On 04/16/2013 09:31 AM, Asher Baker wrote:
On Tue, Apr 16, 2013 at 4:44 PM, Ook <ooksser...@zootal.com> wrote:
movq is supposed to be an MMX instruction, and this cpu supports MMX so I'm
not sure why it is dying here. It may be that what the dissembler calls
"movq" is in fact be an SSE2 instruction, and the cpu chokes on it because
of that.
There are separate MMX and SSE2 variants of it, and the fact it's
using the xmm registers means that it is indeed the SSE2 one.
http://ref.x86asm.net/coder32.html#x0F78
Nice, thank you! So there is our smoking gun, and the definitive answer
to the SSE2 question. No more running our servers on Socket-A boards.
Judging by the screaming on this list (or in this case the lack there
of, am I really the only odd man out?) I don't suppose Valve can be
convinced to turn SSE2 off, even for old mods like OP4?
The end result is that my OP4 servers, which I've run for....ten years?
Has it really been that long? ... are dead in the water until I get
around to forking out the bucks for some hardware upgrades. What with
Momma wanting a new fur coat and all of that, it might be a while, I
have four of these socket-A boards but all my newer gear is in use atm...
_______________________________________________
To unsubscribe, edit your list preferences, or view the list archives, please
visit:
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/hlds_linux
