Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
×
AMD Graphics Open Source Upgrades Hardware Linux

AMD's Crimson Radeon Driver For Linux Barely Changes Anything (phoronix.com) 95

An anonymous reader writes: AMD Windows customers were greeted this week to the new "Crimson" Radeon Software that brought many bug fixes, performance improvements, and brand new control panel. While AMD also released this Crimson driver for Linux, it really doesn't change much. The control panel is unchanged except for replacing "Catalyst" strings with "Radeon" and there's been no performance changes but just some isolated slowdowns. The Crimson Linux release notes only mention two changes: a fix for glxgears stuttering and mouse cursor corruption.
This discussion has been archived. No new comments can be posted.

AMD's Crimson Radeon Driver For Linux Barely Changes Anything

Comments Filter:
  • by CaptainJeff ( 731782 ) on Thursday November 26, 2015 @10:18AM (#51008153)
    So...the news here is that there is no news for Linux folks?
    • Re:So... (Score:5, Funny)

      by gstoddart ( 321705 ) on Thursday November 26, 2015 @10:23AM (#51008175) Homepage

      My god man .. they changed the strings from "Catalyst" to "Radeon".

      What more do you want?

      • by Anonymous Coward

        A new retro 8 bit christmas logo.

      • What more do you want?

        Hardware neutrality.

      • My god man .. they changed the strings from "Catalyst" to "Radeon".

        I know, it's a staggeringly complex rewrite. I can only wonder how many developer-months this took to complete.

        • Well, it's not like there is some kind of magical tool that can recursively scan all the files within a directory, and change one string to another one wherever it is found.

          Hey, someone should totally code that up. I bet it would be very useful.

      • My god man .. they changed the strings from "Catalyst" to "Radeon".

        I'd love to be there when they do the year-end performance reviews in a few weeks..... "I personally updated 632 source files to conform with our new marketing paradigm".

        • You laugh ... but years ago at a different job, the marketing people decided to rename/re-brand a product.

          We literally had to stop everything, and build an entire release which had the name changed; which ended up having to finalize other things or roll them back to add later.

          Never underestimate how much marketing can screw up a dev schedule.

    • I didn't follow this closely but is second rate support for Linux by AMD anything new??

      • Re:So... (Score:4, Interesting)

        by MyAlternateID ( 4240189 ) on Thursday November 26, 2015 @10:44AM (#51008295) Homepage

        I didn't follow this closely but is second rate support for Linux by AMD anything new??

        Since the late 90s I've always used nVidia for Linux systems* and I've never regretted that choice. I would prefer having nVidia open-source their own driver (nouveau has made progress but just isn't there yet) but this is not a big deal to me. I run a source-based distro (Gentoo) so I compile my own kernels anyway; it's no big deal to add "&& emerge --oneshot nvidia-drivers" at the end of that command line. That's the most I've ever had to do. Unfortunately some binary distros are more "purist" out of either ideology or fear of legal action so they make users jump through a few hoops to get proprietary drivers and codecs.

        The more I heard about first ATI and now AMD driver quality on Linux, the more convinced I am that I made the right choice. I hope AMD gets their shit together in this area because competition is a good thing.

        * That's on my desktop system which has a discrete card. My netbook with Intel graphics runs Linux Mint which "just works" (I believe that driver *is* open source, MIT licensed IIRC).

        • by vivian ( 156520 )

          I used to use nVidia cards until I had a bad run of luck with a few overheating and dying prematurely.
          I currently have an ATI and wish I'd stuck with Nvidia - monitors don't wake up from sleep properly - I have to toggle the video source buttons or turn them off and on again (generally only one or the other wakes - sometimes neither wake, very rarely both wake - it seems to be random), and there are still corruption issues with Chrome and some opengl apps- though this is better since the last (not current

          • This is a known issue right now with many of their GPU families and affects all of the major OSes, but is especially prevalent on the latest version of OSX, Win8, Win8.1, Win10, and various kernel versions & distros of Linux.

            They are also having pretty big issues with video acceleration (which might be your issue with Chrome corruption) & subtitle rendering (either not showing subtitles at all, or rendering the entire screen black when the subtitle is supposed to display) during video playback.

            You a

  • " a fix for glxgears stuttering "

    so they only FIXED a TEST !!!
    and one that is a YES / NO test
    glxgears should never be used as a speed test

    the FPS in glxgears is mostly meaningless

    • " a fix for glxgears stuttering "

      so they only FIXED a TEST !!!
      and one that is a YES / NO test

      I know, it's like Christmas came early with this amazing code revision.

      It's such an awesome fix, it's almost incomprehensible how goshdarn difficult this must have been for them to implement.

      "Hey Bill, set $glxgears_stutter to "off" in the shit_module_stuff() function and mark it as complete in the fix list."

  • AMD is prioritizing Windows with respect to these changes. That should not be surprising since the market is dominated by Windows. Now if these changes aren't reflected in their Linux drivers down the road, then yes there will be reason for concern.

    • AMD is prioritizing Windows with respect to these changes. That should not be surprising since the market is dominated by Windows. Now if these changes aren't reflected in their Linux drivers down the road, then yes there will be reason for concern.

      Your argument only holds water if you're a Windows user. If you're a Linux user, this is cause for concern, because the Linux drivers are shit. They offer dramatically less performance than the Windows drivers. This is just further proof that AMD doesn't give a Fuck about Linux users. Frankly, they are very poor compared to intel about Linux support in general; hell, there's Still no good PM support in Linux for Mobile Athlon 64 Cool'n'Quiet, and the R690 chipset has never been properly supported. They don'

      • by MacTO ( 1161105 )

        I use Linux on my desktop. It has an AMD card and uses AMD drivers. So I do have a vested interest in better performance and stability in their products.

        That being said, I am also a realist. AMD is a business, and they're there to make money. If they have to invest too much money into making Linux drivers, it hurts their profitability. If they have to divert too many resources away from Windows and towards Linux, hurts their profitability. I am not arguing that these are excuses for poor support, or f

        • I use Linux on my desktop. It has an AMD card and uses AMD drivers. So I do have a vested interest in better performance and stability in their products. That being said, I am also a realist. AMD is a business, and they're there to make money. If they have to invest too much money into making Linux drivers, it hurts their profitability.

          Yes but don't you have a vested interest in getting the best stability and performance you can get for your money, and in not rewarding a company that has decided you're not a priority? I want to like AMD. I want to have to really think about the next card I purchase because I want to have more than one really good option. That would be a healthy market. But I've stuck with nVidia and I've been very glad I did. That's been the single most effective way for me to avoid all of the problems I keep hearing

  • by Anonymous Coward

    Sounds like It's AMD's turn for a good old Linus Torvalds Blasting.

  • them getting rid of that horrible, horrible .Net interface.

    What I'm really wondering is what the bleep were they doing before. I read this:

    AMD shifted their development process for the Catalyst driver set, focusing on delivering feature updates in fewer, larger updates while interim driver releases would focus on bug fixes, performance improvements, and adding new cards.

    And my first thought was, how the hell else do you develop software? You put out one or two big releases a year and then fix and
    • AMD shifted their development process for the Catalyst driver set, focusing on delivering feature updates in fewer, larger updates while interim driver releases would focus on bug fixes, performance improvements, and adding new cards.

      And my first thought was, how the hell else do you develop software? You put out one or two big releases a year and then fix and patch up in between. What the hell was AMD doing before Crimson? Where they completely re-writing their driver stack 3 or 4 times a year?

      Well, the way AMD has been doing it is that they make some minor changes occasionally, and once in a while they make the driver configuration GUI bigger and more bloated and increment the major version. Then you have to wait for someone (e.g. DnA) to hack the drivers up to make them not crash your system. At least, that's my experience of ATI graphics on Windows. I still have one machine with integrated Radeon, and it is by far the biggest PITA of everything I own. Making it work right on Windows is difficu

  • I'm not sure guys, on a lot of boxes glxgears is the only work that the video card gets to do. Fixing that helps the needs of the many over the needs of the few...

  • Reading reviews & comments on a couple other tech sites with similar article are calling B.S., there really is performance improvement even though certain promised features didn't appear. Hmmm, fanboys of other vendors?

    • Reading reviews & comments on a couple other tech sites with similar article are calling B.S., there really is performance improvement even though certain promised features didn't appear. Hmmm, fanboys of other vendors?

      I don't know about you, but I tend to be disappointed when promises are broken (especially by people who had the resources to keep them) and I rightly tend to take a dim view of people who make promises they won't keep. This doesn't make someone a "fanboy", it makes them a person with good judgment.

      • No, I'm only calling "fanboy" on those say no performance improvements; certainly it is bad AMD didn't deliver on promised features

  • stop spreading Michael's bullshit. The driver has 0% gain in ANY platform, any means every platform. Stop giving hits to every *biased* article that guy makes.
  • While it may bring performance enhancement, it also brings new bugs. The compass in Fallout 4, for example, becomes unusable.

    • While it may bring performance enhancement, it also brings new bugs. The compass in Fallout 4, for example, becomes unusable.

      Leave it to AMD to fail to test their driver on the most important game of the year, and possibly the decade.

  • Considering ATI/AMD has more or less never once produced a stable driver for video on any operating system. I don't want to read the article because I prefer to believe all the changes were purely just string replace Catalyst to Crimson and nothing more.
  • a fix for glxgears stuttering and mouse cursor corruption.

    See? Now Linux is unstoppable when it comes to high-performance graphics.

  • Why would they bother with Linux at? Especially if they are forced to GPL their code.

    • by jopsen ( 885607 )

      Why would they bother with Linux at? Especially if they are forced to GPL their code.

      They are not... there is so many work arounds it's crazy...

      Anyways, they should do it to get more competition on a market they depend on.

    • by spauldo ( 118058 )

      They're not forced to GPL their code.

      Linus interprets the GPL in a way that allows for binary drivers (I'm sure RMS doesn't agree, but it's not RMS' software). AMD has a binary driver, which is what this article is about.

      NVidia has a proprietary driver as well, and it's basically the same code on both Windows and Linux (and FreeBSD, I believe). There is some GPL shim code that allows the kernel to talk to the binary blob, but the driver itself is closed source.

      Binary drivers aren't just for video cards.

  • ...6 years ago, and haven't needed to look back since. For as closed source as they might be, at least stuff usually works like it should. AMD on the other hand has a long history of burning its users. I find intel processors more reliable too. Sure I'll pay a little more, but I'd rather not find the gotchas that always crop up with AMD hardware later.

Every nonzero finite dimensional inner product space has an orthonormal basis. It makes sense, when you don't think about it.

Working...