Subject: Re: Kernel copyin/out optimizations for ARM...
To: David Laight <firstname.lastname@example.org>
From: Richard Earnshaw <email@example.com>
Date: 03/15/2002 09:54:15
> > I've written a new copyin this afternoon. 1 byte xfers are
> > 15 instructions - 2 jumps, no register saves.
> ok, copyin, copyout and kcopy all written.
> kcopy is tested, the other should just work - code is the same.
> see www.l8s.co.uk then netbsd/bcopyinout.S
> I#m out till tomorrow - anyone feeling brave?
Looks pretty good, though I haven't tried it.
One thing I've noticed by inspection is that we are using ldr(t) + masking
for in_short. This may start to fail on future kernels 'cos we may well
turn on strict alignment checking. Can you fix that bit to use ldrb(t)
plus an orr?