Mesa hm2/hm2_7i96s.0: error finishing read
20 Oct 2022 12:23 #254598
by bensttech
I have also now tried the 6 series kernel that is in experimental and that has same issues on my hardware, (yet to test on all 3 boxes).
I am compiling a spreadsheet of all my testing which i will share soon, I want to try and find some kind of pattern. I think next steps is we need to take a case to the preempt_RT patch maintainers and see if they can shed so light on this, issue will be i doubt any of them use linuxCNC or are likly to have a mesa ethernet board to reproduce the issue, I dont know enough about the workings of the preempt rt kernel to know how else we can compare/profile the versions in a way that might be relevant to this issue.
Replied by bensttech on topic Mesa hm2/hm2_7i96s.0: error finishing read
Thats interesting, i did try the r8168-dkms driver on 5.10 but it did not seem to make much difference on my hardware.He pointed me to this post on the Odroid wiki demonstrating the kernel drivers are 15-30% slower than the native realtek driver. He said that he experinced following errors while jogging with the 5.10 kernel which was resolved after compiling the realtek 8125 driver!wiki.odroid.com/odroid-h3/hardware/insta...net_driver_on_h3plus
I have also now tried the 6 series kernel that is in experimental and that has same issues on my hardware, (yet to test on all 3 boxes).
I am compiling a spreadsheet of all my testing which i will share soon, I want to try and find some kind of pattern. I think next steps is we need to take a case to the preempt_RT patch maintainers and see if they can shed so light on this, issue will be i doubt any of them use linuxCNC or are likly to have a mesa ethernet board to reproduce the issue, I dont know enough about the workings of the preempt rt kernel to know how else we can compare/profile the versions in a way that might be relevant to this issue.
The following user(s) said Thank You: arvidb, tommylight
Please Log in or Create an account to join the conversation.
21 Oct 2022 00:34 - 21 Oct 2022 00:36 #254638
by PCW
Replied by PCW on topic Mesa hm2/hm2_7i96s.0: error finishing read
I don't seem to have an issue on my 8300 USDT and a 5.18 kernel
(This is after a couple hours with a youtube video running)
(This is after a couple hours with a youtube video running)
Attachments:
Last edit: 21 Oct 2022 00:36 by PCW.
Please Log in or Create an account to join the conversation.
21 Oct 2022 01:05 #254640
by rodw
Replied by rodw on topic Mesa hm2/hm2_7i96s.0: error finishing read
Interesting. what NIC card does HP use? This should report itI don't seem to have an issue on my 8300 USDT and a 5.18 kernel
sudo ethtool -i eth0Please Log in or Create an account to join the conversation.
21 Oct 2022 01:23 #254642
by PCW
Replied by PCW on topic Mesa hm2/hm2_7i96s.0: error finishing read
Intel 82579LM
Please Log in or Create an account to join the conversation.
21 Oct 2022 01:35 #254643
by rodw
That sums it up. The 5.1x kernels are crap with Realtek NIC's with real time.I wonder how we could report that back through Debian or the kernel maintainers?
It seems a bit silly to bring preempt_RT into the mainstrema kernel to be let down by non-realtime network access.
Replied by rodw on topic Mesa hm2/hm2_7i96s.0: error finishing read
Thanks, I thought HP would use Intel hardware.Intel 82579LM
That sums it up. The 5.1x kernels are crap with Realtek NIC's with real time.I wonder how we could report that back through Debian or the kernel maintainers?
It seems a bit silly to bring preempt_RT into the mainstrema kernel to be let down by non-realtime network access.
Please Log in or Create an account to join the conversation.
21 Oct 2022 05:50 - 21 Oct 2022 05:56 #254648
by bensttech
Replied by bensttech on topic Mesa hm2/hm2_7i96s.0: error finishing read
Intel 82579LM
Thanks, I thought HP would use Intel hardware.
That sums it up. The 5.1x kernels are crap with Realtek NIC's with real time.I wonder how we could report that back through Debian or the kernel maintainers?
More data points but i think there must be something else going on as i have issue on all 3 of the below PC`s and 2 have intel NIC, all are resolved by rolling back 4.19 kernel. One thing i have spotted is that PCW is using mint and I am on deb-11, I know they have similar based but the user space programs are likily to be very different. What version of Mint is it and what spec is your 8300?
All of these running 5.10.0-18-rt-amd64 and linucCNC 2.9 on deb-11 give me big spikes in read.time
HP Compaq Elite 8300:-
CPU:- i5-3470
NIC:- Intel 82579LM
IMD150D :-
CPU:- Intel(R) Celeron(R) CPU J1900,
NIC:-RTL8111/8168/8411
MF94 :- Intel(R) Celeron(R) CPU J3455
NIC:- Intel Corporation I211
Last edit: 21 Oct 2022 05:56 by bensttech. Reason: table not showing
Please Log in or Create an account to join the conversation.
- tommylight
-
- Away
- Moderator
-
Less
More
- Posts: 19188
- Thank you received: 6430
21 Oct 2022 08:35 #254654
by tommylight
Replied by tommylight on topic Mesa hm2/hm2_7i96s.0: error finishing read
Preempt-RT kernel is used by musicians for anything audio related, probably more using Linux for audio than CNC.
So they do not care much for ethernet as long as internet works.
Although i am sure till 4.n.n kernels they used some other realtime kernel, not the RT we use.
I did test it back then, several of those, did not work for CNC.
Maybe something changed, so testing Ubuntu Studio is a good idea since LinuxCNC can be installed from the Debian repo.
Time, need time....
So they do not care much for ethernet as long as internet works.
Although i am sure till 4.n.n kernels they used some other realtime kernel, not the RT we use.
I did test it back then, several of those, did not work for CNC.
Maybe something changed, so testing Ubuntu Studio is a good idea since LinuxCNC can be installed from the Debian repo.
Time, need time....
Please Log in or Create an account to join the conversation.
21 Oct 2022 11:07 #254660
by rodw
Replied by rodw on topic Mesa hm2/hm2_7i96s.0: error finishing read
Its been a while since PCW convinced me to compile the PREMPT_RT kernel patch on Mint 17,3 so I could use my 7i76e well before it was supported by PNCCONF. Thre was no other solution back then.
My recolllection is the musos were quite happy with a low latency kernel which floated around on some muso sites back then. PREEMPT_RT attracted interest from the automotive industry (BMW springs to mind). This industry threw their not insubstantial development resources at PREEMPT_RT. After all when an air bag has to be deployed you don't want any latency do you?
Recently I wonder if the automotive guys use ethernet for their CANBUS bus through a vehicle. Maybe not, I have no idea.
But by the end of the day (eg. today), PREEMPT_RT won out over the low lateny kernel and is headed to be fully incorporated into the standard LInux kernel one day in the not to distant future.
I suspect that the kernel guys got something wrong by not considering realtime when they combined all of the Realtek drivers into a few software drivers.
I somehow think we have a responsibility to take the issue forward now we are part of the Debian ecosystem. I'm just not sure of the best way....
My recolllection is the musos were quite happy with a low latency kernel which floated around on some muso sites back then. PREEMPT_RT attracted interest from the automotive industry (BMW springs to mind). This industry threw their not insubstantial development resources at PREEMPT_RT. After all when an air bag has to be deployed you don't want any latency do you?
Recently I wonder if the automotive guys use ethernet for their CANBUS bus through a vehicle. Maybe not, I have no idea.
But by the end of the day (eg. today), PREEMPT_RT won out over the low lateny kernel and is headed to be fully incorporated into the standard LInux kernel one day in the not to distant future.
I suspect that the kernel guys got something wrong by not considering realtime when they combined all of the Realtek drivers into a few software drivers.
I somehow think we have a responsibility to take the issue forward now we are part of the Debian ecosystem. I'm just not sure of the best way....
The following user(s) said Thank You: tommylight
Please Log in or Create an account to join the conversation.
21 Oct 2022 11:54 #254662
by rodw
Replied by rodw on topic Mesa hm2/hm2_7i96s.0: error finishing read
So I reported it. Please feel frre to add your own observations to the issue report.
bugs.debian.org/cgi-bin/bugreport.cgi?bug=1022170
bugs.debian.org/cgi-bin/bugreport.cgi?bug=1022170
The following user(s) said Thank You: tommylight
Please Log in or Create an account to join the conversation.
- tommylight
-
- Away
- Moderator
-
Less
More
- Posts: 19188
- Thank you received: 6430
21 Oct 2022 13:04 #254677
by tommylight
Replied by tommylight on topic Mesa hm2/hm2_7i96s.0: error finishing read
I doubt it will become standard, ever, as it has to stop plenty of functions used for power saving, and since the era of PC domination is over, laptops and tablets rule now.
Using Preempt-RT on laptops a lot, so i know they use much more power and heat a lot more with it. On most when RT is in use running LinuxCNC the processor will remain at max speed (not turbo) all the time.
Using Preempt-RT on laptops a lot, so i know they use much more power and heat a lot more with it. On most when RT is in use running LinuxCNC the processor will remain at max speed (not turbo) all the time.
Please Log in or Create an account to join the conversation.
Time to create page: 0.193 seconds