OpenBSD ports changes summary for 2017-06-04 ============================================
devel/boost devel/grcs devel/mingw graphics/asymptote graphics/krita infrastructure/mk lang/clang lang/clisp lang/gcc lang/go lang/go-bootstrap lang/tcl net/dictd net/libmaxminddb net/olsrd net/packetq net/py-botocore productivity/tdl security/p5-Crypt-OpenSSL-EC sysutils/amazon-ecs-cli sysutils/awscli sysutils/ec2-api-tools sysutils/rdiff-backup www/man2web www/netsurf www/py-gunicorn x11/kde4 x11/rxvt x11/tk == devel ============================================================= 01/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/devel boost ~ Makefile ~ files/user.hpp > the tests are wrong, but it's simpler to define the annoying constant to 0 > (espie@) grcs ~ Makefile > mandoc groks .ec and .eo now; drop USE_GROFF and bump (schwarze@) mingw ~ Makefile ~ patches/patch-x86-mingw32-build_sh_conf > THIS configuration fragment *is* clang specific (espie@) == graphics ========================================================== 02/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/graphics asymptote - patches/patch-genv_cc - patches/patch-prc_writePRC_h ~ Makefile ~ distinfo + patches/patch-settings_cc > update to asymptote-2.41 > tests pass on amd64/gcc and amd64/ports clang. > aarch64/clang not tested as it hanged again (pv).. (sthen@) ~ Makefile > remove stray COMPILER= setting from testing (sthen@) ~ Makefile > Add a "NOT_FOR_ARCHS=${CLANG_ARCHS}" to asymptote for now, with a > comment explaining what's not working (I would have just let it fail > so it's easier to spot, but it hangs during build). this is a terminal > port, not knocking out others. > This does not fail with ports clang though.. (sthen@) krita + patches/patch-plugins_extensions_CMakeLists_txt > unbreak with clang, no reason to insist on gcc (espie@) == infrastructure ==================================================== 03/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/infrastructure mk ~ bsd.port.mk > Write out wrapper scripts for compilers using -B to set the toolchain > search path to ${WRKDIR}/bin. Ensures that the ld wrapper script needed > for USE_WXNEEDED gets used (clang was ignoring it otherwise). > Rework ccache support and gcc4/clang MODULES to use the same wrapper > generator as above, setting up a list of command names (cc, c++, etc) and > destinations (egcc, eg++, etc) as the port is processed, and writing the > wrappers in one place in bsd.port.mk, with an optional wrapper (e.g. > ccache), and with an echo to make it clear on screen/in log which > compiler binaries were chosen (easier than inspecting the wrapper > scripts) and whether ccache is used. > Initial wrapper using -B from espie, ccache bits from me, discussed > with/ok espie. Been through a bulk on i386. (sthen@) == lang ============================================================== 04/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/lang clang ~ clang.port.mk > Write out wrapper scripts for compilers using -B to set the toolchain > search path to ${WRKDIR}/bin. Ensures that the ld wrapper script needed > for USE_WXNEEDED gets used (clang was ignoring it otherwise). > Rework ccache support and gcc4/clang MODULES to use the same wrapper > generator as above, setting up a list of command names (cc, c++, etc) and > destinations (egcc, eg++, etc) as the port is processed, and writing the > wrappers in one place in bsd.port.mk, with an optional wrapper (e.g. > ccache), and with an echo to make it clear on screen/in log which > compiler binaries were chosen (easier than inspecting the wrapper > scripts) and whether ccache is used. > Initial wrapper using -B from espie, ccache bits from me, discussed > with/ok espie. Been through a bulk on i386. (sthen@) clisp ~ Makefile > Disable the use of groff. > Otherwise, the PostScript versions of the man pages may > be created or not during build, depending on wether groff > is installed or not, which may lead to consusion when > updating this port. > Suggested by Ingo a few days ago. (kili@) gcc ~ 4.6/gcc4.port.mk ~ 4.9/gcc4.port.mk ~ 6/gcc4.port.mk > Write out wrapper scripts for compilers using -B to set the toolchain > search path to ${WRKDIR}/bin. Ensures that the ld wrapper script needed > for USE_WXNEEDED gets used (clang was ignoring it otherwise). > Rework ccache support and gcc4/clang MODULES to use the same wrapper > generator as above, setting up a list of command names (cc, c++, etc) and > destinations (egcc, eg++, etc) as the port is processed, and writing the > wrappers in one place in bsd.port.mk, with an optional wrapper (e.g. > ccache), and with an echo to make it clear on screen/in log which > compiler binaries were chosen (easier than inspecting the wrapper > scripts) and whether ccache is used. > Initial wrapper using -B from espie, ccache bits from me, discussed > with/ok espie. Been through a bulk on i386. (sthen@) go ~ Makefile ~ patches/patch-src_cmd_link_internal_ld_lib_go + patches/patch-src_cmd_go_go_test_go + patches/patch-src_cmd_link_internal_ld_ldelf_go > Make lang/go work with clang. > This requires three changes: > - OpenBSD clang produces __guard_local symbols that are marked as hidden, > which causes problems with the go linker. Workaround this by explicitly > making __guard_local symbols visible when reading in an ELF object. > - OpenBSD clang currently fails to correctly report the compiler runtime > via --print-libgcc-file-name. Workaround this issue by hardcoding it > for the time being. > - The TestCgoConsistentResults test passes successfully outside of ports, > however fails under a ports build. This requires further investigation > but we can just skip the test for now. (jsing@) go-bootstrap ~ Makefile ~ pkg/PLIST > Build lang/go-bootstrap without enabling cgo. > This functionality is not required for bootstrapping lang/go - disabling > it reduces future maintenance and allows it to build with clang. (jsing@) tcl ~ 8.5/Makefile ~ 8.6/Makefile > mandoc groks .mc now; drop USE_GROFF and bump (schwarze@) == net =============================================================== 05/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/net dictd ~ Makefile > mandoc now copes with blank lines in TP heads; drop USE_GROFF and bump > (schwarze@) libmaxminddb ~ Makefile + patches/patch-configure + patches/patch-include_maxminddb_h + patches/patch-src_maxminddb_c > remove the problematic shit that doesn't compile with our old gcc > because the combo of -fms-extensions -std=c99 > okay fcambus@ (espie@) olsrd ~ Makefile ~ distinfo ~ patches/patch-Makefile_inc ~ patches/patch-make_Makefile_obsd ~ patches/patch-src_bsd_kernel_routes_c ~ patches/patch-src_bsd_net_c ~ patches/patch-src_cfgparser_local_mk ~ patches/patch-src_olsr_random_h ~ patches/patch-src_plugin_loader_c ~ patches/patch-src_unix_ifnet_c ~ pkg/PLIST-main > Update olsrd (again) to the actual latest stable release 0.9.6.1. > ok phessler@ sthen@ (stsp@) packetq ~ Makefile ~ distinfo > Update packetq to 1.3.1. (fcambus@) py-botocore ~ Makefile ~ distinfo ~ pkg/PLIST > Update to py-botocore-1.5.60. (ajacoutot@) == productivity ====================================================== 06/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/productivity tdl ~ Makefile ~ patches/patch-Makefile_in + patches/patch-tdl_1 > Don't spoil the tables by running GNU tbl(1) at build time, > just install the source and drop USE_GROFF. > While here, fix wrong escaping that may lose information. (schwarze@) == security ========================================================== 07/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/security p5-Crypt-OpenSSL-EC ~ Makefile ~ distinfo > Update to p5-Crypt-OpenSSL-EC-1.31. (benoit@) == sysutils ========================================================== 08/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/sysutils amazon-ecs-cli ~ Makefile ~ distinfo + patch-ecs-cli_vendor_github_com_docker_docker_pkg_system_stat_openbsd_go > Update to ecs-cli-0.5.0. (ajacoutot@) awscli ~ Makefile ~ distinfo + patches/patch-awscli_help_py > Update to awscli-1.11.97 and drop the requirement on groff. (ajacoutot@) ec2-api-tools ~ Makefile > It looks like groff is not needed anymore. (ajacoutot@) rdiff-backup ~ Makefile > whitespace differences only, drop USE_GROFF and bump (schwarze@) == www =============================================================== 09/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/www man2web ~ Makefile > whitespace differences only, drop USE_GROFF and bump (schwarze@) netsurf ~ libnsfb/Makefile + libnsfb/patches/patch-Makefile > disable hidden dep on libvncserver (sthen@) py-gunicorn ~ patches/patch-requirements_test_txt > regen patch (danj@) ~ Makefile ~ pkg/PLIST > unbreak, found by semarie (danj@) == x11 =============================================================== 10/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/x11 kde4 ~ cantor/Makefile > bump, this lost and regained some fortran-related lib deps at some point > run into by naddy and aja (sthen@) rxvt ~ Makefile > Rather than making the manual page unusable for mandoc by preprocessing > it with GNU tbl(1) at build time, simply install the man(7)/tbl(7) > source unmangled --> identical formatting, drop USE_GROFF. (schwarze@) tk ~ 8.5/Makefile ~ 8.6/Makefile > mandoc groks .mc now; drop USE_GROFF and bump (schwarze@) =============================================================================== _______________________________________________ odc mailing list [email protected] http://www.squish.net/mailman/listinfo/odc
