Messages in this thread | | | Date | Mon, 24 Aug 1998 09:59:35 -0700 (PDT) | From | Linus Torvalds <> | Subject | Re: pre egcs-1.1 testing and Linux 2.1.x |
| |
On Mon, 24 Aug 1998, Philip Blundell wrote: > > >The regparm thing has been singled out because it makes it easier to show > >the bugs, not because of any technical reasons. And that's why it's so > >wrong to disable it, and hiding the real problem. > > Precisely, and my objection to Andrea's patch was that it just disabled more > optimisations, again to hide the bug.
Note that disabling "regparm" does not fix the bug - it just makes it harder to trigger. Much harder. But you can still make it happen.
Andrea's patch makes the bug go away completely, and doesn't remove any user-visible features.
As such, Andrea's patch is, on purely technical grounds, the _much_ more correct thing to do.
And thus I maintain that _if_ something needs to be disabled, then Andrea's patch is the right thing to do. Disabling "regparm" just because it happens to show the bug is bogus. Andrea's patch goes to the heart of the matter and disables the bug itself, even though obviously it does so with a sledgehammer.
Obviously, the _best_ solution is to just fix the bug in the first place, and I don't think you'll get any argument about that from anybody. But David Miller and others have claimed that an "interim" solution would be to disable regparms, and I'm just telling you that the much better "interim" solution is the one that Andrea suggested.
Yes, Andrea's patch will also just hide the bug, but by God, it at least does it properly, instead of the cop-out that people have been suggesting.
So no, I certainly hope his patch won't be needed, but as an alternative to disabling "regparm" it makes much more sense.
Linus
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html
| |