On Thu, 8 May 2003, Jens Axboe wrote:
> On Thu, May 08 2003, Linus Torvalds wrote:
> > On Thu, 8 May 2003, Jens Axboe wrote:
> > >
> > > Maybe a define or two would help here. When you see drive->addressing
> > > and hwif->addressing, you assume that they are used identically. That
> > > !hwif->addressing means 48-bit is ok, while !drive->addressing means
> > > it's not does not help at all.
> >
> > Why not just change the names? The current setup clearly is confusing, and
> > adding defines doesn't much help. Rename the structure member so that the
> > name says what it is, aka "address_mode", and when renaming it you'd go
> > through the source anyway and change "!addressing" to something more
> > readable like "address_mode == IDE_LBA48" or whatever.
>
> Might not be a bad idea, drive->address_mode is a heck of a lot more to
> the point. I'll do a swipe of this tomorrow, if no one beats me to it.
Good idea.
> > (Anyway, I'll just drop all the 48-bit patches for now, since you've
> > totally confused me about which ones are right and what the bugs are ;)
:-)
> I think we can all agree on the last one (attached again, it's short) is
> ok. The 'only use 48-bit when needed' can wait until Bart gets the
> taskfile infrastructure in place, until then I'll just have to eat the
> overhead :)
Okay for me.
btw. Jens, do you have any benchmarks of using 1MiB requests
and/or removing 48-bit overhead?
-- Bartlomiej- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu May 15 2003 - 22:00:28 EST