Description of problem: Koschei tells me that builds of the why3 package just started failing in Rawhide, due to a pdflatex failure. The build log shows this: /usr/include/c++/12/bits/stl_vector.h:1141: std::vector<_Tp, _Alloc>::const_reference std::vector<_Tp, _Alloc>::operator[](size_type) const [with _Tp = std::pair<std::__cxx11::basic_string<char>, Object>; _Alloc = std::allocator<std::pair<std::__cxx11::basic_string<char>, Object> >; const_reference = const std::pair<std::__cxx11::basic_string<char>, Object>&; size_type = long unsigned int]: Assertion '__n < this->size()' failed. I can reproduce this by building why3 in mock, waiting for the failure, launching a mock shell, changing to the doc/latex directory, and running "pdflatex manual.tex". Version-Release number of selected component (if applicable): texlive-latex-20210325-46.fc36.noarch How reproducible: Always Steps to Reproduce: 1. Build the why3 package in Rawhide 2. 3. Actual results: pdflatex fails an assertion Expected results: Successful completion Additional info:
Jerry, I'm closing this as a dupe of 2042421, but if you have any time to spare and help figuring out what's failing here and how to fix it, it would be appreciated. I am buried in my day job right now. It could be: 1. Something new in GCC 2. I stopped passing -std=c++11 everywhere, instead only passing -std=c++17 in the pdflatex bits that need the new poppler (thanks poppler) 3. Something broken in the patch I did to get pdflatex to build with the new poppler. If not, no worries, I will get to it as soon as I can.
*** This bug has been marked as a duplicate of bug 2042421 ***