On Fri, 2005-04-08 at 15:07 -0500, Kylene Jo Hall wrote: > Basically, what I need to figure out is how to solve both issues > simultaneously. I need to not register a pci_driver as I would be > taking over an ID that is not unique to my device as well as get the > hotplugging correct (which i don't know how to do with out a pci_remove > function). Perhaps it makes sense to create a new lpc subsystem with an lpc bus type. Then the PCI driver would register a new lpc bus and LPC device drivers would register themselves with that. Or if the LPC is independent of any PCI device the LPC "bridge" driver would just be another driver to be loaded. It sounds like it might end up similar to the i2c subsystem for example. Ian. -- Ian Campbell Every time I think I know where it's at, they move it.
Attachment:
signature.asc
Description: This is a digitally signed message part
- References:
- Re: [PATCH] Add TPM hardware enablement driver
- From: Jeff Garzik <[email protected]>
- Re: [PATCH] Add TPM hardware enablement driver
- From: Greg KH <[email protected]>
- Re: [PATCH] Add TPM hardware enablement driver
- From: Jeff Garzik <[email protected]>
- Re: [PATCH] Add TPM hardware enablement driver
- From: Greg KH <[email protected]>
- Re: [PATCH] Add TPM hardware enablement driver
- From: Kylene Jo Hall <[email protected]>
- Re: [PATCH] Add TPM hardware enablement driver
- From: Kylene Jo Hall <[email protected]>
- Re: [PATCH] Add TPM hardware enablement driver
- Prev by Date: Re: HELP:porting linux PXA audio driver to RTLinux(RTLinux core driver)
- Next by Date: Re: Kernel SCM saga..
- Previous by thread: Re: [PATCH] Add TPM hardware enablement driver
- Next by thread: [PATCH scsi-misc-2.6 00/08] scsi: small fixes & cleanups
- Index(es):
 |