[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: enlarging a partition

From: pgs
Subject: Re: enlarging a partition
Date: Wed, 26 Apr 2006 15:53:29 +0200
User-agent: Mutt/

Hi leslie,

I've come back to my work for which I mailed this list some weeks ago.
I have remade some tests and looked deeper into the code.

First of all, I have resolved my problem by issuing a call to lilo and a reboot 
after having enlarged my partition with parted.
I believed it didn't work with the kernel that I am using but it 
seems that I made a mistake in my previous tests, and in fact it does work.
So I do with a reboot, which was not exactly what I first wanted to do but 
finally I can accept to do like this.

I also figure out much better why I am confronted to this problem :
in fact, the kernel doesn't want to take into account a change made on a 
partition which is in use because it considers that the partition could have 
been deleted which would generate problems.
And what made me thinking the opposite was because I confused between "a 
mounted partition" and a "partition in use".
I believed that if the partition was unmounted (with an umount -l in 
occurence), the kernel accepted to reflect the modification, but it's not true, 
the partition can be unmounted and still in use, which is the case for me.
So it could probably be arranged by patching the kernel to add a new action to 
the ioctl BLKPG allowing to reflect the changes on an enlarged partition, but 
this costs some work and I cannot afford doing it now.

So thank you a lot for your help, don't hesitate if you have any question or 
suggestion about my patch.


address@hidden a écrit, le Wed 19 Apr 2006 à 06:55:55PM :
> Hi there,
> On Fri, Apr 14, 2006 at 05:38:36PM +0200, Pascal GREGIS wrote:
> > HOwever, if you just can tell me if my patch is correct or not, it
> > would always be helpfull.
> I can't see any problem with your patch.
>   But if it helps, this might be a kernel bug that surfaced earlier and
> could be solved by running "partprobe" after each boot.  So you might
> call it somehow (initrd, small root on another disk) before attempting
> to mount the problematic partitions.
>   Let me know if I can provide further help.
>     Leslie
> -- 
> gpg --keyserver pgp.mit.edu --recv-keys 0x52D70289
> http://nic-nac-project.de/~skypher/

reply via email to

[Prev in Thread] Current Thread [Next in Thread]