Hi, Tobias,
Sorry for the late reply (just came back from a long vacation after Cauldron).
And thank you for reporting this issue.
Please see my reply embedded below:
> On Sep 25, 2023, at 2:24 PM, Tobias Burnus <tob...@codesourcery.com> wrote:
>
> Hi all,
>
> I stumbled over this as I found the wording in the release notes rather
> unclear.is.
>
>
> First, the following gives only a -pedantic warning and not a
> -Wflex-array-member-not-at-end:
>
> struct t { int b; int x[]; };
> struct q { int b; struct t a[2]; int c; };
>
> warning: invalid use of structure with flexible array member [-Wpedantic]
>
> If I remove the "[2]", it shows additionally:
> warning: structure containing a flexible array member is not at the end of
> another structure [-Wflex-array-member-not-at-end]
I think that the above behavior is correct as Richard mentioned previously. -:)
First, by C99, a structure with flexible array member cannot be an element of
an array,
Therefore, struct t a[2] is an incorrect usage, the warning:
warning: invalid use of structure with flexible array member [-Wpedantic]
is complaining about this standard violation. Though the diagnositic message
might need to be more specific like the following:
Warning: invalid use of structure with flexible array member as array element
[-Wpedantic]
Then, after fixing this standard violation issue, the testing case is as
following:
struct t { int b; int x[]; };
struct q { int b; struct t a; int c; };
adding -Wflex-array-member-not-at-end will report the expecting message:
/home/opc/Install/latest/bin/gcc -O -Wflex-array-member-not-at-end t.c -S
t.c:2:29: warning: structure containing a flexible array member is not at the
end of another structure [-Wflex-array-member-not-at-end]
2 | struct q { int b; struct t a; int c; };
| ^
So, I think that GCC’s behavior is correct.
However, we might need to make the diagnostic message more accurate here (I can
submit a small patch to improve this).
>
> It seems as if it should print latter warning also inside the struct.
>
> Qing? Joseph? Thoughts?
>
> * * *
>
> Secondly, if this is deprecated, shouldn't then the warning enabled by, e.g.,
> -Wall or made
> otherwise more prominent? (-std=?) - Currently, one either has to find the
> new flag or use
> -pedantic.
Yes, agreed, However, I think that it might be better to delay this to next GCC
release by giving users plenty time to fix all the
-Wflex-array-member-not-at-end warnings. As I know, linux kernel exposed a lot
of warnings when adding -Wflex-array-member-not-at-end, and kernel people are
trying to fix all these warnings in the source base.
>
> Or is this not really regarded as deprecated? But then (IMHO) we should not
> really claim so and just
> add the warning without deprecation.
I think that our final goal is to deprecate this ambiguous extension from GCC
completely, but we need time to mitigate users step by step.
>
> BTW; clang-15 prints the -Wgnu-variable-sized-type-not-at-end warning by
> default.
>
> Joseph, all: Thoughts?
>
> * * *
>
> Cross ref: The patch adding the new warning is r14-2197-g070a6bf0bdc6761
> https://gcc.gnu.org/pipermail/gcc-cvs/2023-June/385730.html (cf. previously
> in this thread)
>
>
> * * *
>
> Regarding the changes.html wording:
>
> On 07.08.23 16:22, Qing Zhao via Gcc-patches wrote:
>
>> Comparing to the 1st version, the only change is to address Richard's
>> comment on refering a warning option for diagnosing deprecated behavior.
> ...
>> +++ b/htdocs/gcc-14/changes.html
>> @@ -30,7 +30,18 @@ a work-in-progress.</p>
>> <!-- .................................................................. -->
>> <h2>Caveats</h2>
>> <ul>
>> - <li>...</li>
>> + <li><strong>C:</strong>
>> + Support for the GCC extension, a structure containing a C99 flexible
>> array
>> + member, or a union containing such a structure, is not the last field
>> of
>> + another structure, is deprecated. Refer to
>> + <a href="https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html";>
>> + Zero Length Arrays</a>.
>
> ...
>
> I find the first sentence difficult to read. What do you think of the
> following?
> (It is hard to come up with some good wording.)
>
> --- a/htdocs/gcc-14/changes.html
> +++ b/htdocs/gcc-14/changes.html
> @@ -31,9 +31,10 @@ a work-in-progress.</p>
> <h2>Caveats</h2>
> <ul>
> <li><strong>C:</strong>
> - Support for the GCC extension, a structure containing a C99 flexible
> array
> - member, or a union containing such a structure, is not the last field
> of
> - another structure, is deprecated. Refer to
> + Support for the GCC extension that a structure containing a C99
> flexible
> + array (and any union containing a member of such structure) can be a
> + member of a structure has been deprecated for the case that it is not
> + the last member. Refer to
> <a href="https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html";>
> Zero Length Arrays</a>.
> Any code relying on this extension should be modifed to ensure that
I felt the above is more difficult to understand… how about below:
> + Support for the GCC extension that a structure containing a C99
> flexible
> + array, or a union containing a member of such structure, is not the
> last field
> + of another structure, has been deprecated. Refer to
>
thanks.
Qing
> Tobias
>
> PS: C17 has:
> "A structure or union shall not contain a member with incomplete or function
> type (hence, a structure
> shall not contain an instance of itself, but may contain a pointer to an
> instance of itself), except that
> the last member of a structure with more than one named member may have
> incomplete array type;
> such a structure (and any union containing, possibly recursively, a member
> that is such a structure)
> shall not be a member of a structure or an element of an array."
>
> -----------------
> Siemens Electronic Design Automation GmbH; Anschrift: Arnulfstraße 201, 80634
> München; Gesellschaft mit beschränkter Haftung; Geschäftsführer: Thomas
> Heurung, Frank Thürauf; Sitz der Gesellschaft: München; Registergericht
> München, HRB 106955
