Service manager crashing after launch

[Neo]

After boot when the init launches the service manager it exits quickly
and this happens a few more times which causes the system to reboot.
What could be the cause of this crash? I am running android with
2.6.28 kernel but don't have all the configs enabled as listed in the
goldfish_defconfig. Is there anything in particular it depends on? I
have left out yaffs, crypto and few other options which I feel weren't
important for android to come up.

Thanks,
Neo

[Neo]

Here are the boot messages-

<5>RAMDISK: Compressed image found at block 0
VFS: Mounted root (ext2 filesystem) readonly.
<6>Freeing init memory: 116K
<3>init: cannot open '/initlogo.rle'
Clock CLCDCLK: setting VCO reg params: S=1 R=99 V=98
<3>binder_mmap: 680 40008000-40028000 bad vm_flags failed -1
<6>warning: `rild' uses 32-bit capabilities (legacy support in use)
<3>binder_mmap: 687 40008000-40108000 bad vm_flags failed -1
<3>init: cannot find '/system/bin/playmp3', disabling 'bootsound'
<6>eth0: link up
<3>binder_mmap: 702 40008000-40108000 bad vm_flags failed -1
<3>init: untracked pid 686 exited
<3>init: untracked pid 687 exited
<6>request_suspend_state: wakeup (3->0) at 532299389005 (2009-06-17
04:46:07.398654002 UTC)
<3>binder_mmap: 704 40008000-40028000 bad vm_flags failed -1
<3>init: service servicemanager crashed
<3>binder_mmap: 706 40008000-40108000 bad vm_flags failed -1
<6>request_suspend_state: wakeup (0->0) at 534892029004 (2009-06-17
04:46:09.991274001 UTC)
<3>binder_mmap: 708 40008000-40028000 bad vm_flags failed -1
<3>init: service servicemanager crashed
<3>binder_mmap: 710 40008000-40108000 bad vm_flags failed -1
<6>request_suspend_state: wakeup (0->0) at 539711141004 (2009-06-17
04:46:14.810384001 UTC)
<3>binder_mmap: 712 40008000-40028000 bad vm_flags failed -1
<3>init: service servicemanager crashed
<3>binder_mmap: 715 40008000-40108000 bad vm_flags failed -1
<3>init: untracked pid 713 exited
<6>request_suspend_state: wakeup (0->0) at 546137671005 (2009-06-17
04:46:21.236914002 UTC)
<3>binder_mmap: 717 40008000-40028000 bad vm_flags failed -1
<3>init: service servicemanager crashed
<3>init: critical process 'servicemanager' exited 4 times in 4
minutes; rebooting into recovery mode
<0>Restarting system with command 'recovery'.

[mizmit1222]

Hi,

If your .config has CONFIG_ANDROID_PMEM = y, then try
comment it out and rebuild a kernel. As far as I know, pmem
driver works only on MSM cpu as OpenGL accelerator I/F.

Good luck,

[Neil Nguyen]

I had the same problem with Kernel 2.6.29.
And I do not think it is PMEM that causes it. I did not have the HIMEM reserved for pmem anyway. So in a way when some process open /dev/pmem, it will fail anyway. I move back to its original 2.6.27 (cupcake branch),
and everything works fine. I boot Android successfully.

That is what I learned but I could be wrong.
Regards,
Neil

--- On Thu, 6/18/09, mizmit1222 <mizmi...@gmail.com> wrote:

[Neo]

On Jun 19, 5:31 am, mizmit1222 <mizmit1...@gmail.com> wrote:
> Hi,
>
> If your .config has CONFIG_ANDROID_PMEM = y, then try
> comment it out and rebuild a kernel. As far as I know, pmem
> driver works only on MSM cpu as OpenGL accelerator I/F.
>
> Good luck,
>
> On 6月17日, 午後1:49, Neo <zingafri...@yahoo.com> wrote:

No, this also didn't help.

Also what do the following do in the init.rc file ?
*********************************************************
# mount mtd partitions
# Mount /system rw first to give the filesystem a chance to save a
checkpoint
mount yaffs2 mtd@system /system
mount yaffs2 mtd@system /system ro remount

# We chown/chmod /data again so because mount is run as root +
defaults
mount yaffs2 mtd@userdata /data nosuid nodev
chown system system /data
chmod 0771 /data

# Same reason as /data above
mount yaffs2 mtd@cache /cache nosuid nodev
chown system cache /cache
chmod 0770 /cache

# This may have been created by the recovery system with odd
permissions
chown system system /cache/recovery
chmod 0770 /cache/recovery
***********************************************************************
I have replaced the yaffs2 with ext2 because I have built the
filesystem with ext2.

[Neo]

> <3>binder_mmap: 717 40008000-40028000 bad vm_flags failed -1
> <3>init: service servicemanager crashed
> <3>init: critical process 'servicemanager' exited 4 times in 4
> minutes; rebooting into recovery mode
> <0>Restarting system with command 'recovery'.

Just got over this error: in binder.c I had to modify the
'FORBIDDEN_MMAP_FLAGS' from (VM_WRITE | VM_EXEC) to (VM_WRITE) only.