tag:blogger.com,1999:blog-6469704299235308349.post282973505016029251..comments2024-02-09T16:24:46.087+11:00Comments on -ck hacking: BFS and -ck delays for linux-3.5.0ckhttp://www.blogger.com/profile/02904761195451530213noreply@blogger.comBlogger34125tag:blogger.com,1999:blog-6469704299235308349.post-56337631467812232662012-08-16T06:20:44.042+10:002012-08-16T06:20:44.042+10:00I used your patch for bfs kernel 3.5, and it works...I used your patch for bfs kernel 3.5, and it works perfectly, is very good indeed, also put the locks and urw ukms and bld, and BFQ, compiled the kernel, and it works very fast, I used my atlon64 le-with 1GB of memory, under kde 4.9 on debian, my congratulations.betohttps://www.blogger.com/profile/01619792691883766545noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-32914691465340121132012-08-14T19:59:14.806+10:002012-08-14T19:59:14.806+10:00Hi Manuel,
yes I have tested with coreduo. Only on...Hi Manuel,<br />yes I have tested with coreduo. Only one cpu might not profit from my settings!<br /><br />CONFIG_RCU_BOOST_PRIO > 1 (default) is needed to not get time overflows with tools like htop.<br /><br />Manuel, to your previous observations, I also get this: <br />linux-3.4.8 is buggy. linux-2.4.7 was much better with BFS!<br /><br />Ralph UlrichAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-24382835739018767522012-08-13T21:43:50.781+10:002012-08-13T21:43:50.781+10:00@Maunel
That's not the reason to seethe — I&#...@Maunel<br /><br />That's not the reason to seethe — I've just posted bugreport.<br /><br />@_sid_<br /><br />Thanks, will give it some try.Oleksandr Natalenkohttps://www.blogger.com/profile/12098091624630953604noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-5144977478084611312012-08-13T07:09:00.367+10:002012-08-13T07:09:00.367+10:00@post-factum
Thank you for letting me know. Here ...@post-factum<br /><br />Thank you for letting me know. Here is an incremental patch that fixes this issue.<br /><br />http://www.file-upload.net/download-4666029/3.5-sched-bfs-424-_sid_-2-3.patch.html_sid_noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-72622749551043657492012-08-13T07:02:53.960+10:002012-08-13T07:02:53.960+10:00@post-factum:
There's area for "staging&q...@post-factum:<br />There's area for "staging". Whatever that should mean at all: These drivers are more unstable than BFS ever was.<br /><br />-ck has never been included into kernel. But kept stability over releases!!<br /><br />Private kernel patching goes onto own personal risk. See below. <br />Using unauthorized patches for kernels that Con haven't checked would not benefit either.<br /><br />MaunelAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-57461330598881985702012-08-13T06:22:43.254+10:002012-08-13T06:22:43.254+10:00@Kelvin
No excuse — if there's an option to c...@Kelvin<br /><br />No excuse — if there's an option to choose whether to enable BFS, kernel must compile in any case.Oleksandr Natalenkohttps://www.blogger.com/profile/12098091624630953604noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-49289452083566374542012-08-13T05:51:55.784+10:002012-08-13T05:51:55.784+10:00But CONFIG_HZ_300 makes way better scores in world...But CONFIG_HZ_300 makes way better scores in worldcommuniygrid, howewer.<br /><br />;-) Just to mention...<br /><br />ManuelAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-1999709938826439192012-08-13T04:42:21.387+10:002012-08-13T04:42:21.387+10:00@ Ralph Ulrich,
your recently posted settings for ...@ Ralph Ulrich,<br />your recently posted settings for BFS-kernels have up- and downsides.<br /><br />CONFIG_HZ_1000 works better for 1 CPU in favour of interactivity than _300.<br /><br />The following seems to be really only cosmetic:<br />CONFIG_RCU_BOOST_PRIO=14<br />(default=?)<br />CONFIG_RCU_BOOST_DELAY=440<br />(default=500)<br />in my opinion. <br /><br />MPlayer lags in Sound vs. Video. If I revert, it's vice versa.<br />These take up Con's considerations about interacivity versus throughput (AGP, PCI, etc.)<br /><br />Best regards,<br />Manuel Krause<br /><br /><br />You proposed @20120720:<br /># CONFIG_NO_HZ is not set<br />...<br />CONFIG_HZ_300=y<br /># CONFIG_HZ_1000 is not set<br />CONFIG_HZ=300<br />...<br /># RCU Subsystem<br />#<br />CONFIG_TREE_PREEMPT_RCU=y<br />CONFIG_PREEMPT_RCU=y<br />CONFIG_RCU_FANOUT=64<br /># CONFIG_RCU_FANOUT_EXACT is not set<br /># CONFIG_TREE_RCU_TRACE is not set<br />CONFIG_RCU_BOOST=y<br />CONFIG_RCU_BOOST_PRIO=14<br />CONFIG_RCU_BOOST_DELAY=440Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-5403225295019244622012-08-13T03:46:54.354+10:002012-08-13T03:46:54.354+10:00Con, are you completely distracted by contract wor...Con, are you completely distracted by contract work, now?<br /><br />I had been running patched to 3.4.8 for 25h and it hardlocked again, then.<br />openSUSE 3.4.6 + BFS single + mm-drop_swap_cache_aggressively.patch + BFQ + inc. patches. <br />3.4.7 went well for many many many days.<br /><br />I want to highly encourage you to improve your patch!<br /><br />Manuel KrauseAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-15360120494913270242012-08-13T00:13:39.646+10:002012-08-13T00:13:39.646+10:00@post-factum
If you do not enable BFS, why do you...@post-factum<br /><br />If you do not enable BFS, why do you use this patch?Kelvinhttps://www.blogger.com/profile/15478327761974929811noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-72463210375005870192012-08-12T23:12:41.228+10:002012-08-12T23:12:41.228+10:00This patch breaks compilation with BFS disabled.This patch breaks compilation with BFS disabled.Oleksandr Natalenkohttps://www.blogger.com/profile/12098091624630953604noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-68270632614907301362012-08-12T18:51:32.245+10:002012-08-12T18:51:32.245+10:00It works for me as well (with CONFIG_NO_HZ). Thank...It works for me as well (with CONFIG_NO_HZ). Thanks _sid_!Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-59658082072915973552012-08-12T00:56:54.248+10:002012-08-12T00:56:54.248+10:00Thank you very much for testing and reporting back...Thank you very much for testing and reporting back. Nice to hear that it works for you._sid_noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-22682304567179319192012-08-12T00:43:43.382+10:002012-08-12T00:43:43.382+10:00Thank you very much for testing and reporting back...Thank you very much for testing and reporting back. Nice to hear that it works for you._sid_noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-66371387340624708002012-08-11T22:38:13.759+10:002012-08-11T22:38:13.759+10:00Thank you very much! It really works!Thank you very much! It really works!Kelvinhttps://www.blogger.com/profile/15478327761974929811noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-24202090094482213012012-08-11T15:24:12.002+10:002012-08-11T15:24:12.002+10:00At least it works here.At least it works here._sid_noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-1764104753113033782012-08-11T12:57:01.786+10:002012-08-11T12:57:01.786+10:00Can it compile with CONFIG_NO_HZ? CONFIG_NO_HZ can...Can it compile with CONFIG_NO_HZ? CONFIG_NO_HZ cannot be enabled since kernel 3.4.6. Thanks.Kelvinhttps://www.blogger.com/profile/15478327761974929811noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-44501926637937175932012-08-11T05:41:54.048+10:002012-08-11T05:41:54.048+10:00Second try, this time with CONFIG_NUMA=y hopefully...Second try, this time with CONFIG_NUMA=y hopefully correctly taken into account. Compiled with CONFIG_NUMA=y, CONFIG_NUMA_EMU=y and booted on a non-NUMA machine. No problems here so far. And no guarantees for you.<br /><br />http://www.file-upload.net/download-4659064/3.5-sched-bfs-424-sid-2.patch.html_sid_noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-5457477651221610522012-08-11T04:51:41.834+10:002012-08-11T04:51:41.834+10:00Please make a diff before doing the porting work. ...Please make a diff before doing the porting work. This is very important. Scheduler domain for ALLNODES option has been deprecated with mainline. So, I have fixed this on the RIFS releasesXhttps://www.blogger.com/profile/08096886089147792760noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-2281785164571931022012-08-11T03:48:38.452+10:002012-08-11T03:48:38.452+10:00I'm running this patch for a few hours on top ...I'm running this patch for a few hours on top of Linux 3.5.1 without any problems, but contrary to you I have CONFIG_NUMA unset since it is not needed on my machine._sid_noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-11267101905509191452012-08-10T20:29:40.292+10:002012-08-10T20:29:40.292+10:00Also, after applying your patch, there is the foll...Also, after applying your patch, there is the following error when compiling:<br />kernel/sched/bfs.c: In function 「sd_init_ALLNODES」:<br />kernel/sched/bfs.c:6233:1: Error: 「SD_ALLNODES_INIT」 undeclared (first use in this function)<br />kernel/sched/bfs.c:6233:1: Note: each undeclared identifier is reported only once for each function it appears in<br />kernel/sched/bfs.c: In function 「sd_init_NODE」:<br />kernel/sched/bfs.c:6234:1: Error: 「SD_NODE_INIT」 undeclared (first use in this function)Kelvinhttps://www.blogger.com/profile/15478327761974929811noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-24256599051361388602012-08-10T18:26:43.776+10:002012-08-10T18:26:43.776+10:00Have you tried this patch yourself? This should no...Have you tried this patch yourself? This should not work. The error will be the same as applying patch-3.4-ck3 to kernel 3.4.6 or above. (I am not sure. I have not tried)Kelvinhttps://www.blogger.com/profile/15478327761974929811noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-88491710940005553842012-08-10T08:56:36.556+10:002012-08-10T08:56:36.556+10:00Just my simple port of BFS 424 to Linux 3.5 and 3....Just my simple port of BFS 424 to Linux 3.5 and 3.5.1. No guarantees.<br /><br />http://www.file-upload.net/download-4655945/3.5-sched-bfs-424.patch.html_sid_noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-42669936386900877272012-08-10T08:28:04.169+10:002012-08-10T08:28:04.169+10:00My simple port of BFS 424 to Linux 3.5 and 3.5.1. ...My simple port of BFS 424 to Linux 3.5 and 3.5.1. No guarantees.<br /><br />http://www.file-upload.net/download-4655945/3.5-sched-bfs-424.patch.html_sid_noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-25240074223277791512012-08-07T21:20:37.277+10:002012-08-07T21:20:37.277+10:00no risk, no fun.no risk, no fun.Anonymousnoreply@blogger.com