Re: generic_file_write code segment in 2.2.18

[Andi Kleen]

On Thu, Jan 04, 2001 at 02:58:58PM -0800, Asang K Dani wrote:
> --- Andi Kleen <[e-mail]> wrote:
> > On Thu, Jan 04, 2001 at 10:42:34PM +0000, Stephen C. Tweedie wrote:
> > > No, because then you'd be skipping the updatepage() call if we
> > took a
> > > fault mid-copy after copying some data.  That would imply you had
> > > dirtied the page cache without an updatepage().
> > > 
> > > The current behaviour should just result in a short IO, which
> > should
> > > be fine.
> > 
> > The problem is that the short write is not reported to the caller,
> > even when only zero bytes are copied (the page is corrupted anyways
> > though because cfu  zeros the uncopied rest).  
> 
> I think it will be reported to caller, because when cfu copies 0
> bytes,
> 
>    bytes -= copy_from_user(dest, buf, bytes);
> 
> will make 'bytes' zero. Since 'bytes' is 'zero' updatepage will not
> be called and status retains value '-EFAULT' and it breaks out of
> the while loop immediately.

Right for zero it is handled, but not for 1 byte copied but the rest zeroed
(which is a severe IO error) 

-Andi