Testing 2.6.26 on a GTA01

Andy Green andy at openmoko.com
Sun Aug 3 11:01:55 CEST 2008


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Somebody in the thread at some point said:

| To see if I can base my cpufreq work on 2.6.26, I compiled a 2.6.26
| kernel and successfully booted it both on both qemu and a real GTA01Bv4.
|
| I used:
| - The "August snapshot" u-boot[1]
| - A kernel based on the stable-2.6 branch[2] with the following changes
| to defconfig-2.6.26[3] (which I used as the .config):
|   CONFIG_S3C_LOWLEVEL_UART_PORT=0
|   CONFIG_MTD_NAND_S3C2410_DEBUG is not set
|   CONFIG_DEBUG_S3C_UART=0
| - An openmoko-devel-image from today (compiled from the org.openmoko.dev
| branch)[4]
|
| I haven't tested the peripheral chips (GSM, BT, GPS); however, sound
| works (this defconfig has everything built-in instead of using modules,
| which is much better than having sound as a module).

Seeing your later posts too this is looking pretty good.  I guess GPS
shouldn't make trouble since it's the same code in neo1973-pm-gps.c.  SD
is broken from s3cmci somehow you later found.

| root at om-gta01:/sys/class/i2c-adapter/i2c-0/0-0008# cat battvolt
| 4037
|
| After some time (while typing this message in fact), battemp is now at 40.

I guess the paths in the /sys didn't change?

| I have attached the dmesg output (copied via ssh over usb ethernet)[5],
| which is probably more informative than anything else on this message
;-) .

dmesg looks normal AFAICT.  That bug in the timestamp for printk was
introduced in 2.6.25 I think and has never gone away since.

- -Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iEYEARECAAYFAkiVdAMACgkQOjLpvpq7dMrDYACbBuE2Q3v8gCwvg1YthIX/Fwv3
O+IAnAscnjbPBeaU1KISRJWbQMEAcO0a
=s0bE
-----END PGP SIGNATURE-----




More information about the openmoko-kernel mailing list