|
Message
From: Matjaz Breskvar<phoenix@o...>
Date: Fri Feb 11 15:19:21 CET 2005
Subject: [openrisc] or1ksim stable_0_1_0
Hi all!The first series of Nog patches are now in cvs and tested. Since there
will be more drastical changes comming I tagged this version of or1ksim with
stable_0_1_0, and also created branch stable_0_1_x that will be used
to track any bugfixes. The HEAD trunk will be used for development and may
not be stable all the time.
The most visible change is 15%-20% faster simulation times:
or1ksim: 11. 02. 2005
Exception 0x00000900 (Data TLB Miss) at 0xc004c978, EA: 0xc1d88110, ppc:
0xc004c978, npc: 0xc004c97c, cycles 2650520152, #109247582
exit(-1042613780)
@reset : cycles 0, insn #0
@exit : cycles 2650534782, insn #109248113
diff : cycles 2650534782, insn #109248113
real 4m56.850s
user 4m13.323s
sys 0m5.595s
or1ksim: 24. 12. 2004
Exception 0x900 (Data TLB Miss) at 0xc004c978, EA: 0xc1d88110, ppc:
0xc004c978, npc: 0xc004c97c, cycles 2650520152, #109247582
exit(-1042613780)
@reset : cycles 0, insn #0
@exit : cycles 2650534782, insn #109248113
diff : cycles 2650534782, insn #109248113
real 5m51.364s
user 5m7.375s
sys 0m5.301s
Nog i looked at your patches and you may commit all of the
peripheral cleanups.
Regarding FAST_SIM i'd really be interested in some numbers. Does it
really speed up execution by some meaningful margin (As far as i know
faster simulation time is the only it exists) ? If it's
negligable i'm all for throwing that one out.
best regards,
p.
|
 |
Browse
