Tiny Core Linux
Tiny Core Base => TCB Talk => Topic started by: jls on September 25, 2009, 04:35:31 AM
-
:o
-
It is run with the 'p' flag (automatic repair, no questions asked)
Now you don't want that?
-
on my ext2 formatted filesystems sometimes -p fails. Should I change to ext3?
-
In my opinion ext3 would be even more intense, as it wants to write a journal.
If my disk was failing, aparently, beyond repair I would look to replace it.
-
but e2fsck on my ext2 partion without any option passed, apart form the device of course, works.
So I'm suggesting to drop to the shell instead of auto reboot the machine in this case
-
I agree, if the file system check fails, the failure should at least be shown to the user. Perhaps wait for a keypress in a failure case. I'm not sure what can be done in a shell if a -p check fails though, so just showing the error might be enough.
-
the error is shown for a sec already.
-
Yes, but if the user is not watching the screen, it could get missed.
-
tc-config could be pretty simply made to reboot only on success of e2fsck, if that's a desired option.
EXIT CODE
The exit code returned by e2fsck is the sum of the following conditions:
0 - No errors
1 - File system errors corrected
2 - File system errors corrected, system should
be rebooted
4 - File system errors left uncorrected
8 - Operational error
16 - Usage or syntax error
32 - E2fsck canceled by user request
128 - Shared library errorNot sure if all of these apply in TC; this is from the man page.
-
Alright already, I will add a pause upon non-zero result.
-
Not to be an annoy, but 1 and 2 look like successful returns as well. I don't really understand what they mean by "the sum of the following conditions", though.
-
Alright already, I will add a pause upon non-zero result.
pause means that when the user presses any button the system will be rebooted?
If yes then I don't agree, I think the system should drop into a shell to let the user redo the e2fck without the -p option
-
And a new user would know to do that with the cursor sitting at a shell prompt?
Perhaps I should re-run e2fsck without the p option? Then I will get a complaint that the user has to answer y hunderds of times.
-
I will make a pause with option to go to shell. But it is a very limited shell.
-
I will make a pause with option to go to shell. But it is a very limited shell.
That might work; the last few times I ran as a boot option, it couldn't fix the inode problems (maybe it was some FS handle issue.) Running e2fsck on the unmounted device fixed it without issue. (I don't know if that was because of the 'p' option).
-
On second thought, perhaps I will just drop this option all together.
It is really not needed as booting with base norestore provides a fully booted base system no drives mounted. You can then pick and choose whatever and which ever and run as many times e2fsck as desired. This option is a carry over from compressed knoppix image booting days and is really not required when the base system loads entirely into ram.
-
Now that you can load the *core* extensions and wm and "startx" with microcore after booting with "base norestore", it's as easy to run e2fsck as use the "checkfs" boot code.
Is there a way to invoke the boot routine to load all extensions in /tce after "base norestore" rather than "tce-load -i" one-by-one?
-
not at present. I usually type loop command at prompt to batch load.
-
On second thought, time to implement it.