Date: Mon, 13 Apr 2020 16:22:34 -0700 From: Chris <bsd-lists@BSDforge.com> To: Pete Wright <pete@nomadlogic.org> Cc: FreeBSD X11 <x11@freebsd.org>, Niclas Zeising <zeising@freebsd.org> Subject: Re: Ars Technica article Message-ID: <5d7dd8e81ae934adf7dfdf389faa5d15@udns.ultimatedns.net> In-Reply-To: <9fb362dd-0b3b-2c09-8ae9-26167f9d42ff@nomadlogic.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 13 Apr 2020 14:40:35 -0700 Pete Wright pete@nomadlogic=2Eorg said > =C2=A0What direction change are you talking about? > > As alluded to earlier; the importation of so much Linux code=2E On one > > hand; yes it shortens the time-to-implementation=2E But in the broader > > scope; it's more work (and time) in the long term for it's removal, > > and replacement -- assuming that day ever arrives=2E >=20 > this misses the key point that there is literally *zero* people being=20 > paid full-time to implement graphics drivers for FreeBSD, whereas at=20 > both Intel and AMD developers are being paid to develop drivers for the= =20 > linux kernel=2E=C2=A0 They are also getting access to documentation and oth= er=20 > resources on how these chips are implemented which I am not certain we=20 > have access to either=2E ??? why ??? This has *always* been what's said in response to th(is|ese) topics=2E Look I'm not saying this to be a bitch/whiner/troll, and I already have enough bikesheds to last a lifetime, thank you=2E But why do we not have at *least* the documentation? Is this something I can purchase, and *donate* to the project? Is it so cost prohibitive? If so, can a pool be created to acquire the needed funds? >=20 > as such it seems like a good opportunity for us to leverage this work=20 > that is being done for the linux kernel (warts and all) to get better=20 > coverage to modern GPU's on FreeBSD=2E >=20 > > The Kpi is also a kludge, and with it comes a performance hit=2E > How is it a kludge, and what is the performance hit in real numbers? It's *perceptible*=2E I don't need to cobble up a script, or install a benchmark tool to tell be when my finger is burnt, and I don't need one to see the impact, when I experience it=2E :) I meant *l*Kpi above; BTW=2E :) > Regarding perf numbers there is no data to back this up because there=20 > has not been enough work to get the testing & benchmarking suites=20 > working in a reliable state on FreeBSD=2E See (just) above=2E >=20 > As a counterpoint, I periodically run OpenBSD which as gone in a=20 > different direction of implementing their own drivers for i915=2E=C2=A0 I w= ould=20 > say subjectively the performance with their implementation is several=20 > orders of magnitude less performant than FreeBSD's - but you know what,= =20 > that is OK!=C2=A0 They have different objectives and approaches which is= =20 > totally healthy IMHO=2E Agreed=2E Who doesn't like having choices? It also helps add for perspective=2E >=20 > We just need to be honest that their are trade offs that will be taken=20 > with either approach - and perf is one of the most obvious and noisy area= s=2E Sure, and *that/those* arguments almost *always* end up in much hair-splitting -- very *non* productive=2E >=20 > > > > Is there really that little interest in the Graphics area/dept=2E that > > what we've currently been using couldn't be sustained/improved? > > >=20 > I would say yes! >=20 > Until this work began we had support for older i915 graphics but that=20 > development had stalled while hardware most definitely had *not*=20 > stalled=2E=C2=A0 The situation that we are at now is a direct result of thi= s -=20 > someone stood up and got things working, entropy took over and support=20 > was added and improved=2E >=20 > There is nothing preventing others from standing up and implementing=20 > non-linux derived graphics drivers though!=C2=A0 I would just suggest tak= ing=20 > a moment to understand how much of a lift this work is from a dev=20 > perspective, let alone support after bits land=2E=C2=A0 At the end of the d= ay=20 > most people just assume graphics to work so they can get on with their=20 > real work they need to accomplish=2E I fully appreciate what you're saying here, and I couldn't be more grateful for any, and all the time, and dedication put into this project=2E I'm simply *dumbfounded* that everyone *else* can manage *native* support=2E But we *can't*=2E I'm *really* interested in discovering *why* so I can perhaps initiate *change*=2E >=20 > Not trying to start a flame, just offering perspective from things i've= =20 > observed through this process=2E=2E=2E And that perspective is *greatly* appreciated, pete=2E :) > -pete >=20 > --=20 > Pete Wright > pete@nomadlogic=2Eorg > @nomadlogicLA --Chris -- UNIX is like Ice Cream=2E It comes in several flavors=2E But in the end, it's still Ice Cream=2E
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?5d7dd8e81ae934adf7dfdf389faa5d15>