tag:blogger.com,1999:blog-6469704299235308349.post3140366824671801779..comments2024-03-28T15:50:13.644+11:00Comments on -ck hacking: SMT Nice 6ckhttp://www.blogger.com/profile/02904761195451530213noreply@blogger.comBlogger51125tag:blogger.com,1999:blog-6469704299235308349.post-47574507695462237552014-08-20T10:01:50.972+10:002014-08-20T10:01:50.972+10:00@Manuel Sorry that I can't remember which emai...@Manuel Sorry that I can't remember which email to LKML, but we are looking at the same bug on freedesktop.org. As I google search today, there are still similar issues reported in 3.16. Alfred Chenhttps://www.blogger.com/profile/03164306846702841944noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-50564464145236279252014-08-18T18:15:45.039+10:002014-08-18T18:15:45.039+10:00@Manuel
Gpu issues are very odd, when it happens o...@Manuel<br />Gpu issues are very odd, when it happens on 3.15, 3 of my GM45 machine have different behaviors. Re-test on 1 of them using 3.16, mpv playback, suspend/resume, mpv playback again works good. Or you mean suspend while playing back?Alfred Chenhttps://www.blogger.com/profile/03164306846702841944noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-13423066226810523402014-08-17T06:34:22.305+10:002014-08-17T06:34:22.305+10:00@Alfred Chen: Of course, it's very off topic o...@Alfred Chen: Of course, it's very off topic on here. I only wanted to leave a little message. With vanilla 3.16.1 video playback in SMplayer or VLC will not display anything other than black playback content with sound(!) after a resume from hibernation, and it's not related to TOI. The result with 3.15.y-gc is provoking a gpu ring reset (what I meant with "not healing" the issue) on my GM45 graphics and I'm somekind of proud to have managed to replicate this behaviour with my reworked patch without breaking the rest of the kernel ^^. :-)<br />I'm a bit annoyed/tired to file another bug in maybe another bugzilla, where the assignees don't seem(!) to fix -- but are able to push 'masses' of new code into the kernel in the meantime. Please, don't read this as flames, maybe I'm just too spoiled by the experience of this -ck related community: Just see the fast efforts that were possible with Con's SMTnice.<br /><br />Best regards, and to all, please, keep up your good work,<br />ManuelAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-52299905097810879722014-08-16T16:58:47.197+10:002014-08-16T16:58:47.197+10:00@Manel
It's a little off topic about bfs. The ...@Manel<br />It's a little off topic about bfs. The i915 revert patch I added in 3.15 b/c it breaks my GM45 chipset machines, I found the first introduction is to fix an issue but finally turns out it hasn't, but maintainer pull it to upstream as a fix. I have no idea about the gpu setup codes but have to revert it, lucky this works. That's the sorry in 3.15. In 3.16, vanilla kernel works without drm issue here, so I doesn't cherry-pick that revert patch from 3.15.<br />I don't use TOI, suspend/resume works fine with my machines in 3.16.Alfred Chenhttps://www.blogger.com/profile/03164306846702841944noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-88776299029849253942014-08-16T10:38:08.851+10:002014-08-16T10:38:08.851+10:00@ post-factum: I'm embarrassed, I haven't ...@ post-factum: I'm embarrassed, I haven't looked at the end of Nigel's repo. My understandable excuse is: It doesn't provide a proper sort by date... <br /><br />@ Alfred Chen: I've now spent several hours to adjust this i915 related revert patch, step-wise, as Im no developer. The result still doesn't heal the issue but brings back the known behaviour of your 3.15.y-gc patch. E.g. getting after a resume from hibernation: <br />[ 218.708017] [drm] stuck on render ring<br />[ 218.709466] [drm] GPU HANG: ecode 0:0x00000000, in Xorg [793], reason: Ring hung, action: reset<br /><br />But that at least has no impact and reenables me to work on the resumed machine.<br />Patch: http://pastebin.com/ufGe87LM<br />I hope I haven't forgotten something for the complete patch.<br /><br />BTW, I'm now using your full patch collection for 3.16.y-gc and it's working fine, so far. <br /><br />Best regards, Manuel KrauseAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-841525463427638632014-08-15T23:30:51.051+10:002014-08-15T23:30:51.051+10:00No, thank you for your tireless testing and quick ...No, thank you for your tireless testing and quick reporting back which helped this code mature much faster than it otherwise would :)ckhttps://www.blogger.com/profile/02904761195451530213noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-72890766151382945922014-08-15T23:19:02.044+10:002014-08-15T23:19:02.044+10:00Yeah, that did the trick everything is working nic...Yeah, that did the trick everything is working nicely now thx for your effort.Anonymoushttps://www.blogger.com/profile/06732867168326047410noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-34857544881713204882014-08-15T22:49:47.122+10:002014-08-15T22:49:47.122+10:00SMT nice 6 incremental and full patches uploaded. ...SMT nice 6 incremental and full patches uploaded. Thanks to all those that keep testing and finding bugs! This one has to be close to release quality now.ckhttps://www.blogger.com/profile/02904761195451530213noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-74013580403861242112014-08-15T22:37:08.496+10:002014-08-15T22:37:08.496+10:00Found a bug. Will post an update shortly.Found a bug. Will post an update shortly.ckhttps://www.blogger.com/profile/02904761195451530213noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-33545356555540394832014-08-15T15:45:17.765+10:002014-08-15T15:45:17.765+10:00Why to port TOI manually if there's official o...Why to port TOI manually if there's official one?Oleksandr Natalenkohttps://www.blogger.com/profile/12098091624630953604noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-10270181535530070322014-08-15T12:44:29.701+10:002014-08-15T12:44:29.701+10:00Without your i915 related revert, not shipped with...Without your i915 related revert, not shipped with your current collection, I'm not able to survive a suspend-to-disk with open videos in SMPLAYER. I'm also lacking brain, to port the related patch myself to 3.16.y-gc.<br /><br />What I've done successfully is, to port/"edit" TuxOnIce to 3.16. Completely unofficial: and pastebin doesn't let it be posted as of greater than 500kB.*grrr*<br /><br />So, I'd say good night, <br />Manuel KrauseAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-91486316645422364702014-08-15T08:34:35.382+10:002014-08-15T08:34:35.382+10:00Thanks. All those threads are niced. I'm guess...Thanks. All those threads are niced. I'm guessing ffmpeg is a combination of niced and un-niced processes and the un-niced ones are making the niced ones go to sleep which is pretty much how smt nice is supposed to work. If it's not time critical then you can run the whole of ffmpeg as sched idleprio so they're all treated equally. If it is, then it's unusual to nice the encoding threads so heavily. At the moment the baseline bias is 100% and decreases with decreasing nice levels. On the next incarnation I'm planning on making it 75% (to account for the increased cpu power overall) and configurable, but the extra code/decision making in a configurable version is what has put me off doing it so far.<br /><br />smt_should_schedule needs to be called on every single scheduling decision so that's expected but I guess since it's such a hot spot it needs to be ultra optimised.<br /><br />Thanks for your testing so far.ckhttps://www.blogger.com/profile/02904761195451530213noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-53602776517300439422014-08-15T03:47:35.285+10:002014-08-15T03:47:35.285+10:00Ok, the thread readout looks pretty much the same:...Ok, the thread readout looks pretty much the same:<br />http://paste.ubuntu.com/8047135/<br />But the overall from top looks like this:<br />%Cpu(s): 0.8 us, 0.0 sy, 99.0 ni, 0.0 id, 0.0 wa, 0.1 hi, 0.0 si, 0.0 st<br />One thing in addition perf top shows nothing specific to the scheduler in the higher up functions without smt nice but if I turn it on it shows smt_should_schedule in the top 5 functions.Anonymoushttps://www.blogger.com/profile/06732867168326047410noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-3667757489415341132014-08-15T03:32:07.096+10:002014-08-15T03:32:07.096+10:00I did some more testing of the SMT nice problem wi...I did some more testing of the SMT nice problem with threads (ffmpeg)<br />Here is the thread readout from ps:<br />http://paste.ubuntu.com/8047001/<br />top shows this as overall:<br /> %Cpu(s): 6.9 us, 0.0 sy, 36.2 ni, 53.4 id, 0.0 wa, 3.4 hi, 0.0 si, 0.0 st<br />I also played around with SCHED_ISO and the nice levels but nothing changes seems like some threads are starving I will also test this without smtnice in a couple of minutes...Anonymoushttps://www.blogger.com/profile/06732867168326047410noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-36833740270074006902014-08-14T10:27:05.506+10:002014-08-14T10:27:05.506+10:00@Manuel Krause
smtnice is not yet be ported in my ...@Manuel Krause<br />smtnice is not yet be ported in my linux-3.16.y-gc for two reason, firstly, smtnice is new and still under testing in this thread, secondly I don't have smt hardware to test it.<br /><br />>did you have changed much from 3.15 to 3.16? <br />For the 3 commits I have posted here to port 0449 to 3.16, I don't change bfs logic besides the tsk_is_polling thing. Most works are sync up mainline changes in core.c, and for 3.16, those are topology level/sd related and none for the core code. <br />The rest bfs commits on linux-3.16.y-gc are ported from linux-3.15.y-gc, I have posted and explain them in previous threads, nothing new.<br /><br />I do have multiple queue locking code for bfs, but it need to rebase to 3.16 then retest.<br /> <br />>exist & imply problems on 3.15 ? <br />Both code works on 3.15 and 3.16 for me, so I can't tell if problems caused by this.Alfred Chenhttps://www.blogger.com/profile/03164306846702841944noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-90462487806584864642014-08-14T06:13:32.928+10:002014-08-14T06:13:32.928+10:00changes for 3.15 to 3.16:
https://bitbucket.org/a...changes for 3.15 to 3.16:<br /><br />https://bitbucket.org/alfredchen/linux-gc/commits/22dc036748dcd2a5effc7c233e1bd287c24363b4?at=linux-3.16.y-gc<br /><br />kernelOfTruthnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-76553877444112952912014-08-14T06:05:53.647+10:002014-08-14T06:05:53.647+10:00****, sorry for forgetting half of my question to ...****, sorry for forgetting half of my question to Alfred Chen...<br />did you have changed much from 3.15 to 3.16? <br /><br />And does the following<br /><br />>Alfred Chen13 August 2014 22:46<br />><br />>@ck<br />>One remarkable thing when I port 0449 to 3.16 is tsk_is_polling define in the >original bfs.c<br />>-#ifndef tsk_is_polling<br />>-#define tsk_is_polling(t) 0<br />>which total disabled tsk_is_polling function and always make >smp_send_reschedule(cpu) called in resched_task().<br />><br />>Is this the bfs intend design?<br />><br />>When porting, I remove the defines and use the mainline new routines.<br /><br />exist & imply problems on 3.15 ? <br />Thx&sorry.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-9690909558641977472014-08-14T05:56:28.625+10:002014-08-14T05:56:28.625+10:00I'd really like to read something from Con abo...I'd really like to read something from Con about the _additional_ proposed patches by Alfred Chen (not only for 3.16 as in this thread here). He already advertised most/all of his BFS related patches for 3.15 when I tried Con's URWlocks revisitation IIRC. <br /><br />So, all said and asked in a friendly tone: Con, what's your opinion? And Alfred Chen: Are your patches ck-smtnice-5-ready already?<br /><br />Best regards, Manuel KrauseAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-25874370065358221062014-08-14T03:11:02.244+10:002014-08-14T03:11:02.244+10:00I'll upload an incremental branch shortlyI'll upload an incremental branch shortlykernelOfTruthnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-12524287848563203552014-08-14T03:10:04.428+10:002014-08-14T03:10:04.428+10:00@Anonymous:
thanks - now I understand what that m...@Anonymous:<br /><br />thanks - now I understand what that meant ;)<br /><br />that current patch from github works for you, right ?kernelOfTruthnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-84919215536700827682014-08-14T02:13:33.264+10:002014-08-14T02:13:33.264+10:00I have informed you that your patches can not be i...I have informed you that your patches can not be integrated<br /><br />http://ck-hack.blogspot.de/2014/08/smthyperthreading-nice-and-scheduling.html<br /><br />see Anonymous9 August 2014 02:39<br /><br />/scratches head?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-73097050451186522222014-08-14T02:06:37.844+10:002014-08-14T02:06:37.844+10:00and here the patch for your convenience:
https://...and here the patch for your convenience:<br /><br />https://github.com/kernelOfTruth/linux/commit/5c452befc4438d9422e74606f05ffedef5ef5b31.diff<br /><br />(adding a .diff to the commit number/address)kernelOfTruthnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-5834298561144133842014-08-14T02:05:14.631+10:002014-08-14T02:05:14.631+10:00ctrl + a, ctrl + v
probably not "good enough...ctrl + a, ctrl + v<br /><br />probably not "good enough" :P<br /><br />anyway - haven't worked with git (github) for some time<br /><br />here's the branch:<br /><br />https://github.com/kernelOfTruth/linux/commits/linux-3.16.0-ck1-smtnice5<br /><br /><br />that should be more easy to fork and work withkernelOfTruthnoreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-88103227245214054642014-08-14T01:49:59.960+10:002014-08-14T01:49:59.960+10:00In fact, -Naur didn't help, and your patch is ...In fact, -Naur didn't help, and your patch is still broken :). How do you paste it?Oleksandr Natalenkohttps://www.blogger.com/profile/12098091624630953604noreply@blogger.comtag:blogger.com,1999:blog-6469704299235308349.post-75449500137122804242014-08-14T00:18:52.542+10:002014-08-14T00:18:52.542+10:00thank you ! :)thank you ! :)kernelOfTruthnoreply@blogger.com