On Tue, Jul 19, 2022 at 03:09:35PM +0000, David Laight wrote: > From: Stafford Horne > > Sent: 19 July 2022 13:24 > > > > On Tue, Jul 19, 2022 at 01:55:03PM +0200, Arnd Bergmann wrote: > > > On Tue, Jul 19, 2022 at 12:55 PM Stafford Horne <shorne@xxxxxxxxx> wrote: > > Option 3: > > > > > Or we could try to keep the generic definition in a global header > > > like linux/isa-dma.h. > > > > Perhaps option 3 makes the whole patch the most clean. > > Isn't there a define that can be used inside an if? > So you could do: > if (!IS_CONFIG_X86_32 || !isa_dma_bridge_buggy) > disable_dma(); > (but I can't remember the name!) I think you probably mean: if (IS_ENABLED(CONFIG_X86_32) ... ) That could work too, but we still want to move the definition of isa_dma_bridge_buggy out of architectures and into the global header. I have gone with option 3 for now. -Stafford
