Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

sparty411

macrumors 6502a
Nov 13, 2018
552
499
This was inevitable. Unless someone forks the kernel, all legacy hardware will eventually be reduced to a paperweight.
 

Macbookprodude

Suspended
Jan 1, 2018
3,306
898
What a stupid move by them - no, not only will we be forced to PCI, but it’s a WARNING SIGN - Linux is ending PowerPC support - while my Quad G5 has Nvidia and it’s still a PIA to set up (continue to get blank splash screen or black screen), 2 of my PPC machines - a PB G4 DLSD and G4 upgraded Pismo use Radeon cards.

No, what the developers are doing is forcing us to stop using PPC for Linux.
 

z970

macrumors 68040
Jun 2, 2017
3,580
4,502
@Macbookprodude This has nothing to do with PowerPC directly. AGP cards will remain completely usable - the AGP slot will effectively just turn into 66mhz PCI (to my current understanding).

In fact, PowerPC might not even change at all, as one user explains in the replies section ...

If a AGP card is in a linux powerpc system has not in recent time used the AGP GART/MMU system always driven the card in PCI MMU mode with accelerated graphics this is the mode that is going to remain. AGP GART/MMU usage has been restricted to x86 platforms. Before this patch particular x86 systems by Linux quirks(as in we detected hardware issue here we will do special behaviour) when you put AGP card in them they also only drive the card in PCI MMU mode as well and also by quirks particular AGP cards will only drive in PCI MMU no matter what X86 system they in.

The reality is you have a x86 system with a AGP card in it running Linux you have less than a 50% chance before this patch that is using AGP GART/MMU and have greater than 50% chance is being driven in PCI MMU mode. If you have a powerpc system before this patch you have a 0% chance that you are using AGP GART/MMU and 100% that you have been using PCI MMU mode on the AGP card this complete time.

The reality here is quite a few people for quite a long time have been using their AGP cards in PCI MMU mode and have not reported performance problems or any other issues.

Now even recently people have been reporting AGP cards in AGP GART/MMU mode causing system wide lock-ups you don't get much worse of a performance problem when you have just lost all your work to driver issue.

So this is not no acceleration removing the AGP modes. Yes in particular programs AGP GART/MMU mode may be faster but this comes at a major risk to system stability. Developers having to maintain a growing quirk entries with more and more entries reading disable AGP GART use PCI MMU mode for system stability is not really a productive usage of developers time when there is not a major documented performance difference. The difference is small enough that users have not notice or complain when all the past quirks that disabled AGP GART have been set heck some reported result in faster speeds after the prior quirk fixs that were just use PCI MMU because the person stops having system wide MMU stalls.

Its one thing if the feature worked with documented performance benefit large enough to justify the risk. Powerpc kernel developers way back in early days decided that AGP GART/MMU was not worth the problems due to not enough performance gain to justify it this was 1998 one year into AGP existance.

Powerpc users have not had the AGP GART/MMU feature now removed from x86 complete since 1998 when it was removed from the powerpc kernel code so anyone with a powerpc system this recent change in fact changes nothing because the same change for your was over a 2 decade ago.
 
Last edited:

ADunsmuir

Contributor
Mar 26, 2020
29
31
What a stupid move by them - no, not only will we be forced to PCI, but it’s a WARNING SIGN - Linux is ending PowerPC support - while my Quad G5 has Nvidia and it’s still a PIA to set up (continue to get blank splash screen or black screen), 2 of my PPC machines - a PB G4 DLSD and G4 upgraded Pismo use Radeon cards.

No, what the developers are doing is forcing us to stop using PPC for Linux.
Wrong.

As usual Phoronix has a tendency for overblown flame-bait headlines. In this case the developer proposing the change used VERY poor wording on the subject line, and Michael just ran with that.

i posted in the driver mailing list and the reply clarified that they are looking to use the PCI GART (MMU) on the card instead of the AGP bus GART. This is already the default mode of operation on PPC.

Today’s follow up article explained that AGP bus still is used for the full benefit of faster data transfers, but the caching and DMA operations will tend to be saner using the PCI bus GART. It is worth the read of the comments.

Some folks on the mailing lists do have the useless “If it ain’t new and shiny you should throw it away” attitude. Others understand the need to keep older hardware running... and especially running well. The tech leads seem sane and do want to move forwards and keep the older hardware support. They believe this change will simplify the code, improve stability and enable future improvements.

The consensus was that the PPC code was in better shape in this area than the generic (x86) code in terms of optimization and performance. The goal will be to come up with one optimized version and move forwards with improvements without the memory access errors that the AGP GART tends to induce. One gent’s comment today on Phoronix indicated for one Intel chipset the CPU load was a bit higher, throughput (frame rate) the same, but the Intel motherboard chipset + x86 PCI code had stability issues.

In the longer term, this has the potential to correct things like PPC Nouveau incompatibility with PPC64 page sizes >4KB.

The PPC Mac community does need to help on an ongoing basis, not just consume. Support your distributions, test, report bugs and help validate fixes.

I’ve built up a collection of PC, Mac and AIX video cards (often both flavours so endian bugs can be identified and fixed) and plan to get involved with driver testing and eventually driver development.

Hope this helps
Al Dunsmuir
[automerge]1589860436[/automerge]
@Macbookprodude This has nothing to do with PowerPC directly. AGP cards will remain completely usable - the AGP slot will effectively just turn into 66mhz PCI (to my understanding).

PowerPC might not change at all, as one user explains in the replies section ...
The default is to not use the AGP GART on PPC, but there is code still there to support it. Since it is apparently different than the x86 code AND not used much, it tends not to work well when enabled - the worst of both worlds.

The more folks who know this area of hardware and software talk, the more it sounds like the RFC is reasonable and likely a win in the end for PPC as more code becomes common across platforms.

Al Dunsmuir
 
Last edited:

ADunsmuir

Contributor
Mar 26, 2020
29
31
What a relief :) sorry for overblowing my mind there.

It's not the first time a Phoronix headline has caused me to yell "What the everloving Frack" at the top of my lungs. It likely won't be the last, either.

Best approach is for us all to try to keep a close eye on the latest kernel and DRM... and help fix things via bug reports, and fix testing if something does go sideways.

Al
 
  • Like
Reactions: Amethyst1
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.