findVictim() will return an invalid block to be written to without doing an
unnecessary replacement.

On Tue, Mar 8, 2011 at 4:03 PM, Adam Jacobvitz <[email protected]> wrote:

> So how do you handle the first write into the cache before there is
> anything to replace? I'm also noticing a lot of "misses" in the caches for
> which there is no replacement event.
>
> Thanks,
>
> Adam
>
> On 3/8/2011 6:42 PM, Steve Reinhardt wrote:
>
>>
>>
>> On Tue, Mar 8, 2011 at 10:07 AM, Adam Jacobvitz <[email protected] <mailto:
>> [email protected]>> wrote:
>>
>>    Two Questions:
>>
>>    1) How are blocks written into the L1 or L2 cache when there is no
>>    block to replace? I see this thing called a "tempBlock" but I
>>    can't seem to find the code that writes it into the cache...
>>
>>
>> It isn't written in... that's the point, if there's no block to replace,
>> then you can't write it into the cache.  In practice I don't know if that
>> ever happens though.
>>
>>    2) In the M5 Cache structure (not Ruby) I can't seem to find any
>>    notion of a "way". Does the cache structure know what a "way" is?
>>
>>
>> Not really, the code is organized based on sets rather than ways.  See
>> src/mem/cache/tags/cacheset.hh.
>>
>> Steve
>>
>>
>>    Thanks!
>>
>>    --     Adam Jacobvitz
>>    Graduate Student
>>    Duke Computer Architecture Group
>>    Electrical and Computer Engineering Department
>>    Duke University, North Carolina
>>    http://arch.cs.duke.edu/
>>    _______________________________________________
>>    m5-users mailing list
>>    [email protected] <mailto:[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
>
_______________________________________________
m5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/m5-users

Reply via email to