Rainer Orth <r...@cebitec.uni-bielefeld.de> writes:
> Hi Jakub,
>
>> On Fri, Apr 11, 2025 at 10:50:25AM +0200, Rainer Orth wrote:
>>> 2025-04-08 Rainer Orth <r...@cebitec.uni-bielefeld.de>
>>>
>>> gcc/cobol:
>>> PR cobol/119217
>>> * dts.h (csub_match): Initialize rm_so, rm_eo fields explicitly.
>>>
>>
>>> # HG changeset patch
>>> # Parent 6f227ddea0046a0164509bdb8069e96184304054
>>> cobol: Initialize regmatch_t portably [PR119217]
>>>
>>> diff --git a/gcc/cobol/dts.h b/gcc/cobol/dts.h
>>> --- a/gcc/cobol/dts.h
>>> +++ b/gcc/cobol/dts.h
>>> @@ -33,7 +33,8 @@ namespace dts {
>>> : input(input)
>>> , first(NULL), second(NULL), matched(false)
>>> {
>>> - static regmatch_t empty = { -1, -1 };
>>> + static regmatch_t empty;
>>> + empty.rm_so = empty.rm_eo = -1;
>>
>> Why not static regmatch_t empty = { .rm_so = -1, .rm_eo = -1 };
>> ?
>
> I tried that initially, but got warnings:
>
> In file included from /vol/gcc/src/hg/master/cobol/gcc/cobol/lexio.h:208,
> from /vol/gcc/src/hg/master/cobol/gcc/cobol/lexio.cc:36:
> /vol/gcc/src/hg/master/cobol/gcc/cobol/dts.h: In constructor
> ‘dts::csub_match::csub_match(const char*)’:
> /vol/gcc/src/hg/master/cobol/gcc/cobol/dts.h:36:60: warning: missing
> initializer for member ‘regmatch_t::rm_sp’ [-Wmissing-field-initializers]
> 36 | static regmatch_t empty = { .rm_so = -1, .rm_eo = -1 };
> | ^
> /vol/gcc/src/hg/master/cobol/gcc/cobol/dts.h:36:60: warning: missing
> initializer for member ‘regmatch_t::rm_ep’ [-Wmissing-field-initializers]
> /vol/gcc/src/hg/master/cobol/gcc/cobol/dts.h:36:60: warning: missing
> initializer for member ‘regmatch_t::rm_ss’ [-Wmissing-field-initializers]
> /vol/gcc/src/hg/master/cobol/gcc/cobol/dts.h:36:60: warning: missing
> initializer for member ‘regmatch_t::rm_es’ [-Wmissing-field-initializers]
>
>> While the FE needs to be valid C++14, the library is GNU++17 (compiled
>> by the just built GCC and not built with -pedantic*, so can use GCC
>> extensions or features from C++20 that are known to be implemented there).
>>
>> Ah, I see dts.h include from the FE. So that would need to be
>> #ifndef IN_GCC_FRONTEND
>> static const regmatch_t empty = { .rm_so = -1, .rm_eo = -1 };
>> #else
>> regmatch_t empty = {};
>> empty.rm_so = empty.rm_eo = -1;
>> #endif
>
> Isn't this overkill for such a minor issue?
The question remains: how to handle this, especially for the gcc-15
branch where this issue is one of several breaking Solaris/amd64
--enable-languages=all bootstrap.
Rainer
--
-----------------------------------------------------------------------------
Rainer Orth, Center for Biotechnology, Bielefeld University
