> Katsuhiko Isono wrote:
> >> -----Original Message-----
> >> Katsuhiko Isono wrote:
> >>> I use linux-2.6.26 + xenomai-2.4.10 with custom arm(i.mx31) board.
> >>>
> >>> When realtime application was executed on my environment,
> >>> application caused the SIGSEGV fault about once every
> several hours.
> >>>
> >>> However, when I applied "arm: fix mm switching" patch of
> >> xenomai-2.5 to
> >>> my environment, SIGSEGV fault did not happen.
> >>>
> >>> There is a following question.
> >>> - Is this patch effective also for 2.4.10 ?
> >>> - What trouble is corrected with this patch ?
> >> The version of the I-pipe patch you are using has another
> >> known bug than
> >> the one fixed by the "arm: fix mm switching" patch (by the
> way, to be
> >> sure that we are talking about the same patch, I would
> prefer an URL
> >> pointing to xenomai's gitweb), the bug mentioned here:
> >>
> >> https://mail.gna.org/public/xenomai-core/2009-08/msg00021.html
> >>
> >> And the fix for this bug in the I-pipe patches which comes
> >> with Xenomai
> >> 2.4.10 should fix both bugs.
> >>
> >> So, in short, do you oberve the segmentation faults when using the
> >> I-pipe patches which ship with Xenomai 2.4.10 rather than
> >> whatever patch
> >> you are using for linux 2.6.26?
> >>
> >
> > Thank you for your reply.
> >
> > I use linux-2.6.26 of Freescale BSP because I develop custom board.
>
> Actually, I do not really understand this argument. It looks as easy to
> develop a custom board support with the mainline kernel as with a vendor
> version. And at least, with the mainline kernel, you are not stuck with
> whatever version the vendor chose.
The reason for using FreeScale BSP is that functions(SDMA,..etc) were richer
than mainline at the time of transplanted to our board.
_______________________________________________
Xenomai-help mailing list
[email protected]
https://mail.gna.org/listinfo/xenomai-help
