Re: [PATCH v5 3/5] p1689r5: initial support

[Jason Merrill <jason@redhat.com>]

On 5/12/23 10:24, Ben Boeckel wrote:
> On Tue, Feb 14, 2023 at 16:50:27 -0500, Jason Merrill wrote:
>> I notice that the actual flags are all -fdep-*, though some of them are
>> -fdeps-* here, and the internal variables all seem to be fdeps_*.  I
>> lean toward harmonizing on "deps", I think.
> 
> Done.
> 
>> I don't love the three separate options, but I suppose it's fine.  I'd
>> prefer "target" instead of "output".
> 
> Done.
> 
>> It should be possible to omit both -file and -target and get reasonable
>> defaults, like the ones for -MD/-MQ in gcc.cc:cpp_unique_options.
> 
> `file` can be omitted (the `output_stream` will be used then). I *think*
> I see that adding:
> 
>      %{fdeps_file:-fdeps-file=%{!o:%b.ddi}%{o*:%.ddi%*}}

%{!fdeps-file: but yes.

> would at least do for `-fdeps-file` defaults? I don't know if there's a
> reasonable default for `-fdeps-target=` though given that this command
> line has no information about the object file that will be used (`-o` is
> used for preprocessor output since we're leaning on `-E` here).

I would think it could default to %b.o?

I had quite a few more comments on the v5 patch that you didn't respond 
to here or address in the v6 patch; did your mail client hide them from you?

Jason