> On 2010-10-15 17:33:04, Steve Reinhardt wrote:
> > The first sentence of the commit message is pretty confusing 
> > (self-contradicting if you read it wrong).
> > 
> > I don't object to this particular change, but isn't it just working around 
> > a design flaw in TimingSimpleCPU, and wouldn't it be better to fix that 
> > instead?
> > 
> > Also, assuming the answer there is no, shouldn't the same fix be applied in 
> > the x86 table walker too?
> 
> Ali Saidi wrote:
>     I think it's pretty reasonable that the table walker will take a cycle 
> after it gets the required data to provide it to the CPU. I don't know that 
> we want to convolute the code in the timing "simple" cpu model anymore.

Yea, my hope was that we might find a way to fix this while simplifying 
TimingSimpleCPU... I agree, I have no desire to make it more complicated.


- Steve


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviews.m5sim.org/r/257/#review379
-----------------------------------------------------------


On 2010-10-02 19:20:00, Ali Saidi wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://reviews.m5sim.org/r/257/
> -----------------------------------------------------------
> 
> (Updated 2010-10-02 19:20:00)
> 
> 
> Review request for Default.
> 
> 
> Summary
> -------
> 
> ARM: Don't return the result of a table walk the same cycle it's returned.
> 
> The L1 cache may have been accessed to provide this data, which confuses
> it, if it ends up being accesses twice in one cycle. Instead wait 1 tick
> which will force the timing simple CPU to forward to its next clock cycle
> when the translation completes.
> 
> Also prevent multiple outstanding table walks from occuring at once.
> 
> 
> Diffs
> -----
> 
>   src/arch/arm/table_walker.hh e78b6bba67ca 
>   src/arch/arm/table_walker.cc e78b6bba67ca 
> 
> Diff: http://reviews.m5sim.org/r/257/diff
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Ali
> 
>

_______________________________________________
m5-dev mailing list
[email protected]
http://m5sim.org/mailman/listinfo/m5-dev

Reply via email to