On Fri, Aug 26, 2005 at 05:54:15PM +0100, Paul Jakma wrote: > On Thu, 25 Aug 2005, Uwe Schindler wrote: > >This is a software problem that lies in the ALPHA RISC processor > >architecture. The processor cannot access memory that is not > >aligned at boundaries conforming to the datatype to be > >read/written. > Urmm, don't you mean: > "It's a software problem which lies in the laxity of the x86 > architecture" > x86 being one of the /few/ architectures which doesn't fault on > unaligned accesses. Though, even on x86 they still are /slower/ than > aligned accesses (the CPU is doing the fixups in hardware rather than > trapping). > SPARC definitely traps unaligned. PPC I /think/ does too. > Just Linux which generally tries to cover up for bad software. > >because alpha is not the main platform and most developers do not > >have such a processor. > Try PPC. Unless they're running a fully ppc64 system (which is contraindicated by kernel and glibc upstream), they're not going to get the same unaligned traps on ppc as on alpha. Anyway, I've never seen trap-related errors from ppc; the other arch that tends to generate these a lot is ia64, which rivals sparc for least-efficient fix-ups. -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. vorlon@debian.org http://www.debian.org/
Attachment:
signature.asc
Description: Digital signature