On May 8, 7:43 pm, "Howard M. Harte" <hhar...@gmail.com> wrote:
> I also saw this problem. I'm not using the ADP1, so my hardware is
> different. Anyway, modifying uevent.c in system/core/vold/vold.c to
> use PHYSDEVICEPATH seems to work. Here is the patch, it is obviously
> a hack, but it does seem to work.
>
> --- a/vold/uevent.c
> +++ b/vold/uevent.c
> @@ -309,6 +309,7 @@ static int handle_block_event(struct uevent
> *event)
>
> truncate_sysfs_path(event->path, n, mediapath, sizeof
> (mediapath));
>
> + strncpy(mediapath, get_uevent_param(event, "PHYSDEVPATH"), 255);
> if (!(media = media_lookup_by_path(mediapath, false))) {
> #if DEBUG_UEVENT
> LOG_VOL("No backend media found @ device path '%s'",
> mediapath);
>
> This is the vold.conf I'm using:
> ## vold configuration file for the emulator/SDK
>
> volume_sdcard {
> ## This is the direct uevent device path to the SD slot on the
> device
> media_path /class/mmc_host/mmc0/
> media_type mmc
> mount_point /sdcard
> ums_path /devices/platform/usb_mass_storage/lun0
>
> }
>
> -Howard
I've tried this, plus a few other variations, but still no luck. Was
there more to your hack? I still have issues with vold looking for...
/sys/class/mmc_host/mmc0/mmc0:b368/block
but target fs only has...
/sys/class/mmc_host/mmc0/mmc0:b368/block:mmcblk0
so mmc_bootstrap_mmcblk_partition() ends up failing because it doesn't
find a "block" dir.
Any ideas?
twebb
--~--~---------~--~----~------------~-------~--~----~
unsubscribe: android-porting+unsubscr...@googlegroups.com
website: http://groups.google.com/group/android-porting
-~----------~----~----~----~------~----~------~--~---
