RobertCNelson / linux-dev

MIT License
151 stars 96 forks source link

am33x-v3.8 Issues #9

Open ghulands opened 11 years ago

ghulands commented 11 years ago

Hi, This is using the following built image 3.8.0-rc5-bone1 using the build_kernel.sh script

pwm interface doesn't show up in /dev/ or /sys/class/pwm

dmesg shows

[ 0.461436] platform 48300200.ehrpwm: alias tbclk already exists [ 0.461640] ehrpwm 48300200.ehrpwm: unable to select pin group [ 0.462654] platform 48302200.ehrpwm: alias tbclk already exists [ 0.462843] ehrpwm 48302200.ehrpwm: unable to select pin group [ 0.463655] platform 48304200.ehrpwm: alias tbclk already exists [ 0.463841] ehrpwm 48304200.ehrpwm: unable to select pin group

Full dmesg is here http://pastebin.com/r84aJmLs

A quick grep of the .config shows that the pwm code should be building into the kernel

CONFIG_PWM_TIEHRPWM=y CONFIG_PWM_TIPWMSS=y

Edit: Also the ain interfaces aren't showing up in /sys/devices/platform

Thanks, Greg

ghulands commented 11 years ago

Hi Robert, Does bone5 provide the pwm interface yet?

Thanks, Greg

JoseGoncalves commented 11 years ago

I've build a 3.8.4-bone9 kernel and found that most of the issues that I've detected in the am33x-v3.7 branch are now solved!

The only issues still remaining (that I've detected so far) are;

Best regards, José Gonçalves

inindev commented 11 years ago

Robert, I am running a bone A3 and tried the 3.8.6-bone12 kernel which fails with "unrecognized/unsupported machine ID (r1 = 0x00000e05)" -- here are the highlights if the install & boot.

http://rcn-ee.net/deb/quantal-armhf/v3.8.6-bone12/install-me.sh

Installing [3.8.6-bone12-dtbs.tar.gz] /tmp/deb/dtb/am335x-boneblack.dtb' ->/boot/uboot/dtbs/am335x-boneblack.dtb' /tmp/deb/dtb/am335x-bone.dtb' ->/boot/uboot/dtbs/am335x-bone.dtb' /tmp/deb/dtb/am335x-evm.dtb' ->/boot/uboot/dtbs/am335x-evm.dtb' /tmp/deb/dtb/am335x-evmsk.dtb' ->/boot/uboot/dtbs/am335x-evmsk.dtb' /tmp/deb/dtb/am335x-tester.dtb' ->/boot/uboot/dtbs/am335x-tester.dtb' /tmp/deb/dtb/omap2420-h4.dtb' ->/boot/uboot/dtbs/omap2420-h4.dtb' /tmp/deb/dtb/omap3-beagle.dtb' ->/boot/uboot/dtbs/omap3-beagle.dtb' /tmp/deb/dtb/omap3-beagle-xm.dtb' ->/boot/uboot/dtbs/omap3-beagle-xm.dtb' /tmp/deb/dtb/omap3-evm.dtb' ->/boot/uboot/dtbs/omap3-evm.dtb' /tmp/deb/dtb/omap3-tobi.dtb' ->/boot/uboot/dtbs/omap3-tobi.dtb' /tmp/deb/dtb/omap4-panda-a4.dtb' ->/boot/uboot/dtbs/omap4-panda-a4.dtb' /tmp/deb/dtb/omap4-panda.dtb' ->/boot/uboot/dtbs/omap4-panda.dtb' /tmp/deb/dtb/omap4-panda-es.dtb' ->/boot/uboot/dtbs/omap4-panda-es.dtb' /tmp/deb/dtb/omap4-sdp.dtb' ->/boot/uboot/dtbs/omap4-sdp.dtb' /tmp/deb/dtb/omap4-var-som.dtb' ->/boot/uboot/dtbs/omap4-var-som.dtb' /tmp/deb/dtb/omap5-evm.dtb' ->/boot/uboot/dtbs/omap5-evm.dtb'


Image Name: 3.8.6-bone12 Created: Sun Apr 7 12:04:25 2013 Image Type: ARM Linux Kernel Image (uncompressed) Data Size: 3098840 Bytes = 3026.21 kB = 2.96 MB Load Address: 80008000

Entry Point: 80008000

Image Name: initramfs Created: Sun Apr 7 12:04:27 2013 Image Type: ARM Linux RAMDisk Image (uncompressed) Data Size: 2598475 Bytes = 2537.57 kB = 2.48 MB Load Address: 00000000

Entry Point: 00000000

<boot>

Starting kernel ...

Uncompressing Linux... done, booting the kernel.

Error: unrecognized/unsupported machine ID (r1 = 0x00000e05).

Available machine support:

ID (hex) NAME ffffffff Generic OMAP4 (Flattened Device Tree) ffffffff Generic AM33XX (Flattened Device Tree) ffffffff Generic OMAP3-GP (Flattened Device Tree) ffffffff Generic OMAP3 (Flattened Device Tree) 0000060a OMAP3 Beagle Board 00000a9d IGEP OMAP3 module 00000928 IGEP v2 board 00000ae7 OMAP4 Panda board

Please check your kernel config and/or bootloader.

inindev commented 11 years ago

I just tried the 3.8.4-bone9 kernel José reported working with am33x above. Could the A3 am33x bone be reporting 0x00000e05 while A4, A5, and A6 be reporting 0xffffffff?

RobertCNelson commented 11 years ago

Hi John, this happens when you switch from a v3.2.x based "board" file kernel to the v3.8.x "device tree" kernel.. You need to make a few changes to the uEnv.txt boot script:

Edit uEnv.txt in the boot partition...

Add: initrd_high=0xffffffff fdt_high=0xffffffff

Change... From: xyz_mmcboot=run xyz_load_image; run xyz_load_initrd; echo Booting from mmc ... To: xyz_mmcboot=run xyz_load_image; run xyz_load_initrd; run xyz_load_dtb; echo Booting from mmc ...

Change... From: loaduimage=run xyz_mmcboot; run device_args; bootz 0x80300000 0x81600000:${initrd_size} To: loaduimage=run xyz_mmcboot; run device_args; bootz 0x80300000 0x81600000:${initrd_size} 0x815f0000

Going forward, the "uEnv.txt" will show hints on what needs to be changed: https://github.com/RobertCNelson/omap-image-builder/commit/975f6cecbe21c826ceb7dc5edf277c48828fe68a

JoseGoncalves commented 11 years ago

Hi Robert. When I was building a 3.7 and 3.8 kernel I've discovered the need for the two last changes on uEnv.txt in order to boot a kernel with a device tree. Nevertheless I did not add the 'initrd_high' and 'fdt_high' variables. Missing setting these variables could be the cause for the issues that I reported earlier (I have no BeagleBone near to test it now)?

RobertCNelson commented 11 years ago

Hi Jose,

While re-browsing the setup_sdcard.sh source, i noticed i was setting those two variables by default in the dtbs case.. They might be already set in the bootloader, thus adding them to uEnv.txt does nothing.. So to be safe i added them here..

Regards,

inindev commented 11 years ago

Thanks Robert -- I guess it would have been better to post over on www.eewiki.net where others would have been more likely to see it. I also see you detail these instructions over there as well... I can't tell you how nice it is to run 3.8 on the beaglebone!

root@bone:/# uname -a Linux bone 3.8.6-bone12 #1 SMP Sat Apr 6 06:36:28 UTC 2013 armv7l armv7l armv7l GNU/Linux

John Clark

ghulands commented 11 years ago

Hi Robert, I have just upgraded to 3.8.13-bone24 on both my beaglebone and beaglebone black and I see the pwm interface is still not showing up. If you can give me some pointers of where to look I can try to take a stab at this to get it working again.

Thanks, Greg