Re: pci-sysfs resource mmap broken (and PATCH)

[Jesse Barnes]

David S. Miller <davem <at> davemloft.net> writes:
> The only problem could me getting the generic mmap() code to
> properly pass the flag down into the driver, I seem to recall
> that it either does an -EINVAL or masks out any flags which
> are not in the standard set.

But it would be a relatively clean solution, if a bit arch specific (i.e. some
arches would allow MAP_WRITECOMBINE or somesuch, while others have different
sorts of batching attributes).

> But then again this conflicts with what I remember seeing in the
> XFree86 PCI support, in that IA64 passed in such a mmap() flag
> to indicate a framebuffer like mapping that didn't need a guard-like
> bit to be set.

It used an ioctl last time I looked.

Jesse


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/