[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] fpu/softfloat.c licensing
From: |
Peter Maydell |
Subject: |
Re: [Qemu-devel] fpu/softfloat.c licensing |
Date: |
Tue, 3 Jun 2014 12:42:38 +0100 |
On 3 June 2014 12:38, Alexey Kardashevskiy <address@hidden> wrote:
> On 06/03/2014 07:45 PM, Peter Maydell wrote:
>> On 3 June 2014 04:52, Alexey Kardashevskiy <address@hidden> wrote:
>>> Is there anything I can do to help with this? Chase someone down? :)
>>
>> Confirming that your lawyers are happy with the approach
>> Anthony proposed to take to fixing this would probably
>> be useful :-)
>
> Let me refresh. Anthony suggested this -
> http://marc.info/?l=qemu-devel&m=136725946312880&w=4
>
> To complete this, we need ack from a lot of people and three of them (
> Fabrice Bellard, Jocelyn Mayer, Thiemo Seufer) have not ack'ed yet so we
> either have to revert their changes (mwahaha) or we are stuck here.
Essentially, yes. In particular it's not clear to me
exactly how "clean room" our reimplementation of the
bits of code we need to revert and rewrite has to be.
Also is it sufficient to say "I started with the -2a
codebase and applied a lot of patches to it, and ended
up with a git commit which looks like "just change the
comment style", or does our git history really need to
look like "remove the -2b code, apply the -2a code,
reapply intervening changes"? The former has the
benefit of not breaking bisection, but if you look
at the commit it looks a bit odd that we can remove
the -2b licensing taint without actually touching most
of the lines of code in the files...
thanks
-- PMM