[Bug 1001453] CAN IO package: wider flags field, flag to report return to 'error active' mode

[bugzilla-daemon at bugs dot ecos dot sourceware dot org]

Please do not reply to this email. Use the web interface provided at:
http://bugs.ecos.sourceware.org/show_bug.cgi?id=1001453

--- Comment #16 from Bernard Fouchà <bernard.fouche@kuantic.com> 2012-01-26 08:50:35 GMT ---
(In reply to comment #15)
> [snip]
> > LPC2XXX driver I'll patch without my other changes, so every driver
> > will be kept coherent with the CAN IO package.
> 
> I see. Sorry that you did not find volunteers that have the eCos CAN
> hardware.

Instead of updating the LPC2XXX CAN driver, I'll make a LPC17XX driver that can
still be compiled for LPC2XXX with a minimal amount of work. Anyway what stops
me to provide hardware drivers at this time is bug #1001456 since its solution
seems to be some API evolution regarding interrupt management, at least for
Cortex-M systems.

> By the way, we may on occasion remove all the warnings in builds of CAN
> tests. I found such annoying warnings (of course, they did exist before
> the patching):
> 
>   cc1: warning: command line option "-Woverloaded-virtual" is valid for
> C++/ObjC++ but not for C
> 
> Medicine: s/\$\(CFLAGS\)/$(ACTUAL_CFLAGS)/g for make rules in CAN config
> files
> 
>   io/can/current/cdl/io_can.cdl
>   devs/can/m68k/mcf52xx/current/cdl/can_mcf52xx.cdl
>   devs/can/arm/lpc2xxx/current/cdl/can_lpc2xxx.cdl
> 
> and may be this can be fixed in another patch-set. Your decision.

Why not wait after the release of the new toolchain? Maybe other compilation
flags will have to be updated in different places?

> 
> Bernard, your work looks great for me. In particular, thank you for
> updated documentation.  I understand (agree with) your arguments. I
> plan to accept all changes after further testing, if nobody objects.

There is still something missing: having the application to be able to get the
values of TXREC/RXREC but this means more modifications to all hw drivers and
I'm not sure how to do this properly. Ideally the counter values would be
provided in events having flags raised like ERR_ACTIVE/ERR_PASSIVE, or they
could be made available by a call to _get_config(). The first solution would
increase the event size (could be an option at the CAN IO package level), the
second solution returns values that are already obsolete by the time the
application get them. So I'll leave this to the next volunteer ;-)

-- 
Configure bugmail: http://bugs.ecos.sourceware.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.