timo bugs.debian.org/cgi-bin/bugreport....ug=1059502 - we should probably move some stuff out of the ldlibs config var if we have ecosystem modules just taking those flags to get "a working compiler commandline" like NativeLibs does 00:11
actually, taking variables that were used for building moarvm in order to build modules on the system moar is installed on is Very Wrong in the first place 00:12
i'm sure there is plenty of prior art to learn from / be horrified by 00:18
00:37 MasterDuke joined
MasterDuke timo: how do i get your 3rdparty/dynasm update? 00:42
a configure gives `Updating submodules .................................... fatal: remote error: upload-pack: not our ref 847ea5d910c706cc7b69bffe0a5f0567fd2b15bd` 00:43
timo oh? does `git submodule update` help? 00:47
i think that's what Configure tries to do
huh the github actions also failed 00:49
wow, i just totally flubbed that huh 00:50
try again please
MasterDuke good now 01:06
Geth MoarVM/reprotest_tile_compiler: cd78d3ee4c | (Timo Paulssen)++ | build/Makefile.in
Try running tiler table generator without hash randomization
01:21
MoarVM/reprotest_tile_compiler: 2a6f79dad3 | (Timo Paulssen)++ | build/Makefile.in
Try building tile table without hash randomization

To ensure that we get identical outputs, make a .one and a .two file and git diff --no-index them so if anything were different, it shows up in the build log.
MoarVM: timo++ created pull request #1835:
Reprotest tile compiler
01:26
timo dev.azure.com/MoarVM/MoarVM/_build...&l=788 oh wow oh crap 01:31
can we have something that immediately flags moarvm panics in any of the outputs from builds as something to look closer at? 01:32
MasterDuke well, ideally our CI is green, and those would turn it red because they'll probably cause a test to fail... 01:37
it was green for a little while, but the rakudo jvm build is broken again
07:58 sena_kun joined 08:09 sena_kun left 13:58 sena_kun joined 22:03 sena_kun left