m5-patches set

last time, i ported m5 and i use 2.6.22 kernel version. as we know, there are a lots differences between m3 and m5, especially in binder driver. to get android running i did modification with kernel source.

runtime complained that some function, or DEFINED value missing in 2.6.22, try to do little investigation in which parts should be adjusted and make a diff between 2.6.22 phytec kernel source and 2.6.23 m5 kernel for certain file, that solve my problem. now android m5 running well.

these are logs before the patch applied :

FromNodes+0x1f0/0x27c)

[] (binder_transaction_ConvertFromNodes+0x0/0x27c) from [] (

binder_thread_ReturnTransaction+0x58/0x50c)

[] (binder_thread_ReturnTransaction+0x0/0x50c) from [] (bind

er_thread_Read+0xb9c/0xfcc)

[] (binder_thread_Read+0x0/0xfcc) from [] (binder_thread_Con

trol+0x174/0x438)

[] (binder_thread_Control+0x0/0x438) from [] (binder_unlocke

d_ioctl+0x168/0x1b0)

[] (binder_unlocked_ioctl+0x0/0x1b0) from [] (do_ioctl+0x34/

0xd0)

 r7:bed11518 r6:c0186201 r5:c7efdae0 r4:c7efdae0

[] (do_ioctl+0x0/0xd0) from [] (vfs_ioctl+0x2c0/0x2e8)

 r7:c7efdae0 r6:bed11518 r5:0000000a r4:c7efdae0

[] (vfs_ioctl+0x0/0x2e8) from [] (sys_ioctl+0x40/0x64)

 r9:c2bda000 r8:c0026024 r6:c0186201 r5:bed11518 r4:0000000a

[] (sys_ioctl+0x0/0x64) from [] (ret_fast_syscall+0x0/0x2c)

 r7:00000036 r6:0013f7b0 r5:00000000 r4:c0186201
[] (dump_stack+0x0/0x14) from [] (binder_transaction_Convert
FromNodes+0x1f0/0x27c)
[] (binder_transaction_ConvertFromNodes+0x0/0x27c) from [] (
binder_thread_ReturnTransaction+0x58/0x50c)
[] (binder_thread_ReturnTransaction+0x0/0x50c) from [] (bind
er_thread_Read+0xb9c/0xfcc)
[] (binder_thread_Read+0x0/0xfcc) from [] (binder_thread_Con
trol+0x174/0x438)
[] (binder_thread_Control+0x0/0x438) from [] (binder_unlocke
d_ioctl+0x168/0x1b0)
[] (binder_unlocked_ioctl+0x0/0x1b0) from [] (do_ioctl+0x34/
0xd0)
 r7:befef518 r6:c0186201 r5:c7a3b400 r4:c7a3b400
[] (do_ioctl+0x0/0xd0) from [] (vfs_ioctl+0x2c0/0x2e8)
 r7:c7a3b400 r6:befef518 r5:0000000a r4:c7a3b400
[] (vfs_ioctl+0x0/0x2e8) from [] (sys_ioctl+0x40/0x64)
 r9:c24ee000 r8:c0026024 r6:c0186201 r5:befef518 r4:0000000a
[] (sys_ioctl+0x0/0x64) from [] (ret_fast_syscall+0x0/0x2c)
 r7:00000036 r6:0013f7b0 r5:00000000 r4:c0186201
binder_release(c76cfa40 c7a3b400) (pid 2965) pd c2e3b000
binder_release(c76cfa40 c7e6df00) (pid 2942) pd c3439000
binder_release(c76cfa40 c7201180) (pid 2929) pd c3b7b000
[1]   Killed                  /system/old/runtime
#
[] (binder_unlocked_ioctl+0x0/0x1b0) from [] (do_ioctl+0x34/
0xd0)
Backtrace:
[] (kmem_cache_create+0x0/0x450) from [] (init_binder+0x38
x2d0)
[] (init_binder+0x0/0x2d0) from [] (kernel_init+0xb4/0x26c
 r6:00000000 r5:00000000 r4:c001f000
[] (kernel_init+0x0/0x26c) from [] (do_exit+0x0/0x90c)
Code: e59f13ac e51b203c ebfee03f e3a03000 (e5833000)
Kernel panic - not syncing: Attempted to kill init!

unfortunately I lost my backup files which are the patches reside. But it seems no more sense now
to count on "dirty" hacking since Android is open sourced!

One Comment

  1. yonathana
    Posted November 21, 2008 at 3:36 am | Permalink

    Hi there, Been trying the same thing and wondering if any patches are available for me to try.


Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*