linuxppc-dev.lists.ozlabs.org archive mirror
 help / color / mirror / Atom feed
* Why all the SAVE_ macros in transfer_to_handler (head.S)?
@ 2000-08-08  0:36 William Kucharski
  0 siblings, 0 replies; only message in thread
From: William Kucharski @ 2000-08-08  0:36 UTC (permalink / raw)
  To: linuxppc-embedded


This is likely a FAQ, but does anyone know of any particular reason
transfer_to_handler() (in head.S) uses the three SAVE_ macros (SAVE_4GPRS,
SAVE_8GPRs, SAVE_8GPRS) rather than just the single instruction

    stmw  r8,(GPR8)r21

Yeah, I know the green book says this can have greater latency and take MUCH
longer to execute than separate instructions, but I've yet to find a CPU for
which this is actually true.

So, the question stands, anyone know the original rationale?  Historical
reasons?  To avoid case(s) in which it is true?  To avoid the POWER heritage of
the instruction?  Coding style differences?

					Thanks,
					    William Kucharski
					    wkucharski@lucent.com

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2000-08-08  0:36 UTC | newest]

Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2000-08-08  0:36 Why all the SAVE_ macros in transfer_to_handler (head.S)? William Kucharski

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).