Announcement

Collapse
No announcement yet.

R600g Gallium3D HyperZ Defeats Developers

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • R600g Gallium3D HyperZ Defeats Developers

    Phoronix: R600g Gallium3D HyperZ Defeats Developers

    Jerome Glisse has published a new patch to enable HyperZ support for the AMD Radeon (R600g) Gallium3D driver. While this patch could be pushed to Mesa, it's not being enabled by default as it's still causing some GPU lock-ups and developers can't seem to figure out the cause. Jerome is now moving onto other work...

    http://www.phoronix.com/vr.php?view=MTEzOTg

  • #2
    Post was strangely worded, it sounded like 'Jerome will be leaving radeon work behind him forever'.
    Last edited by oliver; 07-13-2012, 03:38 PM.

    Comment


    • #3
      I also got the impression that Jerome will move away from radeon. Which i hope is wrong.

      Comment


      • #4
        Marek will fix it.

        Comment


        • #5
          The impression I got was that he may just be burnt out from trying to solving the lockup problem. I think he's just publicly announcing his work so far and if anyone wants to fix it, it's up to them because jerome is off to work on some other feature (not leave the radeon entirely)

          Comment


          • #6
            To me it sounds like that stuff is not properly documented. Why should he else have to look at fglrx?

            Comment


            • #7
              Not surprising because the hardware sucks.

              Comment


              • #8
                Originally posted by Rabauke View Post
                Marek will fix it.
                You mean like R300g?
                Unfortunately it seems hyper-z is a terrible beast
                ## VGA ##
                AMD: X1950XTX, HD3870, HD5870
                Intel: GMA45, HD3000 (Core i5 2500K)

                Comment


                • #9
                  Hyperz is here since R100... should be fixable.

                  Comment


                  • #10
                    Originally posted by Rabauke View Post
                    Marek will fix it.
                    Why you say that? r300g still has open bugs against HyperZ and it doesn't look like there is many people working on it... HyperZ locked up a *LOT* under r300g hardware in he past for random and seemingly unexplainable reasons.

                    When two very skilled devs can't figure it out, it's because AMD isn't doing their job in releasing the specs and/or they're not helping enough to implement HyperZ in the open source driver. I'd definitely say it's AMD's fault and it's probably a waste of time / effort to work on it more until AMD puts out some better specs or helps in some way.

                    Hyper-Z has always been a clever hardware hack to get a bit of extra performance. From what I've seen over the last 2+ years of HyperZ patches, it's just a matter of getting all the planets to align under the night sky with a full moon, and then you have to throw a little fairy dust down from the highest mountain top... If you don't have all that at the same time, it just doesn't work.

                    Comment


                    • #11
                      I'm personally a strong supporter in the support that AMD (nee ATi) has been giving to the OSS community. They have released specs and helped on the OSS driver where possible.

                      In this _specific_ case however, I have to agree, AMD has dropped the ball or may if it does not help a little out. If a extremely skilled OSS dev can't get it to work after 7 months, it's time that AMD step up to the plate and help out. This would not be expected of course, they have no 'real' obligation to do so, but it would confirm and show their commitment.

                      We are not asking here that AMD do all the work in getting Hyper-Z support in the OSS drivers, nor do we ask them to break any IP bs. They would merely be helping the already existing code out.

                      Having said all that, with the code being pushed only now, I would expect that it is only now that someone is even able to help, so my hopes are high at this moment.

                      Comment


                      • #12
                        Originally posted by darkbasic View Post
                        You mean like R300g?
                        Unfortunately it seems hyper-z is a terrible beast
                        No, I meant it wouldn't be the first time that Marek fixes up Glisse's code (no offense).

                        Comment


                        • #13
                          BTW there is a TODO at http://dri.freedesktop.org/wiki/R600ToDo just in case anyone can tackle those tasks.

                          Comment


                          • #14
                            Originally posted by oliver View Post
                            In this _specific_ case however, I have to agree, AMD has dropped the ball or may if it does not help a little out. If a extremely skilled OSS dev can't get it to work after 7 months, it's time that AMD step up to the plate and help out. This would not be expected of course, they have no 'real' obligation to do so, but it would confirm and show their commitment.

                            We are not asking here that AMD do all the work in getting Hyper-Z support in the OSS drivers, nor do we ask them to break any IP bs. They would merely be helping the already existing code out.
                            I believe that has already been happening. Will check again on Monday but I don't think we have the answers yet either, unfortunately.

                            Comment


                            • #15
                              Originally posted by allquixotic View Post
                              Not surprising because the hardware sucks.
                              the Nvidia equivalent of this technology is called "Lightspeed Memory Architecture". So is this technology already implemented in the Noveau driver or does Nvidia hardware suck as well?

                              Comment

                              Working...
                              X