Re: Ubuntu 8.10 significantly slower than previous versions
On Fri, Nov 07, 2008 at 08:03:19AM +0100, Markus Hitter wrote: Well, newer gcc's are meant to produce faster code, aren't they? Faster code? No, GCC doesn't rewrite code. Streamline the compiled binary to make efficient use of system calls? Yes. Different GCC versions can have dramatic effects on binaries of the exact same code. -- ,-O Aaron Toponce O } Ubuntu Member `-O http://www.ubuntu.com signature.asc Description: Digital signature -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
I wonder how this discussion is able to drift so much away from the actual subject on both ubuntu-devel and ubuntu-devel-discuss. Many people do not want to believe results or just point out one or two of them are meaningless (like NVIDIA graphics performance with closed drivers is not that interesting). 2008/11/6 mr [EMAIL PROTECTED]: According to the recent benchmarking article by Phoronix, the previous two releases of Ubuntu are significantly slower than Feisty Fawn. In some cases this can be seen as up to 50% performance drop with certain desktop tasks. Yep. I found three important points I list below, sound encoding (video might be because of changed default parameters, dunno), SQLite, compiling. 1. See eg. the page: http://www.phoronix.com/scan.php?page=articleitem=ubuntu_bench_2008num=4 It takes about double time to encode mp3:s, ogg:s or flac:s In Ubuntu 8.04/8.10 vs. earlier. Fedora seems to have been affected by the problem all the time: http://www.phoronix.com/scan.php?page=articleitem=fedora_test_2008num=2 2. SQLite is over two times slower in Ubuntu 8.10 vs. earlier: http://www.phoronix.com/scan.php?page=articleitem=ubuntu_bench_2008num=5 3. Compiling has been almost two times slower in Ubuntu than Fedora after Ubuntu 7.04: http://www.phoronix.com/scan.php?page=articleitem=fedora_test_2008num=2 The 1. and 3. is what I'd worry about. 1. in general and 3. because Fedora is almost two times faster. (from later in the thread) The default scheduler is optimised for general desktop usage, where you have a large number of simultaneously running applications, applets, etc. and each one needs to be responsive. Please, everybody, do not take this as granted. Question it, test it, feel it etc. With the default CFQ, I could not do about _anything_ when I did eg. svn update or Firefox churned through its enormous databases on a laptop hard drive. After changing to elevator=deadline these cases work _much_ smoother without visible regressions elsewhere. -Timo -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Scott James Remnant wrote on 06/11/08 22:44: ... I've always thought it would be interesting to be able to influence the scheduler on a per process basis - and do that from the Window Manager. ie. deliberately give the user's foreground process the majority of the time, and fair schedule the rest. ... http://brainstorm.ubuntu.com/idea/13718/ Windows Vista does this. It also lets programs register multimedia threads that are then given higher priority to avoid playback glitches. http://arstechnica.com/reviews/os/vista-under-the-hood.ars/4 http://technet.microsoft.com/en-us/magazine/cc162494.aspx - -- Matthew Paul Thomas http://mpt.net.nz/ -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAkkUDdgACgkQ6PUxNfU6ecrteQCgrk/VBZjynKZmsRFZX8qdo5/l BJgAoIr08zBVftvTXpqjBQLkTJhBsv36 =UNNU -END PGP SIGNATURE- -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
Markus Hitter wrote: Am 06.11.2008 um 20:21 schrieb Dan Colish: They're using very different gcc versions between the os's. Well, newer gcc's are meant to produce faster code, aren't they? Quite a few GCC optimisations are for specific CPUs. 32bit ubuntu uses very conservative options so that it works on everything back to i586 (original pentium). The core duo used in the test could potentially do better if it took advantage of things like SSE instructions. I remember the ubuntu devs saying the past that they were unconvinced of the advantage of building packages optimised for newer CPUs, but if someone could show good benchmarks they might consider it. If GCC is now better at things like automatic vectorisation it might be a good time to make some new bench marks. https://wiki.ubuntu.com/GenBunToo is probably a good resource to start at if you want to take on this task. Sam Tygier -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
Matthew Paul Thomas wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Scott James Remnant wrote on 06/11/08 22:44: ... I've always thought it would be interesting to be able to influence the scheduler on a per process basis - and do that from the Window Manager. ie. deliberately give the user's foreground process the majority of the time, and fair schedule the rest. ... http://brainstorm.ubuntu.com/idea/13718/ Windows Vista does this. It also lets programs register multimedia And all previous versions of NT. The thread owning the foreground window has always gotten a boost to its priority. -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
2008/11/6 mr [EMAIL PROTECTED] Hi, According to the recent benchmarking article by Phoronix, the previous two releases of Ubuntu are significantly slower than Feisty Fawn. In some cases this can be seen as up to 50% performance drop with certain desktop tasks. I can confirm that this is true in that my girlfriends desktop used to be quite capable of playing a 1080p x264 video but since upgrading to gutsy and then hardy it has become unwatchable, even mplayer reports that YOUR COMPUTER IS TOO SLOW I think that the reasons behind this reduction in performance across the board needs some serious investigation and work done to reverse this trend. At the moment I am faced with either running an old distro or upgrading hardware. Any discussion on this is welcome :) Thanks, Alan Phoronix article: http://www.phoronix.com/scan.php?page=articleitem=ubuntu_bench_2008num=1 -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss See here: https://lists.ubuntu.com/archives/ubuntu-devel/2008-October/026794.html -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
On Thu, Nov 06, 2008 at 03:58:51PM +0100, mr wrote: Hi, According to the recent benchmarking article by Phoronix, the previous two releases of Ubuntu are significantly slower than Feisty Fawn. In some cases this can be seen as up to 50% performance drop with certain desktop tasks. I can confirm that this is true in that my girlfriends desktop used to be quite capable of playing a 1080p x264 video but since upgrading to gutsy and then hardy it has become unwatchable, even mplayer reports that YOUR COMPUTER IS TOO SLOW I think that the reasons behind this reduction in performance across the board needs some serious investigation and work done to reverse this trend. Indeed, half the reason I suggested Phoronix do these tests is to stimulate more investigation into performance issues. We've had anecdotal evidence of performance reductions since Gutsy at least, and Phoronix presented a good opportunity to get some solid numbers. I've spoken with upstream about -intel performance previously. They've indicated their focus is on the current git version of the driver, and so would ask that anyone wishing to provide feedback on performance to first run the git-head version of the driver. This would enable the user to update and give swift feedback to the developers on any performance changes they are experimenting with. Beyond that, I'd encourage anyone wishing to help improve -intel performance on Ubuntu to join the ubuntu-x mailing list to discuss it in additional detail. Bryce -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
Whoops, I thought you were talking about the recent article about -intel performance on x45 chips. But I see you're actually talking about an earlier article about Ubuntu performance in general: http://www.phoronix.com/vr.php?view=13022 Note that in that article they looked only at the proprietary -nvidia driver's performance, and did not find any noteworthy regressions in that. So depending on what video driver you're using, it may not have much relevance to your issue. Bryce On Thu, Nov 06, 2008 at 03:58:51PM +0100, mr wrote: Hi, According to the recent benchmarking article by Phoronix, the previous two releases of Ubuntu are significantly slower than Feisty Fawn. In some cases this can be seen as up to 50% performance drop with certain desktop tasks. I can confirm that this is true in that my girlfriends desktop used to be quite capable of playing a 1080p x264 video but since upgrading to gutsy and then hardy it has become unwatchable, even mplayer reports that YOUR COMPUTER IS TOO SLOW I think that the reasons behind this reduction in performance across the board needs some serious investigation and work done to reverse this trend. At the moment I am faced with either running an old distro or upgrading hardware. Any discussion on this is welcome :) Thanks, Alan Phoronix article: http://www.phoronix.com/scan.php?page=articleitem=ubuntu_bench_2008num=1 -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
I'm not convined those Phoronix test are really that accurate, especially after reading this one: http://www.phoronix.com/scan.php?page=articleitem=ubuntu_macosxnum=1 It looks like they are not really comparing apples to apples, especially when it comes to java benchmarking. They're using very different gcc versions between the os's. Anyway, it does look like linux wins in the end. On Thu, Nov 6, 2008 at 1:58 PM, Bryce Harrington [EMAIL PROTECTED]wrote: Whoops, I thought you were talking about the recent article about -intel performance on x45 chips. But I see you're actually talking about an earlier article about Ubuntu performance in general: http://www.phoronix.com/vr.php?view=13022 Note that in that article they looked only at the proprietary -nvidia driver's performance, and did not find any noteworthy regressions in that. So depending on what video driver you're using, it may not have much relevance to your issue. Bryce On Thu, Nov 06, 2008 at 03:58:51PM +0100, mr wrote: Hi, According to the recent benchmarking article by Phoronix, the previous two releases of Ubuntu are significantly slower than Feisty Fawn. In some cases this can be seen as up to 50% performance drop with certain desktop tasks. I can confirm that this is true in that my girlfriends desktop used to be quite capable of playing a 1080p x264 video but since upgrading to gutsy and then hardy it has become unwatchable, even mplayer reports that YOUR COMPUTER IS TOO SLOW I think that the reasons behind this reduction in performance across the board needs some serious investigation and work done to reverse this trend. At the moment I am faced with either running an old distro or upgrading hardware. Any discussion on this is welcome :) Thanks, Alan Phoronix article: http://www.phoronix.com/scan.php?page=articleitem=ubuntu_bench_2008num=1 -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
Anyway, it does look like linux wins in the end. I do not believe that is a good thing; Just because Gnu/Linux can be faster than windows vista doesn't automatically mean we are serving our users well. The good news always comes from the users directly who never complain about slowness. When you start to hear complaints, that's when you have a problem. I've noticed Savage2 doesn't work as well any more. But I was cutting it thin with 1GB of RAM with that game but fortunately 8.04 was just slim enough to run it well. Now however the extra bulk in 8.10 has made the game cache more often and me die is horribly messy ways. :-P Regards, Martin -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
On Thu, 2008-11-06 at 14:38 -0500, Martin Owens wrote: Anyway, it does look like linux wins in the end. I do not believe that is a good thing; Just because Gnu/Linux can be faster than windows vista doesn't automatically mean we are serving our users well. Yes, the response on /. to Ubuntu 8.10 is faster than Vista was generally so what? One guy said his father in law with a slide rule, graph paper, and a pencil was faster than Vista. The consensus was faster than Vista isn't hard, we want it to be faster than XP because remember, that's what most people are running. Why would they switch to Ubuntu if it's going to make their machine slower? -- Mackenzie Morgan http://ubuntulinuxtipstricks.blogspot.com apt-get moo signature.asc Description: This is a digitally signed message part -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
Quoting Bryce Harrington [EMAIL PROTECTED]: On Thu, Nov 06, 2008 at 03:58:51PM +0100, mr wrote: Hi, According to the recent benchmarking article by Phoronix, the previous two releases of Ubuntu are significantly slower than Feisty Fawn. In some cases this can be seen as up to 50% performance drop with certain desktop tasks. I can confirm that this is true in that my girlfriends desktop used to be quite capable of playing a 1080p x264 video but since upgrading to gutsy and then hardy it has become unwatchable, even mplayer reports that YOUR COMPUTER IS TOO SLOW I think that the reasons behind this reduction in performance across the board needs some serious investigation and work done to reverse this trend. Indeed, half the reason I suggested Phoronix do these tests is to stimulate more investigation into performance issues. We've had anecdotal evidence of performance reductions since Gutsy at least, and Phoronix presented a good opportunity to get some solid numbers. I've spoken with upstream about -intel performance previously. They've indicated their focus is on the current git version of the driver, and so would ask that anyone wishing to provide feedback on performance to first run the git-head version of the driver. This would enable the user to update and give swift feedback to the developers on any performance changes they are experimenting with. Beyond that, I'd encourage anyone wishing to help improve -intel performance on Ubuntu to join the ubuntu-x mailing list to discuss it in additional detail. The disk IO performance decrease from Gutsy to Hardy is anything but anecdotal. This ( http://groups.google.com/group/zumastor/browse_thread/thread/7e413960ddc22811# ) bug report in the Zumastor project has some (quite scarce) info, although if you read the IRC logs, you'll realize how much pain it caused (it made Zumastor unusable due to slowness). It seems that the scheduler included by default since kernel 2.6.23 cuts performance in half for certain IO operations. Same thing for the CPU scheduler. For instance, here VMware Server is barely usable when high disk IO is requied because for some reason, after a few seconds of good performance (which implies high CPU usage) the kernel starts throttling CPU and VMware starves for a few seconds. What surprises me about this CPU throttling is it achieves nothing: VMware is left with 10% of CPU and the other 90% is idling! I'm forward-porting kernels 2.6.20 (Feisty's) and 2.6.22 (Gutsy's) to Hardy to verify it is the scheduler what is causing trouble. It will be available in my PPA ( http://launchpad.net/~pgquiles/+archive ) tomorrow, in case anyone is interested. -- Pau Garcia i Quiles http://www.elpauer.org (Due to my workload, I may need 10 days to answer) -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
faster than Vista isn't hard, we want it to be faster than XP because remember, that's what most people are running. Why would they switch to Ubuntu if it's going to make their machine slower? I think performance is a very relative term. Slow for games can be great for a database. I am a lot more interested in baseline comparisons between identical systems. I think Ubuntu will make systems faster, but it also make some systems slower. It depends on what you mean by speed. In any case, purpose-built will always beat one size fits all. On Thu, Nov 6, 2008 at 2:41 PM, Mackenzie Morgan [EMAIL PROTECTED] wrote: On Thu, 2008-11-06 at 14:38 -0500, Martin Owens wrote: Anyway, it does look like linux wins in the end. I do not believe that is a good thing; Just because Gnu/Linux can be faster than windows vista doesn't automatically mean we are serving our users well. Yes, the response on /. to Ubuntu 8.10 is faster than Vista was generally so what? One guy said his father in law with a slide rule, graph paper, and a pencil was faster than Vista. The consensus was faster than Vista isn't hard, we want it to be faster than XP because remember, that's what most people are running. Why would they switch to Ubuntu if it's going to make their machine slower? -- Mackenzie Morgan http://ubuntulinuxtipstricks.blogspot.com apt-get moo -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
On Thu, 2008-11-06 at 20:41 +0100, Pau Garcia i Quiles wrote: The disk IO performance decrease from Gutsy to Hardy is anything but anecdotal. This ( http://groups.google.com/group/zumastor/browse_thread/thread/7e413960ddc22811# ) bug report in the Zumastor project has some (quite scarce) info, although if you read the IRC logs, you'll realize how much pain it caused (it made Zumastor unusable due to slowness). It seems that the scheduler included by default since kernel 2.6.23 cuts performance in half for certain IO operations. Same thing for the CPU scheduler. For instance, here VMware Server is barely usable when high disk IO is requied because for some reason, after a few seconds of good performance (which implies high CPU usage) the kernel starts throttling CPU and VMware starves for a few seconds. What surprises me about this CPU throttling is it achieves nothing: VMware is left with 10% of CPU and the other 90% is idling! Remember that you can change the scheduler on the fly. The default scheduler is optimised for general desktop usage, where you have a large number of simultaneously running applications, applets, etc. and each one needs to be responsive. It performs badly at single operations that wish to consume all of the CPU or IO resource available. That includes disk copies, VMware, and funnily enough - benchmarks ;P The scheduler would fair extremely well if you compared, say, 20 simultaneously running benchmark suites between earlier releases and this one. All 20 would show fair results. I've always thought it would be interesting to be able to influence the scheduler on a per process basis - and do that from the Window Manager. ie. deliberately give the user's foreground process the majority of the time, and fair schedule the rest. I'm forward-porting kernels 2.6.20 (Feisty's) and 2.6.22 (Gutsy's) to Hardy to verify it is the scheduler what is causing trouble. It will be available in my PPA ( http://launchpad.net/~pgquiles/+archive ) tomorrow, in case anyone is interested. Why do you need to forward-port? The same kernel binary will just work. Also you can just fiddle on a per-disk basis, e.g.: echo -n deadline /sys/block/sda/queue/scheduler Scott -- Scott James Remnant [EMAIL PROTECTED] signature.asc Description: This is a digitally signed message part -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: Ubuntu 8.10 significantly slower than previous versions
Am 06.11.2008 um 20:21 schrieb Dan Colish: They're using very different gcc versions between the os's. Well, newer gcc's are meant to produce faster code, aren't they? MarKus - - - - - - - - - - - - - - - - - - - Dipl. Ing. Markus Hitter http://www.jump-ing.de/ -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
