Announcement

Collapse
No announcement yet.

Ubuntu Is Going After A New Linux Kernel API

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

  • #16
    mmap()

    hmm .. as this can only be used with static or regenerateable data, whats the benefit over using a mmap()'ed file (readonly) and let the kernel cache do the magic?

    Comment


    • #17
      Originally posted by Delgarde View Post
      On, the contrary - the benefit is about caching. Lots of apps benefit from loading stuff into memory and keeping it there in case it's needed again - if the memory is available, it's a good tradeoff to avoid the cost of loading it from disk next time you need it. This mechanism is a way for the kernel to tell a process "hey, we're running short of memory... get rid of that stuff you were caching".

      Currently, the only options are to prevent processes from using that memory at all, or to kill them when memory runs out. This is a middle ground, letting them use memory if it's available, but avoiding the need for a hard kill when it's not.
      i do know this, my point is you need a real great number of concurrent applications or a very bad coding to starve the memory, specially since default this days is at least 1gb, i mean my full gentoo system with kde-4.11 + akonadi+nepomuk+ zram + ramdisk for everything + firefox 10tabs+ calligra + kernel caches don't hit the 1gb barrier and we are talking about a full fledged x86 desktop system with a bazillion deps and entire set of libraries preloaded on ram.

      i can understand this predicament with java/gavik or .NET garbage collectors eating ram to optimize real crappy loops but ubuntu phone will be using c++ and QML with qt5 right? which both are very ram lightweight if coded right. so base on this i can only imagine those 3 posibilities in my previous posts to actually need something like this.

      Comment


      • #18
        vmpressure_fd

        Sounds like the want the inverse of vmpressure_fd

        Comment


        • #19
          Originally posted by Delgarde View Post
          On, the contrary - the benefit is about caching. Lots of apps benefit from loading stuff into memory and keeping it there in case it's needed again - if the memory is available, it's a good tradeoff to avoid the cost of loading it from disk next time you need it. This mechanism is a way for the kernel to tell a process "hey, we're running short of memory... get rid of that stuff you were caching".
          Why not just read from disk every time? If there's lots of free RAM, it will be in the disk cache, if not, the operating system can drop it from the disk cache automatically to make space.

          Comment


          • #20
            Originally posted by jrch2k8 View Post
            i mean my full gentoo system with kde-4.11 + akonadi+nepomuk+ zram + ramdisk for everything + firefox 10tabs+ calligra + kernel caches don't hit the 1gb barrier and we are talking about a full fledged x86 desktop system with a bazillion deps and entire set of libraries preloaded on ram.
            You don't play games, do you?

            Comment


            • #21
              *COUGH* *COUGH* http://www.phoronix.com/scan.php?pag...tem&px=MTA0MjM

              Comment


              • #22
                Originally posted by arashbm View Post
                You don't play games, do you?
                i doubt you find a game that demanding running in ARM for a while, is not like battefiled 4 will run on ubuntu phone anytime soon

                Comment


                • #23
                  Originally posted by jrch2k8 View Post
                  i doubt you find a game that demanding running in ARM for a while, is not like battefiled 4 will run on ubuntu phone anytime soon
                  Well, that's not the point is it? I think the point is to avoid a catch22. An "If you build it they will come" type of thing.

                  Comment


                  • #24
                    Let the User Decide

                    I fail to understand why Linux doesn't let the user make the decision in low memory situations. How hard is it to just pause everything and display a dialog to the user letting him pick the application to kill/swap out/etc? This would be far more effective, and far easier to implement, than what Canonical is proposing.

                    Comment


                    • #25
                      Originally posted by Vistaus View Post
                      Ah well, what does it matter? Half of Phoronix will find a way to bash this idea anyway just because it comes from Canonical.
                      Nope, if they work with upstream, this would be a very useful and welcome feature. And yes, I did bash Unity and Mir.

                      Comment


                      • #26
                        Originally posted by duby229 View Post
                        Well, that's not the point is it? I think the point is to avoid a catch22. An "If you build it they will come" type of thing.
                        nope, ARM problem is not "OMG linux don't let me" or "OMG games will crash because linux is evil", the problem is more visceral and simple, ARM don't have the bandwith requirements to do so at hardware level, before anyone come yelling ARM is close to the raw performance of PS3 try upload an multiset of 720p+ texture in ARM and then try in a PS3 and you will understand why there is no grand turismo in Android/iOS

                        Comment


                        • #27
                          maybe future revision of ARM8 AArch64 will fix this tho

                          Comment


                          • #28
                            Originally posted by jrch2k8 View Post
                            nope, ARM problem is not "OMG linux don't let me" or "OMG games will crash because linux is evil", the problem is more visceral and simple, ARM don't have the bandwith requirements to do so at hardware level, before anyone come yelling ARM is close to the raw performance of PS3 try upload an multiset of 720p+ texture in ARM and then try in a PS3 and you will understand why there is no grand turismo in Android/iOS
                            I never said anything about linux being evil, that's definitely not my opinion.

                            And of course it's not going to compete with the PS3, it has the same dedicate hardware for years and is thoroughly optimized. But, who said anything about competing with the PS3? It wasnt me. For example Qualcomm's Snapdragon processor with Adreno GPU is a decent gaming experience for a handset. It doesnt need to be 720p, it can simply be the native resolution of the screen. But games easily max out its available RAM. Any help in that department is acceptable.

                            Comment


                            • #29
                              Originally posted by fscan View Post
                              hmm .. as this can only be used with static or regenerateable data, whats the benefit over using a mmap()'ed file (readonly) and let the kernel cache do the magic?
                              Originally posted by movieman View Post
                              Why not just read from disk every time? If there's lots of free RAM, it will be in the disk cache, if not, the operating system can drop it from the disk cache automatically to make space.
                              Because it might not be on disk... it might be info that can be re-calculated if needed, but which can be kept in memory for improved performance. E.g a browser might be keeping some rendering/layout structures for tabs that aren't currently visible, to save time if the user switches to them. Easily regenerated if the cache has to be cleared, but keeping it saves a second or two when it's needed.

                              Comment


                              • #30
                                Originally posted by duby229 View Post
                                I never said anything about linux being evil, that's definitely not my opinion.

                                And of course it's not going to compete with the PS3, it has the same dedicate hardware for years and is thoroughly optimized. But, who said anything about competing with the PS3? It wasnt me. For example Qualcomm's Snapdragon processor with Adreno GPU is a decent gaming experience for a handset. It doesnt need to be 720p, it can simply be the native resolution of the screen. But games easily max out its available RAM. Any help in that department is acceptable.
                                well my point is gavick/java/.Net garbage collector punish very hard the RAM and the I/O and i can agree that android could starve ram like no tomorrow in more visual appelaing games with GPU that share RAM but in C++ would be a different story entirely since you can't just let object all over the place without survelaince and expect the garbage collector to deal with it. now if ubuntu is offering stuff like Mono/Java then would make more sense this idea.

                                Comment

                                Working...
                                X