you need to look at `dmesg` to see if the kernel provided any useful info in there about the mount operation

you probably should also run `sudo partx -u /dev/sdb` after rewriting the partition table but before trying to use /dev/sdb2 to make sure the kernel is in sync with the layout on disk

Folow dmesg when i used usb (adapter) to Micro SDCARD
[ 5667.168060] usb 1-2.2: new high-speed USB device number 4 using xhci-hcd
[ 5667.217149] usb 1-2.2: New USB device found, idVendor=0bda, idProduct=0326
[ 5667.217163] usb 1-2.2: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[ 5667.217171] usb 1-2.2: Product: USB3.0 Card Reader
[ 5667.217178] usb 1-2.2: Manufacturer: Realtek
[ 5667.217184] usb 1-2.2: SerialNumber: 201404081410
[ 5667.224976] usb-storage 1-2.2:1.0: USB Mass Storage device detected
[ 5667.225388] scsi0 : usb-storage 1-2.2:1.0
[ 5668.233608] scsi 0:0:0:0: Direct-Access     Generic- USB3.0 CRW   -SD 1.00 PQ: 0 ANSI: 4
[ 5668.246211] scsi 0:0:0:1: Direct-Access     Generic- USB3.0 CRW   -SD 1.00 PQ: 0 ANSI: 4
[ 5669.303259] sd 0:0:0:1: [sdb] 131072000 512-byte logical blocks: (67.1 GB/62.5 GiB)
[ 5669.304516] sd 0:0:0:1: [sdb] Write Protect is off
[ 5669.304526] sd 0:0:0:1: [sdb] Mode Sense: 2f 00 00 00
[ 5669.306392] sd 0:0:0:0: [sda] Attached SCSI removable disk
[ 5669.307921] sd 0:0:0:1: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA  ******** LOOK HERE
[ 5669.352415] Alternate GPT is invalid, using primary GPT.
[ 5669.352452]  sdb: sdb1 sdb2
[ 5669.356623] sd 0:0:0:1: [sdb] Attached SCSI removable disk
[ 5737.809500]  mmcblk0: p1 p2 p3 p4 p5 p6 p7 p8 p9 p10 p11 p12
[ 5742.552301] Alternate GPT is invalid, using primary GPT.
[ 5742.552333]  sdb: sdb1 sdb2
[ 5742.602219] Alternate GPT is invalid, using primary GPT.
[ 5742.602255]  sdb: sdb1 sdb2

is that the dmesg after you ran `mount` ?

you can run `cgpt repair` on the device to fix up the alternate GPT too

 vapier ,

       Whenever I clean up i start again the device (64GB sdcard) either it's usb (Adapter) or directly by the micro Sdcard creating the gpt table, partitions 1 & 2 , as you saw previously describe by the bug after I insert the MOUNT command the gpt table backup corrupts
        Even using gpt options to repair the backup table as main table as good as you see in dmesg after i insert again the command MOUNT the system CRASH my GPT TABLE in the device sdcard > 64GB 

sorry, but i don't understand what you're trying to say

please run these series of commands and copy & paste the exact terminal session

  sudo -i
  dmesg -c
  fdisk -l /dev/sdb
  dd if=/dev/zero of=/dev/sdb bs=4M count=1
  cgpt add -i 1 -b 64 -s 32768 -S 1 -P 5 -l KERN-A -t "kernel" /dev/sdb
  cgpt add -i 2 -b 65600 -s 122617759 -l ROOT-A -t "rootfs" /dev/sdb
  cgpt repair /dev/sdb
  cgpt show /dev/sdb
  sync
  partx -u /dev/sdb
  mkfs.ext4 /dev/sdb2
  mkdir -p /mnt/rar
  mount /dev/sdb2 /mnt/rar
  dmesg -c

Ok , a lot of work to prove your eyes

     I exclude the variable (USB-adapter to micro-sdcard) of this equation and running this command DIRECT to micro-sdcard reader of this machine HP Chromebook 14 G3 ---.as knew in the system dev/mmcblk1.
     Ok follow the file documment in .pdf as you wish

Deleted: MicroSDCARD high-thanorequal 64GB BUG step by step.pdf 147 KB

MicroSDCARD high-thanorequal 64GB BUG step by step.pdf 147 KB Download

i'm not sure the partition/filesystem aspects are relevant.  the kernel log & cgpt output makes it look like the card is seeing data corruption for some reason.  once you do `cgpt repair`, `cgpt show` shouldn't list the GPT entries as invalid.

are you only seeing problems with this card ?  do other cards have issues ?

what card are you using exactly ?

This card not have any problem when i use format  fat32 file system and use the same card reader HW and the same machine with chromeOS HP 14 chromebook G3 nyan-blaze 
This card is the same model and speed  and brand (SANDISK Ultraplus RED)  with 32 GB of capacity i have and  work with ext4 Filesystem with the same partitions/filesystem you call relevant using the same HW.
The difference between one and the other is the capacity of 64 GB

Follow proof about the model / cards / 

Deleted: sdcards brothers.jpg 517 KB

	sdcards brothers.jpg 517 KB View Download

being able to format it with a different FS format isn't proof that the driver/hardware is working correctly.  every FS has wildly different I/O behavior, so if ext4 formatting/mounting pushes the cards harder than fat32, then it's perfectly understandable that you'd see bugs using one but not the other.

Ok you agree the system has a bug !

i agree it's a possibility

can you file feedback (Alt-Shift-i) and in the text, mention this bug by URL/number ?  that'll gather system logs for us to look at for further debugging.

I already did

https://support.google.com/chromebook/answer/3142217?visit_id=1-636646906021333333-260581873&p=feedback_confirmation&rd=1

that URL is only valid for you.  it doesn't work for others.

i found it here (for internal users): https://listnr.corp.google.com/report/85501637240

Vapier,

    When the Status: 	Unconfirmed will change ?

Thanks !!!

#6, You ran:
mount /dev/mmcblk1 /mnt/rar

while the right command is:
mount /dev/mmcblk1p2 /mnt/rar


I found that when creating the partition table we also need to create a PMBR. I am wondering it the kernel is using an old partition table to present mmcblk1p2 device.

What works for me on 3.10 and 4.4 on a 128GB SD card:

dmesg -c
fdisk -l /dev/mmcblk1
dd if=/dev/zero of=/dev/mmcblk1 bs=4M count=1
cgpt create /dev/mmcblk1
cgpt boot -p /dev/mmcblk1
cgpt add -i 1 -b 64 -s 32768 -S 1 -P 5 -l KERN-A -t "kernel" /dev/mmcblk1
cgpt add -i 2 -b 65600 -s 250131871 -l ROOT-A -t "rootfs" /dev/mmcblk1
cgpt repair /dev/mmcblk1
cgpt show /dev/mmcblk1
sync
partx -u /dev/mmcblk1
mkfs.ext4 -E nodiscard /dev/mmcblk1p2
mkdir -p /tmp/rar
mount /dev/mmcblk1p2 /tmp/rar
dmesg -c

Deleted: screen-exchange 59.0 KB

screen-exchange 59.0 KB View Download

Hi Gwenda !!!

 Like Dummy I follow your isntructions : see 

localhost ~ # fdisk -l /dev/mmcblk1
Disk /dev/mmcblk1: 62.5 GiB, 67108864000 bytes, 131072000 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: gpt
Disk identifier: DA125270-7E88-4EF4-9792-58105EA2E0F1

Device         Start       End   Sectors  Size Type
/dev/mmcblk1p1    64     32831     32768   16M ChromeOS kernel
/dev/mmcblk1p2 65600 122683358 122617759 58.5G ChromeOS root fs
localhost ~ # dd if=/dev/zero of=/dev/mmcblk1 bs=4M count=1
1+0 records in
1+0 records out
4194304 bytes (4.2 MB, 4.0 MiB) copied, 1.53163 s, 2.7 MB/s
localhost ~ # cgpt create /dev/mmcblk1
WARNING: Primary GPT header is invalid
localhost ~ # cgpt repair /dev/mmcblk1
localhost ~ # cgpt create /dev/mmcblk1
localhost ~ # cgpt boot -p /dev/mmcblk1
00000000-0000-0000-0000-000000000000
localhost ~ # cgpt add -i 1 -b 64 -s 32768 -S 1 -P 5 -l KERN-A -t "kernel" /dev/mmcblk1
localhost ~ # cgpt add -i 2 -b 65600 -s 250131871 -l ROOT-A -t "rootfs" /dev/mmcblk1
ERROR: Entry outside of valid region
ERROR: -i 2 -l ROOT-A -b 65600 -s 250131871 -t 3CB8E202-3B7E-47DD-8A3C-7FF2A13CFCEC 
localhost ~ # cgpt add -i 2 -b 65600 -s 122617759 -l ROOT-A -t "rootfs" /dev/mmcblk1
localhost ~ # cgpt repair /dev/mmcblk1
localhost ~ # cgpt show /dev/mmcblk1
       start        size    part  contents
           0           1          PMBR
           1           1          Pri GPT header
           2          32          Pri GPT table
          64       32768       1  Label: "KERN-A"
                                  Type: ChromeOS kernel
                                  UUID: 301EF5D8-A033-104A-919F-BAFAA892E53F
                                  Attr: priority=5 tries=0 successful=1 
       65600   122617759       2  Label: "ROOT-A"
                                  Type: ChromeOS rootfs
                                  UUID: E75C54D7-9BE2-1346-8ADB-38263E17AFFD
   131071967          32          Sec GPT table
   131071999           1          Sec GPT header
localhost ~ # sync
localhost ~ # partx -u /dev/mmcblk1
localhost ~ # mkfs.ext4 -E nodiscard /dev/mmcblk1p2
mke2fs 1.43.6 (29-Aug-2017)
/dev/mmcblk1p2 contains a ext4 file system
        created on Mon Jun 18 13:15:39 2018
Proceed anyway? (y,N) y
Creating filesystem with 15327219 4k blocks and 3833856 inodes
Filesystem UUID: 6a601fa2-903b-4054-8afc-f325fe056aa9
Superblock backups stored on blocks: 
        32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208, 
        4096000, 7962624, 11239424

Allocating group tables: done                            
Writing inode tables: done                            
Creating journal (65536 blocks): done
Writing superblocks and filesystem accounting information: done   

localhost ~ # mkdir -p /tmp/rar
localhost ~ # mount /dev/mmcblk1p2 /tmp/rar
mount: wrong fs type, bad option, bad superblock on /dev/mmcblk1p2,
       missing codepage or helper program, or other error

       In some cases useful info is found in syslog - try
       dmesg | tail or so.
localhost ~ #echo ' DO USE THE RIGHT CHROMEBOOK MODEL TO TEST ?'

Deleted: text.txt 75.9 KB

text.txt 75.9 KB View Download

Guenda ,


    When the Status: 	Unconfirmed will change ?

Thanks !!!

text.txt shows a lo of IO errors on mmcblk1, try another sd card.

This error of IO start in [   19.871996] when I start put the sdcard to the card reader and stops in [   91.049385] mmc2: card 2145 removed when I remove from

I use adapter (MicroSd to USB) in   [112.636498]  sdb: sdb1 sdb2
[  112.642898] sd 0:0:0:1: [sdb] Attached SCSI removable disk
to clean up the micro sdcard.

                    AND ....
In [  493.759683] usb 1-2.2: USB disconnect, device number 4 I remove from

                        NOW ...
After that start you procedure you gave to me you see in

 [ 503.337478] mmc2: host does not support reading read-only switch. assuming write-enable.
[  503.341010] mmc2: new high speed SDXC card at address 2145
[  503.364626] mmcblk1: mmc2:2145 APPSD 62.5 GiB 
[  503.371187] Alternate GPT is invalid, using primary GPT.
[  503.371205]  mmcblk1: p1 p2
[  951.352289] JBD2: no valid journal superblock found
[  951.352302] EXT4-fs (mmcblk1p2): error loading journal

end of you procedure you gave to me... Its not sdcard !!! 

Guenda ,


    When the Status: 	Unconfirmed will change ?

Thanks !!!

To confirm is not sd card related, use a different SD card. Resend the whole output of script in #16. 

Thanks
Gwendal.

SanDisk Ultra PLUS 64 GB its not compatible with ext4 FS Linux Chromebook 3.10 for sure.. I test with Ubuntu Linux Kernel 4.4 and  when I using fat32/vfat  flying OK both Kernels!!!. I dont have resource to be testing sdcards with another Brands to discover what the kernel its compatible to run ext4 > 32GB
DAMM !!!

259/5000
Hi gwendal,

Finally I was able to buy a new sdcard for testing on chromebook chormeOS 3.10 follows the attached photo.
I performed all the script in #16 by changing the bytes of parttition 2 for the new size script.txt attached and dmesg -c in dmesg.txt
 
Its a B    .       U        .     G

When the Status: 	Unconfirmed will change ?

Thanks !!!

I spend 24.99 Dollars for this new SDCARD  brand new Sdcard.jpg

Deleted: Script.txt 3.3 KB

Script.txt 3.3 KB View Download

Deleted: dmesg.txt 113 bytes

dmesg.txt 113 bytes View Download

Deleted: Brand New SDCARD 128G .jpg 566 KB

	Brand New SDCARD 128G .jpg 566 KB View Download

The way I go around in the journal error it's format with this command:

mkfs.ext4 -E nodiscard -t ext4 -O ^has_journal /dev/mmcblk1p2