Messages in this thread | | | From | (Alan Cox) | Subject | Re: de4x5.c patch against 2.1.117 | Date | Thu, 27 Aug 1998 02:56:27 +0100 (BST) |
| |
> The patches I'd actually accept would be something that made sense on a > larger scale, and were able to support PnP as a side effect. > > Quite frankly, PnP as a standard sucks. Anything that just reads the PnP > standard and tries to force that on Linux will be most definitely ignored.
I actually took the ISAPnP patches and tried to beat them into that after the last time Linus made his comments.
If you try this you'll discover two things
1. Most of ISAPnP can be done in user space
2. The only stuff that can't is (with one exception I know of)[*] related entirely to hot reconfiguration.
As things stand right now we have a problem over all with the following related issues
Hot Swap - PCMCIA, Cardbus, USB,...
Physical Resource Management
Mostly related to the above. We need to be able to do dynamic reallocations, partial decode and track memory and other resources
Device restoration
When we swap a device in we need to restore its state, PCMCIA does this via a daemon but its not yet generic just a very good basis.
The last one along with state saving is needed for ACPI suspend to disk as well.
In the end I found I had a choice of (2.3ish) doing the whole resource stuff, or avoiding the resource stuff, which reduced to the "isapnp tools needs a decent front end" problem not a kernel one.
Alan [*] Folks putting mac level addresses in the ISAPnP card serial number
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html
| |