#linuxcnc-devel Logs

Jan 13 2020

07:34 AM jepler: eep that's mine. fixing
11:03 AM linuxcnc-build: build #3494 of 1903.clang-wheezy-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1903.clang-wheezy-amd64/builds/3494 blamelist: Dewey Garrett <dgarrett@panix.com>
11:10 AM linuxcnc-build: build #3493 of 1902.clang-wheezy-rtai-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1902.clang-wheezy-rtai-i386/builds/3493 blamelist: Dewey Garrett <dgarrett@panix.com>
11:11 AM linuxcnc-build: build #5562 of 1301.rip-precise-rtai-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1301.rip-precise-rtai-i386/builds/5562 blamelist: Dewey Garrett <dgarrett@panix.com>
11:15 AM linuxcnc-build: build #4503 of 1403.rip-wheezy-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1403.rip-wheezy-amd64/builds/4503 blamelist: Dewey Garrett <dgarrett@panix.com>
11:16 AM linuxcnc-build: build #6341 of 1300.rip-precise-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1300.rip-precise-i386/builds/6341 blamelist: Dewey Garrett <dgarrett@panix.com>
11:19 AM linuxcnc-build: build #1359 of 1630.rip-stretch-rtpreempt-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1630.rip-stretch-rtpreempt-amd64/builds/1359 blamelist: Dewey Garrett <dgarrett@panix.com>
11:20 AM linuxcnc-build: build #4502 of 1400.rip-wheezy-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1400.rip-wheezy-i386/builds/4502 blamelist: Dewey Garrett <dgarrett@panix.com>
11:20 AM linuxcnc-build: build #6344 of 1306.rip-precise-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1306.rip-precise-amd64/builds/6344 blamelist: Dewey Garrett <dgarrett@panix.com>
11:47 AM linuxcnc-build: build #2971 of 1520.rip-jessie-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1520.rip-jessie-amd64/builds/2971 blamelist: Dewey Garrett <dgarrett@panix.com>
11:51 AM linuxcnc-build: build #4706 of 1404.rip-wheezy-rtpreempt-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1404.rip-wheezy-rtpreempt-amd64/builds/4706 blamelist: Dewey Garrett <dgarrett@panix.com>
11:51 AM linuxcnc-build: build #2969 of 1500.rip-jessie-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1500.rip-jessie-i386/builds/2969 blamelist: Dewey Garrett <dgarrett@panix.com>
12:00 PM linuxcnc-build: build #2971 of 1530.rip-jessie-rtpreempt-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1530.rip-jessie-rtpreempt-amd64/builds/2971 blamelist: Dewey Garrett <dgarrett@panix.com>
12:01 PM linuxcnc-build: build #4021 of 1402.rip-wheezy-rtpreempt-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1402.rip-wheezy-rtpreempt-i386/builds/4021 blamelist: Dewey Garrett <dgarrett@panix.com>
12:03 PM linuxcnc-build: build #4173 of 1401.rip-wheezy-rtai-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1401.rip-wheezy-rtai-i386/builds/4173 blamelist: Dewey Garrett <dgarrett@panix.com>
12:04 PM linuxcnc-build: build #1359 of 1610.rip-stretch-rtpreempt-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1610.rip-stretch-rtpreempt-i386/builds/1359 blamelist: Dewey Garrett <dgarrett@panix.com>
12:04 PM linuxcnc-build: build #2969 of 1510.rip-jessie-rtpreempt-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1510.rip-jessie-rtpreempt-i386/builds/2969 blamelist: Dewey Garrett <dgarrett@panix.com>
12:04 PM linuxcnc-build: build #6358 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/6358 blamelist: Dewey Garrett <dgarrett@panix.com>
01:58 PM CaptHindsight: jthornton_: sounds right for Craigslist, if your item is still on for more than a few weeks the scammers stop
02:00 PM CaptHindsight: first few days are always the worst
02:03 PM CaptHindsight: look at those angle plates! https://springfieldil.craigslist.org/tls/d/modesto-horizontal-boring-mill-fixture/7041070861.html
02:03 PM CaptHindsight: 36”x48” 96” tall
02:04 PM CaptHindsight: 3 tons ea
02:05 PM CaptHindsight: woops meant this for the other LCNC channel
03:17 PM skunkworks: pcw_home: https://pastebin.com/rZuc6QjL
03:37 PM skunkworks: with isolcpus=1,2,3 - it seems to run a 1khz thread - atleast for 10 minutes so far
03:38 PM jepler: blech it feels awful giving up 75% CPU to LinuxCNC
03:38 PM jepler: but if it gets you to 1kHz maybe it's worth it
03:39 PM jepler: If you are willing it would be very interesting to benchmark with isolcpus=1,3 and 2,3
03:39 PM skunkworks: it isn't too bad... My previous experience was very unusable.. But ether I was doing something wrong - or something was updated to run better on 1 core
03:40 PM skunkworks: oh - I bet it was the force_turbo.. Initally only running 800mhz
03:40 PM jepler: you're running with my most recent image, or something else?
03:40 PM skunkworks: I don't think your latest - the one pre-ntp
03:40 PM jepler: okay
03:40 PM jepler: only difference was installing ntp
03:41 PM skunkworks: right
03:41 PM skunkworks: running good
03:41 PM skunkworks: .00003 following error
03:41 PM jepler: is that better than at 500Hz?
03:41 PM skunkworks: yes
03:41 PM CaptHindsight: LCNC usually tested best on dual core cpu's
03:42 PM skunkworks: but I should tweek the pid too - it is running a P of 500
03:44 PM skunkworks: definatly not an i5 but usable for sure
03:45 PM skunkworks: heh - half the following error with twice the P
03:47 PM jepler: I suppose that isolcpus=[all but 0] means that RT competes with the fewest things for L2 cache and main memory bandwidth
05:07 PM memfrob: skunkworks, you still there?
05:10 PM memfrob: isolcpus=3 would probably give you the best performance between LCNC and non-RT tasks (allowing non-RT tasks the ability to migrate across cores) but still isolating CPU 3 for LCNC (nothing runs on CPU 3 except LCNC)
05:11 PM memfrob: If isolcpus=3 gives worse latency than isolcpus=1,2,3 then there's cache thrashing or something going on due to poor CPU microarchitecture or firmware.
05:13 PM memfrob: CPUs that do a bad job at SMP can have worse latency if you're only isolating the CPU specific to your RT thread. isolcpus=1,2,3 disables CPU migration completely, forcing everything except LCNC to run on only one core, essentially turning the CPU into a single core, plus 1 for LCNC.
05:15 PM memfrob: Sometimes when non-RT threads migrate, even though it's a completely isolated from LCNC, somehow those migrating non-RT threads screw up the latency on a complete unrelated CPU that those threads can't even touch.
05:40 PM skunkworks: memfrob: yes - realtime is a bit worse with just 3..
06:59 PM memfrob: skunkworks, how much worse?
07:03 PM skunkworks: I would have to run both way to know exactly.. Comes down to 1ms servo thread vs 2ms
07:07 PM memfrob: so negligible?
09:27 PM burklefoo: skunkworks: did you try turning off irq coalescing ?