IRC channel logs

2020-01-21.log

back to list of logs

<fossy>yo oriansj i just tried updating all the submodules in mescc-tools-seed and now blood-elf segfaults on M2.M1
<fossy>"Replace NOP with NULL because that is what is really ment by it" in M2-Planet broke the compiling of hex2 (Phase 7 of mescc-tools-seed)
<oriansj>fossy: probably need to hae cc_x86 and cc_amd64 export a NULL instead of a NOP
<oriansj>as cc_x86 and cc_amd64 have not been updated yet
<oriansj>trivial to fix, just change the strings
<fossy>ah.
<oriansj>(probably want to fix the NASM versions as well)
<oriansj>also update the x86_defs.M1 and amd64_defs.M1
<oriansj>actually it looks like cc_* don't need to support the exporting of NOP to NULL but the *.defs do need to be updated with NULL instead of NOP
<oriansj>also Target label FUNCTION_in_set is not valid is the error I get; which indicates it did not add in_set.c to it's source dependencies
<oriansj>I'll have the fix up shortly
<fossy>i think i fixed tat
<fossy>in my fork
<fossy>but my fork is un-ready
<oriansj>no worries, I'll fix things up a bit and have the patch up shortly
<fossy>FUNCTION_builtin_keywordp
<fossy>^ what'd that be?
<oriansj>fossy: it would be a FUNCTION named builtin_keywordp
<fossy>yeah, ofc. but would you know what might define that
<oriansj>fossy: it would be defined in a C source file, in that case specifically in mes_keyword.c
<fossy>ah.
<oriansj>in mes-m2
<fossy>danke
<fossy>yep
<oriansj>grep -R builtin_keywordp was the fastest way to find it
<oriansj>(after a git clean of course)
<fossy>ah
<oriansj>fastest way to lookup a function in the set is simply grep -R name and look for the one without a ;
<fossy>yep
<oriansj>also if you run M2-Planet without arguments it throws you into a REPL which evaluates when you ctrl-d
<oriansj>for example just type: int bar = 12; and then ctrl-d
<oriansj>(if you use --architecture, it'll customize the output to fix the architecture of your choosing)
<fossy>ah; nice
<oriansj>it is collectively designed to be used as a tool for learning; to not only bootstrap our infrastructure but also people's understanding of how it all works
<oriansj>fossy: patch is up, hopefully it solves the problem you had with updating the submodules
<fossy>:) thanks
<fossy>I hate trying to make things work on a wide variety of system
<oriansj>fossy: understandably, that is why M2-Planet and mescc-tools are designed to have consistent behavior across different architectures. So you only need your work to function on a single system and getting to work on other systems becomes a trivial job for anyone who wants to do the porting work.
***ng0_ is now known as ng0
<dddddd>hi
<dddddd>oh! Hi dzho (:
<dzho>lol, hi!
<dddddd>you'll find very(!) nice people here. Don't hesitate to ask or share whatever is on your mind.
<dddddd>welcome!
<dzho>dddddd: ok, thanks!
<dzho>I mostly idle but hopefully will remember to save this to my baseline set of channels I load up
<oriansj>mes-m2 now passes 8hours of fuzzing without a single segfault or hang
<oriansj>and tomorrow, I'll start fuzzing M2-Planet
<fossy>Im currently fuzzing kaem
<fossy>Been going for 30 minutes and 0/0