[beagleboard] 2013.05.08 image

It worked for me yesterday. I downloaded it and sued it as well. I will try a fresh download and see.

Gerald

I did finally get a download from it that was not corrupted. I wiped the sd card that I used successfully with the prior images but when I try and flash the BB with the new image when I hold down the user boot button it starts to come up, all the LED’s almost immediately turn on and after that nothing happens.

It works fine for me. Make sure you hold the Boot Button down until the first user LED comes on. You must apply power while holding the button down in order t get it to boot from the uSD.

Sounds like you are booting from eMMC and it is hanging because it cannot find the uENV.txt file on the uSD card used for flashing…

Gerald

just a few minutes ago, i "dd"ed the eMMC flasher 2013.05.08 image
to an SD card, powered off my BBB, inserted the SD card, pressed the
boot button, then plugged in 5V, waited for first blue LED, and let go
the boot button. as of right now, it's flashing madly away according
to the LEDs.

  you're holding down the correct "boot" button, right? the one next
to the SD card?

rday

It worked for me yesterday. I downloaded it and sued it as well. I will

I never took you for a lawyer Gerald!

David

That was a Freudian slip. I was thinking about the software!!!

Gerald

I wrote the image to the sd card again and this time around it did start flashing the eMMC. I tried several times before rewriting the image and I know I was holding the user boot buton prior to applying power. Interesting thing is that I had no success with the older images using a 16gb micro sd card. I went yesterday and got a 8 gig micro sd card and was able to get that to boot, but had issues with the HDMI then which I am hoping this build will fix.

Which site is the official builds for the BBB ? I was surprised yesterday when I got it to reflash ( after I had bricked it trying to use the opkg update before finding this forum ). Depending on what path you go down there are different images now all over the place. It’s a little confusing.

It seems that the quality and possibly sd card sizes are playing an issue is reflashing the boards. Yesterday I had to sprinkle some garlic and recite the lyrics to “Talula does the Hula in Hawaii” before it would flash the board.

http://circuitco.com/support/index.php?title=BeagleBoneBlack#Updating_the_eMMC_Software

There is a new image that fixes the update issue as well as HDMI and Flashing…

Gerald

Are you powering the board over USB during your flashing attempt?

Gerald

I agree there should be. I listed the issues fixed on the update page.

It would also be nice to have a way to tell which version of the image was actually flashed on a board. I asked for that several weeks ago and the crickets were so loud, I just left the room.

Maybe the SW folks will jump in and answer that question.

Gerald

That may work, assuming the hub violates the specification and lets’ more than 500mA go it on a single port. If the Power LED turns off, that is typical if the power is shutdown by the PC or hub.

Gerald

There is:

1) ID.txt on the vfat partition of the flasher
2) version info in /usr/bin/emmc.sh of the flasher
3) ID.txt on the vfat partition on the eMMC
4) /etc/dogtag in the root partition of the eMMC
5) /etc/angstrom-version in the root partition of the eMMC
6) /etc/angstrom-build-info in the root partition of the eMMC

And then there's the circumstantial bits like kernel and bootloader build dates.

Awesome! The SW people are still there!!! I will get this add into the Wiki so people can read it if they wander over that way.

Thank you sir!

Gerald

I am not sure if this right or wrong, but after I updated to the 5/3 build and likewise with the 5/8 build, when I do a dmesg command I am getting tons of repeating messages, that look like:
[ 66.168867] gen_ndis_query_resp: RNDIS_OID_GEN_RCV_NO_BUFFER
[ 66.172709] gadget: rndis reqa1.01 v0000 i0000 l4096
[ 66.172803] gadget: rndis req21.00 v0000 i0000 l36
[ 66.176707] gadget: rndis reqa1.01 v0000 i0000 l4096
[ 66.176946] gadget: rndis req21.00 v0000 i0000 l36
[ 66.180585] gadget: rndis reqa1.01 v0000 i0000 l4096
[ 66.180806] gadget: rndis req21.00 v0000 i0000 l36
[ 66.180862] gen_ndis_query_resp: RNDIS_OID_GEN_RCV_NO_BUFFER
[ 66.184705] gadget: rndis reqa1.01 v0000 i0000 l4096
[ 66.184923] gadget: rndis req21.00 v0000 i0000 l36
[ 66.188708] gadget: rndis reqa1.01 v0000 i0000 l4096
[ 66.188976] gadget: rndis req21.00 v0000 i0000 l36
[ 66.192709] gadget: rndis reqa1.01 v0000 i0000 l4096
[ 66.192952] gadget: rndis req21.00 v0000 i0000 l36
[ 66.192976] gen_ndis_query_resp: RNDIS_OID_GEN_RCV_NO_BUFFER

Is this normal, or is there something I should do?

Thanks
Kurt

This is one for the SW people.

Gerald

> I agree there should be. I listed the issues fixed on the update page.
>
> It would also be nice to have a way to tell which version of the
> image was actually flashed on a board. I asked for that several
> weeks ago and the crickets were so loud, I just left the room.

There is:

1) ID.txt on the vfat partition of the flasher
2) version info in /usr/bin/emmc.sh of the flasher

  however, that's in the *flasher* image, and may no longer be
available on a running system. i think both gerald and i were
specifically interested in gleaning version and build info solely from
what we could find on a running system.

3) ID.txt on the vfat partition on the eMMC
4) /etc/dogtag in the root partition of the eMMC

  useful info, but on my system, the above have exactly the same info:

  "Cloud9 GNOME Image 2013.05.08"

so if that's always the case, there's no point looking at ID.txt on
the FAT partition if you can get precisely the same info from
/etc/dogtag, which is fine.

5) /etc/angstrom-version in the root partition of the eMMC
6) /etc/angstrom-build-info in the root partition of the eMMC

  *now* we're talking -- that's the good stuff:

# cat /etc/angstrom-version
Angstrom v2012.12 (Core edition)
Built from branch: danny
Revision: abdca5ed77a803e9eb8fdacd4977956755276c3b
Target system: arm-angstrom-linux-gnueabi