On Mon, Apr 14, 2014 at 3:47 PM, Stefan Wallentowitz <[email protected]> wrote: > No, of course it would be nice if you pull it after review.
I didn't really find anything to complain about, so I've applied the patches with only a small change to a comment referring to "OPTIMSOC_I/DCACHE_WAYS". > I will also finalize the smal change of the multicore core identifier > including or1ksim and or1k-src extensions as soon as possible. > Afterwards I will integrate the snooping cache (still working on a nice > LSU solution), so maybe we can reuse parts of this for the LL/SC > implementation you proposed recently. > Yes, I think the snooping unit for cache coherency and atomicity should be able to share a lot of logic. I've actually added the LL/SC functionality to mor1kx, but I'm holding that off until we are sure about the details. If you're going to poke around the LSU, you might want to base your work upon it, I don't think the final version will be a lot different. https://github.com/skristiansson/mor1kx/commit/3f77e56e8e89ac1de2b77c996d0104ccae69b52a Stefan _______________________________________________ OpenRISC mailing list [email protected] http://lists.openrisc.net/listinfo/openrisc
