Hi Arnd,
On Fri, Apr 5, 2024 at 1:26 PM Arnd Bergmann <arnd@xxxxxxxxxx> wrote:
On Fri, Apr 5, 2024, at 12:16, Geert Uytterhoeven wrote:
On Wed, Apr 3, 2024 at 8:35 PM Arnd Bergmann <arnd@xxxxxxxxxx> wrote:
Michael is worried about non-ISA drivers using inb() and friends.
At some point in time (i.e. eons ago), we were told it was better to
use in[bwl]()/read[bwl]() instead of directly dereferencing volatile
pointers...
It's definitely still better to use an abstraction layer for MMIO
accesses using inline asm instructions than open-coding the
volatile pointer dereferences. Over time we have gotten better
at defining which of the available abstractions should be used
for a given bus, so inb()/outb() is now only really used for
things derived from ISA in some form, including e.g. PCI and LPC.
Indeed. Nowadays the consensus is that readl() and friends should
always perform little-endian accesses, thus implying byte-swapping on
big-endian platforms, but most of the m68k drivers were written long
before that enlightenment...
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
