From: Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org>
To: Nero Fernandez <grimlynch@domain.hid>
Cc: xenomai@xenomai.org
Subject: Re: [Xenomai-core] [xeno_switchtest] missing linux-dev registration?
Date: Wed, 21 Apr 2010 13:59:40 +0200 [thread overview]
Message-ID: <4BCEE8AC.4080101@domain.hid> (raw)
In-Reply-To: <k2xbfcfb7e81004210425l88dbbb99gb8b1ec4dd254300d@domain.hid>
Nero Fernandez wrote:
> This is regarding the pervasive error "Unable to open switchtest device"
> while
You are using Xenomai 2.5.2 right? switchtest is not working in Xenomai
2.5.2 because RTDM syscalls were switchted do __xn_exec_conforming. This
gave birth to a long discussion, and finally a commit to fix the issue.
Xenomai 2.5 current head contains this change.
> trying to test context-switch timings using 'switchtest' testsuite
> application.
switchtest does not test context switch timing.
>
> Looking at the testsuite code, i found that
> - 'latency' (which can successfully operate the xeno_switchtest driver )
> uses rtdm_dev_xxx calls whereas
No, latency uses the timerbench driver. The fact that the device almost
has the same name is misleading and has been fixed in Xenomai 2.5
current head too.
> - 'switchtest' application uses the posix 'open'/'close' system-calls
> for driver-interaction
>
> Looking at ksrc/drivers/testing/switchtest.c, i found that there is no
> registration
> of switchtest under linux-device-driver subsystem (even as a misc driver
> like rtnet or rtheap do).
True. But there is a registration as an RTDM device.
> Thus, there doesnt seem to be a way through which a linux-process (non-RT
> mode)
> may access xeno_switchtest driver through /dev/rttest<x>
> (or the new /dev/rttest-switchtst<x>
> ref: https://mail.gna.org/public/xenomai-git/2010-04/msg00073.html).
No. It should work. The device is registered as an RTDM device, and the
switchtest application opens the same device using open, which is turned
into the corresponding rtdm syscall by the posix skin.
>
> In case there is such a registration required, i would be more than willing
> to contribute.
>
> In case, the 'open' system-call is handled by one of the skins
> (i have native, posix,rtai and rtdm enabled in my kernel-config),
> kindly point out the code and enlighten.
>
> Thanks for running this awesome project and all the help/pointers.
It works here, are you sure you enabled the switchtest driver?
--
Gilles.
next prev parent reply other threads:[~2010-04-21 11:59 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-04-21 11:25 [Xenomai-core] [xeno_switchtest] missing linux-dev registration? Nero Fernandez
2010-04-21 11:54 ` Jan Kiszka
2010-04-21 11:59 ` Gilles Chanteperdrix [this message]
2010-04-21 16:05 ` Nero Fernandez
2010-04-21 16:08 ` Gilles Chanteperdrix
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=4BCEE8AC.4080101@domain.hid \
--to=gilles.chanteperdrix@xenomai.org \
--cc=grimlynch@domain.hid \
--cc=xenomai@xenomai.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.