[lowrisc-dev] Re: lowRISC kernel changes rebased over stock 4.19

Bdale Garbee bdale at gag.com
Wed Oct 31 16:49:09 GMT 2018


Dr Jonathan Kimmitt <jrrk2 at cam.ac.uk> writes:

> Obviously it is a pain to keep retesting everything.

As interest in the Debian riscv port continues to grow, throwing some
manpower and resources behind an effort at continuous integration
testing here might help?  We've certainly done similar things in the
early stages of other ports in the past. 

> FPGA code as part of the "firmware" or even the "software" might 
> disagree with upstreaming.

Kinda depends on what the toolchain is for building the bitstream.  The
work on things like icestorm is really encouraging.  If we ever get to
the point where a credible device can be placed entirely using a free
software toolchain, that'd rock *my* world hard and make the whole
upstreaming thing simpler in some ways.

Sure would be nice if Xilinx would finally go all-in and help the
community build and maintain a free software toolchain for their
parts... 

> If you want a usable Linux then the tendency is to make
> the core larger and add fancy features.

I don't think trying to win a smallest design competition is something
I'd ever be interested in, but if there ends up being a difference
between the scale achievable with a completely free software toolchain
and that achievable with secret-source tools... I could easily get
excited about working on whatever the biggest, baddest design is that we
can achieve without leaving the free software space.

Regards,

Bdale
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 832 bytes
Desc: not available
Url : http://listmaster.pepperfish.net/pipermail/lowrisc-dev-lists.lowrisc.org/attachments/20181031/743caad5/signature.sig


More information about the lowrisc-dev mailing list