>
> request/release regions is what Linux drivers are supposed to do to
> mark that they have a resource in use. There's a pci_request_region()
> available that hides whether it is IO or memory so I should switch to
> that one. All drivers should implement these functions. You can't do
> this as part of add/initmap since they may only map subsets of the
> regions.

Do they have to be per-driver, or can the core work it out from the pci
address space and do it itself?, if they are Linux specific they need to
go in the radeon_drv.c or a linux specific file at any rate.. or do we
just abstract them out for different OSes..

Dave.

 >
> pci_read_config_dword(RADEON_AGP_COMMAND_PCI_CONFIG) is in the
> category of user space code shouldn't be setting variables into the
> driver for things user space has no control over. The driver should
> know whether it's hardware is AGP or PCI. It shouldn't rely on a
> command from user space to tell it what hardware it has. This code
> should be on all platforms.
>
> Over time we should remove all cases in all drivers of user space code
> telling the devices drivers what hardware they have. That's backwards,
> the hardware drivers should be telling user space, not the other way
> around. Eliminating code like this removes one of the reasons X needs
> to run as root.
>
>

-- 
David Airlie, Software Engineer
http://www.skynet.ie/~airlied / airlied at skynet.ie
pam_smb / Linux DECstation / Linux VAX / ILUG person



-------------------------------------------------------
This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
Project Admins to receive an Apple iPod Mini FREE for your judgement on
who ports your project to Linux PPC the best. Sponsored by IBM. 
Deadline: Sept. 13. Go here: http://sf.net/ppc_contest.php
--
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel

Reply via email to