log☇︎
16000+ entries in 0.124s
bvt: however i don't understand if this is possible only to achieve only for gnat components, and when i do it globally, i get 'undefined symbol' when linking with libgcc
bvt: i still have no solution for this, afaik musl authors solved the problem for fortran and c++, but gnat seems to lack equivalent knob
bvt: i.e. with static linking, all locks are compiled into noops.
bvt: (still tested only with gnat2017, but this is a different story; i see no reason to believe that ave1gnat does not have the same issue)
bvt: http://btcbase.org/log/2019-02-12#1895598 << i'd like to briefly report on this: at this point i'm sure that there is a race condition. i'll do a post tomorrow morning on it, but short summary is ☝︎
diana_coman: asciilifeform, thank you! and yes, I know it's not enough to know what it does exactly but it's still useful.
mircea_popescu: well, i did what the article proposes, set my referrer to trilema permanently, so i always see trilema pics.
asciilifeform: http://btcbase.org/log/2019-02-16#1897225 << the former link eggogs, but i think this is by design, on acct of mircea_popescu's anti-hotlinker pill ☝︎
asciilifeform: ( i dun currently use, but it's definitely going into the final revs of ffa )
asciilifeform: ( to add insult to injury, erry time i post 1 of these, i gotta manually switch font to monospace and remove the winturd line endings, cuz the joys of closedsores proggy )
asciilifeform: ^ near as i can tell, it uses exactly same scheme as the old gnat's sjlj-mode handlers
diana_coman: I added a delay in there before the final check of tasks
diana_coman: i.e I want to see what did it do
diana_coman: asciilifeform, they abort; we are sawing them open because I don't trust it anymore re "what does it actually do exactly"
diana_coman: mircea_popescu yes + I must say I like the idea of "rip'em open after tea" - sounds great; but other than that: I'll try objdump too since after this gnat-adventure I'd rather have some way to look inside it
asciilifeform: mircea_popescu: i hesitate to addict folx to ida, it's a gnarly ball of proprietary liquishit, and needs a toilet box with 'wine' to run etc
mircea_popescu: asciilifeform i think she means to learn how it's done.
asciilifeform: diana_coman: if they dun abort and you want another detailed vivisect, feel free to post'em, i'ma rip'em open after tea
BingoBoingo: mircea_popescu: I'll have to dig, but it sounds very familiar
diana_coman: asciilifeform, how did you vivisect that pair of binaries? I'd rather do it to see what came out since otherwise I have only this "oh, now it DOES kill tasks!"
diana_coman: after all the struggle I'm still looking suspiciously at it, lol
asciilifeform: diana_coman: could plz tar up that built gnat and post ? i have a suspicion ( to get ljmp with old gnat, i had to set include paths on my box )
diana_coman: ave1, ah, thanks! I'll set it to false and try again then
ave1: I'm currenlty building with it set to False, but it is True in the distributed files
diana_coman: when I say "it built" above I mean specifically those versions: http://ave1.org/2018/building-gnat-on-musl-no-more-usrincludex86_64-linux-gnu/ and http://ave1.org/2018/building-gnat-on-musl-now-with-partial-and-parallel-build-support/ i.e. both static only and previous
diana_coman: rts-sjlj with its own adainclude/ and so on; so I am back to my previous obv that this flag per se is apparently not enough
diana_coman: update re builds: it built fine with --enable-sjlj-exceptions in place, checked it in the log and yes, it's set; but the result still seems to be build ada code with zcx in fact (i.e. my test code with tasks is STILL hung waiting for them to abort) and if I specify --rts=sjlj to gprbuild it complains that there is no native compiler for ada and so can't do anything; ftr I compared the dirs of my adacore install and it has this specific dir
diana_coman: ave1, it was the link that downloaded the gcc src and stuff for adacore2016 ; after that first fail I just copied my own stuff into tarballs/ , packed properly and never really bothered with the rest i.e. it builds all offline now anyway
ave1: O wait, I tried with: http://ave1.org/2018/building-gnat-on-musl-no-more-usrincludex86_64-linux-gnu/
a111: Logged on 2019-02-15 16:12 diana_coman: hm, full replica might require I upload somewhere the tarballs too since at least 1 link was broken ; anyways, it's with http://ave1.org/2018/building-gnat-on-musl-now-always-static/ + taken out the download script for ada2016 because broken link + added GCC_CONFFLAGS="--enable-sjlj-exceptions" to extraconfig.sh ; set path to point to existing and working adacore 2016 gnat + put all tarballs in their place
ave1: not yet, I'm now trying to get the build up and running
ave1: I'll park that and start running the gnat build script.
ave1: I was working on getting a cuntoo up (which is going slow, I need to relearn to build a kernel, this used to be a lot easier 20 years ago)
ave1: goodmorning, I was digesting te logs
asciilifeform: then they proceeded to wipe out alpha , then hp ( also quite reasonable 'pa-risc' 64-bit irons, i had one ) .
asciilifeform: but as i undertand, arm is permitted to 'just barely work', it Would Not Do (tm) for it to compete with intelism on heavy irons , somehow
asciilifeform: mircea_popescu: i would not go so far as to say 'entirely useless', 2ce the bitness gets you 2ce the rsa speed for same clock.
asciilifeform: i found ~one~, chinese, for ~more~ than the cost of serious opteron with complete filling.
asciilifeform: mircea_popescu: thread was re mid-2000s lappies. ( i have moar recent ones, which have e.g. 32GB in'em etc )
a111: Logged on 2019-02-15 20:02 diana_coman: darn, ofc there are TWO flags to set to the same thing i.e. one for bootstrapper and one for the resulting gnat
asciilifeform: i.e. the latter is pre-alpha quality
mircea_popescu: asciilifeform i suspect this is my problem. i don't mentally distinguish aarch64 from arm correctly.
mircea_popescu: but i mean, eg ida got support for it cca v6 ?
asciilifeform: mircea_popescu: sjlj ? was 'deprecated' Officially, Only A Terrorist(tm) Would Want, was the flavor i got from the http://btcbase.org/log/2019-02-15#1897058 rotters ☝︎
mircea_popescu: i can't imagine it was never written, this long jump thing was standardf for a while neh ?
asciilifeform: mircea_popescu: i sank several hours into the dig and did not find whether 'worked, then broken' or 'never written'
a111: Logged on 2019-02-15 18:36 asciilifeform: this, note, is ~still~ a bug ( or , worse, aarch64 dun know how to longjmp-task ? i.e. not implemented ? ) and will have to be cured. but as i undestand diana_coman is currently interested in x64.
mircea_popescu: http://btcbase.org/log/2019-02-15#1897029 << myeah, eulora suddenly very fucking useful to teh broader community. i have a good mind to add a few months payroll to the goodwill line, because hory shit. ☝︎
diana_coman: I've started it now with both flags, let's see
diana_coman: aha; I remembered the pain from first time when yes, whole night
diana_coman: that yes, I did and it is using
diana_coman: full you mean? I didn't time it, more usual left it and came back to it done, but from scratch ~1hour , possibly a bit more than that (less than 2 though)
diana_coman: but I suspect this is not the end of it because the bootstrap will want additional stuff that it probably doesn't yet have set to have etc
diana_coman: mk, so I'll add then: GCC_BOOTSTRAP_CONFFLAGS="--enable-sjlj-exceptions"
asciilifeform: as i guessed , ha
diana_coman: darn, ofc there are TWO flags to set to the same thing i.e. one for bootstrapper and one for the resulting gnat ☟︎
diana_coman: that's the switch I gave it; let me check in the log too to make sure I didn't dream I actually saw it in there (but for that matter if it didn't, it wouldn't have barfed)
diana_coman: oh hm, on the + it seems at least so far the trouble is indeed on arm only; on the big minus, it seems though that the switch I used is not enough in fact i.e. the result still does not have sjlj rts; so I still need to see what exactly it wants to actually build the sjlj rts too
asciilifeform: i.e. maintainer loudly an' abundantly shat from mouth when asked 'wtf'
BingoBoingo: <PeterL> bingoboingo: when do you plan to do another wff auction? Will you do chunks of 500 wff again? << Next month, I will start at 500 WFF again
diana_coman: and yes, I am currently mainly interested in x64
asciilifeform: this, note, is ~still~ a bug ( or , worse, aarch64 dun know how to longjmp-task ? i.e. not implemented ? ) and will have to be cured. but as i undestand diana_coman is currently interested in x64. ☟︎
asciilifeform: ( 'move_insn' seems to be the peephole optimizer on back end. i gotta wonder if it ended up being fed a x64 .o by the linker, and choked on ~that~ )
asciilifeform: diana_coman: were you attempting to build ~x64~ gnat on this ? ftr i was never able to ( with classic zxc variant, this was long before the current thread )
asciilifeform: diana_coman: i notice you were building 'aarch64' ? this was on rk ?
diana_coman: uhm, and ave1's blog thinks I'm a spammer and won't let me comment...
a111: Logged on 2019-02-15 16:13 diana_coman: anyways, I'll try presently ave1's previous version i.e. http://ave1.org/2018/building-gnat-on-musl-now-with-partial-and-parallel-build-support/
asciilifeform: http://btcbase.org/log/2019-02-15#1896996 << if there exists somewhere a 800+kLoc proggy that ~aint~ a tower o'shit, i have yet to see or even hear of it ☝︎
a111: Logged on 2019-02-15 16:43 mircea_popescu: asciilifeform dood runs windows, what can i tell you.
asciilifeform: http://btcbase.org/log/2019-02-15#1896982 << the... 'smarter' ( i hesitate to use this word, perhaps 'less retarded..' ? ) winblows folx, long ago threw out av.. ☝︎
a111: Logged on 2019-02-15 16:07 diana_coman: but I start suspecting the --enable-sjlj-exceptions + static only -> barf
mircea_popescu: asciilifeform dood runs windows, what can i tell you. ☟︎
diana_coman: for sanity I just checked by the way that the exact same machine DOES in fact build successfully GNAT with those same scripts without the sjlj added requirement
diana_coman: anyways, I'll try presently ave1's previous version i.e. http://ave1.org/2018/building-gnat-on-musl-now-with-partial-and-parallel-build-support/ ☟︎
diana_coman: hm, full replica might require I upload somewhere the tarballs too since at least 1 link was broken ; anyways, it's with http://ave1.org/2018/building-gnat-on-musl-now-always-static/ + taken out the download script for ada2016 because broken link + added GCC_CONFFLAGS="--enable-sjlj-exceptions" to extraconfig.sh ; set path to point to existing and working adacore 2016 gnat + put all tarballs in their place ☟︎
asciilifeform: diana_coman: plox to briefly describe how you set up this build, so i can replicate locally
diana_coman: but I start suspecting the --enable-sjlj-exceptions + static only -> barf ☟︎
asciilifeform: i didn't even see any mention of 'crypto', other than a 2017 piece where he gripes about uninvited js minerism
mircea_popescu: (for thread completness : dood's a "crypto skeptic" for the simple reason that i told 'em to get in back in 2011, when ro speakers were going through a phase of our-homegoat-just-as-good-as-mp's-mercedes, and so of course they "knew better" then. romanians are dumb enough to get locked into these permamently, just like any other rural morons.)
mircea_popescu: as a smart woman once said, "to have fun you need to organize it". i know, cuz i wrote her.
mircea_popescu: first "global warming", that didn't pan out, then "climate change", that also didn't pan out, now it's "nothing in particular" and i'm supposed to take them seriously ?
diana_coman: fwiw yesterday I had in parallel this recompile task going too, but all sorts of stuff going with it so it's still ongoing (and apparently I'll have afterwards to write-it up as well, on top of the sjlj data )
mircea_popescu: i guess so.
diana_coman: http://btcbase.org/log/2019-02-15#1896874 -> pending getting finally ave1's gnat recompiled with sjlj and then running it on smg test server, I'd say switching to sjlj + no handlers is pretty much the only currently available option anyway ☝︎
mircea_popescu: i select, clicked footnote, select again, it works throughout ?
a111: Logged on 2019-01-18 17:49 asciilifeform: http://btcbase.org/log/2019-01-18#1888188 << oh hey , was this same grandfather as 'i miss stalin, we were younger then' ?
asciilifeform: i would almost say 'i had moar fun with that thing than ever did with pc', but this is prolly http://btcbase.org/log/2019-01-18#1888232 -effect. ☝︎
asciilifeform was given 'mk61' at same age i think diana_coman's kid is nao ☟︎
diana_coman: at any rate, I think it would make a far better item to send kids to investigate than many "projects"
asciilifeform: i.e. exploring undefined instructions
asciilifeform: but requires a bit deeper cut than 'i'ma run $snippet and time'
asciilifeform: this is why most 'i'ma do it in asm!' folx by nao have drunk to death.
asciilifeform: they also dun do much good! if i had a wood furnace, i'dve stoked it with'em by nao
asciilifeform: i.e. when introduced reorderism & pipes.
mircea_popescu: in no case do i know of anyone who has actual data re such things as "ok, so manual claims, but NUMBERS for this penalty"
asciilifeform: for all i know, not even a computable function, lol
asciilifeform would kill for an accurate tick table for opteron. but none exists ( possibly when i have msdos gnat, can bake one..)
mircea_popescu: kinda what i understood, that cmp USED TO BE expensive, but is no longer.
asciilifeform: i was answering q of 'why does gcc put out this odd form for a computed goto'.
asciilifeform: mircea_popescu: do i misread the http://btcbase.org/log/2019-02-14#1896636 run ? ☝︎