Debian Patches
Status for gecode/6.2.0-8
| Patch | Description | Author | Forwarded | Bugs | Origin | Last update |
|---|---|---|---|---|---|---|
| examples-include | =================================================================== | no | ||||
| ppc64el-fbtfs-858132 | =================================================================== | no | ||||
| no-O3 | =================================================================== | no | ||||
| parallel.patch | Fix parallel build failure In parallel builds this "bison && mv" ran twice in parallel, FTBFS if this resulted in the following sequence: bison -t -o gecode/flatzinc/parser.tab.cpp -d gecode/flatzinc/parser.yxx bison -t -o gecode/flatzinc/parser.tab.cpp -d gecode/flatzinc/parser.yxx mv gecode/flatzinc/parser.tab.hpp gecode/flatzinc/parser.tab.hh mv gecode/flatzinc/parser.tab.hpp gecode/flatzinc/parser.tab.hh Running it only once fixes the problem. =================================================================== |
Adrian Bunk <bunk@debian.org> | no | debian | ||
| system-boost | =================================================================== | no | ||||
| bison-parser-tab-ftbfs-966903 | =================================================================== | no | ||||
| configure-ac-978817 | =================================================================== | no | ||||
| const-gcc14-1074982 | no | |||||
| forceinline-virtual-dtor-gcc16-ftbfs-1133488 | Drop always_inline on three virtual destructors that delete polymorphic members (Patch from Claude) . Three destructors in gecode/search/cutoff.hpp (CutoffAppend, CutoffMerge, CutoffRepeat) are marked `forceinline`, which expands under GCC to `inline __attribute__((__always_inline__))`. Their bodies invoke `delete` on members of type `Cutoff*`, a class with a virtual destructor. The inlined call graph therefore folds the recursive destruction of every derived Cutoff subclass into each call site, and under g++-16 the cost model rejects the expansion with . error: inlining failed in call to 'always_inline' 'virtual Gecode::Search::CutoffAppend::~CutoffAppend() noexcept': --param max-inline-insns-single limit reached . With g++-15 the same source compiled because the inliner budget happened to admit the bodies; nothing semantically depends on inlining a virtual destructor that is always invoked through a vtable, so the `always_inline` annotation was never load-bearing. Replace `forceinline` with plain `inline` on the three destructor definitions. `inline` preserves ODR (the bodies sit in a header pulled into ~48 translation units via gecode/search.hh) and lets the compiler decide. The constructors keep `forceinline`; only the three destructors change. =================================================================== |
Kari Pahula <kaol@debian.org> | no | 2026-05-17 |