AMD Catalyst 9.12 For Linux Released

[cutterjohn]

AMD: dump fglrx and help the open source driver(s) instead. You should just give up on the new features and just stabilize and bug fix the existing code and then abandon it eventually.

Had to meet my anti-fglrx pro-OSS quota. ^^

I think everyone should just stop using fglrx so they don't have a reason to continue supporting it, and instead will switch to the OS driver.

fglrx as bridgman keeps harping to us is REALLY for their institutional customers who pay HUGE bucks for the same cards with extra memory(LOTS of extra memory) plus MUCH better support, as in I understand that they get to talk to the driver team directly and request/demand specific bug fixes. The rest of us are just supposed to be happy that we get any linux support as a side effect of this.

So, their new answer (last answer was to change the driver name to catalyst and have monthly releases when their drivers were getting a rep in even windows for being cataclysmic) was to dump out the specs and make an OSS driver framework for the rest of us to use. (I guess that they expect their institutional customers to pony up another several thousand per card every 2-3y of course that's probably the warranty term -- referring to droppage of pre-R500 GPU support in catalyst).

Just glad that I forgot about checking for the fglrx release, so that I didn't have to mess with the hotfix, hopefully. (Did they "fix" the currently available driver download? I'll probably hold off until next weekend when I'll have more time to deal with problems. Losing video on a nb could really suck when you have no other useful machines to ssh(well I have a few that I could resurrect enough to ssh in with, but it would waste some additional time and I don't find it to be nearly as... hmmm... entertaining(?) as I did 10y or so ago when I was regularly re-building my own kernels, and dealing with circular RPM dependency hell amongst other things, etc.) in with and no HDMI monitors when the only video output on the nb is HDMI...)

[bridgman]

The rest of us are just supposed to be happy that we get any linux support as a side effect of this.

I don't think we were saying "you're supposed to be happy", but at the time the fglrx driver was what we had. The choice was to include IDs for consumer GPUs or not to provide any solution for the year or two it would take to get something in place for consumer users, and enabling consumer GPUs on the workstation driver seemed like a better place to start.

So, their new answer (last answer was to change the driver name to catalyst and have monthly releases when their drivers were getting a rep in even windows for being cataclysmic) was to dump out the specs and make an OSS driver framework for the rest of us to use.

Actually, no -- the open source drivers were driven primarily by requests from CPU customers who had a different set of priorities from our graphics customers, so we ended up with the two solutions running in parallel.

The reason for making the Linux driver part of the Catalyst framework was that we were sharing a lot more code and component-level validation than before, so we really needed Linux releases to happen on the same schedule and in the same branches as other OSes.

I don't think anyone expected that adding the "Catalyst" name would improve quality on its own, even if press releases might have suggested that

(I guess that they expect their institutional customers to pony up another several thousand per card every 2-3y of course that's probably the warranty term -- referring to droppage of pre-R500 GPU support in catalyst).

Remember that 3D workstation customers tend to use different distros than consumer users - workstation users tend to acquire hardware, distro and application as a set then keep using them as a set until the hardware is retired. They use enterprise distros so they can get a reliable stream of security updates without having to change anything else and put their day-to-day production at risk. It's primarily consumer users who want or need a new distro/kernel/X every six months.

In other words, 3D workstation users are still happily running their 5xx and earlier products; they will buy new hardware when the old hardware isn't fast enough or won't support a new application they need. This is primarily a pain for consumer users, at least until the Gallium3D/KMS/GEM/TTM stack is ready for general use.

[V!NCENT]

The rest of us are just supposed to be happy that we get any linux support as a side effect of this.

Hey... whoa... wait a second...

AMD releases specs and is helping FLOSS drivers. The 9800 pro (r300) I am using here at a friend of mines can play Half-Life2 under wine with the FLOSS driver at almost the same framerates as it did under XP when I bought that card.

I am happy with that. r600 is on its way and I can run my r600 card with the Catalyst driver.

Also do not forget about the, was it Steam under Wine?, bug that ATI has fixed in the Catalyst driver. So they do care.

Try to relax a little, because AMD's new ATI policy is still fresh and they are putting in efford to Linux. Be at least a little easy on them...

Linux has always had its 3D GPU issues... The situation is pretty well known: nVidia currently own with the proprietary driver, older ATI cards are doing fine under FLOSS. You should have known that. Always Google around first before just buying things. That counts for everything and not just computers.

I do agree that Catalyst for Linux is a bit of consumer hell as it is today...

[phtpht]

.31 is supported just fine.

Yeah well. But .32 is already out there and .33 is in RC stage. Still unsupported xorg server 1.7 is out there and 1.8 is in RC stage. See the pattern?

How hard can it be to be actually WITH or even AHEAD of the releases, not 1-2 (or more) months BEHIND? Both projects have a fairly long feature freeze windows that would fit your release schedule just nicely.

[bridgman]

Um... no

The "evil side" wanted to supply high performance, full 3D feature drivers into the Linux market, which wasn't big enough to cover the development cost unless a lot of code was shared with other OSes. That (code sharing across OSes, combined with a need to keep some secrets for *other* OSes) is why we make closed source drivers.

If you want to think of that as evil, carry on

[V!NCENT]

the evil side (ATI) wana cloused source Catalyst..

the Good side (AMD) wana opensource driver......

Qaridarium, this is totally not the case.

Not always is there a right and a wrong. Good or bad.

You're trying to heavily find a good and bad in places where there isn't any so you can form your own opinion.

Forming your own opinion and finding you own place in this world is good, but you can't find it in objects. That's just materialism

[Kazade]

No. It. Doesn't.

What? Just coz it doesn't happen for you doesn't mean it doesn't exist, have you read the bug report, at all? Here it is: https://bugs.launchpad.net/ubuntu/+s...6?comments=all

357 comments, you go in there and tell them it doesn't exist as you obviously know best - I'm sure they'll be glad of your aggressive, sarcastic and all knowing input.

In the meantime I'm willing to accept that it's a bug on Ubuntu's side - however it is a bug that only occurs with fglrx. That much is a fact.

[V!NCENT]

What? Just coz it doesn't happen for you doesn't mean it doesn't exist, have you read the bug report, at all? Here it is: https://bugs.launchpad.net/ubuntu/+s...6?comments=all

I've read it... It's a bug... It exists...

In the meantime I'm willing to accept that it's a bug on Ubuntu's side - however it is a bug that only occurs with fglrx. That much is a fact.

Which was exactly what I was saying. So your point here is... ?

[d2kx]

bridgman, come right here and take my popcorn before carrying on. It's gonna be a long nigh... thread.

[bridgman]

Yeah......