[BusyBox] Re: [BusyBox-cvs] svn commit: trunk/busybox: include networking

[Rob Landley]

One more try:

On Thursday 28 July 2005 12:23, Ladislav Michl wrote:

> Why does such check exist at all? It seems pretty broken to me. Imagine
> you want to do 'mount -t jffs2 mtd1 /mnt' while in /dev directory.
> Device name gets expanded to /dev/mtd1, which is character device and
> mount will fail.

Okay, going back to your original complaint to try to confirm exactly what the 
problem is, after having reviewed the appropriate section of my mount code to 
see if my mount rewrite will just magically fix this problem:

We need to stat the sucker because if it's a block device, a file, or a 
directory (or a symlink pointing to any of those), we want the absolute path.  
(For a block device, things like /etc/mtab can't have a relative path.  For a 
filename, we record where we attached the loop device as an absolute path.  
For a directory, mount --bind and --move should also record an absolute 
path.)

Now my code will pass through the name unmodified if it can't stat it (for 
comments on -t proc and -t ramfs and such).  But that's not the problem 
you're having.

The problem you're having is there _is_ something with the filename you 
indicated (a char device in the current directory), and mount thinks that's 
what you're trying to use.

Does this seem to be an accurate summary of the problem?

Rob