Opened 31 hours ago
Last modified 31 hours ago
#71297 new defect
postgresql16 @16.4 fails test because of too long path name of socket
Reported by: | ballapete (Peter "Pete" Dyballa) | Owned by: | |
---|---|---|---|
Priority: | Normal | Milestone: | |
Component: | ports | Version: | 2.10.2 |
Keywords: | Cc: | dgilman (David Gilman), vital.had@… | |
Port: | postgresql16 |
Description
gnumake[2]: Nothing to be done for `all'. echo "# +++ regress check in src/test/regress +++" && PATH="/opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_databases_postgresql16/postgresql16/work/postgresql-16.4/tmp_install/opt/local/lib/postgresql16/bin:/opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_databases_postgresql16/postgresql16/work/postgresql-16.4/src/test/regress:$PATH" DYLD_LIBRARY_PATH="/opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_databases_postgresql16/postgresql16/work/postgresql-16.4/tmp_install/opt/local/lib/postgresql16" ../../../src/test/regress/pg_regress --temp-instance=./tmp_check --inputdir=. --bindir= --dlpath=. --max-concurrent-tests=20 --schedule=./parallel_schedule # +++ regress check in src/test/regress +++ # postmaster failed, examine "/opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_databases_postgresql16/postgresql16/work/postgresql-16.4/src/test/regress/log/postmaster.log" for the reason Bail out!gnumake[1]: *** [check] Error 2 gnumake: *** [check] Error 2 Command failed: cd "/opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_databases_postgresql16/postgresql16/work/postgresql-16.4" && /usr/bin/gnumake check Exit code: 2 Error: Failed to test postgresql16: command execution failed Error: See /opt/local/var/macports/logs/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_databases_postgresql16/postgresql16/main.log for details. Error: Follow https://guide.macports.org/#project.tickets if you believe there is a bug.
And /opt/local/var/macports/logs/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_databases_postgresql16/postgresql16/main.log
has exactly this contents:
1 2024-11-12 21:11:17.731 CET postmaster[53649] LOG: starting PostgreSQL 16.4 on x86_64-apple-darwin17.7.0, compiled by Apple LLVM version 10.0.0 (clang-1000.10.44.4), 64-bit 2 2024-11-12 21:11:17.732 CET postmaster[53649] LOG: Unix-domain socket path "/opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_databases_postgresql16/postgresql16/work/.tmp/pg_regress-MeKhhq/.s.PGSQL.61700" is too long (maximum 103 bytes) 3 2024-11-12 21:11:17.732 CET postmaster[53649] WARNING: could not create Unix-domain socket in directory "/opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_databases_postgresql16/postgresql16/work/.tmp/pg_regress-M eKhhq" 4 2024-11-12 21:11:17.732 CET postmaster[53649] FATAL: could not create any Unix-domain sockets 5 2024-11-12 21:11:17.733 CET postmaster[53649] LOG: database system is shut down ~
Couldn't the string "_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports", 83 characters long, be removed that socket related test do not fail? "_opt_local_var_macports_sources" is complete nonsense, echoing a standard. Is someone not able remember a short path name? The RSYNC server can be coded with a number or some Unicode symbol. Does a user have an option to something else than a "release"? Or a "tarball"? YES: ZIP. Or use something else than a "port"?
See also: #71221.
Attachments (1)
Change History (3)
Changed 31 hours ago by ballapete (Peter "Pete" Dyballa)
comment:1 Changed 31 hours ago by ryandesign (Ryan Carsten Schmidt)
comment:2 Changed 31 hours ago by ballapete (Peter "Pete" Dyballa)
Which role plays this string? Isn't MacPorts set up to perform anything below $prefix? So at least these characters do not need to be repeated. The diversification of the file system branch that starts here has its fixed rules. Is there anyone insane enough to change them? Which will hold until the next release of MacPorts…
Main.log from High Sierra 10.13.6