> Øyvind Harboe wrote:
> > On Tue, Feb 2, 2010 at 6:59 PM, David Brownell <davi...@pacbell.net>
wrote:
> >> On Tuesday 02 February 2010, Fredrik Hederstierna wrote:
> >>> This is the patch that fix the timeout issue:
> >> I think we'll hold this until the 0.5 merge window opens,
> >> to make sure it gets enough testing.
> >
> > I'm loathe to add codepaths to particular targets unless there
> > is a *huge* upside and it is sure to get enough testing...
> >
> > W.r.t. testing there is evidence that this target does not
> > get enough testing to justify the luxury of extra optimized
> > codepaths... especially for some aging part.... perhaps
> > a rock solid & slow code path is the way to go while this
> > part rides into the sunset?
> >
>
> The str9 gets quite a bit of testing, i mentioned in the previous bugfix
> that the time-outs were probably to blame for the error's edgar saw.
>
> I am guessing but it may be that your silicon has had a few erases and
> as such the erase time has extended a little.
>
> But there is no harm in waiting for the next release before merging this
> patch.
I can submit patch again later, no problem.
Unfortunately I'm stuck with crap Lotus Notes mail client here at work,
that's why I submitted patch as both (formatted) text + attachment.
The patch attachment should be still ok, no?
Also I think the patch is valuable, for example when we manufacture
products in factory, time is money.
If we can decrease production erasing/flashing time ~10 secs its very
valuable, the factory yield is important and cost alot of money for
professional use.
And yes the testing is problematic, we dont have any STR912 test farm to
rely on, but we are several developers here flashing several times a day,
and it seems work fine every time. But I cannot guarantee everything is
100.0%, I guess we need ST Micro self do the code in this case...
I think another issue is the ST xpec driver that we need to change boot
banks, unlocking etc, its not working very well all times. Maybe the ST912
drivers could be merged or one could replace the other? Its not really
required to have 2 drivers, no?
Thanks & Regards,
Fredrik Hederstierna
Securitas Direct AB
Malmoe Sweden
_______________________________________________
Openocd-development mailing list
Openocd-development@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/openocd-development
