Re: Update - Re: Open Sourced Glide on Linux PPC

2 messages, 2 authors, 1999-12-28 · open the first message on its own page

Re: Update - Re: Open Sourced Glide on Linux PPC

From: Peter Chang <hidden>
Date: 1999-12-27 16:50:13

Geert Uytterhoeven [off-list ref] writes:
On Sat, 18 Dec 1999, Kostas Gewrgiou wrote:
quoted
On Fri, 17 Dec 1999, Jens Ch. Restemeier wrote:
quoted
- Glide compiles, after disabling the assembler triangle setup code.
quoted
From the documentation I think that P6FENCE should avoid instruction
reordering between I/O accesses. From the information here I think
eieio (sp?) it the PPC instruction to use.
  Yes if you look at the macos defines you see that they use eieio there
they seem to do lots of read/writes without ever doing an eieio but since
it is working in macos i am not too worried.
This sort of depends on teh hw that you're talking to on the other
side. If it is v2 or better then it will take care of doing 'hole
counting' for you so that you don't need explicit fences to make the
hw happy. If you're doing this for sst1 or sst96 then you will need to
do fences so taht the command register on gets written after all of
the command data has been settled.
If you use the *mb() macros instead of eieio(), you don't have to use #ifdefs
and it'll work automagically on e.g. Alpha as well.
This I know less about, but these might fence more often than strictly necessary?

\p
---
The most vulnerable and yet most unconquerable of things is human
vanity; nay, through being wounded its strength increases and can grow
to giant proportions. --- Friedrich Nietzsche

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

Re: Update - Re: Open Sourced Glide on Linux PPC

From: Geert Uytterhoeven <geert@linux-m68k.org>
Date: 1999-12-28 12:35:57

On 27 Dec 1999, Peter Chang wrote:
Geert Uytterhoeven [off-list ref] writes:
quoted
If you use the *mb() macros instead of eieio(), you don't have to use #ifdefs
and it'll work automagically on e.g. Alpha as well.
This I know less about, but these might fence more often than strictly necessary?
wmb() is a barrier for writing.
rmb() is a barrier for reading.
mb() is a barrier for both and thus the `safest'.

The main advantage is that these are defined for all architectures, unlike
eieio().

Gr{oetje,eeting}s,
--
Geert Uytterhoeven -- Linux/{m68k~Amiga,PPC~CHRP} -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help