On Thu, 2008-05-22 at 14:14 +0200, Holger Macht wrote: > On Thu 22. May - 14:33:01, Shaohua Li wrote: > > docking station releted fixes > > patch 1, 2: fix > > patch 3, 4, 5: merge bay driver into dock driver. dock driver can handle > > battery and bay. > > patch 6, 7: make libata work with dock driver > > patch 8: cleanup > > > > I tested in a T43 with bay and battery hotplug, please review. > > Have you seen this patch? Are those changes incorporated? > http://marc.info/?l=linux-acpi&m=121129188316720&w=2 I thought my patch already handles this. > What are the advantages against just letting libata-acpi doing the _EJ0? The idea is letting dock driver handle bay/battery/dock station hotplug. This simplified things. > I gave your patch set a short test and have encountered the following > issues: > > 1. No uevent on a ACPI_NOTIFY_EJECT_REQUEST giving no chance to userspace > to clean up filesystems etc. > > 2. When a bay device is removed, the device (e.g. /dev/sr0) stays visible > for userspace, even after accessing it and getting a rescan and > 'klogd: ata5.00: disabled' So the two are one issue, uevent isn't sent. I added a .uevent method for libata, it should be called on ACPI_NOTIFY_EJECT_REQUEST. Also dock_notify() has a check if the dock is a bay. If yes, it just sent a uevent, not directly handle eject. Userspace should write 1 to dock's 'undock' sysfile to actually remove bay. Thanks, Shaohua -- To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
