more eeh

[Nathan Fontenot]

Hope I don't spoil anyones dinner with another dose of eeh :)

I have attached a patch that generates the hotplug event in the
kernel. At least it's supposed to do that.  This eliminates the
need for any kind of an eeh daemon and any /proc usage (two good
things).

The code works by calling disable_slot() in the rpaphp driver
code (rpaphp_core.c) to have a pci device removed.  Looking through
the kernel it appears that this is the routine that drives hotplug
remove requests that come in through sysfs.

The issue I kept running into with this code is that it would
either DSI in schedule() or the partition would quit responding
when an eeh event is injected.  Unfortunately, I have not been able to
reproduce the DSI in schedule to give you any debugging info.

Any comments I could get on the code would be greatly appreciated.
Namely, is this the correct way to generate a hotplug event, or
is there another interface I should use.  And, is this safe? This
causes the removal of a slot in the middle of a read on the slot.

thanks.

Nathan F.

--
-------------- next part --------------
A non-text attachment was scrubbed...
Name: eeh_khotplug.patch
Type: text/x-patch
Size: 4888 bytes
Desc: not available
Url : http://ozlabs.org/pipermail/linuxppc64-dev/attachments/20040316/73eb434e/attachment.bin