Branch: refs/heads/master
Home: https://github.com/perl5-dbi/dbi
Commit: adcea5faa4cf2b445a12fbc63a17bb5d263c7d25
https://github.com/perl5-dbi/dbi/commit/adcea5faa4cf2b445a12fbc63a17bb5d263c7d25
Author: Paul Johnson <[email protected]>
Date: 2026-07-16 (Thu, 16 Jul 2026)
Changed paths:
M DBI.xs
A t/44prof_caller.t
Log Message:
-----------
Fix !Caller2 caller loss when $^P is set
dbi_caller_cop() mirrors perl's pp_caller but omits its check that
&DB::sub is actually defined before treating a frame as a debugger
frame. When $^P is set without &DB::sub being defined, as happens
under Devel::Cover and similar tools, PL_DBsub is initialised but
GvCV(PL_DBsub) is NULL. For an eval or do FILE context blk_sub.cv
aliases blk_eval.old_eval_root, which is NULL at the top level, so
such frames were skipped and the "via ..." part of !Caller2 profile
paths and of trace output was silently dropped. Perl core fixed the
same bug in pp_caller in 2005 (RT 35059, core commit f2a7f2982c).
Fixes #177
Commit: 317625e076237041d2f4b94b105e08d42d698ab5
https://github.com/perl5-dbi/dbi/commit/317625e076237041d2f4b94b105e08d42d698ab5
Author: H.Merijn Brand <[email protected]>
Date: 2026-08-15 (Sat, 15 Aug 2026)
Changed paths:
M DBI.xs
A t/44prof_caller.t
Log Message:
-----------
Merge pull request #184 from pjcj/GH-177-caller2-dbsub-guard
Fix !Caller2 caller loss when $^P is set
Compare: https://github.com/perl5-dbi/dbi/compare/6afdea97fe88...317625e07623
To unsubscribe from these emails, change your notification settings at
https://github.com/perl5-dbi/dbi/settings/notifications
