Re: BUG #19095: Test if function exit() is used fail when linked static

Hi Andres and hackers,
I came up with the solution
— short follow-up with what I changed and how the Meson check actually runs.

Summary of what I did
- Added a Meson custom_target in src/interfaces/libpq/meson.build that
scans libpq's object files for direct exit() references.
- Added pthread_exit to the Makefile whitelist (the Makefile test
already existed; this just updates the whitelist).
-Added the custom target in the top level meson.build

Where the Meson check lives and when it runs
- I put the custom_target in src/interfaces/libpq/meson.build,
immediately after the libpq shared/static library targets are
defined and after the declare_dependency for libpq. That is the
correct location in the libpq build file.
- The custom_target declares `depends: [libpq_st, libpq_so]`. This
is important: it tells Meson to build those targets first so the
.o files are present in the build directory before the check runs.
The check itself does not scan libpq_st or libpq_so; it scans
files in build/src/interfaces/libpq/*.o.
- The command uses `find <builddir>/src/interfaces/libpq -name '*.o'`
and runs `nm -u` on each .o, piping through `grep -v` for the
whitelisted names and finally checking for exit.
- The check is skipped for:
* cross-builds
* when coverage (b_coverage) is enabled
* on Windows (no nm/grep in the same form there) as Andrew questioned

Why pthread_exit is whitelisted
- pthread_exit can legitimately appear in a few build/runtime
configurations (thread runtimes or link-time glue), and the
Makefile test was updated to whitelist it. The Meson check has the
same whitelist so both build systems behave the same.
- Whitelisting pthread_exit doesn't remove the value of the test:
it only avoids false positives for legitimate thread shutdown code.
We still catch direct exit()/ _exit()/abort()-style calls as tom said.

How to reproduce locally
- From repo root:
rm -rf build
meson setup build
cd build
ninja
The custom_target runs as part of the normal build and will fail
the build if any .o contains an un-whitelisted exit() reference.

On Fri, Nov 14, 2025 at 7:31 PM Andres Freund <andres(at)anarazel(dot)de> wrote:

But more generally: If we allow pthread_exit(), what's the point of this
> test?
> That's one of the functions we better avoid calling, no?
>
I agree it's worth questioning which functions we allow. The
current choice (whitelist pthread_exit) mirrors the Makefile
behavior, avoids false positives, and keeps the test focused on
direct process-termination calls authored in our code. If the
community prefers a stricter policy so we can adjust the whitelist.

> ISTM that if we do want to continue having this test, the issue is that
> we're
> testing the shared library - which will have already linked against static
> libraries like the sanitizer ones or in this case libcrypto. What we ought
> to
> do is to test the .o files constituting libpq.so, rather than the already
> linked .so. That way we will find our own calls to exit etc, but not ones
> in
> static libraries.
>

TBH After so many test runs I have finalized
Why scan .o files (not the final .so)?
- A shared library is usually linked with other static libraries
(libcrypto, sanitizer runtimes). If we scan the final .so we'll
see references that originate in those static libraries and produce
false positives.
- If we scan the .o files that make up libpq, we only inspect our
own compilation units and will catch only exit() calls introduced
by our code.

HTH! I attached the patch and also added the meson test output before
/after of any file containing an 'exit' explicitly; it fails the
build[ninja].

Regards,
Vasuki M
BharatDB[CDAC chennai]