Re: [2.6] smbfs & "du" illness

[Jeremy Allison]

On Sat, Sep 25, 2004 at 12:20:20PM -0700, Linus Torvalds wrote:
> 
> If the number doesn't make sense, then we shouldn't look at it. The 
> _only_ thing that number makes sense for is st_blocks as far as the Linux 
> client is concerned. 
> 
> In other words, the Linux client is a hell of a lot better off just taking 
> "(filesize + 511) >> 9", as far as I can tell. It's more accurate than the 
> random number you have.

At present this is true, but just because our implementation of the spec
is broken (and yes it's complicated by the fact that we're the only
ones implementing the server side of this at the moment) doesn't mean
that the client should depend on this.

After all, now I know about it I'll fix it for the next release
and eventually modern clients and servers will be consistent on
this issue. But if you want to have a fallback hack that does
this at present then that's fine by me. It'd be nice if it was
temporary though.

Jeremy.
-
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/