Thanks for you answer!
I'm almost sure that I can handle what I have to if I could make RTAI works on
TS-7260.
What I need is just a 10us resolution to change the logic level on a specif pin
of DIO1. I don't mean that there will be a process interrupting every 10us, but
there will be a process that could interrupt after 1,01ms or 1,02ms and so.
That's why I need usleep with the resolution of microsecond, but with RTAI it
is completely different and I think I shouldn't use usleep anymore.
I'm having troubles to compile the new kernel. I just could with an old gcc
version in cross toolchains (3.3.4).
--- In wrote:
> > I'm trying to put RTAI kernel in a TS-7260 and I'm getting the
> > following kernel panic:
> >
> > Kernel panic: VFS: Unable to mount root fs on 1f:01
> >
> > The command that I'm passing to RedBoot is:
> >
> > exec -c "console=ttyAM0,115200 root=/dev/mtdblock1"
> >
> > I have already tried many different versions of RTAI kernel from
> > Embedded ARM (ts8, ts9 and ts11) and everyone show the same error.
>
> I had similar issue with getting rtai up and running on 7250.
>
> In my case it had to do with the on board NAND flash. The pre-compiled
> rtai kernels from ts web site didn't include the patch to use the flash,
> so I had to get the kernel source and then patch with the supplied rtai
> patch AND the mtd 2k flash patch (I think all the flash patch did was set
> some options for the MTD device in the .config file for the kernel). The
> kernel sends the panic message because it doesn't find a file system
> because it doesn't know how to read the NAND flash.
>
> This was all many moons ago, and I am not really sure the 7260 uses the
> same flash (I switched production to the 7200 so haven't repeated the
> process).
When you mean mtd 2k flash patch is to enable the option CONFIG_MTD2K=y, isn't
it?
Another thing that I think to be strange is that in the initial configuration
proposed for TS-7200, the support for YAFFS file system was configured to be
built as a module. I think I have to compile the kernel with this support
built-in, is it correct?
> > I'm also having the problem already described at this forum about
> > the resolution of usleep (10ms). But I really need something that
> > sleep my program with a resolution of, at least, 10us, because I
> > have to generate a very specif wave at DIO1.
>
> On the performance of 10us sleep... this might be pushing it on the
> platform, although I've never really tried it. I have done 40usec timed
> interrupt routines on rtai, but those were using P3 intel processors and
> more modern 2.6 kernels. Since 2.6 kernels are better with interrupts,
> rtai is better (although it could be limitations of the hardware also... I
> don't know enough to really say). usleep -- at least with later rtai
> versions -- is improved but I don't know if it will give you hard
> determination at that rate.... but again I never attempted this.
>
Thank you for the help,
Ronan.
------------------------------------
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/ts-7000/
<*> Your email settings:
Individual Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/ts-7000/join
(Yahoo! ID required)
<*> To change settings via email:
<*> To unsubscribe from this group, send an email to:
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|