>Sent: Monday, April 27, 2020 10:17 AM
>
>Hi Meng:
>
>> We use Ftrace to check if Cobalt thread is rightly sched; I can help to
>> setup Ftrace env on your machine, if u interested,
>
>Yes, if you have any guideline to help me setup this?
We plan to add Ftrace guide on:
https://github.com/intel/linux-stable-xenomai/wiki/Guide-of-Xenomai-Build-and-Installation-on-Top-of-Ubuntu
Before May 15th, is that OK timeline for you?
>
>________________________________
>
>Best Regards
>Huang Mianbo
>Softlink Inc
>电话:+8613682612959
>
>
> From: Meng, Fino <mailto:fino.m...@intel.com>
> Date: 2020-04-27 10:08
> To: huang...@softlinkcloud.cn <mailto:huang...@softlinkcloud.cn>
> CC: xenomai@xenomai.org <mailto:xenomai@xenomai.org>
> Subject: RE: issue when porting a linux posix applicaiton to a xenomai
> application
>
> >Sent: Sunday, April 26, 2020 7:05 PM
> >
> >Hi all:
> >
> >i have ported a c based application which is a profinet network stack
> and it was running ok in common x86 linux
>platform as a
> >regular process except some realtime issue.
> >
> >For the realtime issue,i have to ported to a xenomai application
> >
> >The profinet application using the posix interface so i use xenomai
> 3.1's posix skin for the porting.
> >
> >But after i had ported it to the xenomai-posix one ,i found it could
> receive network message but it failed to send the
>message
> >out. (by comparison to the regular one using "watch -n 0.1 ifconfig
> eth0"). It seemed that some thread's blocking and i
>did not
> >known how to investegate the issue.
> >
> >The /proc/xenomai/sched/stat was like this:
> >CPU PID MSW CSW XSC PF STAT %CPU
> NAME
> > 0 0 0 0 0 0 00218000 100.0
> [ROOT/0]
> > 1 0 0 0 0 0 00218000 100.0
> [ROOT/1]
> > 2 0 0 119332 0 0 00218000 100.0
> [ROOT/2]
> > 3 0 0 174171 0 0 00218000 99.8
> [ROOT/3]
> > 3 1721 14 2749 2930 0 00248044 0.0
> dksw_linux
> > 3 1723 2 4 7 0 00242042 0.0
> Task_ASYNC_Timer
> > 3 1724 2 4 4 0 00242042 0.0
> Task_OS_Timer
> > 3 1725 4 23 69 0 00242042 0.0
> PrintRS232PostTask
> > 3 1726 2 3 3 0 00242042 0.0
> Pnio_NvData
> > 3 1727 3 5453 5448 0 0024a044 0.0
> Pnio_Idle
> > 3 1728 7 27 188 0 0024a0c0 0.0
> MainAppl
> > 3 1729 2 135222 676095 0 0024a044 0.1
> Pnio_CycleIO
> > 3 1730 3463 3463 3464 0 0024a0c0 0.0
> sniff_packet
> > 3 1731 11 27219 27448 0 0024a044 0.0
> t_IDL
> > 3 1732 6 84 215 0 0024a042 0.0
> t_APP
> > 3 1733 459 4661 8961 0 0024a042 0.0
> t_EDD
> > 3 1734 2 112 166 0 0024a042 0.0
> t_ORG
> > 3 1735 30 1071 7346 0 0024a042 0.0
> t_PNO
> > 3 1736 3 2465 5339 0 0024a042 0.0
> t_TCP
> > 3 1737 2 26 42 0 0024a042 0.0
> t_POF
> > 3 1738 2 10 10 0 00242042 0.0
> t_REM
> > 3 1739 3 53 85 0 0024a042 0.0
> t_STP
> > 3 1740 2 5 30 0 0024a042 0.0
> Pnio_PnpbPost
> > 3 1741 2 25 138 0 0024a042 0.0
> Pnio_Pnpb
> > 0 0 0 165200 0 0 00000000 0.0
> [IRQ2305: [timer]]
> > 1 0 0 12551 0 0 00000000 0.0
> [IRQ2305: [timer]]
> > 2 0 0 893793 0 0 00000000 0.0
> [IRQ2305: [timer]]
> > 3 0 0 2791299 0 0 00000000 0.1
> [IRQ2305: [timer]]
> >the /proc/xenomai/sched/thread was like this:
> >CPU PID CLASS TYPE PRI TIMEOUT STAT NAME
> > 0 0 idle core -1 - R [ROOT/0]
> > 1 0 idle core -1 - R [ROOT/1]
> > 2 0 idle core -1 - R [ROOT/2]
> > 3 0 idle core -1 - R [ROOT/3]
> > 3 1721 rt cobalt 19 698ms369us D dksw_linux
> > 3 1723 rt cobalt 20 - Wr
> Task_ASYNC_Timer
> > 3 1724 rt cobalt 20 - Wr
> Task_OS_Timer
> > 3 1725 rt cobalt 6 - Wr
> PrintRS232PostTask
> > 3 1726 rt cobalt 5 - Wr Pnio_NvData
> > 3 1727 rt cobalt 1 461ms50us Dr Pnio_Idle
> > 3 1728 rt cobalt 4 - Xr MainAppl
> > 3 1729 rt cobalt 19 8ms828us Dr
> Pnio_CycleIO
> > 3 1730 rt cobalt 19 - Xr
> sniff_packet
> > 3 1731 rt cobalt 11 98ms679us Dr t_IDL
> > 3 1732 rt cobalt 10 - Wr t_APP
> > 3 1733 rt cobalt 18 - Wr t_EDD
> > 3 1734 rt cobalt 17 - Wr t_ORG
> > 3 1735 rt cobalt 16 - Wr t_PNO
> > 3 1736 rt cobalt 15 - Wr t_TCP
> > 3 1737 rt cobalt 14 - Wr t_POF
> > 3 1738 rt cobalt 13 - Wr t_REM
> > 3 1739 rt cobalt 12 - Wr t_STP
> > 3 1740 rt cobalt 10 - Wr
> Pnio_PnpbPost
> > 3 1741 rt cobalt 9 - Wr Pnio_Pnpb
> >
> >the dmesg about xenomai:
> >
> >[ 0.000000] Linux version 3.18.20-xenomai-3.1-20200417-02
> (root@cpac) (gcc version 5.5.0 20171010 (Ubuntu
>5.5.0-
> >12ubuntu1) ) #1 SMP Fri Apr 17 16:23:12 HKT 2020
> >[ 0.000000] Kernel command line:
> BOOT_IMAGE=/vmlinuz-3.18.20-xenomai-3.1-20200417-02
>root=UUID=2851ab9b-1766-
> >41a0-b882-3caa9a975bf6 ro isolcpus=1 quiet splash text vt.handoff=7
> >[ 1.257922] [Xenomai] scheduling class idle registered.
> >[ 1.257926] [Xenomai] scheduling class rt registered.
> >[ 1.257969] [Xenomai] SMI-enabled chipset found, but SMI workaround
> disabled
> >[ 1.257969] (see xenomai.smi parameter). You might
> encounter
> >[ 1.258005] I-pipe: head domain Xenomai registered.
> >[ 1.260837] [Xenomai] Cobalt v3.1 [DEBUG]
> >[ 1.371562] usb usb1: Manufacturer: Linux
> 3.18.20-xenomai-3.1-20200417-02 ehci_hcd
> >
> >my application cmake xenomai compiling setting:
> >
> >execute_process(COMMAND xeno-config --posix --cflags OUTPUT_VARIABLE
> XENO_CFLAGS
> >OUTPUT_STRIP_TRAILING_WHITESPACE) execute_process(COMMAND xeno-config
> --posix --ldflags OUTPUT_VARIABLE
> >XENO_LDFLAGS OUTPUT_STRIP_TRAILING_WHITESPACE) set(CMAKE_C_FLAGS
> "${CMAKE_C_FLAGS}
>${XENO_CFLAGS}")
> >set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_SHARED_LINKER_FLAGS}
> ${XENO_LDFLAGS}")
> >
> >my hardware platform and os is
> >
> >Intel(R) Celeron(R) CPU J1900 @ 1.99GHz i686 Ubuntu 14.04 3.18.20
> >
> >can any one help me to find out why this posix porting application
> behaviour abnormaly comparing to the regular one
>using the
> >above data.
> >
> >what measures should i take to make sure it has same thread running
> behaviour after poting to the xenomai-posix one.
> >
> >
> >
> >
> >Best Regards
> >Huang Mianbo
> >电话:+8613682612959
>
> Hi Mianbo,
>
> We use Ftrace to check if Cobalt thread is rightly sched; I can help to
> setup Ftrace env on your machine, if u interested,
>
>
> BR / Fino (孟祥夫)
> Intel – IOTG Developer Enabling
BR / Fino (孟祥夫)
Intel – IOTG Developer Enabling