On Mon, Feb 3, 2025 at 2:22 PM Michael J Gruber <mjg@xxxxxxxxxxxxxxxxx> wrote:
> So, over at upstream they raised the suspicion that some parts of what
> I'm linking are compiled with different flags. To recap: I get
> ```
> /usr/include/c++/15/bits/stl_vector.h:1262: std::vector<_Tp,
> _Alloc>::reference std::vector<_Tp, _Alloc>::operator[](size_type)
> [with _Tp = const char*; _Alloc = std::allocator<const char*>;
> reference = const char*&; size_type = long unsigned int]: Assertion
> '__n < this->size()' failed.
> Aborted (core dumped)
> ```
> in a part of code which basically does:
> ```
> FZ_FUNCTION std::vector<std::string>
> pdf_choice_widget_options2(fz_context* ctx, pdf_annot* tw, int
> exportval)
> {
> int n = pdf_choice_widget_options(ctx, tw, exportval, nullptr);
> std::vector<const char*> opts(n);
> int n2 = pdf_choice_widget_options(ctx, tw, exportval, &opts[0]);
> assert(n2 == n);
> std::vector<std::string> ret(n);
> for (int i=0; i<n; ++i)
> {
> ret[i] = opts[i];
> }
> return ret;
> }
> ```
> Is there anything regarding flags like NDEBUG, GLIBCXX_ASSERTIONS and
> the like which could make the above throw when the .so is compiled
> with different flags compared to the code that calls it? Something
> that appeared with gcc15 or libstdc++ in gcc15?
If n == 0, then &opts[0] is invalid, since opts is an empty vector in
that case. The invalid access to the nonexistent zeroth element
triggers an assertion failure inside operator[] when the code is
compiled with -D_GLIBCXX_ASSERTIONS.
--
Jerry James
http://www.jamezone.org/
--
_______________________________________________
devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx
Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx
Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
