Messages in this thread | | | Date | Fri, 26 Nov 1999 23:32:01 +0100 (CET) | From | Urban Widmark <> | Subject | Re: spin_unlock optimization(i386) |
| |
On Fri, 26 Nov 1999, Ingo Molnar wrote:
> hm, then why does causal.c trigger a causality violation on a 8-way box? > Interestingly it's getting triggered if there are two copies (4 threads > total) running. [which could be simply just randomness added to the system > bus load and thus statistically finds the window]
[snip]
> note that this cannot be reproduced on Andrea's dual-PII box (neither on > my dual-PIII box), but can be reproduced on my 8-way Xeon box, if two > copies of the above are running. [maybe if i added delay timings to the > above then we could trigger it on dual boxes too] 'two copies' means two > independent pairs of the above.
I got something like this:
<1943260> 7223479 996367527 <thread1> BROKE causality! Weakly ordered memory? <15802145> 12115409 998555500 <thread1> BROKE causality! Weakly ordered memory? <3699222> 4779261 999405279 <thread1> BROKE causality! Weakly ordered memory?
I was running a handful 'causal' and I was doing other stuff as well at the same time. (The reason for running so many was that I didn't really check what it did, so I started a few thinking that they had completed without finding anything ...
history: 3 20:17 t causal.c 4 20:17 gcc -O2 -o causal causal.c "normal compile" 5 20:18 ./causal 0 2 6 20:18 ./causal 0 2 7 20:18 ./causal 0 2 8 20:18 ./causal 0 2 9 20:18 ./causal 0 2 10 20:18 ./causal 0 2 11 20:18 ./causal 0 2 12 20:18 ./causal 0 2 13 20:18 ./causal 0 2 14 20:18 ./causal 1 2 15 20:18 ./causal 1 2 16 20:18 ./causal 1 2 17 20:18 ./causal 1 2 18 20:18 gcc -o causal causal.c "maybe it was the -O2 ?" 19 20:18 ./causal 0 2 20 20:18 ./causal 0 2 21 20:18 ./causal 0 2 22 20:18 ./causal 0 2 23 20:18 ./causal 0 2 24 20:18 ./causal 0 2 25 20:18 ./causal 0 2 26 20:18 ./causal 1 2 "still nothing fun happens ..." 27 20:18 t causal.c 28 20:19 t causal.c "read, but not understood" 29 20:21 rm causal* "remove some junk" 30 20:31 killall causal "hmm, system is much less responsive than normal ... oh, they were all still running" )
This on a dual PIII-500.
/Urban
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |