if anyone needs it, was just waiting for any other issue reports.

It's the same way as described in the first post regardless of what you run.

The image I was testing was http://cdimage.debian.org/debian-cd/8.5.0/amd64/iso-cd/debian-8.5.0-amd64-netinst.iso with these commands: sudo rmmod g_cdc sudo modprobe g_file_storage file=/media/fat/tmp/debian-8.5.0-amd64-netinst.iso cdrom=1

I've just tried booting Debian image on my PC (with g_file_storage on pandora) and it booted fine. What were you trying to boot? Maybe it also depends on how good the PC BIOS is and what commands it sends.

Probably not. No changes. probably a coincidence. Hmm never seen that, and I've tested CC quite a lot. Does taking out the battery for 5 minutes or so help? Otherwise it sounds like some hardware issue to me.