Re: [Xenomai] Xenomai tests

[Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org>]

On Wed, Sep 23, 2015 at 07:38:43AM +0200, Gilles Chanteperdrix wrote:
> On Wed, Sep 23, 2015 at 12:41:38PM +1200, vibnwis wrote:
> > On 23 September 2015 at 05:44, Gilles Chanteperdrix <
> > gilles.chanteperdrix@xenomai.org> wrote:
> > 
> > > On Wed, Sep 23, 2015 at 12:32:11AM +1200, vibnwis wrote:
> > > > By regenerating ./configure using autoreconf -fi, I managed to compile
> > > > xenomai-3-3.0-rc7 natively on PandaBoard. All things are put in their own
> > > > place by make install. However, when running latency, it produces
> > > >
> > > > > latency
> > > > 0"000.000| BUG in low_init(): [main] mlockall: Cannot allocate memory
> > >
> > > You should either run the latency test as root, or have a look at
> > > this page:
> > > http://xenomai.org/2014/06/running-a-xenomai-application-as-a-regular-user/
> > >
> > > By root I get this.
> > root@arm:/usr/xenomai# latency
> >    0"000.000| BUG in low_init(): [main] binding failed: Function not
> > implemented
> > 
> > However, the PATH has set the Lib
> 
> Running as root was necessary to avoid the mlockall issue. The issue
> you have now is different: the logs you sent show that the last
> kernel you booted is not patched with Xenomai. The kernel is patched
> with the I-pipe patch, but this is not sufficient, it is not
> prepared with prepare-kernel.sh
> 
> > 
> > root@arm:/usr/xenomai# echo $PATH
> > /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/usr/xenomai/bin:/usr/xenomai/lib:/usr/xenomai/include
> > 
> > 
> > >
> > > > > dmesg | grep -i ipipe
> > > > [    0.000000] Linux version 3.18.20-ipipe (qws@qws-MacBookPro) (gcc
> > > > version 4.9.2 (Ubuntu/Linaro 4.9.2-10ubuntu10) ) #1 SMP Tue Sep 22
> > > 21:04:49
> > > > NZST 2015
> > > > [    0.623763] 3.18.20-ipipe #1 Not tainted
> > > > [    0.623994]   [<c00d3094>] __ipipe_do_sync_stage+0x288/0x2d0
> > > > [    0.624011]   [<c0008634>] __ipipe_grab_irq+0x5c/0x70
> > > > [    0.624321] CPU: 1 PID: 22 Comm: kblockd Not tainted 3.18.20-ipipe #1
> > > > [    0.624479] [<c00843b4>] (trace_hardirqs_on_caller) from [<c05da7f4>]
> > > > (__ipipe_fast_svc_irq_exit+0x4/0x18)
> > > > [    0.624504] [<c05da7f4>] (__ipipe_fast_svc_irq_exit) from [<c001aacc>]
> > > > (ipip_test_and_stall_root+0x2c/0x3c)
> > > > [    0.624524] [<c001aacc>] (ipipe_test_and_stall_root) from [<c0086e58>]
> > > > (lock_acquire+0x44/0x108)
> > > > [    0.728394] Switched to clocksource ipipe_tsc
> > > > [    2.464800] usb usb1: Manufacturer: Linux 3.18.20-ipipe ehci_hcd
> > > > [    6.364516] usb usb2: Manufacturer: Linux 3.18.20-ipipe musb-hcd
> > >
> > > Please post the full kernel logs. You likely enabled a kernel
> > > configuration option which has an issue with CONFIG_IPIPE
> > >
> > Please find it in attachment. Many thanks (very grateful indeed)
> 
> Ok, when I say "please post the full kernel logs", I mean only the
> logs of the kernel currently running, not the logs of previous
> kernel runs, these are irrelevant. Anyway, I think you have
> CONFG_CONTEX_TRACKING enabled, which does not work with
> CONFIG_IPIPE, at least on ARM. Could you post you kernel
> configuration to be sure?

The problematic option is CONFIG_TRACE_IRQFLAGS, not
CONFIG_CONTEXT_TRACKING, and you have it enabled, since without it,
you would not be able to enable CONFIG_LOCKDEP. Please disable
CONFIG_TRACE_IRQFLAGS. You will need to disable all the options
depending on that option, such as CONFIG_LOCKDEP.

-- 
					    Gilles.
https://click-hack.org