[sparc64] fix py-matplotlib build

as(1) does not understand things emitted by gcc/15 with -flto=auto,
which is enabled by b_lto=true in meson.build, resulting in this:

c++ -Iextern/agg24-svn/libagg.a.p -Iextern/agg24-svn -I../extern/agg24-svn -I../extern/agg24-svn/include -fvisibility=hidden -fvisibility-inlines-hidden -flto=auto -fdiagnostics-color=always -DNDEBUG -D_FILE_OFFSET_BITS=64 -Wall -Winvalid-pch -std=c++17 -O3 -fPIC -MD -MQ extern/agg24-svn/libagg.a.p/src_agg_bezier_arc.cpp.o -MF extern/agg24-svn/libagg.a.p/src_agg_bezier_arc.cpp.o.d -o extern/agg24-svn/libagg.a.p/src_agg_bezier_arc.cpp.o -c ../extern/agg24-svn/src/agg_bezier_arc.cpp
/tmp//ccNoBMVq.s: Assembler messages:
/tmp//ccNoBMVq.s:2: Fatal error: unrecognized .section attribute: want a,w,x,M,S,G,T
ninja: build stopped: subcommand failed.

The offending line is

.section .gnu.lto_.profile.5712db33fb59d8ee,"e",@progbits

and the "e" (exclude section from linking) is what trips up as(1), see
obj_elf_parse_section_letters() in gas/config/obj-elf.c.

The below isn't pretty but fixes the build on sparc64. If there are
better ideas, I'm all ears.

Index: Makefile
===================================================================
RCS file: /cvs/ports/graphics/py-matplotlib/Makefile,v
diff -u -p -r1.109 Makefile
--- Makefile 13 Mar 2026 21:57:15 -0000 1.109
+++ Makefile 26 Mar 2026 05:48:43 -0000
@@ -70,3 +70,9 @@ post-install:
cd ${WRKSRC}/galleries/examples && pax -rw * ${EXAMPLESDIR}

.include <bsd.port.mk>
+
+.if ${CHOSEN_COMPILER} == "ports-gcc"
+# as(1) can't deal with port-gcc's output of -flto=auto
+# Fatal error: unrecognized .section attribute: want a,w,x,M,S,G,T
+MODPY_PYBUILD_ARGS+=--config-setting=setup-args=-Db_lto=false
+.endif