On 5/25/2018 1:10 AM, Jan Beulich wrote: >>>> On 24.05.18 at 22:26, <[email protected]> wrote: >> On 05/24/2018 01:57 AM, Jan Beulich wrote: >>>>>> On 24.05.18 at 02:46, <[email protected]> wrote: >>>> --- /dev/null >>>> +++ b/xen/include/xen/linux_compat.h >>> I continue to dislike the idea of having a header with these contents in >> this location. >> As explained previously this header can be used for the any driver that >> we want to port from Linux. This is not arm specific. This seemed like >> the best location for the file. This code was added to every source file that was being ported. I found this explicitly in smmu-v2.c. Eventually this might be needed for the iort code that Manish might pull in.
The function signatures in the header file are consistent and should not change. So, if we plan on porting a lot of Linux code, I would argue that this file is a necessity so that we do not define these functions in every ported source file. (If we want to port Linux code with minimal changes.) The only example I found in the code base is arm smmu driver and the iort code that is being pulled in. So, we can move it to an arm specific location if needed, but this header is needed. Thanks, Sameer > Please take into consideration that with lots of code originally having come > from Linux, in all the years we've never had a need to have such a header. > With that history, it needs good reasoning to introduce one now in such a > global fashion. > >> Which other common location should I pick? > None - that's the whole point of my comment: I don't see this as something > that should go in a common location. > > Jan > > > > _______________________________________________ > Xen-devel mailing list > [email protected] > https://lists.xenproject.org/mailman/listinfo/xen-devel _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
