Hi, On 27/05/17 12:08 PM, Geert Uytterhoeven wrote: > Still, those code patch could be protected by #ifdef CONFIG_HAS_IOMEM, > or better, if (IS_ENABLED(CONFIG_HAS_IOMEM)).
Well I think it would have to be the former seeing the latter would still end up trying to compile the missing function. But having ifdefs inside code is not generally seen as good idea[1]. I'd really like to go forward with either this patch or something like it. The other two arches that have this problem are fine with merging a fix and adding ifdefs to work around a single arch doesn't feel right to me. Thanks, Logan [1] http://yarchive.net/comp/linux/ifdefs.html
