From: Florian Bezdeka <florian.bezdeka@siemens.com>
To: JerryJun <jerryjun123@163.com>
Cc: xenomai@lists.linux.dev
Subject: Re: Re: config-4.19.89-ipipe-9-xenomai-3
Date: Tue, 16 Apr 2024 12:11:03 +0200 [thread overview]
Message-ID: <c60a8f8e8618536f68fd22cd351ede371ae8464f.camel@siemens.com> (raw)
In-Reply-To: <5bb56609.ebf7.18ecdae4495.Coremail.jerryjun123@163.com>
On Thu, 2024-04-11 at 23:02 +0800, JerryJun wrote:
> Yes, we've noticed this change from ipipe to devotail. Although these numbers difference from 4.19 to 5.10, there should be no huge fluctuation in the MSW.
> As you can see in the log below, this phenomenon is sporadic, and the MSW of all real-time threads suddenly becomes larger within 5s, and will continue to grow at this rate. The state code is 00068042 or 00068044 as if these threads are always in secondary mode. The performance of real-time applications cannot be guaranteed after MSW is increased.
> Is this phenomenon related to the config Settings of the kernel or open --enable-lazy-setsched ?
Hard to tell. You might want to trace what the application does when
the problem comes up.
>
> TIMESTAMP CPU PID MSW CSW XSC PF STATE %CPU NAME
> 20240408T211443 - | 2 | 1135 | 1047473 | 2098443 | 1047796 | 0 | 00068044 | 0.4 | threadA === MSW 1047473 === Increase: +1047473
> 20240408T211443 - | 2 | 1138 | 1256963 | 2517148 | 1257187 | 0 | 00068044 | 0.5 | threadB === MSW 1256963 === Increase: +1256963
> 20240408T211443 - | 2 | 1145 | 1054840 | 2085496 | 1082206 | 0 | 00068044 | 2.9 | threadC === MSW 1054840 === Increase: +1054840
> 20240408T211443 - | 2 | 1146 | 1047219 | 2098059 | 1047362 | 0 | 000680c0 | 0.4 | threadD === MSW 1047219 === Increase: +1047219
> 20240408T211443 - | 2 | 1147 | 3088508 | 6187754 | 3088559 | 0 | 00068044 | 1.2 | threadE === MSW 3088508 === Increase: +3088508
> 20240408T211443 - | 2 | 1149 | 1047189 | 2098147 | 1047206 | 0 | 00068044 | 0.4 | threadF === MSW 1047189 === Increase: +1047189
> 20240408T211443 - | 2 | 1352 | 3487 | 5243 | 5267 | 0 | 00068042 | 0.6 | threadG === MSW 3487 === Increase: +3487
>
> 20240408T211448 - | 2 | 1135 | 1063891 | 2131335 | 1064226 | 0 | 00068044 | 0.4 | threadA === MSW 1063891 === Increase: +16418
> 20240408T211448 - | 2 | 1138 | 1276766 | 2556693 | 1276999 | 0 | 00068044 | 0.5 | threadB === MSW 1276766 === Increase: +19803
> 20240408T211448 - | 2 | 1145 | 1072753 | 2112197 | 1112668 | 0 | 00068044 | 3.0 | threadC === MSW 1072753 === Increase: +17913
> 20240408T211448 - | 2 | 1146 | 1063621 | 2130923 | 1063764 | 0 | 000680c0 | 0.4 | threadD === MSW 1063621 === Increase: +16402
> 20240408T211448 - | 2 | 1147 | 3136956 | 6284807 | 3137025 | 0 | 000680c0 | 1.2 | threadE === MSW 3136956 === Increase: +48448
> 20240408T211448 - | 2 | 1149 | 1063619 | 2131068 | 1063636 | 0 | 00068044 | 0.4 | threadF === MSW 1063619 === Increase: +16430
> 20240408T211448 - | 2 | 1352 | 6127 | 9223 | 9227 | 0 | 00068042 | 0.8 | threadG === MSW 6127 === Increase: +2640
>
> 20240408T212029 - | 2 | 1135 | 2183408 | 4374592 | 2183849 | 0 | 00068044 | 0.4 | threadA === MSW 2183408 === Increase: +1119517
> 20240408T212029 - | 2 | 1138 | 2623843 | 5251787 | 2624284 | 0 | 00068044 | 0.5 | threadB === MSW 2623843 === Increase: +1347077
> 20240408T212029 - | 2 | 1145 | 1981556 | 3854421 | 2166359 | 0 | 00068044 | 3.0 | threadC === MSW 1981556 === Increase: +908803
> 20240408T212029 - | 2 | 1146 | 2183315 | 4374666 | 2183478 | 0 | 00068044 | 0.4 | threadD === MSW 2183315 === Increase: +1119694
> 20240408T212029 - | 2 | 1147 | 6433499 | 12888628 | 6437076 | 0 | 00068044 | 1.2 | threadE === MSW 6433499 === Increase: +3296543
> 20240408T212029 - | 2 | 1149 | 2183633 | 4375181 | 2184005 | 0 | 00068044 | 0.4 | threadF === MSW 2183633 === Increase: +1120014
> 20240408T212029 - | 2 | 1352 | 26787 | 43099 | 40217 | 0 | 00068042 | 0.6 | threadG === MSW 26787 === Increase: +20660
>
> 20240408T212034 - | 2 | 1135 | 2199821 | 4407475 | 2200274 | 0 | 00068044 | 0.4 | threadA === MSW 2199821 === Increase: +16413
> 20240408T212034 - | 2 | 1138 | 2643647 | 5291342 | 2644097 | 0 | 00068044 | 0.5 | threadB === MSW 2643647 === Increase: +19804
> 20240408T212034 - | 2 | 1145 | 1999383 | 3880651 | 2197619 | 0 | 00068044 | 2.8 | threadC === MSW 1999383 === Increase: +17827
> 20240408T212034 - | 2 | 1146 | 2199719 | 4407527 | 2199882 | 0 | 00068044 | 0.4 | threadD === MSW 2199719 === Increase: +16404
> 20240408T212034 - | 2 | 1147 | 6481831 | 12985424 | 6485434 | 0 | 00068044 | 1.1 | threadE === MSW 6481831 === Increase: +48332
> 20240408T212034 - | 2 | 1149 | 2200043 | 4408073 | 2200415 | 0 | 00068044 | 0.4 | threadF === MSW 2200043 === Increase: +16410
> 20240408T212034 - | 2 | 1352 | 29725 | 47534 | 44624 | 0 | 00068042 | 0.8 | threadG === MSW 29725 === Increase: +2938
>
>
> At 2024-04-09 19:35:44, "Florian Bezdeka" <florian.bezdeka@siemens.com> wrote:
> > Hi,
> >
> > between 4.19 and 5.10 happened a major change in the underlying
> > infrastructure. We moved from ipipe to dovetail with the first 5.10
> > based release.
> >
> > It might be that the numbers are (completely) different now, but there
> > should be no negative impact.
> >
> > Best regards,
> > Florian
> >
> > On Tue, 2024-04-09 at 11:41 +0800, JerryJun wrote:
> > > hi,xenomai:
> > > We have an issue where the same application was used on different kernels , the MSW became larger in kernel-5.10.199 + xenomai-3.2.4 but not encountered in kernel-4.19.89 + xenomai-3.1
> > > please give me a hint. kernel-config by attached this email
> > >
> > > CPU PID MSW CSW XSC PF STAT %CPU NAME
> > > 0 0 0 9996742 0 0 00018000 99.8 [ROOT/0]
> > > 1 0 0 2024119195 0 0 00018000 95.3 [ROOT/1]
> > > 2 0 0 2586560269 0 0 00018000 96.5 [ROOT/2]
> > > 3 0 0 121156377 0 0 00018000 96.2 [ROOT/3]
> > > 3 2962 0 121156120 0 0 00000046 3.8 [kthead1]
> > > 0 2970 0 9920415 0 0 00000046 0.2 [kthead2]
> > > 1 2975 0 1650262 0 0 00000046 0.0 [kthead3]
> > > 1 2980 0 3099325 0 0 00000046 0.1 [kthead4]
> > > 3 3167 156 181 632 0 000680c0 0.0 XX
> > > 3 3224 9 9 45 0 000680c0 0.0 sysregd
> > > 3 3226 8 8 30 0 000680c0 0.0 sysregd
> > > 3 3233 8 8 30 0 000680c0 0.0 YY
> > > 3 3234 2 3 8 0 00068042 0.0 remote-agent
> > > 1 3236 15 372743488 537438761 0 00068046 1.1 AA_3167
> > > 1 3237 327351466 652251196 327355285 0 00068044 0.9 BB_3167
> > > 1 3238 666888185 1000550068 2334622549 0 00068044 2.6 CC_3167
> > > 1 3239 856330 2163224 1710753 0 00068046 0.0 DD_3167
> > > 1 3240 117 2476148 128 0 00068046 0.0 EE_3167
> > > 1 3241 15 4962912 26 0 00068046 0.0 FF_3167
> > > 2 3249 163093685 326186203 163096062 0 000680c0 0.5 GG
> > > 2 3252 196178072 391844998 196181535 0 00068044 0.5 HH
> > > 2 3259 141521731 260330959 201102670 0 00068044 0.4 II
> > > 2 3260 163141603 326283087 163141808 0 00068044 0.5 JJ
> > > 2 3261 481724449 963372088 481875166 0 00068044 1.3 KK
> > > 2 3263 163130135 326258458 163132953 0 00068044 0.4 LL
> > > 2 3441 8648757 12961988 13009962 0 00068042 0.0 MM
prev parent reply other threads:[~2024-04-16 10:11 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-09 3:41 config-4.19.89-ipipe-9-xenomai-3 JerryJun
2024-04-09 11:35 ` config-4.19.89-ipipe-9-xenomai-3 Florian Bezdeka
2024-04-11 15:02 ` config-4.19.89-ipipe-9-xenomai-3 JerryJun
2024-04-16 10:11 ` Florian Bezdeka [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=c60a8f8e8618536f68fd22cd351ede371ae8464f.camel@siemens.com \
--to=florian.bezdeka@siemens.com \
--cc=jerryjun123@163.com \
--cc=xenomai@lists.linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).