1

I am trying to install Linux (Debian) on a new laptop. This requires special care because the laptop keyboard needs a kernel patch (and I have not compiled a kernel for about 10 years). Also, because 32-bit UEFI.

What I did was:

  • Using a Debian netinst USB drive to be able to mount,
  • On the same drive, install a minimal Debian system (via debootstrap) on the second partition,
  • Use the functional Grub from the first partition to boot the kernel on the second one.

Up to this point, everything work. The second partition contains the custom kernel, so even my keyboard is recognized!

But, the problem is that on the second partition, I am stuck in the initramfs (in which I included a busybox shell). I loaded (manually) the required modules (sg, usb_storage, scsi_mod, libata) to see the USB drive (containing the debootstrap) and the internal hard drive (to install to). The dmesg buffer contains lines such as

[time] sd 1:0:0:0: Attached scsi generic sg1 type 0
[time] sd 1:0:0:0 [sdb] number of blocks etc.

I am typing this by hand from the screen of the not-yet-functional laptop, so I am unable to reproduce the whole log! There are also corresponding lines for sda. This hints that the /dev/sda* and /dev/sdb* block devices should be present; however, they are not, and even when I mknod them, they are not functional:

(initramfs) mount /dev/sdb2 /usb
mount: mounting /dev/sdb2 on /usb failed: No such file or directory

Also, for a simpler example, fdisk does not see /dev/sdb.

I must have missed something really simple, but what ?

Edit: Yes, I missed something really simple, which is the ext4 module. Fixed. Mod delete?

3
  • Are there no partions listed after the line with sdb? Try mkdir -p /media/sdb and then miunt -t auto /dev/sdb /media/sdb. Does that work?
    – ott--
    Sep 3, 2015 at 18:26
  • 1
    Can't you delete it by yuourself?
    – ott--
    Sep 3, 2015 at 18:30
  • 1
    VTC as off-topic: "went away when a typo was fixed" (unlikely to help future readers). The OP isn't interested in further work on the question. And there isn't a full amount of context around the confusing error message. I think it's probably not very useful to try and guess why we specifically saw ENOENT ("no such file") here. I would have expected ENODEV, which is documented as "filesystemtype not configured in the kernel" in man 2 mount.
    – sourcejedi
    Jul 25, 2019 at 8:56

1 Answer 1

0

It seems that the Debian initramfs basic image does only have the capacity to mount NTFS file-systems. If you want to check this, just do:

#> ls /sbin/mount.*

And, you should see the recognized file-systems of the initramfs image.

So, to solve your problem, you just need to find an USB key with an NTFS file-system on it.

2
  • This question turns out to have been about mounting an ext4 filesystem. Not NTFS. I voted to close the question anyway. Since the OP suggested deleting it, and I don't think we can really improve on their self-answer unless they wanted to go back and add more detail.
    – sourcejedi
    Jul 25, 2019 at 8:57
  • I totally agree with you. In fact, I am currently running into these kind of problems and I found the question (without answer) more misleading than helping, so I added my answer to avoid to mislead someone else. But, closing it is also a perfectly valid way of doing (in my humble opinion).
    – perror
    Jul 25, 2019 at 9:08

Not the answer you're looking for? Browse other questions tagged .