> On 16 Aug 2019, at 18:18, Charlene Wendling <juliana...@posteo.jp> wrote:
>
> Hi,
>
>> http://build-failures.rhaalovely.net/sparc64/2019-08-12/devel/angr/py-z3-solver.log
>> http://build-failures.rhaalovely.net/powerpc/2019-07-29/devel/angr/py-z3-solver.log
>
> I've found a fix for this from the "standard" z3 repo [0].
>
> As expected, it builds on macppc [1] and amd64.
>
> OK?
>
> Charlène.
>
>
> [0] https://github.com/Z3Prover/z3/pull/1612
> [1] https://bin.charlenew.xyz/py-z3-solver.log
>
>
> Index: Makefile
> ===================================================================
> RCS file: /cvs/ports/devel/angr/py-z3-solver/Makefile,v
> retrieving revision 1.2
> diff -u -p -u -p -r1.2 Makefile
> --- Makefile 23 Jul 2019 19:50:03 -0000 1.2
> +++ Makefile 16 Aug 2019 16:04:13 -0000
> @@ -5,6 +5,7 @@ COMMENT = efficient SMT solver library
> ANGR_PYTHON_MODULE = z3-solver
> # devel/angr/py-claripy requires this exact version, newer 4.8.5.0 breaks it.
> MODPY_EGG_VERSION = 4.5.1.0.post2
> +REVISION = 0
>
> CATEGORIES += math
>
> Index: patches/patch-core_src_util_lp_permutation_matrix_h
> ===================================================================
> RCS file: patches/patch-core_src_util_lp_permutation_matrix_h
> diff -N patches/patch-core_src_util_lp_permutation_matrix_h
> --- /dev/null 1 Jan 1970 00:00:00 -0000
> +++ patches/patch-core_src_util_lp_permutation_matrix_h 16 Aug 2019
> 16:04:13 -0000
> @@ -0,0 +1,17 @@
> +$OpenBSD$
> +
> +Unbreak with ports-gcc, from:
> +https://github.com/Z3Prover/z3/commit/2d5dd802386d78117d5ed9ddcbf8bc22ab3cb461
> +
> +Index: core/src/util/lp/permutation_matrix.h
> +--- core/src/util/lp/permutation_matrix.h.orig
> ++++ core/src/util/lp/permutation_matrix.h
> +@@ -117,7 +117,7 @@ class permutation_matrix : public tail_matrix<T, X> {
> +
> + unsigned size() const { return static_cast<unsigned>(m_rev.size());
> }
> +
> +- unsigned * values() const { return m_permutation; }
> ++ unsigned * values() const { return m_permutation.c_ptr(); }
> +
> + void resize(unsigned size) {
> + unsigned old_size = m_permutation.size();
OK with me, thanks for the fix.
Cheers,
Jasper
