Arjen de Korte
nut+devel at de-korte.org
Wed Dec 30 13:01:06 UTC 2009
Citeren Arnaud Quette <aquette.dev op gmail.com>:
> the problem at that time was that interrupt were not well handled by
> Sun's libusb.
> which lead me to that crappy sleep hack.
The only driver which uses libusb and for which usage of the interrupt
pipeline is optional, is usbhid-ups. All others either don't use it,
or require it. For the usbhid-ups driver we have the 'pollonly' flag
to deal with broken input reports (it will disable them completely).
> now, in the light of your feedback, this seems not needed anymore, so
> we will have to test the version of Sun's libusb to enable or disable
> interrupt support.
The following is the comment from 'configure.in':
dnl FIXME: Sun's libusb doesn't support timeout (so blocks notification)
dnl and need to call libusb close upon reconnexion
Either interrupt reports are required, not used or can be switched off
by a configuration flag. The only case we haven't dealt with, is the
hal-addons (which I don't mind breaking at this time). I think we can
safely remove the hack (or automatically enable 'pollonly' for Solaris
in the usbhid-ups driver).
Best regards, Arjen
Please keep list traffic on the list
More information about the Nut-upsdev