On 2020-04-17 8:09 a.m., Christian König wrote: > Am 17.04.20 um 12:43 schrieb Michel Dänzer: >> On 2020-04-17 11:22 a.m., Christian König wrote: >>> Agreed, just wanted to reply as well since I think something is not >>> correctly understood here. >>> >>> The cpu_to_be16() and be16_to_cpu() functions work different depending >>> on which architecture/endianess your are. >>> >>> So they should be a NO-OP on x86 if everything is done right. >> The *b*e macros aren't NOPs on little endian architectures like x86, >> they are on big endian architectures. Vice versa for the *l*e macros. > > Yeah, that's what I meant with "if everything is done right" :) > > I usually can't remember what does what with those functions. > > Christian. I think key here is that dmcub FW is little endian, whereas the old dmcu FW was big endian. Hence we had the cpu_to_be conversion here for the old dmcu. Now it looks like we want to reuse the same function for dmcub calls and hence need to ensure we're not converting values to big-endian. The big_endian parameter is specifying the endianness of the FW. The right approach would be to do cpu_to_be for dmcu and cpu_to_le for dmcub. Harry _______________________________________________ amd-gfx mailing list amd-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/amd-gfx
