jffs_mknod: bogus imode when creating sockets

jffs_mknod: bogus imode when creating sockets

[David Vrabel]

Hi,

Creating sockets on JFFS results in
    jffs_mknod: bogus imode (nnnn)
messsages.

In jffs_mknod() in inode-v22.c the code that checks if the inode is a
socket has been commented out.  Everything works fine as it is can I
just uncommented out the socket check?

David Vrabel

ps. What's a imode? :->



To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org

Re: jffs_mknod: bogus imode when creating sockets

[David Vrabel]

David Vrabel wrote:

> Creating sockets on JFFS results in
>     jffs_mknod: bogus imode (nnnn)
> messsages.
>

Never mind.  Thinking about it some more (should do this more often). 
I've realized something else is not setting the mode correctly for
sockets (I'm sure it shouldn't be 140755 but 6).  Hmmm... shouldn't the
socket check be in there anyway?

> ps. What's a imode? :->

Another lack of clues here... Obviously it means inode->i_mode not a
misspelling of inode...

David Vrabel



To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org

Re: jffs_mknod: bogus imode when creating sockets

[David Woodhouse]

dvrabel@arcom.co.uk said:
>  Never mind.  Thinking about it some more (should do this more often).
>  I've realized something else is not setting the mode correctly for
> sockets (I'm sure it shouldn't be 140755 but 6).  Hmmm... shouldn't
> the socket check be in there anyway?

No, 140755 is correct for a rwxr-xr-x socket. And yes, the socket check 
should be there, but it's harmless as it doesn't affect the operation, only 
makes it whinge when it shouldn't. You're not actually getting any 
behavioural problems, are you?

--
dwmw2




To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org