Re: [Xen-devel] [Fwd: Re: Wondering about cirris and stdvga]

[Dario Faggioli <dario.faggioli@xxxxxxxxxx>]

> On Fri, 2016-11-18 at 15:19 -0500, Adam Jackson wrote:
> > On Fri, 2016-11-18 at 13:32 -0500, Konrad Rzeszutek Wilk wrote:
> > On Fri, Nov 18, 2016 at 06:08:57PM +0000, Ian Jackson wrote:
> > > > performance wise, that cirrus is broken and impossible to
> > > > fix (because it is the hardware that it's emulating that was
> > > > broken),
> > > > that stdvga enables better screen resolution in guests, etc.
> > > 
> > > The "broken and impossible to fix" sounds like FUD TBH.
> 
> The "fear" bit, perhaps; it is arguably frightening that we continue
> to
> emulate in 2016 a chip that was inadequate in 1996. In my opinion
> about
> cirrus being broken, however, I am neither uncertain nor in doubt.
> 
Thanks Konrad for poking Adam, and thanks Adam for chiming in so
quickly.

> > > I wouldn't want to change the default just because Wayland is
> > > broken.
> > > Wayland should be fixed.
> 
> You should want to change the default because it is bad, even for X.
> 
In principle, I'm all with Ian's, and I think that's true for pretty
much everyone.

And in fact, I was really hesitant to start a discussion like this
basing on: 1) another project's bug and 2) rumors (as Ian says, and
he's right again) and incomplete knowledge.

Then, poking around a bit more, and talking with Konrad, we had the
feeling that those that were only rumors to us (well, to me), may
actually be more than just that... which indeed seems to be what Adam
is teaching us here. :-)

> > CCing Adam Jackson who I hope can enlighten us on the technical
> > parts of Xorg.
> > 
> > I am really interested in knowing the technical merits
> > of the stdvga vs cirrus and I hope to understand what the Xorg
> > maintainer(s) have in mind.
> 
> The cirrus chip that qemu happens to emulate is trash. It can't do
> 32bpp, it can't do modes larger than 2048 wide, it can't do more than
> 4M of VRAM. We could of course extend the qemu emulation to be more
> capable; doing so would require updating every guest to know about
> these enhancements, though.
> 
> stdvga, as it has none of the above problems, is _today_ uniformly
> better than cirrus. It is still perhaps less awesome than one would
> like (no idea how multihead is expected to work for example), but
> there
> is no world in which stdvga is worse than cirrus.
> 
Ok. These seems very good reasons to me.

Then of course I'm no maintainer, so it's someone else's call...
Hopefully, we at least gathered some useful insights. :-)

Thanks and Regards,
Dario
-- 
<<This happens because I choose it to happen!>> (Raistlin Majere)
-----------------------------------------------------------------
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Senior Software Engineer, Citrix Systems R&D Ltd., Cambridge (UK)

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel