< promag>
off topic, but I see lots of different conventions for commit messages, branch names and PR titles. should there be such conventions in developer notes for instance?
< sipa>
cfields: works fine here
< cfields>
:(
< sipa>
cfields: in other news, your compiler is retarded
< sipa>
0000000000000011movq%rsi, -0x38(%rbp)
< sipa>
000000000000001dmovq-0x38(%rbp), %rsi
< cfields>
erm, that just moves and moves back?
< sipa>
yes
< cfields>
heh
< cfields>
fulfilling a constraint, maybe?
< sipa>
nope
< sipa>
it could leave everything in the original order even
< cfields>
sipa: weird. based on my dinky little test app, it seems to be working
< sipa>
cfields: eh
< cfields>
sipa: not sure yet. confused.
< cfields>
i wrote a little sha256sum clone that just reads stdin and spits out a hash. matches sha256sum so far.
< sipa>
cfields: does it use the sse code?
< cfields>
Assertion failed: (0), function Transform, file crypto/sha256_sse42.cpp, line 16.
< cfields>
yep. just threw ^^ assert in to verify.
< bitcoin-git>
[bitcoin] gmaxwell opened pull request #10831: Batch flushing operations to the walletdb during top up and increase keypool size. (master...topup_batch_flush) https://github.com/bitcoin/bitcoin/pull/10831
< sipa>
BlueMatt: i am confused by the change in #10235
< BlueMatt>
;;later tell promag problem with github tags is you cant have regular folks adding tags (ie you cant limit tag-edits to people who own the issue/pr)
< cfields>
fanquake: ok, so i'm not alone. great. also, shit.
< cfields>
fanquake: thanks for testing!
< fanquake>
cfields no worries. Ping me if you want anything looked at. Should be around for the rest of the afternoon.
< sipa>
cfields: how did you make clang piss off in linux?
< sipa>
can you create and a send me a static binary?
< cfields>
sipa: still fiddling with it. I'm having troubles with boost as well.
< sipa>
cfields: let's have a look at it in tokyo
< sipa>
i can try a static register allocation first too, and see if that works
< sipa>
why am i talking here? i'm getting drunk
< cfields>
haha. are you irc'ing at a bar?
< sipa>
yes
< cfields>
sipa: using clang with -fintegrated-as caused a new crash in linux
< cfields>
my assumption is that it uses gnu as on linux, and llvm's as for apple
< cfields>
but i'll come up with something solid and ping you then. forget about assembly for the night :)
< cfields>
heh, confirmed
< cfields>
unfortunately, it crashes gdb too :(
< bitcoin-git>
[bitcoin] laanwj opened pull request #10832: init: Factor out AppInitLockDataDirectory (master...2017_07_appinitlockdatadirectory) https://github.com/bitcoin/bitcoin/pull/10832
< bitcoin-git>
[bitcoin] laanwj closed pull request #10818: init: Keep track of whether data directory locked, don't cleanup if not (master...2017_05_locked_datadir) https://github.com/bitcoin/bitcoin/pull/10818
< wumpus>
might be some calling convention difference
< wumpus>
or at least platform ABI
< wumpus>
gdb is not very robust against such things (accidental or to step reverse-engineering) gets easily confused if certain registers are not set up as they should
< fanquake>
what is with this spam
< fanquake>
wumpus Any chance you've tested Boost 1.64.0 on openbsd?
< morcos>
wumpus: I assume you knew if you give multiple named arguments for the same argument it's the last one that matters?
< morcos>
I'm not sure how much of an issue that is, but seems like it could cause problems with wallet
< jamesob>
If I wanted to do some kind of analysis on the UTXO set, is my best bet forking and making use of CCoinsView somehow? is https://github.com/bitcoin/bitcoin/pull/7759 worth resurrecting?
< bitcoin-git>
[bitcoin] paveljanik opened pull request #10835: Do not shadow a field of 'reverse_range<T>' (master...20170715_Wshadow_reverseiterator) https://github.com/bitcoin/bitcoin/pull/10835