Related
I'm trying to make an auto install script within the initramfs. I've got everything working, including automatic fdisk, but I cannot get mkfs.ext4 to work on the Android kernel, and there is no version of Busybox with the ext4 applet. Does anyone have a mkfs.ext4 that runs successfully on Android? When I run it from a terminal the output is
Code:
mkfs.ext4: 1: Syntax error: word unexpected (expecting ")")
Here's my code if you want to see what I'm doing.
Code:
mount /dev/mmcblk1p1 /tmp/mnt
if [ -f /tmp/mnt/mysticfw.tar.gz ]; then
$FDISK /dev/mmcblk0 < /home/fdisk.input
$MKFS_EXT4 -O ^huge_file /dev/mmcblk0p2
mount /dev/mmcblk1p2 /tmp/tmproot
mkdir /tmp/tmproot/itworks
tar -xzvf /tmp/mnt/mysticfw.tar.gz -C /tmp/tmproot/
sync
#rm /tmp/mnt/mysticfw.tar.gz
umount /tmp/tmproot
fi
sync
umount /tmp/mnt
It automatically partitions the stock Android block and installs a tar.gz from the internal storage to simplify my unsquashed 2.1.4 filesystem install, and the only part I can't get working is mkfs.ext4.
Found one! It's 3 MBs though, so if anyone has a smaller one, it would be very nice.
msticninja said:
I'm trying to make an auto install script within the initramfs. I've got everything working, including automatic fdisk, but I cannot get mkfs.ext4 to work on the Android kernel, and there is no version of Busybox with the ext4 applet. Does anyone have a mkfs.ext4 that runs successfully on Android? When I run it from a terminal the output is
Code:
mkfs.ext4: 1: Syntax error: word unexpected (expecting ")")
Click to expand...
Click to collapse
Why do you don't use the mke2fs from Uruk for example (the easiest way) asking $auron if it's ok for him. His size is only 49KB.
Find it like the following on Uruk installation:
Code:
[COLOR="DarkRed"]
# mkfs.ext4 -V
mke2fs 1.41.3 (12-Oct-2008)
Using EXT2FS Library version 1.41.3
# find / -name mke2fs | xargs ls -l
-rwxr-xr-x 1 root root 30584 Dec 15 03:46 /system/bin/mke2fs
[B]-rwxr-xr-x 5 root root 49248 4 Jan 15 13:14 /usr/local/sbin/mke2fs[/B]
#/usr/local/sbin/mke2fs -V
mke2fs 1.41.3 (12-Oct-2008)
Using EXT2FS Library version 1.41.3
[/COLOR]
Here's my code if you want to see what I'm doing.
Code:
mount /dev/mmcblk1p1 /tmp/mnt
if [ -f /tmp/mnt/mysticfw.tar.gz ]; then
$FDISK /dev/mmcblk0 < /home/fdisk.input
$MKFS_EXT4 -O ^huge_file /dev/mmcblk0p2
mount /dev/mmcblk1p2 /tmp/tmproot
mkdir /tmp/tmproot/itworks
tar -xzvf /tmp/mnt/mysticfw.tar.gz -C /tmp/tmproot/
sync
#rm /tmp/mnt/mysticfw.tar.gz
umount /tmp/tmproot
fi
sync
umount /tmp/mnt
It automatically partitions the stock Android block and installs a tar.gz from the internal storage to simplify my unsquashed 2.1.4 filesystem install, and the only part I can't get working is mkfs.ext4.
Click to expand...
Click to collapse
and don't forget to add "-l" on FDISK command and change the device mmcblk0 with mmcblk1 on the lines:
Code:
$FDISK /dev/mmcblk0 < /home/fdisk.input
$MKFS_EXT4 -O ^huge_file /dev/mmcblk0p2
Cheers,
shklifo said:
Why do you don't use the mke2fs from Uruk for example (the easiest way) asking $auron if it's ok for him. His size is only 49KB.
Find it like the following on Uruk installation:
Code:
[COLOR="DarkRed"]
# mkfs.ext4 -V
mke2fs 1.41.3 (12-Oct-2008)
Using EXT2FS Library version 1.41.3
# find / -name mke2fs | xargs ls -l
-rwxr-xr-x 1 root root 30584 Dec 15 03:46 /system/bin/mke2fs
[B]-rwxr-xr-x 5 root root 49248 4 Jan 15 13:14 /usr/local/sbin/mke2fs[/B]
#/usr/local/sbin/mke2fs -V
mke2fs 1.41.3 (12-Oct-2008)
Using EXT2FS Library version 1.41.3
[/COLOR]
and don't forget to add "-l" on FDISK command and change the device mmcblk0 with mmcblk1 on the lines:
Code:
$FDISK /dev/mmcblk0 < /home/fdisk.input
$MKFS_EXT4 -O ^huge_file /dev/mmcblk0p2
Cheers,
Click to expand...
Click to collapse
Why didn't I think of that? Thanks.
But regarding mmcblk0/1, I'm replacing the stock Android, so the fdisk.input file contains the commands to delete mmcblk0p2 and p3, and make a new partition in the unused space. I hate using space on my Internal Storage, so I'm using Archos' space.
msticninja said:
Why didn't I think of that? Thanks.
But regarding mmcblk0/1, I'm replacing the stock Android, so the fdisk.input file contains the commands to delete mmcblk0p2 and p3, and make a new partition in the unused space. I hate using space on my Internal Storage, so I'm using Archos' space.
Click to expand...
Click to collapse
If you are using the mmcblk0p2 as rootfs as you say (and you are expanded tar archive on mmcblk1p2), than you have to change the line:
mount /dev/mmcblk1p2 /tmp/tmproot
Click to expand...
Click to collapse
with
mount /dev/mmcblk0p2 /tmp/tmproot
Click to expand...
Click to collapse
shklifo said:
If you are using the mmcblk0p2 as rootfs as you say (and you are expanded tar archive on mmcblk1p2), than you have to change the line:
with
Click to expand...
Click to collapse
I know, that's my current data partition, I'll change it once I'm done testing. The tar file just has a test file in it, so when I boot back into block1, I can see if the IF statement was executed by seeing if it was extracted to block1. I'll also have to change etc/mountpoints once testing is actually finished.
One more question since you're so quick. I think I have everything working, except it needs a reboot in between the fdisk and mke2fs commands to reload the partition table. I'm trying to use partprobe instead of rebooting, but it hasn't been cross compiled to work on Android, AFAIK. Have you seen a way to reload the MBR without rebooting?
msticninja said:
I know, that's my current data partition, I'll change it once I'm done testing. The tar file just has a test file in it, so when I boot back into block1, I can see if the IF statement was executed by seeing if it was extracted to block1. I'll also have to change etc/mountpoints once testing is actually finished.
One more question since you're so quick. I think I have everything working, except it needs a reboot in between the fdisk and mke2fs commands to reload the partition table. I'm trying to use partprobe instead of rebooting, but it hasn't been cross compiled to work on Android, AFAIK. Have you seen a way to reload the MBR without rebooting?
Click to expand...
Click to collapse
I'v been looking at the recovery_lib.sh in the recovery boot image and can't find anything special to re-read the partition table. And yes they also use fdisk to repartition. So I suspect the driver for the block device does not cache the MBR and you can just mke2fs after the partition table is created.
I can't remember from what firmware that recovery boot image was but I think it's from the 2.1.04 and they do some repartitioning there for the swap space.
I'll check it again and get back to you.
wdl1908 said:
I'v been looking at the recovery_lib.sh in the recovery boot image and can't find anything special to re-read the partition table. And yes they also use fdisk to repartition. So I suspect the driver for the block device does not cache the MBR and you can just mke2fs after the partition table is created.
I can't remember from what firmware that recovery boot image was but I think it's from the 2.1.04 and they do some repartitioning there for the swap space.
I'll check it again and get back to you.
Click to expand...
Click to collapse
I just checked the recovery image from 2.1.04 and after the fdisk commands there is nothing to re-read the MBR the next commands executed are mount commands to check if the fs is present I suggest you look at the /etc/scripts/recovery_lib.sh yourself it could give some clues on how to do things.
wdl1908 said:
I just checked the recovery image from 2.1.04 and after the fdisk commands there is nothing to re-read the MBR the next commands executed are mount commands to check if the fs is present I suggest you look at the /etc/scripts/recovery_lib.sh yourself it could give some clues on how to do things.
Click to expand...
Click to collapse
For me too, it have nothing to do with a reboot to load partitions table and access partition to format them with the choised filesystem.
You can delete any partition on linux (except rootfs one ), recreate them and directly format them as you like, reboot isn't necesary.
Thanks for all the replies, very helpful, but I'm stuck. Fdisk seems to use ioctl to reload the partition table, so you don't need a reboot if everything on the device is unmounted before writing the partition table, but I'm having very strange issues with mke2fs now. I've had the whole thing work twice now, but when I flash back to stock, then retry the script, it usually does everything except the formatting. Here's the code:
Code:
mount /dev/mmcblk1p1 /tmp/mnt
if [ -f /tmp/mnt/mysticfw.tar.gz ]; then
umount /dev/mmcblk0p1
umount /dev/mmcblk0p2
umount /dev/mmcblk0p3
umount /dev/mmcblk0p4
fdisk /dev/mmcblk0 < /home/fdisk.input
mv /tmp/mnt/mysticfw.tar.gz /tmp/mnt/mysticf.tar.gz
sync
umount /tmp/mnt
log_and_reboot
fi
if [ -f /tmp/mnt/mysticf.tar.gz ]; then
rm /etc/mtab
touch /etc/mtab
mke2fs -T ext4 -O ^huge_file /dev/mmcblk0p2
mount /dev/mmcblk0p2 /tmp/tmproot
tar -xzf /tmp/mnt/mysticf.tar.gz -C /tmp/tmproot/
sync
mv /tmp/mnt/mysticf.tar.gz /tmp/mnt/mysticdone.tar.gz
umount /tmp/tmproot
rm /etc/mtab
ln -s /proc/mounts /etc/mtab
fi
sync
umount /tmp/mnt
I have it reboot after the fdisk just in case, and the fdisk works perfectly, so the second IF/THEN is the issue. I had to retouch the mtab just to make sure it's empty, as mke2fs fails if mtab doesn't exist(at least in terminal), then I relink it to /proc/mounts as they do in the stock firmware. Here's my mke2fs.conf:
Code:
[defaults]
base_features = sparse_super,filetype,resize_inode,dir_index,ext_attr
blocksize = 4096
inode_size = 256
inode_ratio = 16384
[fs_types]
ext3 = {
features = has_journal
}
ext4 = {
features = has_journal,extent,huge_file,flex_bg,uninit_bg,dir_nlink,extra_isize
inode_size = 256
}
ext4dev = {
features = has_journal,extent,huge_file,flex_bg,uninit_bg,dir_nlink,extra_isize
inode_size = 256
options = test_fs=1
}
small = {
blocksize = 1024
inode_size = 128
inode_ratio = 4096
}
floppy = {
blocksize = 1024
inode_size = 128
inode_ratio = 8192
}
news = {
inode_ratio = 4096
}
largefile = {
inode_ratio = 1048576
blocksize = -1
}
largefile4 = {
inode_ratio = 4194304
blocksize = -1
}
hurd = {
blocksize = 4096
inode_size = 128
}
Is there a way to echo the output from my script to a file like dontpanic so I can see what the error is?
msticninja said:
Is there a way to echo the output from my script to a file like dontpanic so I can see what the error is?
Click to expand...
Click to collapse
Simple append (">>") and "echo" doesn't work to a redirected logfile?
I've only learned what I've needed to learn over the years, usually with endless google searches and reading through man pages, so I've never tried to log outputs, because I could always see the output in a terminal or onscreen during boot. Android is the first time I haven't been able to actually see the boot process.
Once again, thanks for your help. I probably could've googled this, but I kind of asked as an afterthought. I didn't even think about redirecting. So if the mke2fs line is the one I want to log, I just add "2&>> /tmp/mnt/logfile" to the end of it, right?
Also, thanks for your original thread about booting from Internal Storage, I never got around to developing on Android until you posted that, and I realized just how similar Android is to L/unix(e.g. exactly the same).
msticninja said:
I didn't even think about redirecting. So if the mke2fs line is the one I want to log, I just add "2&>> /tmp/mnt/logfile" to the end of it, right?
Click to expand...
Click to collapse
ehm, no.
In your case must be like the following:
Code:
"your command" >> /tmp/mnt/logfile 2>&1
That means redirect all messages from STDERR (2 - standard error) to STDOUT (1 - standard output) and all messaged collected on STDOUT to the redirected log file /tmp/mnt/logfile, or more comprensible redirect all mesage including error ones to the log file.
I'm working in Unix environments and use them often
And a good practice in developing is to use "echo" to the same log file, so you know the exact place the script is running, like:
Code:
echo "I'm just before the formatting step of the ..." >> /tmp/mnt/logfile
shklifo said:
ehm, no.
In your case must be like the following:
Code:
"your command" >> /tmp/mnt/logfile 2>&1
That means redirect all messages from STDERR (2 - standard error) to STDOUT (1 - standard output) and all messaged collected on STDOUT to the redirected log file /tmp/mnt/logfile, or more comprensible redirect all mesage including error ones to the log file.
I'm working in Unix environments and use them often
And a good practice in developing is to use "echo" to the same log file, so you know the exact place the script is running, like:
Code:
echo "I'm just before the formatting step of the ..." >> /tmp/mnt/logfile
Click to expand...
Click to collapse
Strange. The google searching lead me to believe the "2>&1" was before the location, and replaced ">>". Once again, thanks for the help.
any chance of you getting this uploaded? iterested in this since I'm not that good with linux
TjaXanK said:
any chance of you getting this uploaded? iterested in this since I'm not that good with linux
Click to expand...
Click to collapse
I could probably finish it, but I'm waiting for a new version of Uruk first, as his install script already gets rid of the linux steps, it just doesn't give you the choice to install to the internal memory yet. Once he does that, I'll add my changes to make it install to the Archos partition. The latest Uruk is also a bit too big for the archos partition. I had to cut my data partition down to ~150 megs.
msticninja said:
I could probably finish it, but I'm waiting for a new version of Uruk first, as his install script already gets rid of the linux steps, it just doesn't give you the choice to install to the internal memory yet. Once he does that, I'll add my changes to make it install to the Archos partition. The latest Uruk is also a bit too big for the archos partition. I had to cut my data partition down to ~150 megs.
Click to expand...
Click to collapse
ok, I'm currently running urk with the new install system and it's brillant but it would be perfect if we could run it without cutting down on our storage space
So after a failed attempt to upgrade from CyanogenMod 10.1.3 to 10.2, I was unable to access /data or /sdcard because both systems were encrypted. I ended up having to factory reset my phone because it refused to co-operate or let me access my files. However, before I did that, I was able to run
Code:
adb shell "dd if=/dev/block/mmcblk0p2" > data.img
and
Code:
adb shell "dd if=/dev/block/mmcblk0p3" > sdcard.img
, which appears to have copied the raw partition images from the phone (at least, they're the right sizes).
According to my reading, Android (and, I'm inferring, CyanogenMod) encrypts filesystems using dm-crypt, with a AES-CBC ESSIV:SHA256 cipher, with the key being derived from the password using PBKDF2. Knowing the precious little I do about encrypted file systems, my guess is that if I configure the image in cryptsetup to create a drive mapping, I can mount the mapped drive and recover the data from the images.
According to /fstab.herring on my ahem, fresh, install of Android, the /data partition is in ext4 format whereas the /sdcard partition is vFAT. So, once I've gotten through the encryption on the partition images, they should mount normally, right?
I know that dm-crypt accepts plain, LUKS, LoopAES and TrueCrypt device formats. I'm inferring from the PBKDF2 extension that Android goes the LUKS route for encrypting. Is this conclusion correct?
Could someone explain whether it's possible to decrypt a dumped android image? I'm really hoping that the cypher information is stored on the file system and not on some key file that I nuked in the factory reset. If it can, in theory, be decrypted, am I using the right tools to approach the matter? If so, I'll continue fiddling with cryptsetup and mount, but no sense in wasting time if it's an impossible task.
Never did get a response to this question, so I'll try it again, but start with a simpler question:
If someone dds an Android (specifically Cyanogenmod 10.x) partition to an img file, is there any way to read that image from, say a Linux laptop? I dumped the contents of the /system partition using
Code:
adb shell "dd if=/dev/block/mmcblk0p1" > system.img
I expected system.img to be a normal ext4 partition. However, attempting to loopback mount it with
Code:
sudo mount -t ext4 -o loop,ro system.img ~/android/system
Gave me errors about corrupt group descriptors, bad magic numbers and other maladies indicative of a thoroughly corrupted file system. I'm assuming that:
/data has the same ext4 partition structure as /system; and
The process to mount /storage would be no different to mounting /system with the exception that the former uses vFAT as its file system
However, as my Android is currently working normally (well, as well as one can hope for Android to work), I know I don't have a corrupted file system.
So what's going on? Does Android use a special version of ext4 that other Linuxes don't recognise? Am I not dd-ing correctly? Is there a block-size issue I ignored to my peril?
Borden Rhodes said:
So after a failed attempt to upgrade from CyanogenMod 10.1.3 to 10.2, I was unable to access /data or /sdcard because both systems were encrypted. I ended up having to factory reset my phone because it refused to co-operate or let me access my files. However, before I did that, I was able to run
Code:
adb shell "dd if=/dev/block/mmcblk0p2" > data.img
and
Code:
adb shell "dd if=/dev/block/mmcblk0p3" > sdcard.img
, which appears to have copied the raw partition images from the phone (at least, they're the right sizes).
According to my reading, Android (and, I'm inferring, CyanogenMod) encrypts filesystems using dm-crypt, with a AES-CBC ESSIV:SHA256 cipher, with the key being derived from the password using PBKDF2. Knowing the precious little I do about encrypted file systems, my guess is that if I configure the image in cryptsetup to create a drive mapping, I can mount the mapped drive and recover the data from the images.
According to /fstab.herring on my ahem, fresh, install of Android, the /data partition is in ext4 format whereas the /sdcard partition is vFAT. So, once I've gotten through the encryption on the partition images, they should mount normally, right?
I know that dm-crypt accepts plain, LUKS, LoopAES and TrueCrypt device formats. I'm inferring from the PBKDF2 extension that Android goes the LUKS route for encrypting. Is this conclusion correct?
Could someone explain whether it's possible to decrypt a dumped android image? I'm really hoping that the cypher information is stored on the file system and not on some key file that I nuked in the factory reset. If it can, in theory, be decrypted, am I using the right tools to approach the matter? If so, I'll continue fiddling with cryptsetup and mount, but no sense in wasting time if it's an impossible task.
Click to expand...
Click to collapse
Can you give the result of the "file sdcard.img" and "file data.img" commands?
You are quite right. With regular LUKS container/partition, you would do (being root) the following. With the following commands, you can create a container named "safe", setup it, then format its content in ext3 and mount the partition:
Code:
dd if=/dev/zero bs=1M count=50 of=safe
losetup /dev/loop0 safe
cryptsetup luksFormat -c aes -h sha256 /dev/loop0
cryptsetup luksOpen /dev/loop0 safe
mkfs.ext3 /dev/mapper/safe
(losetup /dev/loop0 safe)
(cryptsetup luksOpen /dev/loop0 safe)
mkdir mnt
mount -t ext3 /dev/mapper/safe mnt
//HERE: do whatever you want in your mounted encrypted filesystem
umount mnt
cryptsetup luksClose safe
losetup -d /dev/loop0
For details, you can go there: http://blog.theglu.org/index.php/20...-couteau-suisse-du-chiffrement-de-partitions/
Sorry, the article is in French but you can translate it if you need to.
Here, using "hexdump", you can see the "safe" file has a LUKS magic at the beginning. And doing a "file safe" command, you can check it detects it as a "LUKS encrypted file".
If doing "file" on your .img files does not give you the same result, you may not be able to directly use the "cryptsetup" command and need to adapt it.
Finally: usually in Android the header containing the key is stored on another partition so you may have lost it when wiping your phone, sorry.
---------- Post added at 02:44 PM ---------- Previous post was at 02:41 PM ----------
Borden Rhodes said:
Never did get a response to this question, so I'll try it again, but start with a simpler question:
If someone dds an Android (specifically Cyanogenmod 10.x) partition to an img file, is there any way to read that image from, say a Linux laptop? I dumped the contents of the /system partition using
Code:
adb shell "dd if=/dev/block/mmcblk0p1" > system.img
I expected system.img to be a normal ext4 partition. However, attempting to loopback mount it with
Code:
sudo mount -t ext4 -o loop,ro system.img ~/android/system
Gave me errors about corrupt group descriptors, bad magic numbers and other maladies indicative of a thoroughly corrupted file system. I'm assuming that:
/data has the same ext4 partition structure as /system; and
The process to mount /storage would be no different to mounting /system with the exception that the former uses vFAT as its file system
However, as my Android is currently working normally (well, as well as one can hope for Android to work), I know I don't have a corrupted file system.
So what's going on? Does Android use a special version of ext4 that other Linuxes don't recognise? Am I not dd-ing correctly? Is there a block-size issue I ignored to my peril?
Click to expand...
Click to collapse
Can you give the result of the "file system.img" command?
Thanks, saidlike, for your reply:
saidelike said:
Can you give the result of the "file sdcard.img"...
Click to expand...
Click to collapse
sdcardPartitionDump.img: data
saidelike said:
... and "file data.img" commands?
Click to expand...
Click to collapse
data.img: data
saidelike said:
Can you give the result of the "file system.img" command?
Click to expand...
Click to collapse
system.img: Linux rev 1.0 ext4 filesystem data, UUID=57f8f4bc-abf4-655f-bf67-946fc0f9f25b (needs journal recovery) (extents) (large files)
Again, attempting to run
Code:
mount -t ext4 -o loop systemimg mountpoint/
yields
mount: wrong fs type, bad option, bad superblock on /dev/loop0,
missing codepage or helper program, or other error
In some cases useful info is found in syslog - try
dmesg | tail or so
Click to expand...
Click to collapse
Ignoring the results of data.img and sdcard.img for the time being, the fresh dump of the system partition shows that it's an EXT4 filesystem, but that it's heavily corrupted. fsck.ext4 on that partition basically asks me to fix every single inode, so it's not a simple unclean journal issue. Therefore, is it fair to conclude that CyanogenMod (and maybe AOSP too) have modified the ext4 partiiton type?
@Borden Rhodes
Maybe, my reply is too late, but you could try to make the same experiment with backup of your current data.
If you get the same results as with the old pre-wipe backup, then you still have a hope.
Hy!
I have a mi2s and this phone is come to separated partitions in its internal drive. It has separated data and sdcard partition. My sdcard partition not mounted for some reason.
I want to keep this partition system, I just want to either mount the sdcard partition, or resize them without loseing data. (I can delete the sdcard partition but I want the data partition untouched, I had a long fight till this rom started to work with google play store, and I dont really want to remach it after all my apps are installed... Fun thing that after the first boot both partitions were mounted, after my first reboot only the data.)
I tried:
adb mount - adb sees it Android not sees it
write it to the fstab.qalcom - its on the / if I reboot the phone its loaded from somewhere again (I know its a ramdisk), my modifications are not permanent on there
I have basic linux knowlage and I started to dig into it, but I cant google out a general solution.
My questions:
How can I mount a fs like the usb otg from adb/android shell?
Can I edit the fstab file in its permanent store on an installed rooted device? And if I can where?
If I place new lines to the fstab on rootfs how can I tell the system to "reload" it?
Can I extend an ext4 partition from adb without loseing its data? *
* I have the required tools like parted from xiaomi forum, I cant post the link but you can google it with "Mi2S extending size of storage partition stillka".
Any help appreciated, and sorry for my english I'm not native.
So the basics:
If you can mount it from adb its a half win!
Try search the correct block partition and mount it with -t, add the correct file system and don't try auto it.
After you can mount it, you need to start an sdcard process its in /system/bin/sdcard. I had to see the custom rom implementation for that, in cm u need to param it "sdcard from to 1023 1023", but in samsung devices the to is hardcoded, and you nedd to do some sed magic.
After that your android programs will see it as a valid sdcard partition.
The harder way:
Wrap it to a startup script.
Add this script somewhere to run at bootup.
I'm still working on it, but I'm closer and closer. After I have the final solution I will write here once more.
I get so much help from there:
http://forum.xda-developers.com/showthread.php?t=2467048
If somebody want to do this:
After few hours of trying to mount the filessystem in boottime (in CM 12.1 its a hard work), i gave up, and went to a repartitioning way.
BE CAREFUL YOU CAN BRICK YOUR DEVICE IF YOU HAVE NO IDEA WHAT IM TALKING ABOUT!
I merged 2 tutorials:
reboot phone into CWM, connect phone to PC
connect to phone over adb and check if you are root
mount system
umount cache
umount data
copy content of partition_tools.zip into /system/bin and add executable attributes if necessary
Run parted on your device: parted /dev/sdX
Change display unit to sectors: unit s
Print current partition table and note the start sector for your partition: p
Delete your partition (won't delete the data or filesystem): rm <number>
Delete your partition (the second one we will delete data from there): rm <number>
Recreate the partition with the starting sector from above: mkpart primary <start> <end>
Recreate partition 27 (the last) mkpartfs primary ext2 3070 15758
name 26 userdata #we have to set back partition labels
name 27 storage
Exit parted: quit
Check the filesystem of 26: sudo e2fsck -f /dev/sdXX
Resize filesystem 26: sudo resize2fs /dev/sdXX
restore partition 27 with:
tune2fs -j /dev/block/mmcblk0p27
e2fsck -fDp /dev/block/mmcblk0p27
tune2fs -O extents,uninit_bg,dir_index /dev/block/mmcblk0p27
e2fsck -fDp /dev/block/mmcblk0p27
Of course in parted print you can see your original partition layout and this case it is possible that you have other partition numbers (my 26 partition is labeld by userdata and 27 with storage, and I gave more space to userdata from storage without loseing any data from userdata).
You can download the partition_tools.zip from the original miui forum, try to search to mi2s extending size of storage partition. (yes it will work with other devices too)
Hi, I'd like to ask a general question about fstab and init.rc. I'd like to select an alternative fstab from the boot partition, between two different fstab files, one of these with emmc partitions and the other with partitions on external sdcard, in order to boot either emmc or sdcard, alternatively.
I've read the init documentation at: https://android.googlesource.com/platform/system/core/+/master/init/readme.txt about init.rc configuration file.
Reading my init.rc file a doubt has been emerging: if the init executable looks at fstab (via mount_all) by default, why does init.rc contain mount commands related mounting /system, /data, and /cache partitions, if they are already defined in fstab file? Not enough the fstab existence in order to mount partitions, or do I need always to operate mounts in init.rc?
Any ideas?
cristian_c said:
Hi, I'd like to ask a general question about fstab and init.rc. I'd like to select an alternative fstab from the boot partition, between two different fstab files, one of these with emmc partitions and the other with partitions on external sdcard, in order to boot either emmc or sdcard, alternatively.
I've read the init documentation at: https://android.googlesource.com/platform/system/core/+/master/init/readme.txt about init.rc configuration file.
Reading my init.rc file a doubt has been emerging: if the init executable looks at fstab (via mount_all) by default, why does init.rc contain mount commands related mounting /system, /data, and /cache partitions, if they are already defined in fstab file? Not enough the fstab existence in order to mount partitions, or do I need always to operate mounts in init.rc?
Any ideas?
Click to expand...
Click to collapse
The fstab file is just a reference..
The init.rc is actually mounting them
The fstab file is just a reference..
The init.rc is actually mounting them
Click to expand...
Click to collapse
Thanks for the help! I'll try to work with this stuff
Romeotamizh said:
The fstab file is just a reference..
The init.rc is actually mounting them
Click to expand...
Click to collapse
I have a problem related to this and i can't figure it out. Basically my rom is stuck in a plash screen and i have narrowed it down to this:
Code:
<11>[ 3.142415] fs_mgr: Cannot open file /fstab.qcom
<11>[ 3.142425] init: unable to read fstab /fstab.qcom: No such file or directory
so it looks like init.rc is sending fs_mgr to look for fstab in the root directory /,
but it is actually in /etc, i can't understand why is this happening, I think it may have to do with other previous errors that occur a few lines before
Code:
<11>[ 2.887007] init: /init.rc: 11: invalid command 'setcon'
<11>[ 2.887037] init: /init.rc: 57: invalid command 'load_all_props'
<11>[ 2.887088] init: could not import file '/init.recovery.logd.rc' from '/init.rc'
<13>[ 2.887160] init: (Parsing /init.recovery.usb.rc took 0.00s.)
<11>[ 2.887177] init: could not import file '/init.recovery.qcom.rc' from '/init.rc'
<13>[ 2.887186] init: (Parsing /init.rc took 0.00s.)
heres my code: https://github.com/edTheGuy00/android_device_zuk_z2pro
any help would be appreciated.
Maybe you should paste your init.rc here in order to spot the issue.
Solved
I've found a way to select alternative partitions during boot process. It's possible to use busybox ln commands inside a shell script (placing busybox binary and shell script into ramdisk root directory of boot.img) and run the script by busybox ash command from init.rc. The commands inside shell script should look like as the following:
Code:
/busybox ln -sf /dev/block/mmcblk1p2 /[email protected]
/busybox ln -sf /dev/block/mmcblk1p3 /[email protected]
/busybox ln -sf /dev/block/mmcblk1p4 /[email protected]
These commands should be put into an if...then block inside the shell script. The conditional statement should looks like this:
Code:
if [ -b /dev/block/mmcblk1 ]
That way, if external sdcard is detected, shell script will change targets of [email protected] symbolic links from internal emmc partitions to external sdcard ones. So, init will use /system, /data and /cache mountpoints with the latter in place of the default ones during mount.
I'm a mechanical engineer, not an IT guy. I can fix machines, perhaps, but not bricked phones. So try anything at your own extreme risk. This is NOT a step by step guide.
DEVICE:
QMobile Z8, same as Wikio Ridge 4G, (MSM8916).
Running Android 5.0.2, SuperSU rooted.
Kernel v 3.10.49
Thanks to @ASAZING for TWRP 3.0.2-0
PROBLEM:
So the screen started blinking and locking / unlocking automatically like UI resetting. And there was no SIM. At first I thought it's launcher or SuperSU causing problem. But it got worse over days. So I decided a factory flash since I didn't have untouched flashable zip.
Flashed firmware using QFIL but no success. Rebooted to recovery and TWRP was still there.
/data partition was locked and TWRP doesn't support decryption. So I did a factory reset and the message came: /data not mounted. Invalid Argument
Formatted /data from "Repair or Change Filesystem" option in TWRP and as a result /data and /cache both couldn't be mounted.
Formatted /cache, and /system too not mounted.
Manually formatted using 'make_ext4' and tried 'fastboot format:ext4 userdata' as well. Both succeeded apparently but mount still failed.
Run 'e2fsck' and that showed: "Bad magic number in super block" and "The superblock could not be read."
Run 'mke2fs -n' for alternate super blocks, run again 'e2fsck' but no success. Images are attached.
'sgdisk --verify' gives this error log:
Code:
sgdisk --verify mmcblk0p1
[COLOR="Red"]***************************************************************
Found invalid GPT and valid MBR; converting MBR to GPT format in memory.
***************************************************************
Exact type match not found for type code 7200; assigning type code for 'Linux filesystem'
Exact type match not found for type code 6500; assigning type code for 'Linux filesystem'
Exact type match not found for type code 7900; assigning type code for 'Linux filesystem'
Exact type match not found for type code 0D00; assigning type code for 'Linux filesystem'
Warning! Main partition table overlaps the first partition by 34 blocks!
You will need to delete this partition or resize it in another utility.
Warning! Secondary partition table overlaps the last partition by 3805778618 blocks!
You will need to delete this partition or resize it in another utility.
Problem: partitions 2 and 1 overlap:
Partition 2: 168689522 to 2104717761
Partition 1: 778135908 to 1919645538
Problem: partitions 3 and 1 overlap:
Partition 3: 1869881465 to 3805909656
Partition 1: 778135908 to 1919645538
Problem: partitions 3 and 2 overlap:
Partition 3: 1869881465 to 3805909656
Partition 2: 168689522 to 2104717761
Problem: partition 1 is too big for the disk.
Problem: partition 2 is too big for the disk.
Problem: partition 3 is too big for the disk.
Problem: partition 4 is too big for the disk.
Warning! Main partition table overlaps the first partition by 34 blocks!
You will need to delete this partition or resize it in another utility.
Warning! Secondary partition table overlaps the last partition by 3805778618 blocks!
You will need to delete this partition or resize it in another utility.
Identified 9 problems![/COLOR]
==========================
sgdisk --verify mmcblk0p16
[COLOR="red"]Creating new GPT entries.
Problem: GPT claims the disk is larger than it is! (Claimed last usable sector is 18446744073709551584, but backup header is at 1 and disk size is 2 sectors.
The 'e' option on the experts' menu will probably fix this problem
Identified 1 problems!
[/COLOR]
==========================
sgdisk --verify mmcblk0p17
[COLOR="red"]Creating new GPT entries.
Problem: GPT claims the disk is larger than it is! (Claimed last usable sector is 18446744073709551598, but backup header is at 15 and disk size is 16 sectors.
The 'e' option on the experts' menu will probably fix this problem
Identified 1 problems![/COLOR]
==========================
sgdisk --verify mmcblk0p22
[COLOR="red"]Creating new GPT entries.
Problem: GPT claims the disk is larger than it is! (Claimed last usable sector is 18446744073709551614, but backup header is at 31 and disk size is 32 sectors.
The 'e' option on the experts' menu will probably fix this problem
Identified 1 problems![/COLOR]
'parted rm' and 'fastboot erase' didn't work either. Partition was still there.
Then I tried to flash stock recovery through TWRP. And recovery too gone.
Now, device boots directly to bootloader (fastboot mode) and is halted there. Have to 'fastboot boot recovery.img' or 'fastboot boot boot.img' each time.
Download Mode (QDLoader 9008) is also accessible.
FLASHING FACTORY FIRMWARE:
Now left only with fastboot and EDL, tried once again QFIL flasher, Wiko official flasher, QDownloader. Log says: "Read back verify failed at sector ...." for partitions misc, system, cache, persist, recovery, userdata (6 partitions) and 2 partition table *.bins
Hence proved, eMMC is malfunctioning and device now can't boot on its own due to no partition table.
Tried 'sgdisk --backup' and 'sgdisk --load-backup' options for partition table. It gives error: "Warning! Current disk size doesn't match that of backup." and "Problem: Partition 28 ends before it begins." etc.
'fastboot flash partition *.bin' also failed with error: "remote: failed to write partition".
'dd if=gpt_main0.bin of=/dev/block/mmcblk0' apparently succeeded but comparing octal dump ('od') files of 34 sectors at start shows no difference, means file is not written to eMMC.
SOLUTION SUMMARY:
Partition SD card according to already existing partition table on internal eMMC.
Flash partition images from factory firmware to newly created partitions.
Modify kernel (boot.img) and recovery to boot from sd card instead of internal memory.
Boot kernel or recovery through fastboot.
SECTION 1
PARTITION SD CARD:
Here comes Google. Following the footsteps of @lexelby at this, I created gpt (parted command) on 16GB C-10 sd card using Ubuntu virtual machine.
Created first partition for external_sd card and 6 more of same size as original ones (size checked by parted and from rawprogram_unsparse.xml). Filesystems: system, userdata, cache & persist of ext4 while misc, recovery of linux-swap (though 'dd' will overwrite them).
Then I unsparsed userdata, system and cache images from factory firmware (on Windows used packsparseimg.exe binary). Sparsed images can only be flashed through fastboot?
Copied 5 prtitions images: userdata, system, cache, persist and misc using dd command to /dev/block/mmcblk1p*.
MODIFYING BOOT & RECOVERY:
Now coming to the changes in mount paths of boot and recovery (fstab and init.*.rc).
Extracted boot.img and then ramdisk using "Image Studio for Android". 'unpackbootimg' and 'abootimg' don't extract all files on Ubuntu. 'mkbootimg' makes smaller boot.img file without boot.img-dtb. Perhaps I'm doing it wrong.
Anyway, then did 'grep dev/block' on all extracted files. Results are attached for reference.
Made changes in "fstab.qcom" and "init.target.rc". For details on changes made, please read on RE-MODIFYING BOOT & RECOVERY.
Repacked boot.img
Similarly extracted recovery.img, did 'grep dev/block' on all extracted files. And made changes in "recovery.fstab".
Repacked recovery.img
COPYING IMAGES TO PARTITIONS AND BOOTING:
'fastboot flash boot boot.img' and 'dd if=recovery.img of=dev/block/mmcblk1p*' (though useless, have to boot from fastboot)
Rebooted to recovery by 'fastboot boot recovery.img'
userdata, persist and cache couldn't be mounted in TWRP. Tried 'mount -t ext4 -o loop *.img' on Ubuntu but there too not mounted. Googled and using commands 'file', 'fdisk', 'sfdisk', 'e2fsck' and finally 'resize2fs -f /*.img' resolved the problem "bad geometry: block count xxx exceeds size of device...".
Also unsparsed userdata too large to handle and only a few MBs data inside, that too useless. Therefore, did 'make_ext4fs' on cache & userdata.
Now booted kernel by 'fastboot boot boot.img'
And.......... it boots. But very very slow (due to slow write speed of sd card obviously). Took almost half an hour at first boot.
UNRESOLVED PROBLEMS:
There is no sound. Because of /persist not mounted? And still no SIM, means radio firmware isn't readable from eMMC or this too due to /persist absent? After all that contains drivers. And also Wi-Fi and bluetooth not working.
SECTION 2
RE-PARTITION SD CARD:
So re-created gpt on sd card (using parted and fdisk) and in a hope to utilize all necessary partitions, 100% replicated all partitions (except larger userdata) including space required at start and end of eMMC for partition table. Partition tables of both mmcblk0 and mmvblk1 are attached.
RE-MODIFYING BOOT & RECOVERY:
Made following changes in boot.img:
DEVICE BOOTS ALSO WITHOUT MAKING ANY CHANGES TO BOOT.IMG.
I don't know why but 'bootdevice' is automagiacally changed from 7824900.sdhci (eMMC) to 7864900.sdhci (external SD card). It seems there is some auto-detection mechanism.
Code:
########## ./ramdisk/fstab.qcom ##########
#/dev/block/bootdevice/by-name/system /system ext4 ro,barrier=1,discard wait
#/dev/block/bootdevice/by-name/userdata /data ext4 nosuid,nodev,barrier=1,noauto_da_alloc,discard wait,check,encryptable=footer
#CHANGED TO
/dev/block/[B]mmcblk1p24[/B] /system ext4 ro,barrier=1,discard wait
/dev/block/[B]mmcblk1p32[/B] /data ext4 nosuid,nodev,barrier=1,noauto_da_alloc,discard wait,check,encryptable=footer
#/devices/soc.0/7864900.sdhci/mmc_host /storage/sdcard1 vfat nosuid,nodev wait,voldmanaged=sdcard1:auto,noemulatedsd
#[B]disabled[/B]
Code:
########## ./ramdisk/init.target.rc ##########
on fs
mount_all fstab.qcom
#wait /dev/block/bootdevice/by-name/cache
#mount ext4 /dev/block/bootdevice/by-name/cache /cache nosuid nodev barrier=1
#CHANGED TO
wait /dev/block/[B]mmcblk1p26[/B]
mount ext4 /dev/block/[B]mmcblk1p26[/B] /cache nosuid nodev barrier=1
#wait /dev/block/bootdevice/by-name/persist
#mount ext4 /dev/block/bootdevice/by-name/persist /persist nosuid nodev barrier=1
#CHANGED TO
wait /dev/block/[B]mmcblk1p25[/B]
mount ext4 /dev/block/[B]mmcblk1p25[/B] /persist nosuid nodev barrier=1
#wait /dev/block/bootdevice/by-name/modem
#mount vfat /dev/block/bootdevice/by-name/modem /firmware ro context=u:object_r:firmware_file:s0,shortname=lower,uid=1000,gid=1000,dmask=227,fmask=337
#CHANGED TO
wait /dev/block/[B]mmcblk1p1[/B]
mount vfat /dev/block/[B]mmcblk1p1[/B] /firmware ro context=u:object_r:firmware_file:s0,shortname=lower,uid=1000,gid=1000,dmask=227,fmask=337
on charger
#wait /dev/block/bootdevice/by-name/system
#mount ext4 /dev/block/bootdevice/by-name/system /system ro barrier=1
#CHANGED TO
wait /dev/block/[B]mmcblk1p24[/B]
mount ext4 /dev/block/[B]mmcblk1p24[/B] /system ro barrier=1
Code:
########## ./split_img/boot.img-cmdline ##########
#console=ttyHSL0,115200,n8 androidboot.console=ttyHSL0 androidboot.hardware=qcom user_debug=30 msm_rtb.filter=0x3F ehci-hcd.park=3 androidboot.bootdevice=7824900.sdhci lpm_levels.sleep_disabled=1
#CHANGED TO
console=ttyHSL0,115200,n8 androidboot.console=ttyHSL0 androidboot.hardware=qcom user_debug=30 msm_rtb.filter=0x3F ehci-hcd.park=3 [B]androidboot.bootdevice=7864900.sdhci[/B] lpm_levels.sleep_disabled=1
And following changes in recovery.img:
Code:
########## ./ramdisk/etc/recovery.fstab ##########
#/cache ext4 /dev/block/bootdevice/by-name/cache flags=display=Cache
#/system ext4 /dev/block/bootdevice/by-name/system flags=display=System
#/data ext4 /dev/block/bootdevice/by-name/userdata flags=encryptable=footer;length=-16384
#/persist ext4 /dev/block/mmcblk0p25 flags=backup=1;display=Persist
#/boot emmc /dev/block/bootdevice/by-name/boot flags=display=Boot
#/recovery emmc /dev/block/bootdevice/by-name/recovery flags=backup=1;display=Recovery
#/misc emmc /dev/block/bootdevice/by-name/misc /misc flags=backup=1;display=Misc
#/firmware vfat /dev/block/mmcblk0p1 flags=backup=1;display=Modem
#/splash emmc /dev/block/mmcblk0p18 flags=backup=1;display=Splash
#/fsg emmc /dev/block/mmcblk0p20 flags=backup=1;subpartitionof=/oem
#/aboot emmc /dev/block/mmcblk0p4 flags=backup=1;display=Aboot
#/abootbak emmc /dev/block/mmcblk0p5 flags=subpartitionof=/aboot;backup=1
#/hyp emmc /dev/block/mmcblk0p10 flags=backup=1;display=Firmware-update
#/sbl1 emmc /dev/block/mmcblk0p2 flags=backup=1;subpartitionof=/hyp
#/rpm emmc /dev/block/mmcblk0p6 flags=backup=1;subpartitionof=/hyp
#/tz emmc /dev/block/mmcblk0p8 flags=backup=1;subpartitionof=/hyp
#/hypbak emmc /dev/block/mmcblk0p11 flags=backup=1;subpartitionof=/hyp
#/sbl1bak emmc /dev/block/mmcblk0p3 flags=backup=1;subpartitionof=/hyp
#/rpmbak emmc /dev/block/mmcblk0p7 flags=backup=1;subpartitionof=/hyp
#/tzbak emmc /dev/block/mmcblk0p9 flags=backup=1;subpartitionof=/hyp
#/modemst1 emmc /dev/block/mmcblk0p13 flags=backup=1;display=EFS
#/modemst2 emmc /dev/block/mmcblk0p14 flags=backup=1;subpartitionof=/modemst1
#/oem emmc /dev/block/mmcblk0p30 flags=backup=1;display=OEM
#/DDR emmc /dev/block/mmcblk0p20 flags=backup=1;subpartitionof=/oem
#/fsc emmc /dev/block/mmcblk0p16 flags=backup=1;subpartitionof=/oem
#/ssd emmc /dev/block/mmcblk0p17 flags=backup=1;subpartitionof=/oem
#/pad emmc /dev/block/mmcblk0p12 flags=backup=1;subpartitionof=/oem
#CHANGED TO
/cache ext4 /dev/block/[B]mmcblk1p26[/B] flags=display=Cache
/system ext4 /dev/block/[B]mmcblk1p24[/B] flags=display=System
/data ext4 /dev/block/[B]mmcblk1p32[/B] flags=encryptable=footer;length=-16384
/persist ext4 /dev/block/[B]mmcblk1p25[/B] flags=backup=1;display=Persist
/boot emmc /dev/block/[B]mmcblk1p23[/B] flags=display=Boot
/recovery emmc /dev/block/[B]mmcblk1p27[/B] flags=backup=1;display=Recovery
/misc emmc /dev/block/[B]mmcblk1p15[/B] flags=backup=1;display=Misc
/firmware vfat /dev/block/[B]mmcblk1p1[/B] flags=backup=1;display=Modem
/splash emmc /dev/block/[B]mmcblk1p18[/B] flags=backup=1;display=Splash
/fsg emmc /dev/block/[B]mmcblk1p21[/B] flags=backup=1;subpartitionof=/oem
/aboot emmc /dev/block/[B]mmcblk1p4[/B] flags=backup=1;display=Aboot
/abootbak emmc /dev/block/[B]mmcblk1p5[/B] flags=subpartitionof=/aboot;backup=1
/hyp emmc /dev/block/[B]mmcblk1p10[/B] flags=backup=1;display=Firmware-update
/sbl1 emmc /dev/block/[B]mmcblk1p2[/B] flags=backup=1;subpartitionof=/hyp
/rpm emmc /dev/block/[B]mmcblk1p6[/B] flags=backup=1;subpartitionof=/hyp
/tz emmc /dev/block/[B]mmcblk1p8[/B] flags=backup=1;subpartitionof=/hyp
/hypbak emmc /dev/block/[B]mmcblk1p11[/B] flags=backup=1;subpartitionof=/hyp
/sbl1bak emmc /dev/block/[B]mmcblk1p3[/B] flags=backup=1;subpartitionof=/hyp
/rpmbak emmc /dev/block/[B]mmcblk1p7[/B] flags=backup=1;subpartitionof=/hyp
/tzbak emmc /dev/block/[B]mmcblk1p9[/B] flags=backup=1;subpartitionof=/hyp
/modemst1 emmc /dev/block/[B]mmcblk1p13[/B] flags=backup=1;display=EFS
/modemst2 emmc /dev/block/[B]mmcblk1p14[/B] flags=backup=1;subpartitionof=/modemst1
/oem emmc /dev/block/[B]mmcblk1p30[/B] flags=backup=1;display=OEM
/DDR emmc /dev/block/[B]mmcblk1p20[/B] flags=backup=1;subpartitionof=/oem
/fsc emmc /dev/block/[B]mmcblk1p16[/B] flags=backup=1;subpartitionof=/oem
/ssd emmc /dev/block/[B]mmcblk1p17[/B] flags=backup=1;subpartitionof=/oem
/pad emmc /dev/block/[B]mmcblk1p12[/B] flags=backup=1;subpartitionof=/oem
#/external_sd auto /dev/block/mmcblk1p1 /dev/block/mmcblk1 flags=display="MicroSD Card";storage;wipeingui;removable
#CHANGED TO
# None. [B]External sd disabled[/B].
Code:
########## ./ramdisk/uneventd.rc ##########
#/dev/block/bootdevice/by-name/config 0660 system system
#CHANGED TO
/dev/block/[B]mmcblk1p29[/B] 0660 system system
Code:
########## ./split_img/recovery.img-cmdline ##########
#console=ttyHSL0,115200,n8 androidboot.console=ttyHSL0 androidboot.hardware=qcom user_debug=30 msm_rtb.filter=0x3F ehci-hcd.park=3 androidboot.bootdevice=7824900.sdhci lpm_levels.sleep_disabled=1 androidboot.selinux=permissive
#CHANGED TO
console=ttyHSL0,115200,n8 androidboot.console=ttyHSL0 androidboot.hardware=qcom user_debug=30 msm_rtb.filter=0x3F ehci-hcd.park=3 [B]androidboot.bootdevice=7864900.sdhci[/B] lpm_levels.sleep_disabled=1 androidboot.selinux=permissive
Repacked boot.img and recovery.img.
RE-COPYING IMAGES TO PARTITIONS AND BOOTING:
Copied (dd) all available (15) images to (20) partitions on sd card.
Copied (dd) the 10 images not found in factory firmware from mmcblk0 to mmcblk1. (Not sure if successful).
2 partitions (/data and /cache) already formatted in ext4.
'fastboot boot recovery.img'. All partitions are mounted now. No horrible error lines.
'fastboot boot boot.img'
ROM booted successfully WITH sounds, SIM, Wi-Fi and Bluetooth. All seems working well so far.
SECTION 3
Continued on post 3...
hello hi
i have xiaomi redmi 2 chinesse version with same problem with your device. stuck logo, only still can access recovery TWRP via fastboot boot trwp.img.
twrp cant wipe, cant format, internal storage 0mb, "failed argument ".cant flash stock rom with flash tools "failed write partition", . try terminal parted rm not solve. try to many google same issue not solve. i think emmc or hardware issue
i never using linux and linux command so
please help me.make step by step guide , boot from sdcard .
- make partition sd card to be like emmc partition block
- can i using windows os or using small linux distro
- how to modif image stock rom ,kernel ,and flashing to sdcard
- how to boot from sdcard
many thank you
Continued from OP...
SECTION 3
QUERIES:
UNCERTAIN PARTITIONS
But there are no images available for these 10 partitions in factory firmware:
pad, modemst1, modemst2, fsc, ssd, DDR, keystore, config, oem & devinfo.
These seem to be very essential for OS, also containing IMEI if I'm not mistaken? I'm not sure of their contents. How system working without them? All are useless?
HOW TO COMPLETELY BOOT FROM SD CARD
In boot.img, "fstab.qcom" contains mount paths for system & userdata. While "init.target.rc" contains only mount paths for cache, persist and modem. In total 5 partitions which are mounted (checked by 'mount').
Code:
[email protected]:/ $ mount | grep mmcblk1
/dev/block/mmcblk1p24 /system ext4 rw,seclabel,relatime,discard,data=ordered 0 0
/dev/block/mmcblk1p32 /data ext4 rw,seclabel,nosuid,nodev,relatime,discard,noauto_da_alloc 0 0
/dev/block/mmcblk1p26 /cache ext4 rw,seclabel,nosuid,nodev,relatime 0 0
/dev/block/mmcblk1p25 /persist ext4 rw,seclabel,nosuid,nodev,relatime,data=ordered 0 0
/dev/block/mmcblk1p1 /firmware vfat ro,context=u:object_r:firmware_file:s0,relatime,uid=1000,gid=1000,fmask=0337,dmask=0227,codepage=437,iocharset=iso8859-1,shortname=lower,errors=remount-ro 0 0
[email protected]:/ $
So the primary question is:
How to change mount source of other partitions from mmcblk0 to mmcblk1? Or how to force OS to read the essentially required partitions from mmcblk1 instead of mmcblk0?
Need to modify any other files in ramdisk or kernel-zimage or in /system or to modify init.d scripts or create new scripts? Any help?
Other than 10 partitions mentioned above, these "not mounted" partitions also include modem, sbl1, aboot, rpm, tz and hyp and fsg. Modem contains bootable code of MBR and following 5 are also executable binaries. I think these are all part of bootloader i.e. loading in initial booting process and not required by OS. But what about the fsg and ten others? Where are those used? Here is a partition detail.
Another primary issue is:
I think it's almost impossible to make Boot ROM (CPU embedded) hand over charge to bootloader at "mmcblk1". "mmcblk0" must be hardcoded in Boot ROM.
So, how to make bootloader load "kernel" and "rootfs" from mmcblk1p* instead of mmcblk0p*? Like there are switches in testing devices to optionally boot from different memories. Can we modify "aboot" (the little kernel) or "emmc_appsboot.mbn" ELF binary for this purpose? It must be complicated as bootloaders are signed by vendor (Qualcomm) and involve low-level programming as discussed here. Right?
Or in other words, how to force bootloader to read partition table from dev/mmcblk1 instead of dev/mmcblk0?
If we can't do this, system doesn't know how to boot in the absence of eMMC. That would have to be done through fastboot everytime we need to. Because boot chain will be stuck at bootloader.
Multi-booting solutions are also dependent on a fully working /boot partition on eMMC because they (one way or the other) re-flash/replace modified boot image every time a ROM is to be switched. EFIDroid is a secondary bootloader but that too replaces /boot and/or /recovery.
I have gone through this, this, this and this. But they only address partial booting from sd card e.g. dual booting in which only /system, /data and /cache are involved. None has discussed complete boot from sd. Is it really impossible? This link gives a little hope but it points to a ready made solution (bootloader) which boots kernel from SD card. But it gives no explanation how.
I have also come across a few threads discussing Samsung (and HTC too) booting from SD Card as a fix to QHSUSB_DLOAD mode or bricked-bootloaader state. They extracted "debrick" file from a working phone and flashed that to the start of SD Card. Debrick file seems to be a single bootloader file containing all bootloaders in it as explained here and here. So after flashing the bootloader(s) with its accompanying partitions to SD Card, when device was powered on, it automatically booted from SD Card. If it's that simple for all devices with Qualcomm SoC, the only thing I have to do is :laugh:
Code:
[COLOR="Red"]dd if=/dev/zero of=/dev/block/mmcblk0 bs=1m count=200[/COLOR]
Any suggestions? I believe this must be possible as they are discussing here.
Edit: Related quote from [GUIDE][9008][EDL|QDL][QUALCOMM ONLY] Unbrick via external sdcard (no QFIL!):
On eMMC devices, the boot path is /dev/block/mmcblk0. If you have a 9008 brick, the SD card is seen as /dev/block/mmcblk0 so the phone will boot from it on an eMMC device.
Click to expand...
Click to collapse
Some secondary questions:
HOW ARE PARTITIONS IDENTIFIED BY BOOT-ROM WITHOUT PARTITION TABLE ON eMMC
If there is no readable partition table on a bricked eMMC, how Boot ROM (primary bootloader on SoC) switches control to secondary bootloader or bootable modem partition or other partitions used by processors? Means how SoC / Processors locate modem, sbl, rpm, tz or aboot (the little kernel's offspring) on eMMC? Also, why 'parted /dev/block/mmcblk0 p' and 'sgdisk --print /dev/block/mmcblk0'show partitions if there is no table?
Though parted-2.2 shows warning:
Code:
[COLOR="Red"]Error: Both the primary and backup GPT tables are corrupt. Try making a fresh
table, and using Parted's rescue feature to recover partitions.[/COLOR]
Or I'm thinking in wrong direction? This link discusses the issues but I'm not clear how it works.
Once the a device is powered on it starts code from a know location (ROM) and looks for the first stage bootloader in a specific block.
Click to expand...
Click to collapse
How is this "specific block" located by cpu ROM?
It's talking about some "low-level" and "high-level" partition tables. How they differ? How can we manipulate the former?
And finally...
HOW TO SPEED UP SD CARD
Other than using a UHS-III or the most recent and expensive App Performance Class (A1) sd card, what changes we can make to kernel to boost read/write speed? Otherwise, it's almost useless with too slow speed, frequent ANRs, hangs and laggings.
Default I/O scheduler being used on QMobile Z8 is cfq with default tune-able settings. I think it's one of best schedulers for higher throughput. Na? Try other? Details here:
Code:
[email protected]:/ # cat /sys/block/mmcblk0/queue/scheduler
noop deadline row [cfq]
[email protected]:/ # for fyle in $(find /sys/block/mmcblk0/queue/iosched/ -type f); do echo $fyle; cat $fyle; done;
/sys/block/mmcblk0/queue/iosched/fifo_expire_async
50
/sys/block/mmcblk0/queue/iosched/group_idle
0
/sys/block/mmcblk0/queue/iosched/quantum
20
/sys/block/mmcblk0/queue/iosched/slice_async
40
/sys/block/mmcblk0/queue/iosched/slice_idle
10
/sys/block/mmcblk0/queue/iosched/slice_sync
100
/sys/block/mmcblk0/queue/iosched/low_latency
0
/sys/block/mmcblk0/queue/iosched/fifo_expire_sync
50
/sys/block/mmcblk0/queue/iosched/back_seek_max
16384
/sys/block/mmcblk0/queue/iosched/target_latency
300
/sys/block/mmcblk0/queue/iosched/back_seek_penalty
2
/sys/block/mmcblk0/queue/iosched/slice_async_rq
2
[email protected]:/ # cat /sys/block/mmcblk1/queue/scheduler
noop deadline row [cfq]
[email protected]:/ # for fyle in $(find /sys/block/mmcblk1/queue/iosched/ -type f); do echo $fyle; cat $fyle; done;
/sys/block/mmcblk1/queue/iosched/fifo_expire_async
50
/sys/block/mmcblk1/queue/iosched/group_idle
0
/sys/block/mmcblk1/queue/iosched/quantum
20
/sys/block/mmcblk1/queue/iosched/slice_async
40
/sys/block/mmcblk1/queue/iosched/slice_idle
10
/sys/block/mmcblk1/queue/iosched/slice_sync
100
/sys/block/mmcblk1/queue/iosched/low_latency
0
/sys/block/mmcblk1/queue/iosched/fifo_expire_sync
50
/sys/block/mmcblk1/queue/iosched/back_seek_max
16384
/sys/block/mmcblk1/queue/iosched/target_latency
300
/sys/block/mmcblk1/queue/iosched/back_seek_penalty
2
/sys/block/mmcblk1/queue/iosched/slice_async_rq
2
[email protected]:/ #
Tried different cache values (read_ahead_kb) from 64 to 4048. Makes no difference apparently.
Also disabled jounalling using 'tune2fs -O ^has_journal' and e2fsck checks using 'tune2fs -c -1'.
Changed mount options to for /data and /cache:
Code:
[email protected]:/ # mount | grep -E "/cache|/data"
/dev/block/mmcblk1p32 /data ext4 rw,seclabel,[B]noatime,discard,nobarrier,noauto_da_alloc,commit=60[/B] 0 0
/dev/block/mmcblk1p26 /cache ext4 rw,seclabel,noatime,discard,nobarrier,noauto_da_alloc,commit=60 0 0
[email protected]:/ #
Seems useless so far. Any ideas? Or it's a hardware limitation of device?
Is there a way to get rid of FUSE and use ext4 in true sense for whole /data (only possible if someone is willing to quit using MTP), though it doesn't matter much for Android's internal operations? But it's a real pain for I/O operations on external media.
Edit: Speed much improved by using a more certain branded SD Card; Sandisk C-10.
@yoAeroA00 Sir need your special attention for kernel part. You have a good history with kernel tweaking and multibooting.
I try to manual flash commands, one by one read from flash_all.bat. everything is okay finish, except file "gpt_both0.bin" and "sec.dat"
jeksparo said:
I try to manual flash commands, one by one read from flash_all.bat. everything is okay finish, except file "gpt_both0.bin" and "sec.dat"
Click to expand...
Click to collapse
If flasher is unable to flash partitions, then flashing manually won't make any difference. "gpt_both0.bin" contains partition tables; main and backup. First sector is protective mbr for legacy partitioning tools and next 33 sectors contain gpt partition table. A backup of partition table is stores on last 33 sectors of disk or emmc in our case. Total 67 sectors make 33.5 KiB size which is same as that of gpt_both0.bin. Let me have a look at partition table for further clarity. Run these from twrp to save your partition table.
Code:
sgdisk -p /dev/block/mmcblk0 > pt1
parted /dev/block/mmcblk0 p free > pt2
cant compile sgdisk -p /dev/block/mmcblk0 > pt1 invalid option --p
and parted /dev/block/mmcblk0 > pt2 blank line after entering
my device shell dont have parted command, so i run parted from sd card.
how to created gpt on sdcard using parted and fdisk, if my parted command in sdcard too, it is possible?
jeksparo said:
cant compile sgdisk -p /dev/block/mmcblk0 > pt1 invalid option --p
and parted /dev/block/mmcblk0 > pt2 blank line after entering
my device shell dont have parted command, so i run parted from sd card.
how to created gpt on sdcard using parted and fdisk, if my parted command in sdcard too, it is possible?
Click to expand...
Click to collapse
For sgdisk use --print as help shown in your screenshot.
Code:
sgdisk --print /dev/block/mmcblk0 > /part_table
"> /partition_table" is to save the output in root directory so that you can copy paste it. Otherwise you can also take screenshots in TWRP with PWR + VOL- combination.
'parted' doesn't come bundled with TWRP. You can use the binary from your SD card but you need to copy it somewhere else like '/sbin' if you want to partition you SD card. 'fdisk' can't create GPT, it's legacy tool for MBR partition scheme. You need to use 'parted', 'gdisk' or 'sgdisk' etc. to create partitions. Binaries for Android are with limited functionality. That's why Linux is preferred, but not necessary. Also the copying of partition images will be easy on Linux, though very slow if you connect card reader in virtual machine.
These partitions on your device contain filesystem and will be mounted in ROM
modem vfat
system ext4
cache ext4
persist ext4
userdata ext4
Click to expand...
Click to collapse
Boot and recovery partitions can also be used if your partition table isn't too corrupt to recognize them. Otherwise you'll have to use fastboot on every boot like me.
Simple is to duplicate the whole internal partition table on SD card because rest of the partitions don't occupy much space. It makes partition numbering easy.
But before creating partitions, you need to know exact boundaries in bytes:
Code:
parted /dev/block/mmcblk0
(parted) u b
(parted) p free
(parted) q
please hellp me
Hiii . i have wiko ridge 4g and i have the same problem as you it stuck on wiko logo and when i try to flash it with stock rom from wiko site nothing hapend and i tried to flash it using Qfil in the log i see "Read back verify failed at sector" the same problem as you sooooo please make step by step guid
i can boot to download mode . when i try to boot to recovery it boot to fastboot mode automaticly .....plz help me..... -sorry for my english-
_6ix._.9ine said:
Hiii . i have wiko ridge 4g and i have the same problem as you it stuck on wiko logo and when i try to flash it with stock rom from wiko site nothing hapend and i tried to flash it using Qfil in the log i see "Read back verify failed at sector" the same problem as you sooooo please make step by step guid
i can boot to download mode . when i try to boot to recovery it boot to fastboot mode automaticly .....plz help me..... -sorry for my english-
Click to expand...
Click to collapse
On what part you need help? It's all about partitioning an sd card and copying data to it. Then unpack, modify and re-pack boot.img
plllllz help me
mirfatif said:
On what part you need help? It's all about partitioning an sd card and copying data to it. Then unpack, modify and re-pack boot.img
Click to expand...
Click to collapse
i've been trying to understand what u wrote for the last 7 days and i couldn't understand shiiiit :crying:
i don't know anything about this shiiiit :crying: i'm so sad plz make video and upload it on youtube and show me step by step how did u boot from sd card plllllllllllllllllz _sorry for my english_
MODIFYING BOOT & RECOVERY:
Now coming to the changes in mount paths of boot and recovery (fstab and init.*.rc).
Extracted boot.img and then ramdisk using "Image Studio for Android". 'unpackbootimg' and 'abootimg' don't extract all files on Ubuntu. 'mkbootimg' makes smaller boot.img file without boot.img-dtb. Perhaps I'm doing it wrong.
Anyway, then did 'grep dev/block' on all extracted files. Results are attached for reference.
Made changes in "fstab.qcom" and "init.target.rc". For details on changes made, please read on RE-MODIFYING BOOT & RECOVERY.
Click to expand...
Click to collapse
In that part I used the "ABOOTIMG". To work, do the following:
Code:
$ sudo abootimg -x boot.img ramdisk ramdisk kernel kernel
or
Code:
#abootimg -x boot.img ramdisk ramdisk kernel kernel
You will find 3 files. The "RAMDISK" comes packaged in "GZIP". Unzip it and enter the folder that will be created. Inside this folder you will have the files to edit as the post follows.
By the way, congratulations for the initiative!
Turkish
Nothing is understood when it is translated. Can a British English translate this to me?
Thanks you very much for your great story and for a lot of informations!
Peace & Respect
mirfatif,
thanks for this interesting and promising information!
I'd like, though, get an additional explanation: you use "fastboot" to handle your smartphone. Does it mean,
that it was bootable when you've started all this stuff with booting from SD-card?
I'm asking because I'm trying to boot my samsung galaxy GT-N7100 from sd-card with completely dead emmc.
igorbounov said:
you use "fastboot" to handle your smartphone. Does it mean,
that it was bootable when you've started all this stuff with booting from SD-card?
Click to expand...
Click to collapse
My phone doesn't have "completely" dead eMMC. Booting process works up to bootloader (aboot) and it's related partitions. So fastboot works (as it's managed by bootloader). But after that, bootloader can't load boot image (kernel) from boot partition. Neither recovery partition is readable. Thus I have to do it manually using fastboot. And the remaining OS related partitions are read from SD card.
mirfatif said:
My phone doesn't have "completely" dead eMMC..
Click to expand...
Click to collapse
Now I get the idea... Nevertheless it looks like now I should stop attempts recovering smartphone because while trying I've turned
(somehow) my available 64Gb Samsung SD-card to a write-protected state (while partitioning). So my further experiments seem
not worth it - now it looks like buying a new Galaxy Note with a new SD-card is more cost-effective.
igorbounov said:
I've turned
(somehow) my available 64Gb Samsung SD-card to a write-protected state (while partitioning).
Click to expand...
Click to collapse
Are you unable to create a new partition table using parted/fdisk/gdisk?
mirfatif said:
Are you unable to create a new partition table using parted/fdisk/gdisk?
Click to expand...
Click to collapse
No, I've used everything - even Windows-oriented utilities for low level formatting. All of them complain that
this SD card is write-protected. It has stuck in a strange state - a gpt table created, but no partitions.
And some programs (sfdisk or sgdisk, and even diskpart from Windows) find there some inconsistences.
Perhaps the inner electronics thinks that this errors correspond to a worn state - and sets this read-only attribute.
When I partitioned this sd-card, I've first created the new gpt table, then for a long time speculated about which
partition of what type and size should be created. In this process I've opened two or maybe more parted and
gparted sessions, and then I've saved partitions from one session, then maybe from other... and now this
memory card is in read-only state. Perhaps it has decided that this is the most safe way.
igorbounov said:
No, I've used everything - even Windows-oriented utilities for low level formatting....
Click to expand...
Click to collapse
How did you connect SD card to PC? I mean USB card reader, SD card slot etc. Sometimes card reader drivers are causing the problems. Are you using Linux / Windows natively or on a VM? Did you try creating partition table on Android phone? Usually phones can handle SD cards better. Try card slot or OTG, in TWRP or from ROM, using arm or aarch64 binaries of parted, fdisk and gdisk. Command line tools are preferable for troubleshooting than GUI tools.
mirfatif said:
How did you connect SD card to PC? I mean USB card reader, ...
Click to expand...
Click to collapse
I've used a chip USB card reader. Linux and Windows natively and Windows in a VM (QEMU/KVM). I haven't yet found some other volunteer with Android phone to put my SD card there. My old faithfull Nokia 6131 just don't see this card (and shouldn't, there are no partitions and Nokia 6131 cannot handle sd cards of such size). Yesterday I've used an old card reader, that is built in my daughter's PC, I've used for that purpose a special SD casing for microSD - Windows disk management confirmed that this SD card is read-only.
May be some embedded device could help in this situation - some AVR- or STM32-based device via SPI (than it doesn't matter wether there is some protection or no).