summaryrefslogtreecommitdiff
path: root/pp_ctl.c
diff options
context:
space:
mode:
authorFather Chrysostomos <sprout@cpan.org>2013-08-06 06:08:21 -0700
committerFather Chrysostomos <sprout@cpan.org>2013-08-06 06:08:21 -0700
commit11609d9c96f9c025675f6215051ab94d6735ddd9 (patch)
tree08fa589b758a362e67cfcd1ba2860bc776acca4b /pp_ctl.c
parent1cd19e121cab50d0b5b0e45a0c70cd9294a7ed9a (diff)
downloadperl-11609d9c96f9c025675f6215051ab94d6735ddd9.tar.gz
[perl #119169] index with __PACKAGE__ for 2nd argument
The refactoring of fbm_compile in 66379c06cd to prepare for c72a4eedff1 put in an SvIsCOW check before doing SvPV_force. I sim- ply changed the logic there so that SvPV_force would continue to have its effect but without tripping up on read-only variables for which SvPV_force would not need to make any changes anyway. Now, if a COW scalar is read-only, we can’t call SvPV_force on it, because it will die. It turns out that we don’t actually need to call SvPV_force on COWs. We can just go ahead and attach the BM magic and continue sharing the buffer.
Diffstat (limited to 'pp_ctl.c')
0 files changed, 0 insertions, 0 deletions