There's a bug in pcmcia_disable_device
965 void pcmcia_disable_device(struct pcmcia_device *p_dev) {
966 pcmcia_release_configuration(p_dev);
967 pcmcia_release_io(p_dev, &p_dev->io);
968 pcmcia_release_irq(p_dev, &p_dev->irq);
969 if (&p_dev->win)
^
This will always be true because of the address operator.
970 pcmcia_release_window(p_dev->win);
971 }
972 EXPORT_SYMBOL(pcmcia_disable_device);
I'm not sure what the correct fix for that would be.
Can simply removing the & cause any unwanted sideeffects in drivers?
For example drivers assuming that we always release the window here.
It'd probably best then to remove the whole conditional and release
unconditionally, as it currently is already always true.
--
Greetings Michael.
_______________________________________________
Linux PCMCIA reimplementation list
http://lists.infradead.org/mailman/listinfo/linux-pcmcia
