< hebasto> why bip155 appendix B does not specify that H() is actually SHA3_256 ?
< sipa> hebasto: i assume because it's copied literally from the torv3 spec
< sipa> (where H is defined up front, in a separate section)
< hebasto> is it worth to update bip as it is a design document for implementation code?
< sipa> hebasto: probably better to clarify that
< sipa> it's very unclear now
< hebasto> sipa: thanks! I'll submit a pr
< sipa> hebasto: great
< * luke-jr> glares at libsecp256k1 forcing -O3 even when -O1 is specified
< roconnor> luke-jr: https://gist.github.com/real-or-random/1548239059158370416dfdc6a866329a#gistcomment-3468827 is a patch to instrument GCC 9 (and I think GCC 10). It seems to imply that bitcoin and libsecp256k1 is unaffected by this bug.
< luke-jr> roconnor: the GCC 10 fix patch does generate different code - so you're saying that's just a side effect?
< roconnor> Right. As sipa noticed, the patch doesn't just fix the bug, it totally rewrites the implemenation.
< roconnor> from my reading there was an overall confusing between null-terminated strings and byte arrays that had been slowly accumulating over time, culminating into this final compiler bug form.
< roconnor> and the patch sweeps that all away and tries again.
< roconnor> At least that is my reading of it.
< luke-jr> roconnor: well that's good news
< luke-jr> still rebuilding everything just in case tho :P
< roconnor> I already did on my desktop
< roconnor> I'm now rebuilding everything again with the instrumented gcc 9.
< luke-jr> and logging warnings I assume :p
< sipa> luke-jr: i thought that was recently changed (not overwriting -O CFLAGS)
< luke-jr> sipa: yes, I was building 0.20.1 though
< jb55> rebuilding bitcoin/gcc and all transitive deps with the emit_diagnostic patch... hit a pubsub test in zmq and something in icu so far: https://jb55.com/s/31761d3ea0e88b4e.txt
< jb55> still going
< sanket1729> How can `decoderawtransaction` RPC know reqSigs if the output scriptpubkey is p2sh/p2wsh? I think it is returning `1` everytime.
< bitcoin-git> [bitcoin] decryp2kanon opened pull request #20026: [Trivial] python help message (master...patch-1) https://github.com/bitcoin/bitcoin/pull/20026
< sipa> sanket1729: it only works for bare multisig
< sipa> we should probably just remove it
< bitcoin-git> [bitcoin] sipa opened pull request #20027: Use mockable time everywhere in net_processing (master...202009_mockable_netproc) https://github.com/bitcoin/bitcoin/pull/20027