On Sat, Jan 08, 2022 at 01:40:06AM +0200, Andy Shevchenko wrote: > > > On Saturday, January 8, 2022, Andy Shevchenko <andy.shevchenko@xxxxxxxxx> > wrote: > > > > On Friday, January 7, 2022, Gabriel L. Somlo <gsomlo@xxxxxxxxx> wrote: > > On Fri, Jan 07, 2022 at 10:50:02PM +0200, Andy Shevchenko wrote: > > > > + .of_match_table = of_match_ptr(litex_match), > > > > > > You said "Wrong usage of of_match_ptr()" here, and all I have to go > by > > > is a bunch of other `drivers/mmc/host/*.c` files that use it in a > > > similar way, so can you please clarify and/or provide an example of > how > > > to do it properly? > > > > First of all, you have a dependency to OF, try to remove it and > > compile with OF=n and you will immediately see the issue. You may > also > > go for `git log --no-merges --grep of_match_ptr` and analyze the > > result. > > Ah, grepping for "of_match_ptr" was *never* going to show me the > "right way of using of_match_ptr()" :) > > > > Hmm... this is strange. The above command is grepping in the commit > messages, and not in the code. > > > > https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/log/?qt= > grep&q=of_match_ptr > > Shows a lot of the commits that I referred to. > Right. I was originally looking at *code* for current usage examples. I get that you were referring to grepping through *commits*, which seems to be the new way of figuring out what's canonical (as a lot of the existing code seems to be in need of update :) Thanks, --Gabriel > > > OTOH, grepping for "of_match_table" (and of course the commits > returned by your command line) suggests that I simply should *not* > use it at all, and assign its argument directly to .of_match_table. > > Got it, fixed it, thanks! > > Now, on to the probe ordering issue... :) > > Thanks, > --G > > > > -- > With Best Regards, > Andy Shevchenko > > > > > > -- > With Best Regards, > Andy Shevchenko > >
