github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
01:45 AlexDani` joined 01:49 AlexDaniel left 01:50 AlexDani` is now known as AlexDaniel 03:11 reportable6 left 03:14 reportable6 joined 06:56 AlexDaniel left 06:57 AlexDani` joined 07:52 domidumont joined 11:42 dogbert2_ left 14:19 lucasb joined 15:43 dogbert11 left 16:06 AlexDani` is now known as AlexDaniel 16:58 domidumont left 17:45 dogbert17 joined 18:00 domidumont joined 18:38 lucasb left 19:20 robertle joined 19:36 lucasb joined 19:40 domidumont left 20:37 robertle left 21:22 domidumont joined 21:24 domidumont left
Geth MoarVM: 682ca82f5f | (Stefan Seifert)++ | src/core/nativecall.c
Fix nativecall refresh clearing used child object pointers

We compared the address of the struct member pointing to a child object with the address of the actual child object. Of course this cannot match, so we ended up clearing the child_objs slot all the time. This caused us to garbage collect e.g. a CArray object that was still pointed to by the CStruct.
21:35
MoarVM: ff9666bdba | (Stefan Seifert)++ | src/spesh/optimize.c
Fix predecessor pointers when merging basic blocks in spesh

merge_bbs was deactivated in commit d9a0dd8bf5c9d3ffa2518caf460135d3857d751f because it broke the preds. We now fix up those preds and have them point to the newly merged bb instead of the removed one. With that in place, we can re-enable merge_bbs
MoarVM: 682ca82f5f | (Stefan Seifert)++ | src/core/nativecall.c
Fix nativecall refresh clearing used child object pointers

We compared the address of the struct member pointing to a child object with the address of the actual child object. Of course this cannot match, so we ended up clearing the child_objs slot all the time. This caused us to garbage collect e.g. a CArray object that was still pointed to by the CStruct.
MoarVM: ff9666bdba | (Stefan Seifert)++ | src/spesh/optimize.c
Fix predecessor pointers when merging basic blocks in spesh

merge_bbs was deactivated in commit d9a0dd8bf5c9d3ffa2518caf460135d3857d751f because it broke the preds. We now fix up those preds and have them point to the newly merged bb instead of the removed one. With that in place, we can re-enable merge_bbs
nine Darn....actually didn't want to push that merge_bbs commit yet
lizmat And another Perl 6 Weekly hits the Net: p6weekly.wordpress.com/2019/02/25/...scalaring/ 21:43
23:08 lucasb left