HTTP/2 200 cache-control: max-age=43200 server: Combust/Plack (Perl) vary: Accept-Encoding content-encoding: gzip content-length: 1602 content-type: text/html; charset=utf-8 last-modified: Sat, 11 Oct 2025 12:02:53 GMT traceparent: 2f3ae80bc97c1dfca8030013930a1314 strict-transport-security: max-age=15768000 Front page | perl.perl5.porters | Postings from July 2008

Code review request for the perl guts vodoo in NYTProf v2

In my headlong charge to get Devel::NYTProf v2 out before OSCON I've not
had much time to review the code I'd cooked up to dig around in perl
internals.
For example:
#if (PERL_VERSION < 8) || ((PERL_VERSION == 8) && (PERL_SUBVERSION < 8))
/* If we're using DB::DB() instead of opcode redirection with an old perl
 * then PL_curcop in DB() will refer to the DB() wrapper in Devel/NYTProf.pm
 * so we'd have to crawl the stack to find the right cop. However, for some
 * reason that I don't pretend to understand the following expression works:
 */
#define PL_curcop_nytprof (use_db_sub ? ((cxstack + cxstack_ix)->blk_oldcop) : PL_curcop)
#else
#define PL_curcop_nytprof PL_curcop
#endif
Others include the visit_contexts(), check_context(),
start_cop_of_context() and closest_cop() functions.
I've been playing with magic beyond my powers and have, so far, survived
unscathed (not counting too many late night hacking sessions).
I suspect that the code is suboptimal in places. Perhaps lots of places.
Any feedback gratefully recieved.
Tim.
https://blog.timbunce.org/2008/07/15/nytprof-v2-a-major-advance-in-perl-profilers/
https://code.google.com/p/perl-devel-nytprof/source/browse/trunk/NYTProf.xs

• Code review request for the perl guts vodoo in NYTProf v2 by Tim Bunce