Re: Suspected bug infilesystems (UFS,ADFS,BEFS,BFS,ReiserFS) related to sector_t being unsigned, advice requested

From: Oleg Drokin
Date: Tue Jan 06 2004 - 18:54:50 EST


Hello!

On Wed, Jan 07, 2004 at 02:37:20AM +0300, Hans Reiser wrote:

> >This code was never executing anyway.
> Oleg, I thought you ran a script for finding dead code last fall or
> summer? Any idea why it didn't find this and gcc did? Or did you only
> run it on reiser4?

Actually I found this dead code back then (with gcc as well), though
it was not looked all that serious. I think I decided we may want to have that
just in case sector_t will become signed oneday or something like that.
(in 2.4 the block type is still signed long, for example).

As for why gcc is finding this, but scripts (e.g. smatch) do not is because
scripts generally know nothing about variable types, so they cannot tell
this comparison was always false (and since gcc can do this for long time
already, there is no point in implementing it in scripts anyway).

Bye,
Oleg
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/