metrics: avoid possible segfault after restart twice

[atzm]

Accessing the metrics endpoint after restarting apache twice (using apachectl restart) can result in a segfault due to internal state corruption, especially when using prefork mpm.

[Thu May 22 15:20:51.875701 2025] [core:notice] [pid 9258:tid 9258] AH00051: child pid 9297 exit signal Segmentation fault (11), possible coredump in /usr/local/apache2

#0  0x00007f0dab976e33 in _oidc_metrics_storage_get (s=0x7f0daaf1f100) at src/metrics.c:339
#1  oidc_metrics_handle_request (r=0x7f0d96bd50b0) at src/metrics.c:1585
#2  0x00007f0dab953878 in oidc_content_handler (r=0x7f0d96bd50b0) at src/handle/content.c:60
#3  0x0000558faf66b0bb in ap_run_handler ()
#4  0x0000558faf66bc1b in ap_invoke_handler ()
#5  0x0000558faf68defb in ap_process_async_request ()
#6  0x0000558faf68dfc7 in ap_process_request ()
#7  0x0000558faf689484 in ap_process_http_sync_connection ()
#8  0x0000558faf6895a6 in ap_process_http_connection ()
#9  0x0000558faf67b78b in ap_run_process_connection ()
#10 0x0000558faf67be79 in ap_process_connection ()
#11 0x00007f0dab9e195b in child_main () from /usr/local/apache2/modules/mod_mpm_prefork.so
#12 0x00007f0dab9e1bee in make_child () from /usr/local/apache2/modules/mod_mpm_prefork.so
#13 0x00007f0dab9e1c55 in startup_children () from /usr/local/apache2/modules/mod_mpm_prefork.so
#14 0x00007f0dab9e2319 in prefork_run () from /usr/local/apache2/modules/mod_mpm_prefork.so
#15 0x0000558faf63dde8 in ap_run_mpm ()
#16 0x0000558faf6337ef in main ()

[zandbelt]

thanks for spotting and fixing; could you please add an entry to the ChangeLog and add yourself to the AUTHORS file?

[atzm]

sure. thank you for the quick review!