Announcement

Collapse
No announcement yet.

Intel Linux Graphics Developers No Longer Like LLVM

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

  • #41
    Originally posted by ChrisXY View Post
    His reason about not "importing" llvm into mesa sounds strange. Isn't it that mesa already supports having the llvm libs compiled into it when not using --with-llvm-shared-libs?
    I mean you have buildscripts for your distribution for mesa, couldn't you just clone your favorite llvm revision in that build script and compile it into mesa and so get an updated radeon driver on previous versions of fedoras? Sure, that's a much longer build then, but it should at least be easy...
    He made the "table-flip-rage-quit" statement after the fosdem, so I don't think the importing issue is the blocking factor here.

    Comment


    • #42

      We're absolutely not interested in anything involving LLVM. We've
      tried, several times, to make it work, and it has always ended with a
      table-flip-rage-quit.
      Learning is frustrating.

      Comment


      • #43
        Originally posted by squirrl View Post
        I'd be perfectly fine with just a heavy learning process.... but the constant API breakage is ridiculous! LLVM/CLANG break API so often, they make Wayland look like it's API was chiseled in stone.

        Comment


        • #44
          Originally posted by Marc Driftmeyer View Post
          Reading the thread speaks volumes about Intel staff vs. AMD staff.

          Perhaps Mesa wants the LLVM/Clang project to come in and clean up the code? Or have AMD completely take over the project? Seriously, piss poor code behavior is a product of the developer, not a metric of the tools used.

          Or perhaps Intel is pissed-off Apple won't lift a finger in helping them in their shader compiler work? Who cares? They've got the money and seeing as they have done a half-ass job with their OpenMP 3.1 being ready, and no actual work since the only code dump several months back, I doubt anyone in the LLVM/Clang greater community is going to run to their aid or defense, other than some GCC advocate ready to bend over backwards to swallow Psi into the GCC codebase that Intel couldn't convince anyone in the LLVM/Clang group to just accept.

          You speak of the 'greater' LLVM/Clang community like it's a real organic community, It was never created with an organic community model. Permissively licensed project work is done is by a single company for their self-interest, it's not a value statement, it's just a simple fact. That 'greater community' is not considered at any point, since there is no community beyond company self-interest, it's resulted in the most insane API breakage. I mean it's INSANE! "Don't need the C backend, throw it out..." , "O hey lets change the command lines, why bother making even a simple wrapper, the Cmakelists can break worldwide for all we care"

          Seriously it's undefendable, who can blame anyone? Look at the repositories on the net, full of source-to-source holy grail style transcompilations... but they never can complete a build. It's actually tragic, the best example of why a community project must be reciprocal. O but hey the're all great for that next proprietary project! Sucks that none of it is usable without company resources, since there is no community interests.
          Last edited by techzilla; 11 February 2014, 02:34 AM.

          Comment


          • #45
            When someone explains their programming as desk tossing minutes, you know they really have done their best to like it. It would be awesome to here their particular gripes so others are careful about LLVM.

            james perse | botkier | antik denim

            Comment


            • #46
              Originally posted by zivbrovski View Post
              When someone explains their programming as desk tossing minutes, you know they really have done their best to like it.
              Actually just the opposite, at least from my experience people that express themselves that way haven't put in the effort! Either that or they don't have the intelligence to grasp the system.

              If they had an intelligent explanation to whatever the problem was then I would have a little respect for their opinion. As it is now I have zero respect for their opinion and frankly would discount anything they have to say from this point on. If you have valid concerns express them, if the software comes up short in some way detail the problems, doing things like this does take a bit more effort than a childish response so maybe they don't have it in them.
              It would be awesome to here their particular gripes so others are careful about LLVM.
              No fooling. That is my whole point, this little sharada is as useful as a child's temper tantrum. Based on this sort of non sense there is no reason for somebody to be careful about LLVM, there are far more success stories than there are failures.

              Look at it this way, if you have two mechanics that have worked on your car with one being successful and the other throwing wrenches through the wall, who do you return to for the next repair or opinion. Most people will avoid the wrench thrower and go with the successful talent.

              Comment


              • #47
                Originally posted by wizard69 View Post
                Look at it this way, if you have two mechanics that have worked on your car with one being successful and the other throwing wrenches through the wall, who do you return to for the next repair or opinion. Most people will avoid the wrench thrower and go with the successful talent.
                So you choose your mechanics based on the tools they use, not the end result of their work? It just might be possible that LLVM isn't perfect, and isn't the right tool for every single job out there.

                In the end, who cares if Intel uses LLVM or not, as long as they manage to produce good (and open source) graphics drivers for their chips.

                Comment


                • #48
                  Originally posted by tuubi View Post
                  In the end, who cares if Intel uses LLVM or not, as long as they manage to produce good (and open source) graphics drivers for their chips.
                  Every other user that uses an open source graphics driver that utilizes LLVM cares, because improvements committed by one entity often benefit other entities. Linux wants their products to win. We want our entire ecosystem to win.

                  That said, nothing prevents Intel from coming up with a better implementation than LLVM, and other developers from borrowing from and building upon the work. It's often a challenge to be optimistic though, especially looking at the short-term challenges OSS GFX are facing.

                  Comment


                  • #49
                    Originally posted by russofris View Post
                    Linux wants their products to win.

                    s/Linux/Intel

                    Comment


                    • #50
                      Originally posted by wizard69 View Post
                      Actually just the opposite, at least from my experience people that express themselves that way haven't put in the effort! Either that or they don't have the intelligence to grasp the system.
                      You're replying to a spam bot, here's the original comment:

                      Originally posted by e8hffff View Post
                      When someone describes their coding as table flipping moments, you know they really have done their best to like it.

                      It would be nice to here their specific gripes so others are cautious about LLVM.

                      Comment

                      Working...
                      X