Re: [PATCH] hurd: Add shared mig declarations

[Simon Marchi <simark@simark.ca>]

On 2020-05-31 3:20 a.m., Samuel Thibault wrote:
> Simon Marchi, le sam. 30 mai 2020 21:51:35 -0400, a ecrit:
>> On 2020-05-30 2:23 p.m., Samuel Thibault wrote:
>>> Fixes
>>>
>>> exc_request_S.c:177:24: error: no previous declaration for ‘exc_server’ [-Werror=missing-declarations]
>>>   177 | mig_external boolean_t exc_server
>>>
>>> gdb/ChangeLog:
>>>
>>> 	* config/i386/i386gnu.mn [%_S.o %_U.o] (COMPILE.post): Add
>>> 	"-include gnu-nat-mig.h".
>>> 	* gnu-nat-mig.h: New file.
>>> 	* gnu-nat.c: Include "gnu-nat-mig.h".
>>> 	(exc_server, msg_reply_server, notify_server,
>>> 	process_reply_server): Remove declarations.
>>
>> It took me a while to understand the underlying problem.  My understanding is that
>> gnu-nat.c calls this function exc_server, that is defined in the generated file.  The
>> generated file does not provide a header with declarations, so gnu-nat.c had its own
>> local declaration.  Since we now use the -Wmissing-declarations warning flag, and the
>> definition in the generated exc_request_S.c didn't see a corresponding declaration,
>> it caused that build failure.  Is that correct?  If so, please add that explanation
>> or equivalent to the commit log.
> 
> I have now added
> 
> “
> We are using -Werror=missing-declarations, and the _S.h files generated
> by mig do not currently include a declaration for the server routine.
> gnu-nat.c used to have its own external declarations, but better just
> share them between gnu-nat.c and the _S.c files.
> ”

Thanks, that sounds good.  And this way, I suppose that if for some reason the prototypes
don't match, we'll get a compilation error (which is a good thing).

>> My question now is: that MIG tool appears to generate both a header (%_S.h) and source
>> file (%_S.c) from defs files.  What is this header file used for, if it doesn't contain
>> the declaration for the functions in the source file?
> 
> Mig does include declarations for the functions of the .c files, but
> not for the server routine, I don't know why that was never implemented
> there (this hasn't been touched since the VCS initial import).

Ok.

Simon