Actually the cache will block as soon as the MSHRs are full; see
allocateBufferInternal() in src/mem/cache/base.hh:
if (mq->isFull()) {
setBlocked((BlockedCause)mq->index);
}
Steve
On Tue, Feb 1, 2011 at 8:45 PM, Sheng Li <[email protected]> wrote:
> Hi Guys,
>
> I am a newbie to M5 and studying different levels of cache lockup, say
> allow cache hits to be severed when there is 1 miss, 2 misses, and n misses.
> The MSHR implementation in M5 is a great starting point. I looked at the
> code, and it seems the cache will accept new requests when the MSHR is full,
> as long as the new request is a cache hit. Only when the new request is a
> cache miss and the MSHR is full, the cache access will block. Am I right on
> this? Could anyone give some suggestions If I need to change the cache/MSHR
> so that the cache will lock up when the MSHR is full no matter whether the
> next access is a hit or not?
>
> Thanks!
> -Sheng Li
>
> _______________________________________________
> m5-users mailing list
> [email protected]
> http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
>
_______________________________________________
m5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
