Subject: Re: Kernel copyin/out optimizations for ARM...
To: David Laight <firstname.lastname@example.org>
From: Richard Earnshaw <email@example.com>
Date: 03/12/2002 11:57:25
> On Tue, Mar 12, 2002 at 10:38:39AM +0000, Richard Earnshaw wrote:
> > I've always wondered why that code wasn't written using ldrt/strt for the
> > user-space accesses. That would then use the hardware for permission
> > checking and eliminate the most expensive part of that code (doing the tlb
> > check manually).
> not in ARM v3?
Nope; even the ARM2 had these IIRC.
> > Maybe there is some reason why that would not work (wrong page tables
> > mapped?)
> Isn't there a problem that the cache is checked before the TLBs?
> Maybe bound checking the address is enough?
I don't think so. The instructions are defined to work as though a user
privilaged process had run the instruction.
> Why did the 'move from/to previous address space' instructions that
> the pdp-11 have disappear from most later systems?
Never used a pdp-11.... I'm not THAT old :-)