On 02/07/2025 at 19:01, Geert Uytterhoeven wrote: > Hi Biju, > > On Wed, 2 Jul 2025 at 11:46, Biju Das <biju.das.jz@xxxxxxxxxxxxxx> wrote: >>> From: Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> >>> On Sun, 29 Jun 2025 at 17:04, Biju Das <biju.das.jz@xxxxxxxxxxxxxx> wrote: >>>> Drop unused macros from the rcar_canfd.c. >>>> >>>> Reported-by: Vincent Mailhol <mailhol.vincent@xxxxxxxxxx> >>>> Closes: >>>> https://lore.kernel.org/all/7ff93ff9-f578-4be2-bdc6-5b09eab64fe6@wanad >>>> oo.fr/ >>>> Signed-off-by: Biju Das <biju.das.jz@xxxxxxxxxxxxxx> > >>> #define RCANFD_RFPTR_RFDLC GENMASK(31, 28) >>> >>>> @@ -298,16 +256,10 @@ >>>> #define RCANFD_GSTS (0x008c) >>>> /* RSCFDnCFDGERFL / RSCFDnGERFL */ >>>> #define RCANFD_GERFL (0x0090) >>>> -/* RSCFDnCFDGTSC / RSCFDnGTSC */ >>>> -#define RCANFD_GTSC (0x0094) >>> >>> Note that removed register offsets will become anonymous gaps when the register offsets are replaced >>> by C structs, cfr. commit ab2aa5453bb83d05 >>> ("can: rcar_canfd: Describe channel-specific FD registers using C struct"). >> >> OK. But removing unused reg offset is not an issue at the moment, I guess?? > > No, it is not an issue at the moment, as they are unused. Re-adding > register members to a C struct later is just more tedious, as you have > remove gaps or adjust their sizes (lol, mind the gap ;-). > > IIRC Vincent's biggest gripe is the use of function-like macros, > not the definition of unused register offsets and bits? Correct. My comment really came from the fact that I did not understand why you were only removing RCANFD_GAFLCFG_GETRNC() while keeping the other unused macros in: https://lore.kernel.org/all/7ff93ff9-f578-4be2-bdc6-5b09eab64fe6@xxxxxxxxxx/ To be honnest, I would still rather prefer to see those unused macros removed, but it is something that I am totally fine to leave to your judgement. Keeping the unused macro is totally negotiable if you prefer to do so :) My no-go is indeed on the proliferation of function-like macros. Yours sincerely, Vincent Mailhol
