Hardware latency tests, used PC's

More
6 years 4 months ago #172026 by tommylight
Replied by tommylight on topic Hardware latency tests, used PC's
apt list linuxcnc-uspace | awk '{print $2}'

shows
1:2.9.0~pre0.2228.gcda96a48f
The number behind pre0 is the build version, in my case 2228.

Please Log in to join the conversation.

More
6 years 4 months ago #172029 by seuchato
Replied by seuchato on topic Hardware latency tests, used PC's
OK, so, if the build No. is included in the package version, I suggest to stick with the package version. Any objections?

Please Log in to join the conversation.

More
6 years 4 months ago #172030 by tommylight
Replied by tommylight on topic Hardware latency tests, used PC's
No.
Thank you.

Please Log in to join the conversation.

More
6 years 4 months ago #172058 by seuchato
Replied by seuchato on topic Hardware latency tests, used PC's
New version

Kindly have a look into it and let me know any bugs, amendments, ...

greez
chris

Attachment not found

Please Log in to join the conversation.

More
6 years 4 months ago #172066 by tommylight
Replied by tommylight on topic Hardware latency tests, used PC's
Sooooo close, but no cigar ..... yet.
- Menu works,
- Number of glxgears works,
- Open firefox with vimeo link works, maximise and play work,
- Base thread does not, does not change period and is active even when choosing not to use it. It sticks to 25000 no matter what.
- Open firefox with a youtube link works, but does not play nor maximise.
Nice work.
Do not worry about the link to much, it seems vimeo causes more load on the pc than youtube does, and that is good.

Please Log in to join the conversation.

More
6 years 4 months ago #172068 by tommylight
Replied by tommylight on topic Hardware latency tests, used PC's
Here are the results of a run, choose no base thread, but it is there and cutting the test short due to latency.
1592556201.hw
Code:
General info: ASRockLinuxCNC pc tests, version 20200620 ******************************************************************************** , To Be Filled By O.E.M. Bios version P1.40, dated 07/11/2012 ================================================================================ CPU related data: model name : Intel(R) Celeron(R) CPU G1620 @ 2.70GHz cpu cores : 2 stepping : 9 cache size : 2048 KB ================================================================================ RAM: RAM: Maximum Capacity: 16 GB Size: 2048 MB Size: 2048 MB ================================================================================ GPU related data: product: Xeon E3-1200 v2/3rd Gen Core processor Graphics Controller configuration: driver=i915 latency=0 ================================================================================ Os related data: PRETTY_NAME="Linux Mint 19.3" Desktop: , Session: ================================================================================ Kernel related data: Kernel 5.4.0-0.bpo.4-rt-amd64 GRUB_CMDLINE_LINUX_DEFAULT="quiet splash" Cpu idle driver: intel_idle ================================================================================ Keyboard & Mouse related data: Number of mice: 2 Mouse 0 : PixArt USB Optical Mouse Mouse 1 : ABBAHOME Mouse Mice attached to USB: Keyboards attached to USB: ================================================================================ LinuxCNC related data: LinuxCNC version is: 2.9.0~pre0.2235.g8e7842fa9 Preparing for latency test: Started glxgears No. 1 Started glxgears No. 2 Started glxgears No. 3 .......... Started firefox with 'https://www.youtube.com/watch?v=Na44iMZUcZk' Using glxgears, Started No. ================================================================================ Latency testing loop started Fri Jun 19 10:44:01 CEST 2020, should end after Fri Jun 19 10:54:01 CEST 2020 command for latency test is: su cnc -c "latency-histogram --sbinsize 1000 --nox 2>&1 | tee 1592556201.lat &" done command, retval = 0 Note: Using POSIX realtime 0 secs base min: -8.400 uS max: 6.600 uS sdev: 0.500 uS 0 secs servo min: -4.000 uS max: 5.500 uS sdev: 0.800 uS 2 secs base min: -8.400 uS max: 6.600 uS sdev: 0.500 uS 2 secs servo min: -4.000 uS max: 5.500 uS sdev: 0.800 uS 3 secs base min: -8.400 uS max: 6.600 uS sdev: 0.500 uS 3 secs servo min: -4.000 uS max: 5.500 uS sdev: 0.800 uS 4 secs base min: -9.200 uS max: 8.500 uS sdev: 0.500 uS 4 secs servo min: -4.000 uS max: 5.500 uS sdev: 0.800 uS Unexpected realtime delay on task 0 with period 25000 This Message will only display once per session. Run the Latency Test and resolve before continuing. [0K Remaining 0d 0h:9m:55s, press ^C to abort... 6 secs base min: -24.500 uS max: 26.600 uS sdev: 0.700 uS 6 secs servo min: -41.800 uS max: 41.900 uS sdev: 1.500 uS 7 secs base min: -24.500 uS max: 26.600 uS sdev: 0.700 uS 7 secs servo min: -41.800 uS max: 41.900 uS sdev: 1.500 uS 9 secs base min: -24.500 uS max: 26.600 uS sdev: 0.700 uS 9 secs servo min: -41.800 uS max: 41.900 uS sdev: 1.400 uS 10 secs base min: -24.500 uS max: 26.600 uS sdev: 0.700 uS 10 secs servo min: -41.800 uS max: 41.900 uS sdev: 1.500 uS Unexpected realtime delay on task 0 with period 25000 Useless to proceed, read the relevant part in 1592556201.lat Cleaning up..| ......Closing browser firefox with WID of bye.
And 1592556201.lat
Code:
Note: Using POSIX realtime 0 secs base min: -8.400 uS max: 6.600 uS sdev: 0.500 uS 0 secs servo min: -4.000 uS max: 5.500 uS sdev: 0.800 uS 2 secs base min: -8.400 uS max: 6.600 uS sdev: 0.500 uS 2 secs servo min: -4.000 uS max: 5.500 uS sdev: 0.800 uS 3 secs base min: -8.400 uS max: 6.600 uS sdev: 0.500 uS 3 secs servo min: -4.000 uS max: 5.500 uS sdev: 0.800 uS 4 secs base min: -9.200 uS max: 8.500 uS sdev: 0.500 uS 4 secs servo min: -4.000 uS max: 5.500 uS sdev: 0.800 uS Unexpected realtime delay on task 0 with period 25000 This Message will only display once per session. Run the Latency Test and resolve before continuing. 6 secs base min: -24.500 uS max: 26.600 uS sdev: 0.700 uS 6 secs servo min: -41.800 uS max: 41.900 uS sdev: 1.500 uS 7 secs base min: -24.500 uS max: 26.600 uS sdev: 0.700 uS 7 secs servo min: -41.800 uS max: 41.900 uS sdev: 1.500 uS 9 secs base min: -24.500 uS max: 26.600 uS sdev: 0.700 uS 9 secs servo min: -41.800 uS max: 41.900 uS sdev: 1.400 uS 10 secs base min: -24.500 uS max: 26.600 uS sdev: 0.700 uS 10 secs servo min: -41.800 uS max: 41.900 uS sdev: 1.500 uS

Please Log in to join the conversation.

More
6 years 4 months ago #172078 by seuchato
Replied by seuchato on topic Hardware latency tests, used PC's
Tom

new version:
- removed option to choose webpage, as I could not get youtube to start
+ basethread can now be disabled
+ basinterval can now be set
+ logged swap usage
+ states now, if HD is rotational or not
+ only last two lines containing latency data are now attached to xx.hw file
+ some cleanup and beautifying of outputs

Again: kindly check and report

greez
chris

Attachment not found

Please Log in to join the conversation.

More
6 years 4 months ago #172079 by seuchato
Replied by seuchato on topic Hardware latency tests, used PC's

tommylight wrote: Here are the results of a run, choose no base thread, but it is there and cutting the test short due to latency.
1592556201.hw

Code:
... 4 secs servo min: -4.000 uS max: 5.500 uS sdev: 0.800 uS Unexpected realtime delay on task 0 with period 25000 This Message will only display once per session. Run the Latency Test and resolve before continuing. [0K Remaining 0d 0h:9m:55s, press ^C to abort... 6 secs base min: -24.500 uS max: 26.600 uS sdev: 0.700 uS 6 secs servo min: -41.800 uS max: 41.900 uS sdev: 1.500 uS ... Unexpected realtime delay on task 0 with period 25000 Useless to proceed, read the relevant part in 1592556201.lat Cleaning up..| ......Closing browser firefox with WID of bye.


That is meant to be a feature, not a bug :-).

My strong believe is, any realtime delay renders a candidate machine useless unless the cause of the realtime delay can be found and eliminated. Am I wrong? If so, I will add an option to continue latency test despite realtime delays.
greez
chris

Please Log in to join the conversation.

More
6 years 4 months ago #172080 by tommylight
Replied by tommylight on topic Hardware latency tests, used PC's
- Default run, works, exist on latency over 25000 as it should,
- Run with no base thread exits properly,
- Run with no base thread when pressing ctrl+c doe snot exit, see bellow,
- Run with base thread at 100000 does not exit cleanly,
- Run with base thread at 100000 when pressing ctrl+c does not exit, see below

----This is the bellow part:
After ctrl+c the terminal will continuously show this:
Code:
X Error of failed request: BadWindow (invalid Window parameter) Major opcode of failed request: 3 (X_GetWindowAttributes) Resource id in failed request: 0x5c00002 Serial number of failed request: 18 Current serial number in output stream: 19 X Error of failed request: BadWindow (invalid Window parameter) Major opcode of failed request: 3 (X_GetWindowAttributes) Resource id in failed request: 0x6600002 Serial number of failed request: 18 Current serial number in output stream: 19 X Error of failed request: BadWindow (invalid Window parameter) Major opcode of failed request: 3 (X_GetWindowAttributes) Resource id in failed request: 0x6400002 Serial number of failed request: 18 Current serial number in output stream: 19 X Error of failed request: BadWindow (invalid Window parameter) Major opcode of failed request: 3 (X_GetWindowAttributes) Resource id in failed request: 0x6800002 Serial number of failed request: 18 Current serial number in output stream: 19
It does not exit after several ctrl+c, has to be terminated.
So so close.... :)

Please Log in to join the conversation.

More
6 years 4 months ago #172081 by tommylight
Replied by tommylight on topic Hardware latency tests, used PC's

seuchato wrote:

tommylight wrote: Here are the results of a run, choose no base thread, but it is there and cutting the test short due to latency.
1592556201.hw

Code:
... 4 secs servo min: -4.000 uS max: 5.500 uS sdev: 0.800 uS Unexpected realtime delay on task 0 with period 25000 This Message will only display once per session. Run the Latency Test and resolve before continuing. [0K Remaining 0d 0h:9m:55s, press ^C to abort... 6 secs base min: -24.500 uS max: 26.600 uS sdev: 0.700 uS 6 secs servo min: -41.800 uS max: 41.900 uS sdev: 1.500 uS ... Unexpected realtime delay on task 0 with period 25000 Useless to proceed, read the relevant part in 1592556201.lat Cleaning up..| ......Closing browser firefox with WID of bye.


That is meant to be a feature, not a bug :-).

My strong believe is, any realtime delay renders a candidate machine useless unless the cause of the realtime delay can be found and eliminated. Am I wrong? If so, I will add an option to continue latency test despite realtime delays.
greez
chris

The thing is i did not choose a base thread to run, just the servo thread, but it was still using the base thread.
Now it is working properly as far as this is concerned.
Thank you very much, nice work.

Please Log in to join the conversation.

Time to create page: 0.132 seconds
Powered by Kunena Forum