2016-10-25 14:01:40 +0000ircbrowse(~chrisdone@unaffiliated/chrisdone)
2016-10-25 14:03:59 +0000 <ircbrowse> :-)
2016-10-25 14:17:38 +0000osa1_osa1
2016-10-25 14:17:44 +0000osa1(~omer@156-56-64-180.dhcp-bl.indiana.edu) (Changing host)
2016-10-25 14:17:44 +0000osa1(~omer@haskell/developer/osa1)
2016-10-25 14:19:31 +0000cur8or(~cur8or@72canterbury.cybersmart.co.za) (Ping timeout: 244 seconds)
2016-10-25 14:19:33 +0000quchen(~quchen@ipb21b54ea.dynamic.kabel-deutschland.de) (Remote host closed the connection)
2016-10-25 14:22:58 +0000newhoggy(~newhoggy@ppp121-45-184-142.bras1.syd2.internode.on.net)
2016-10-25 14:23:29 +0000 <siarheit> bgamari: i do, i do
2016-10-25 14:23:42 +0000 <bgamari> siarheit, you do
2016-10-25 14:23:43 +0000 <bgamari> ?
2016-10-25 14:23:47 +0000 <siarheit> https://ghc.haskell.org/trac/ghc/ticket/12585
2016-10-25 14:24:00 +0000 <siarheit> (WRT #12757 failure mechanics)
2016-10-25 14:24:01 +0000 <phaskell> #12757: Compiled program segfaults at -O1 - https://ghc.haskell.org/trac/ghc/ticket/12757
2016-10-25 14:24:11 +0000 <bgamari> ahh
2016-10-25 14:25:18 +0000 <siarheit> basically when bytestring package checks it you are finished traversing bytestring it checks current pointer for '"literal#" + length("literal"#'
2016-10-25 14:25:32 +0000 <bgamari> oh no!
2016-10-25 14:25:36 +0000 <bgamari> I see
2016-10-25 14:25:39 +0000 <siarheit> but it inlines that literal and creates it in new place
2016-10-25 14:25:46 +0000 <bgamari> oh dear
2016-10-25 14:25:54 +0000 <siarheit> and end points to a new literal (and who knows what lies in between)
2016-10-25 14:25:58 +0000 <bgamari> right
2016-10-25 14:26:19 +0000 <bgamari> siarheit, thanks!
2016-10-25 14:26:49 +0000 <siarheit> very scary stuff
2016-10-25 14:28:09 +0000newhoggy(~newhoggy@ppp121-45-184-142.bras1.syd2.internode.on.net) (Ping timeout: 256 seconds)
2016-10-25 14:46:46 +0000 <rwbarton> I thought siarheit reported this about a week ago
2016-10-25 14:47:27 +0000 <rwbarton> oh yeah, #12585
2016-10-25 14:47:27 +0000 <phaskell> #12585: GHC duplicates string literals in rodata section and breaks 'Ptr Addr#' equality - https://ghc.haskell.org/trac/ghc/ticket/12585
2016-10-25 14:48:15 +0000johnw(~johnw@haskell/developer/johnw) (Quit: ZNC - http://znc.in)
2016-10-25 14:48:31 +0000amyers(~quassel@128.177.89.240.available.zip.zayo.com) (Ping timeout: 276 seconds)
2016-10-25 14:50:49 +0000 <rwbarton> siarheit: btw, did you ever try bisecting to find the commit that introduced this issue?
2016-10-25 14:51:02 +0000amyers(~quassel@128.177.89.240.available.zip.zayo.com)
2016-10-25 14:53:16 +0000 <siarheit> rwbarton: i didn't but i have a suspiction it's 498009a904a1e8910f9e0e527f6eb6c8073c8a76
2016-10-25 14:53:32 +0000 <siarheit> ( master's a47b62cb36853d03c77ef63b3208b3d869fb687e )
2016-10-25 14:55:26 +0000 <rwbarton> I see