Re: [PATCH] proc_file_read() (Was: Re: proc_file_read() question)

[Mike Galbraith]

On Tue, 26 Jun 2001, Martin Wilck wrote:

> Hi,
>
> > Shhh ;-)  Last time that hack was mentioned, someone wanted to _remove_
> > it.  It's a very nice little hack to have around, and IKD uses it.
>
> I am not saying it should be removed. But IMO it is a legitimate (if
> not the originally intended) use of "start" to serve as a pointer to
> a memory area allocated in the proc_read () function. This use is broken
> with this hack in its current form, because reading from such a file
> will fail depending on the (random) order of the page and start pointers.
>
> If I understand the "hack" right, legitimate offsets generated for it
> are always between 0 and PAGE_SIZE. Therefore the patch below would
> not break it, while overcoming the abovementioned problem, because
> legitimate page pointers will never be < PAGE_SIZE.

It's dead simple.  My variable length data often didn't quite fit into
the transport vehicle.. so I whacked the excess to avoid the fault.

	-Mike