gnutls-devel 3.8.1 conflicts with legacy-support: benchmark.h: error: static declaration of 'gettime' follows non-static declaration

[barracuda156]

/bin/sh ../libtool  --tag=CC   --mode=compile /opt/local/bin/gcc-mp-12 -DHAVE_CONFIG_H -I. -I..  -I./gl -I./gl -I./../lib/includes -I./../lib/includes -I./../libdane/includes -I./../extra/includes  -I/opt/local/include -I/opt/local/include/LegacySupport -fanalyzer -fno-common -Wall -Wbad-function-cast -Wcast-align=strict -Wdate-time -Wdisabled-optimization -Wdouble-promotion -Wduplicated-branches -Wduplicated-cond -Wextra -Winit-self -Winvalid-pch -Wlogical-op -Wmissing-declarations -Wmissing-include-dirs -Wmissing-prototypes -Wnested-externs -Wnull-dereference -Wold-style-definition -Wopenmp-simd -Wpacked -Wpointer-arith -Wshadow -Wstrict-prototypes -Wsuggest-attribute=cold -Wsuggest-attribute=format -Wsuggest-attribute=malloc -Wsuggest-final-methods -Wsuggest-final-types -Wsync-nand -Wtrampolines -Wuninitialized -Wunknown-pragmas -Wunused-macros -Wvariadic-macros -Wvector-operation-performance -Wwrite-strings -Warray-bounds=2 -Wattribute-alias=2 -Wbidi-chars=any,ucn -Wformat-overflow=2 -Wformat=2 -Wformat-truncation=2 -Wimplicit-fallthrough=5 -Wshift-overflow=2 -Wuse-after-free=3 -Wunused-const-variable=2 -Wvla-larger-than=4031 -Wno-analyzer-malloc-leak -Wno-missing-field-initializers -Wno-unused-parameter -Wno-format-truncation -Wimplicit-fallthrough=2 -Wabi=11 -fdiagnostics-show-option -fno-builtin-strcmp  -pipe -Os -I/opt/local/include/LegacySupport -arch ppc -c -o gnutls-serv-options.lo gnutls-serv-options.c
/opt/local/bin/gcc-mp-12 -DHAVE_CONFIG_H -I. -I..  -I./gl -I./gl -I./../lib/includes -I./../lib/includes -I./../libdane/includes -I./../extra/includes  -I/opt/local/include -I/opt/local/include/LegacySupport -fanalyzer -fno-common -Wall -Wbad-function-cast -Wcast-align=strict -Wdate-time -Wdisabled-optimization -Wdouble-promotion -Wduplicated-branches -Wduplicated-cond -Wextra -Winit-self -Winvalid-pch -Wlogical-op -Wmissing-declarations -Wmissing-include-dirs -Wmissing-prototypes -Wnested-externs -Wnull-dereference -Wold-style-definition -Wopenmp-simd -Wpacked -Wpointer-arith -Wshadow -Wstrict-prototypes -Wsuggest-attribute=cold -Wsuggest-attribute=format -Wsuggest-attribute=malloc -Wsuggest-final-methods -Wsuggest-final-types -Wsync-nand -Wtrampolines -Wuninitialized -Wunknown-pragmas -Wunused-macros -Wvariadic-macros -Wvector-operation-performance -Wwrite-strings -Warray-bounds=2 -Wattribute-alias=2 -Wbidi-chars=any,ucn -Wformat-overflow=2 -Wformat=2 -Wformat-truncation=2 -Wimplicit-fallthrough=5 -Wshift-overflow=2 -Wuse-after-free=3 -Wunused-const-variable=2 -Wvla-larger-than=4031 -Wno-analyzer-malloc-leak -Wno-missing-field-initializers -Wno-unused-parameter -Wno-format-truncation -Wimplicit-fallthrough=2 -Wabi=11 -fdiagnostics-show-option -fno-builtin-strcmp  -pipe -Os -I/opt/local/include/LegacySupport -arch ppc -c -o cli.o cli.c
In file included from cli.c:65:
benchmark.h:38:20: error: static declaration of 'gettime' follows non-static declaration
   38 | inline static void gettime(struct timespec *ts)
      |                    ^~~~~~~
In file included from benchmark.h:32:
./gl/timespec.h:93:6: note: previous declaration of 'gettime' with type 'void(struct timespec *)'
   93 | void gettime (struct timespec *) _GL_ARG_NONNULL ((1));
      |      ^~~~~~~
libtool: compile:  /opt/local/bin/gcc-mp-12 -DHAVE_CONFIG_H -I. -I.. -I./gl -I./gl -I./../lib/includes -I./../lib/includes -I./../libdane/includes -I./../extra/includes -I/opt/local/include -I/opt/local/include/LegacySupport -fanalyzer -fno-common -Wall -Wbad-function-cast -Wcast-align=strict -Wdate-time -Wdisabled-optimization -Wdouble-promotion -Wduplicated-branches -Wduplicated-cond -Wextra -Winit-self -Winvalid-pch -Wlogical-op -Wmissing-declarations -Wmissing-include-dirs -Wmissing-prototypes -Wnested-externs -Wnull-dereference -Wold-style-definition -Wopenmp-simd -Wpacked -Wpointer-arith -Wshadow -Wstrict-prototypes -Wsuggest-attribute=cold -Wsuggest-attribute=format -Wsuggest-attribute=malloc -Wsuggest-final-methods -Wsuggest-final-types -Wsync-nand -Wtrampolines -Wuninitialized -Wunknown-pragmas -Wunused-macros -Wvariadic-macros -Wvector-operation-performance -Wwrite-strings -Warray-bounds=2 -Wattribute-alias=2 -Wbidi-chars=any,ucn -Wformat-overflow=2 -Wformat=2 -Wformat-truncation=2 -Wimplicit-fallthrough=5 -Wshift-overflow=2 -Wuse-after-free=3 -Wunused-const-variable=2 -Wvla-larger-than=4031 -Wno-analyzer-malloc-leak -Wno-missing-field-initializers -Wno-unused-parameter -Wno-format-truncation -Wimplicit-fallthrough=2 -Wabi=11 -fdiagnostics-show-option -fno-builtin-strcmp -pipe -Os -I/opt/local/include/LegacySupport -arch ppc -c gnutls-serv-options.c  -fno-common -DPIC -o .libs/gnutls-serv-options.o
make[3]: *** [cli.o] Error 1
make[3]: *** Waiting for unfinished jobs....
make[3]: Leaving directory `/opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_devel_gnutls/gnutls-devel/work/gnutls-3.8.1/src'
make[2]: *** [all] Error 2

[Schamschula (Marius Schamschula)]

I'm aware of this issue. Unfortunately, I can only guess at workarounds, as I don't have active legacy systems available to me.

[mascguy (Christopher Nielsen)]

I have an old 32-bit MacBook Pro, running both 10.6 and 10.5. Ditto for 64-bit Intel VMs, from 10.5 and up.

But it might help if the ticket included details regarding the macOS version, as well as the hardware...

[barracuda156]

Replying to mascguy:

I have an old 32-bit MacBook Pro, running both 10.6 and 10.5. Ditto for 64-bit Intel VMs, from 10.5 and up.

But it might help if the ticket included details regarding the macOS version, as well as the hardware...

My usual builds are on 10.6 / 10.6.8 for ppc (32 bit), but this issue does not look like anything arch-related. Apparently just defines in the source code do not work well with legacysupport, if it is not just the bug in the source. I am away from Intel machine, but I suspect it will fail on older Intel too. And it looks like it is broken now on El Capitan down: ​https://ports.macports.org/port/gnutls-devel/details

[barracuda156]

In file included from benchmark-cipher.c:33:
./benchmark.h:38:20: error: static declaration of 'gettime' follows non-static declaration
inline static void gettime(struct timespec *ts)
                   ^
./gl/timespec.h:93:6: note: previous declaration is here
void gettime (struct timespec *) _GL_ARG_NONNULL ((1));
     ^
gnutls-serv-options.c:416:56: warning: result of comparison of constant 2147483648 with expression of type 'int' is always false [-Wtautological-constant-out-of-range-compare]
  if (HAVE_OPT(MAXEARLYDATA) && OPT_VALUE_MAXEARLYDATA > 2147483648)
                                ~~~~~~~~~~~~~~~~~~~~~~ ^ ~~~~~~~~~~
1 error generated.

Clang on 10.11: ​https://build.macports.org/builders/ports-10.11_x86_64-builder/builds/231541/steps/install-port/logs/stdio

[barracuda156]

Given this is not a ppc bug, changing priority to high.

[mascguy (Christopher Nielsen)]

Replying to barracuda156:

Given this is not a ppc bug, changing priority to high.

Bear in mind that this is a Devel port though, meaning our users generally aren't using it. So no need to bump the priority to high.

[barracuda156]

Replying to mascguy:

Replying to barracuda156:

Given this is not a ppc bug, changing priority to high.

Bear in mind that this is a Devel port though, meaning our users generally aren't using it. So no need to bump the priority to high.

Agreed, a fair point.

[barracuda156]

Replying to Schamschula:

I'm aware of this issue. Unfortunately, I can only guess at workarounds, as I don't have active legacy systems available to me.

If you suggest a solution, I will test that.

[mascguy (Christopher Nielsen)]

Replying to barracuda156:

If you suggest a solution, I will test that.

I'll take a look on my 10.5 x86_64 VM. Stay tuned...

[barracuda156]

Replying to mascguy:

Replying to barracuda156:

If you suggest a solution, I will test that.

I'll take a look on my 10.5 x86_64 VM. Stay tuned...

Sounds good, thanks.

[mascguy (Christopher Nielsen)]

Replying to mascguy:

I'll take a look on my 10.5 x86_64 VM.

It looks like the build errors are fixed, when Kirill's pending changes for clock_gettime are included in legacy-support:

​PR 58 - clock_gettime: added missing CLOCK_*

No time to dig into the details yet, as my day job is starting. But encouraging for sure.

So @cjones... any thoughts on Kirill's PR? Would be really nice to finally merge that, particularly now that it fixes at least one port...

[mascguy (Christopher Nielsen)]

Replying to mascguy:

Replying to mascguy:

I'll take a look on my 10.5 x86_64 VM.

It looks like the build errors are fixed, when Kirill's pending changes for clock_gettime are included in legacy-support:

​PR 58 - clock_gettime: added missing CLOCK_*

No time to dig into the details yet, as my day job is starting. But encouraging for sure.

So @cjones... any thoughts on Kirill's PR? Would be really nice to finally merge that, particularly now that it fixes at least one port...

Chris just approved the PR, so we're good-to-go. I'll update legacy-support-devel, with Kirill's additions.

[Christopher Nielsen <mascguy@…>]

In d11a55b539a703eb7760a3d68edc8335e942cbff/macports-ports (master):

legacy-support-devel: update to 20230822

• time: add additional CLOCK_* support

See: ​https://github.com/macports/macports-legacy-support/pull/58
See: #67945

[catap (Kirill A. Korinsky)]

BTW I do have some idea how to get reasonable enough CLOCK_THREAD_CPUTIME_ID, see: ​https://github.com/macports/macports-legacy-support/pull/64

[barracuda156]

Replying to mascguy:

Replying to mascguy:

I'll take a look on my 10.5 x86_64 VM.

It looks like the build errors are fixed, when Kirill's pending changes for clock_gettime are included in legacy-support

Yes, all is good now!

[mascguy (Christopher Nielsen)]

Since a new release of legacy-support will be forthcoming over the next few days, closing as fixed.

[Christopher Nielsen <mascguy@…>]

In 92d479b07658094732fee8a4f2f4dc9ae8b136b4/macports-ports (master):

legacy-support: update to 1.0.13

• time: add additional CLOCK_* support

See: ​https://github.com/macports/macports-legacy-support/pull/58
See: #67945