In perl.git, the branch maint-5.20-votes has been updated <http://perl5.git.perl.org/perl.git/commitdiff/f8ae94d505f08de490f4ddf290a3ff28f2770892?hp=b54a6b5663ebad6dffd46e30d3dd9b1fe59fcaa5>
- Log ----------------------------------------------------------------- commit f8ae94d505f08de490f4ddf290a3ff28f2770892 Author: Steve Hay <[email protected]> Date: Sun Jan 25 13:07:36 2015 +0000 cherry-pick-votes: Drop the fixes for #122283 and #122890 from the list Only one voter (arc) came forward for the #122890 fix, and that was on the basis that only some of the series of commits involved be cherry-picked, with a note that they were not likely to apply cleanly either... This is all rather troubling, and hv agreed that it's too complex for maint: http://www.nntp.perl.org/group/perl.perl5.porters/2015/01/msg225074.html So #122890 must be dropped since it's controversial at best. Therefore, #122283 must also be dropped since it's also controversial without the fix for #122890, which #122283 caused: as hv noted, #122283 is merely suboptimal memory usage, whereas #122890 actually causes wrong results: http://www.nntp.perl.org/group/perl.perl5.porters/2015/01/msg225056.html arc also expressed a preference for "both fixes or neither fix" since releasing 5.20.2 with a bug that was known in 5.20.0 and 5.20.1 is less bad than releasing it with a different known bug. ----------------------------------------------------------------------- Summary of changes: Porting/cherry-pick-votes-maint-5.20.xml | 23 ----------------------- 1 file changed, 23 deletions(-) diff --git a/Porting/cherry-pick-votes-maint-5.20.xml b/Porting/cherry-pick-votes-maint-5.20.xml index 933193e..2e49b85 100644 --- a/Porting/cherry-pick-votes-maint-5.20.xml +++ b/Porting/cherry-pick-votes-maint-5.20.xml @@ -44,29 +44,6 @@ This fix has a problem with it that needs to be solved before we can backport it <div><h3>Other Regression Fixes</h3> -<commit votes="steveh, arc, khw" id="a51d618a82a7057c3aabb600a7a8691d27f44a34" ticket="122283" desc="rt 122283 - do not recurse into GOSUB/GOSTART when not SCF_DO_SUBSTR"/> -<commit votes="steveh, arc, khw" id="d9a72fccda5cacaccd2671769c010f3cda59ef8a" ticket="122283" desc="Add tests for a51d618a fix of RT #122283"/> -<commit votes="steveh, arc, khw" id="74f9f9edcc88e2102b767ccefb4c4100eab199bd" ticket="122283" desc="Another perldelta entry"/> - -a51d618a82a7057c3aabb600a7a8691d27f44a34 caused #122890 so do we need this sequence of commits too? -This is hard. #122283 is a genuine maint-worthy regression over 5.18.x, but even though the small, simple change in a51d618a82 fixed that case, it also caused the subsequent regression in #122890 â ... [958 chars truncated] -<commit votes="" id="abd6de04dbc807cb1270d565e5019376c04c77fd" ticket="122890" desc="regcomp.c: struct scan_frame: rename members for clarity"/> -<commit votes="arc" id="c9f0d54c3bad1139a1b0758ca0d999437ed93c95" ticket="122890" desc="regcomp.c: reuse temporary debugging temp sv's"/> -<commit votes="arc" id="3f910e62fcf6ddef8eaffcfa6554e5ff7df08a08" ticket="122890" desc="regcomp.c: eliminate memory leak from GOSUB/GOSTART regops"/> -<commit votes="" id="d615717366f1596d4a1b18ce2ca5a224dc076107" ticket="122890" desc="regcomp.c: study_chunk(): whitespace fixup for legibility"/> -<commit votes="" id="e60d552100fc966cb2917d32cf2e708d20a6427f" ticket="122890" desc="regcomp.c: study_chunk(): improve branch-branch comment"/> -<commit votes="" id="8b9781c905d8bc5e4fbf350df63e212283783324" ticket="122890" desc="regcomp.c: Improve re debug output by showing buffer names if they exist"/> -<commit votes="" id="0a14d816d4f94b3847ccd7cea08086e2244d1e6e" ticket="122890" desc="regcomp.c: Fixup debug output so it matches other messages in context"/> -<commit votes="" id="4420a417a2575a2656eb120d4b4048ac9dfbd46e" ticket="122890" desc="dump.c: Add PERL_PV_PRETTY_EXACTSIZE option"/> -<commit votes="" id="fdfb4f2197409a1d793ee71f18be6e702a5461df" ticket="122890" desc="regcomp.c: Improve advanced re debug output"/> -<commit votes="arc" id="64e04a51361b45dbbb94b038a93c7dc3cca0c7f9" ticket="122890" desc="regcomp.c: study_chunk(): Reset the entire data struct, not just a bit of it."/> -<commit votes="" id="32c35d38c96a76c6f8360ea6854ad3f3fb462725" ticket="122890" desc="regcomp.c: Add comment."/> -<commit votes="arc" id="4286711ac0be1bf8f37eadee54b2e9ebbb08e0d8" ticket="122890" desc="regcomp.c: study_chunk(): Avoid unnecessary recursion."/> -<commit votes="" id="92b05f28a98d1b7301afb8746d17237f1a669174" ticket="122890" desc="Add test names to t/re/regexp.t and friends"/> -<commit votes="" id="890fbec8b209c41a5902022a654cbd782f601bb7" ticket="122890" desc="t/re/re_tests: put comment in correct column"/> -<commit votes="arc" id="f6f22b1fc01b8676e43ef64616ed7899a84bfec2" ticket="122890" desc="t/re/re_tests: Add test for perl #122890"/> -<commit votes="arc" id="efcc8bfdc124f50d33aab3f1d7722b5f9e4de349" ticket="122890" desc="re_tests: More tests for #122890"/> - </div> <div><h3>Configure and hints Fixes (Consider whether these in particular would break binary compatibility)</h3> -- Perl5 Master Repository
