[libre-riscv-dev] TLB Update

[Luke Kenneth Casson Leighton]

---
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68

On Sun, Feb 10, 2019 at 7:46 AM Daniel Benusovich
<flyingmonkeys1996 at gmail.com> wrote:
>
> Evening,
>
> I have been working on the TLB and would like to store what I have
> somewhere in the repository just in case. Is there anywhere in particular
> it should go?

git clone gitolite3 at libre-riscv.org:soc.git (or,
ssh://gitolite3@libre-riscv.org:922/soc.git if you haven't added "Host
libre-riscv.org Port 922" to .ssh/config)

> Also I am thinking about doing a two set associative for the L1 with 8
> entries at the moment but am shopping for better ideas. Nothing yet on the
> L2 of the TLB.  I think that anything is better than nothing until we can
> simulate at least a portion of the types of instructions we will commonly
> see to test the TLB before making any permanent decisions.

 well, if you keep it flexible (parameterised), things can be adjusted
as needed.

> I have been reading that 128 entries are of a decent size if we want to
> upgrade. I read that in a paper found here: Paper
> <https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=10&ved=2ahUKEwjuzPay1LDgAhWTPH0KHUBIBCQQFjAJegQICRAC&url=https%3A%2F%2Fwww.cs.rutgers.edu%2Fresearch%2Ftechnical_reports%2Freport.php%3Freport_id%3D703%26series_id%3D1&usg=AOvVaw2My8IXotmCGVV9b4D01_Id>

 will take a look.

> Most of the paper talks about where the TLB should be placed in a GPU and
> the number of ports a TLB can use before becoming ineffective. Just
> something to chew on and look through, I certainly have been. Let me know
> what you think and if any changes should be made once the TLB goes up.

 well, one thing for sure, we will need some debug / log / analysis
reporting, so that things can be checked during simulations.

> Hope you guys are having a good weekend!

 watching "Need for Speed" from 2014... great film :)

l.