12020-10-10T00:00:02 *** wright has quit IRC
22020-10-10T00:00:22 *** DeanGuss has quit IRC
32020-10-10T00:05:39 *** DeanGuss has joined #bitcoin-core-dev
42020-10-10T00:22:31 *** Jordack has joined #bitcoin-core-dev
52020-10-10T00:35:35 *** davterra has quit IRC
62020-10-10T00:35:55 *** davterra has joined #bitcoin-core-dev
72020-10-10T00:36:09 *** meshcollider has quit IRC
82020-10-10T00:36:29 *** meshcollider has joined #bitcoin-core-dev
92020-10-10T00:58:47 *** bitcoin-git has joined #bitcoin-core-dev
102020-10-10T00:58:47 <bitcoin-git> [bitcoin] leozz37 opened pull request #20116: Implemented GitHub Actions workflow for MacOS and Ubuntu (master...github-actions) https://github.com/bitcoin/bitcoin/pull/20116
112020-10-10T00:58:48 *** bitcoin-git has left #bitcoin-core-dev
122020-10-10T01:02:11 <yanmaani> I'm getting weird errors while building Bitcoin Core, but everything seems to work fine
132020-10-10T01:02:14 <yanmaani> > /usr/bin/ld: error: leveldb/libleveldb.a(libleveldb_a-crc32c.o): <corrupt x86 feature size: 0x8>
142020-10-10T01:02:27 <yanmaani> for what seems to be every .o file
152020-10-10T01:02:35 <yanmaani> didn't find anything about it on the Internet. Is this a known issue?
162020-10-10T01:14:10 *** bitcoin-git has joined #bitcoin-core-dev
172020-10-10T01:14:10 <bitcoin-git> [bitcoin] fanquake closed pull request #20116: Implemented GitHub Actions workflow for MacOS and Ubuntu (master...github-actions) https://github.com/bitcoin/bitcoin/pull/20116
182020-10-10T01:14:11 *** bitcoin-git has left #bitcoin-core-dev
192020-10-10T01:17:42 <sipa> vasild: i was wrong; when making an outbound connection, we overwrite the addrman services with the peer's reported ones
202020-10-10T01:17:53 <sipa> see CAddrMan::SetServices
212020-10-10T01:27:18 *** gloriazhao has joined #bitcoin-core-dev
222020-10-10T01:28:17 *** gloriazhao is now known as glozow
232020-10-10T01:30:28 *** pinheadmz has joined #bitcoin-core-dev
242020-10-10T01:32:49 *** pinheadmz has quit IRC
252020-10-10T01:37:23 *** tralfaz has joined #bitcoin-core-dev
262020-10-10T01:37:55 *** davterra has quit IRC
272020-10-10T01:37:59 *** tralfaz is now known as davterra
282020-10-10T02:03:55 *** kristapsk has quit IRC
292020-10-10T02:04:08 *** kristapsk has joined #bitcoin-core-dev
302020-10-10T02:23:41 <Murch> jonatack: okay, put it on my list
312020-10-10T02:25:09 <fanquake> Murch: I just put something on your list
322020-10-10T02:31:37 <Murch> sorry, I meant that I had put it on my list. I realize now that the conjugations of this particular irregular word make my brevity particularly misleading
332020-10-10T02:32:28 <sipa> lalalala http://ncf.idallen.com/english.html
342020-10-10T02:33:01 <Murch> Yes, let's all learn Esperanto or Klingon
352020-10-10T02:33:42 <Murch> Although, this had nothing to do with homophones or irregular pronunciation :p
362020-10-10T02:34:47 <sipa> put and put are arguably homophones
372020-10-10T02:35:22 <Murch> more importantly for my communication mishap, they're homonyms
382020-10-10T02:35:46 <fanquake> Murch: I wasn't trying to correct your spelling hah, I meant I'd just tagged you in #20040.
392020-10-10T02:35:48 <gribble> https://github.com/bitcoin/bitcoin/issues/20040 | wallet: Refactor OutputGroups to handle fees and spending eligibility on grouping by achow101 · Pull Request #20040 · bitcoin/bitcoin · GitHub
402020-10-10T02:36:11 <Murch> fanquake: I thank ye kindly!
412020-10-10T02:36:32 <achow101> moar refactor
422020-10-10T02:42:26 <fanquake> I've passed on more feedback to GitHub, now that we're seeing even more comment related issues: https://0bin.net/paste/9HgV-Dyl#sZlLldWLuSFQGiT1-yYS5ikMyCIIkRLd5lK3fuAD/ql & https://imgur.com/a/mUVsDSX.
432020-10-10T02:56:32 *** mdunnio has joined #bitcoin-core-dev
442020-10-10T03:00:02 *** Jordack has quit IRC
452020-10-10T03:00:45 *** mdunnio has quit IRC
462020-10-10T03:01:01 *** S3RK has joined #bitcoin-core-dev
472020-10-10T03:05:26 *** S3RK has quit IRC
482020-10-10T03:21:56 *** paxed1 has joined #bitcoin-core-dev
492020-10-10T03:40:45 *** harrigan has quit IRC
502020-10-10T03:42:53 *** harrigan has joined #bitcoin-core-dev
512020-10-10T03:51:16 *** yanmaani has quit IRC
522020-10-10T03:51:53 *** yanmaani has joined #bitcoin-core-dev
532020-10-10T04:07:03 *** kristapsk has quit IRC
542020-10-10T04:10:03 *** k3tan has quit IRC
552020-10-10T04:12:39 *** k3tan has joined #bitcoin-core-dev
562020-10-10T04:49:37 <tryphe> sipa, such a great piece
572020-10-10T04:49:55 *** S3RK has joined #bitcoin-core-dev
582020-10-10T04:54:49 *** S3RK has quit IRC
592020-10-10T05:05:03 *** k3tan has quit IRC
602020-10-10T05:06:58 <tryphe> or how about white night, wight knight
612020-10-10T05:08:50 <tryphe> or white knight and wight night :D
622020-10-10T05:08:57 *** AaronvanW has quit IRC
632020-10-10T05:10:11 *** k3tan has joined #bitcoin-core-dev
642020-10-10T05:13:57 *** neo_1 has joined #bitcoin-core-dev
652020-10-10T05:17:28 *** paxed1 has quit IRC
662020-10-10T05:19:21 *** neo_1 has quit IRC
672020-10-10T05:19:46 *** neo_1 has joined #bitcoin-core-dev
682020-10-10T05:31:51 *** OldMiner has joined #bitcoin-core-dev
692020-10-10T05:48:57 *** AaronvanW has joined #bitcoin-core-dev
702020-10-10T05:53:47 *** AaronvanW has quit IRC
712020-10-10T06:00:02 *** OldMiner has quit IRC
722020-10-10T06:13:06 *** andreacab has joined #bitcoin-core-dev
732020-10-10T06:17:23 *** S3RK has joined #bitcoin-core-dev
742020-10-10T06:21:11 *** ramsey1 has joined #bitcoin-core-dev
752020-10-10T06:22:26 *** AaronvanW has joined #bitcoin-core-dev
762020-10-10T06:29:31 *** Pavlenex has joined #bitcoin-core-dev
772020-10-10T06:29:47 *** S3RK has quit IRC
782020-10-10T06:39:11 *** andreacab has quit IRC
792020-10-10T06:39:40 *** andreacab has joined #bitcoin-core-dev
802020-10-10T06:39:45 *** proofofkeags_ has quit IRC
812020-10-10T06:39:45 *** proofofkeags has quit IRC
822020-10-10T06:44:12 *** andreacab has quit IRC
832020-10-10T06:48:26 *** b10c_ has quit IRC
842020-10-10T06:48:47 *** b10c has joined #bitcoin-core-dev
852020-10-10T06:55:54 *** AaronvanW has quit IRC
862020-10-10T07:16:32 <wumpus> sipa: no, not really since last night, it seems to be a curious case of miscompilation, I still have a broken binary but can't reproduce it anymore when building even exactly the same commit. Things have been working since.
872020-10-10T07:18:21 <wumpus> my only guess is a hardware problem or a weird bug with FreeBSD's clang 8-derived compiler. The binaries are sufficiently different too, it's not like there's a few bytes different so it's easy to compare.
882020-10-10T07:19:23 <sipa> okay
892020-10-10T07:19:28 <sipa> i couldn't reproduce it either
902020-10-10T07:19:37 <sipa> but i don't know the exact circumstancs
912020-10-10T07:20:10 <sipa> if you can't even reproduce it with the same commit (and same peers.dat?), i guess we'll need to assume it was a very local problem
922020-10-10T07:20:28 <wumpus> in any csae, if anyone still wants to figure it out, I two binaries built from the same commit, one with the issue one without
932020-10-10T07:20:53 <sipa> the binaries differ?
942020-10-10T07:21:09 <wumpus> yes, very significantly
952020-10-10T07:21:28 <wumpus> stripped size is the same but everything is in a different place
962020-10-10T07:21:46 <sipa> different compiler/compileflags?
972020-10-10T07:22:09 *** luke-jr has quit IRC
982020-10-10T07:22:28 <wumpus> no, not intentially at least
992020-10-10T07:23:28 <wumpus> might have been some cruft in the build dir, who knows... the curious thing is that it happened with 19954 and not without though. I still can't explain that.
1002020-10-10T07:23:39 *** luke-jr has joined #bitcoin-core-dev
1012020-10-10T07:24:21 <wumpus> but I don't think it's anything to worry about for that PR
1022020-10-10T07:27:42 <wumpus> (no, peers.dat nor anything in the data directory is at fault, it only depends on what binary is run)
1032020-10-10T07:30:47 <sipa> ok
1042020-10-10T07:31:05 *** Pavlenex has quit IRC
1052020-10-10T07:35:17 <wumpus> I guess it will always remain a mystery, running that particular binary still has the same issue: every outgoing peer is disconnected quickly (usually after sending feefilter), no reason logged in debug.log
1062020-10-10T07:39:11 <wumpus> (correction: *always after "sending feefilter (8 bytes) peer=X" there's "disconnecting peer=X", sometimes with "Cleared nodestate for peer=X" in between*)
1072020-10-10T07:39:21 <sipa> anything interesting if you run in valgrind?
1082020-10-10T07:41:51 *** Pavlenex has joined #bitcoin-core-dev
1092020-10-10T07:41:51 <wumpus> havne't tried that yet
1102020-10-10T07:45:05 <wumpus> it could be thinking that every outgong connection is a feeler connection, it's the only fDisconnect I could find without associated NET logging so quickly
1112020-10-10T07:46:51 <wumpus> I wonder what level of debug information FreeBSSD toolchain adds by default
1122020-10-10T07:49:19 <wumpus> given that there's 95MB of size difference between the stripped binary and the plain one, quite a lot would be my guess
1132020-10-10T07:49:39 <wumpus> could try setting some breakpoints
1142020-10-10T08:26:44 *** Processus42 has quit IRC
1152020-10-10T08:26:45 *** _0x0ff has quit IRC
1162020-10-10T08:26:57 *** Processus42 has joined #bitcoin-core-dev
1172020-10-10T08:27:05 *** _0x0ff has joined #bitcoin-core-dev
1182020-10-10T08:45:04 *** tralfaz has joined #bitcoin-core-dev
1192020-10-10T08:45:22 *** davterra has quit IRC
1202020-10-10T08:45:26 *** tralfaz is now known as davterra
1212020-10-10T08:49:25 *** jonatack has quit IRC
1222020-10-10T08:49:31 *** S3RK has joined #bitcoin-core-dev
1232020-10-10T08:53:04 *** AaronvanW has joined #bitcoin-core-dev
1242020-10-10T08:54:03 *** S3RK has quit IRC
1252020-10-10T08:58:23 <wumpus> sipa: ... okay ... after finally figuring out how to set a hardware memory breakpoint on an atomic value (to see when fDisconnect changes), I discovered something shocking: two parts of the program have different CNode layouts
1262020-10-10T08:58:53 <wumpus> sipa: there's an off-by-one where pfrom.fSuccessfullyConnected = true; happens to set pfrom.fDisconnected = true ...
1272020-10-10T09:00:01 *** ramsey1 has quit IRC
1282020-10-10T09:00:47 <wumpus> sipa: so it's clear, a miscompile! I think a stale .o file somehow as linked in, I have --disable-dependency-tracing as it wouldn't build on FreeBSD otherwise
1292020-10-10T09:04:49 *** kexkey has quit IRC
1302020-10-10T09:07:33 <wumpus> vasild: phew, we can ignore the issue, https://github.com/bitcoin/bitcoin/pull/19954#issuecomment-705504702
1312020-10-10T09:14:10 <wumpus> anyhow happy I got to the bottom of this, a "silent link issue" must be one of the most evil things I've ever debugged
1322020-10-10T09:15:32 <hebasto> wumpus: great to know it
1332020-10-10T09:17:04 *** jb55 has quit IRC
1342020-10-10T09:17:32 *** jb55 has joined #bitcoin-core-dev
1352020-10-10T09:21:33 *** MartinAS1 has joined #bitcoin-core-dev
1362020-10-10T09:27:01 *** AaronvanW has quit IRC
1372020-10-10T09:27:45 *** S3RK has joined #bitcoin-core-dev
1382020-10-10T09:29:39 *** jonatack has joined #bitcoin-core-dev
1392020-10-10T09:34:45 *** S3RK has quit IRC
1402020-10-10T09:51:55 *** midnight has quit IRC
1412020-10-10T09:54:55 *** midnight has joined #bitcoin-core-dev
1422020-10-10T09:54:58 *** jonatack has quit IRC
1432020-10-10T09:56:58 *** jonatack has joined #bitcoin-core-dev
1442020-10-10T09:58:45 *** neo_1 has quit IRC
1452020-10-10T09:59:17 *** justanotheruser has quit IRC
1462020-10-10T10:05:38 *** reallll has joined #bitcoin-core-dev
1472020-10-10T10:08:59 *** belcher_ has quit IRC
1482020-10-10T10:09:31 *** mdunnio has joined #bitcoin-core-dev
1492020-10-10T10:10:23 *** vasild has quit IRC
1502020-10-10T10:12:25 *** vasild has joined #bitcoin-core-dev
1512020-10-10T10:13:45 *** mdunnio has quit IRC
1522020-10-10T10:14:09 *** shesek has quit IRC
1532020-10-10T10:18:30 *** Retta31Goodwin has joined #bitcoin-core-dev
1542020-10-10T10:26:58 *** vasild has quit IRC
1552020-10-10T10:27:08 *** vasild has joined #bitcoin-core-dev
1562020-10-10T10:32:28 *** jonatack has quit IRC
1572020-10-10T10:36:51 *** neo_1 has joined #bitcoin-core-dev
1582020-10-10T10:55:46 *** kljasdfvv has quit IRC
1592020-10-10T10:58:18 *** AaronvanW has joined #bitcoin-core-dev
1602020-10-10T11:06:45 *** AaronvanW has quit IRC
1612020-10-10T11:13:01 *** kljasdfvv has joined #bitcoin-core-dev
1622020-10-10T11:31:58 *** kljasdfvv has quit IRC
1632020-10-10T11:34:07 *** Pavlenex1 has joined #bitcoin-core-dev
1642020-10-10T11:35:30 *** Pavlenex has quit IRC
1652020-10-10T11:35:30 *** Pavlenex1 is now known as Pavlenex
1662020-10-10T11:36:10 *** Pavlenex1 has joined #bitcoin-core-dev
1672020-10-10T11:40:14 *** Pavlenex has quit IRC
1682020-10-10T11:41:56 <vasild> sipa: right, so we overwrite the service flags when we outbound-connect to a peer from the received version message, and via gossip we | the flags together if we already have an entry for that addr:port
1692020-10-10T11:42:24 <vasild> wumpus: ccache?
1702020-10-10T11:42:33 *** Pavlenex has joined #bitcoin-core-dev
1712020-10-10T11:43:06 *** reallll is now known as belcher
1722020-10-10T11:43:16 *** Pavlenex1 has quit IRC
1732020-10-10T11:44:12 <wumpus> vasild: no, lack of dependency tracking inside automake
1742020-10-10T11:45:18 <vasild> this is what I wrote to hebasto just a few days ago when discussing #18750: "long time ago I ran into some obscure compilation issues and after long debugging I figured out that it is due to ccache and I have never been using it after that!"
1752020-10-10T11:45:21 <gribble> https://github.com/bitcoin/bitcoin/issues/18750 | build: optionally skip external warnings by vasild · Pull Request #18750 · bitcoin/bitcoin · GitHub
1762020-10-10T11:45:53 *** shesek has joined #bitcoin-core-dev
1772020-10-10T11:46:30 *** Pavlenex1 has joined #bitcoin-core-dev
1782020-10-10T11:46:44 <vasild> wumpus: "I have --disable-dependency-tracing as it wouldn't build on FreeBSD otherwise" -- hmm, I don't use that option and it compiles just fine
1792020-10-10T11:48:48 <vasild> wen cmake?
1802020-10-10T11:49:05 <vasild> :-D
1812020-10-10T11:49:31 *** Pavlenex has quit IRC
1822020-10-10T11:49:31 *** Pavlenex1 is now known as Pavlenex
1832020-10-10T11:54:34 *** Retta31Goodwin has quit IRC
1842020-10-10T11:57:12 *** Pavlenex has quit IRC
1852020-10-10T12:00:02 *** MartinAS1 has quit IRC
1862020-10-10T12:10:09 <fanquake> hopefully no time soon
1872020-10-10T12:10:56 <fanquake> Weâve got lots more important things to deal with than swapping build systems, and god forbid we have to maintain two at once.
1882020-10-10T12:13:43 <wumpus> vasild: you are probably following the new, better build instructions for freebsd that tack on MAKE=gmake instead
1892020-10-10T12:13:57 *** Pavlenex has joined #bitcoin-core-dev
1902020-10-10T12:15:06 <vasild> wumpus: I run ./autogen.sh && ./configure && gmake
1912020-10-10T12:16:45 <wumpus> all build systems have their own set of warts using cmake would just switch one for the other
1922020-10-10T12:16:51 <wumpus> vasild: that doesn't work for me
1932020-10-10T12:17:28 <wumpus> (and for many other people it doesn't either which is why that is in the build instructions)
1942020-10-10T12:18:16 <wumpus> might be a automake/autoconf version thing but in any case I'm not very interested in debugging that now :-)
1952020-10-10T12:18:18 <vasild> I saw some mentions of build failures and --disable-dependency-tracing but never bothered to study that, especially that it works for me :)
1962020-10-10T12:18:45 <vasild> I mean build works for me without --disable-dependency-tracing
1972020-10-10T12:18:49 <wumpus> disabling dependency tracking will result in madness and insanity don't do it, override MAKE intead
1982020-10-10T12:21:14 *** AaronvanW has joined #bitcoin-core-dev
1992020-10-10T12:22:02 *** justmay1 has joined #bitcoin-core-dev
2002020-10-10T12:46:14 *** Guyver2 has joined #bitcoin-core-dev
2012020-10-10T12:46:26 *** ares_ has quit IRC
2022020-10-10T12:46:44 *** ares_ has joined #bitcoin-core-dev
2032020-10-10T12:53:45 *** MasterdonX has quit IRC
2042020-10-10T12:57:20 *** pinheadmz has joined #bitcoin-core-dev
2052020-10-10T13:44:45 *** owowo has quit IRC
2062020-10-10T13:44:59 *** rc_423_ has joined #bitcoin-core-dev
2072020-10-10T13:46:35 *** rc_423 has quit IRC
2082020-10-10T13:49:57 *** owowo has joined #bitcoin-core-dev
2092020-10-10T14:00:37 *** mdunnio has joined #bitcoin-core-dev
2102020-10-10T14:06:06 *** luke-jr has quit IRC
2112020-10-10T14:06:28 *** luke-jr has joined #bitcoin-core-dev
2122020-10-10T14:07:10 *** ZAQfx has joined #bitcoin-core-dev
2132020-10-10T14:08:45 *** jonatack has joined #bitcoin-core-dev
2142020-10-10T14:13:05 *** ZAQfx has quit IRC
2152020-10-10T14:37:32 *** ZAQfx has joined #bitcoin-core-dev
2162020-10-10T14:46:58 <ariard> vasild: I'm still not understanding how a BIP155 node is supposed to discover the list of network IDs supported by its BIP155-peers ?
2172020-10-10T14:47:18 <ariard> for now the table is hardcoded but I assume it will be extended in the future
2182020-10-10T14:50:59 *** proofofkeags_ has joined #bitcoin-core-dev
2192020-10-10T14:50:59 *** proofofkeags has joined #bitcoin-core-dev
2202020-10-10T15:00:01 *** justmay1 has quit IRC
2212020-10-10T15:00:16 *** luke-jr has quit IRC
2222020-10-10T15:01:07 *** luke-jr has joined #bitcoin-core-dev
2232020-10-10T15:03:45 *** mdunnio has quit IRC
2242020-10-10T15:09:48 *** mdunnio has joined #bitcoin-core-dev
2252020-10-10T15:15:04 *** Victorsueca has joined #bitcoin-core-dev
2262020-10-10T15:15:26 *** mdunnio has quit IRC
2272020-10-10T15:22:24 *** hollowleviathan1 has joined #bitcoin-core-dev
2282020-10-10T15:22:29 *** hollowleviathan1 is now known as Guest982
2292020-10-10T15:24:13 *** mol has quit IRC
2302020-10-10T15:30:07 *** proofofkeags_ has quit IRC
2312020-10-10T15:30:07 *** proofofkeags has quit IRC
2322020-10-10T15:30:12 *** mol has joined #bitcoin-core-dev
2332020-10-10T15:38:34 *** Victorsueca has quit IRC
2342020-10-10T15:38:43 *** DeanGuss has quit IRC
2352020-10-10T15:43:00 *** Victorsueca has joined #bitcoin-core-dev
2362020-10-10T16:10:58 *** proofofkeags has joined #bitcoin-core-dev
2372020-10-10T16:10:58 *** proofofkeags_ has joined #bitcoin-core-dev
2382020-10-10T16:12:33 *** Pavlenex has joined #bitcoin-core-dev
2392020-10-10T16:18:30 *** luke-jr has quit IRC
2402020-10-10T16:19:51 *** luke-jr has joined #bitcoin-core-dev
2412020-10-10T16:22:05 *** S3RK has joined #bitcoin-core-dev
2422020-10-10T16:26:25 *** S3RK has quit IRC
2432020-10-10T16:46:31 *** Pavlenex has quit IRC
2442020-10-10T16:53:11 *** Shabra has joined #bitcoin-core-dev
2452020-10-10T16:53:24 *** k3tan has quit IRC
2462020-10-10T16:53:58 *** Shabra has quit IRC
2472020-10-10T16:54:30 *** k3tan has joined #bitcoin-core-dev
2482020-10-10T17:23:46 <hebasto> sipa: could protecting `m_txrequest` with its own mutex rather with `cs_main` be safe and provide more concurrency?
2492020-10-10T17:24:16 <sipa> hebasto: no, because all call sites already hold cs_main
2502020-10-10T17:24:40 <hebasto> sipa: ok, thanks
2512020-10-10T17:24:43 <sipa> it's also pointless, as net_processing is inherently single-threaded (there is only one message handler thread)
2522020-10-10T17:25:45 <sipa> i think net_processing should move to its own locks, instead of cs_main, but that's a larger scope thing, and doing that for just m_txrequest doesn't provide benefits until the rest of the file can make use of it
2532020-10-10T17:26:00 <hebasto> then why protecting `m_txrequest` at all?
2542020-10-10T17:26:14 <sipa> ?
2552020-10-10T17:26:39 <hebasto> if the only thread has access
2562020-10-10T17:26:58 <sipa> because we don't want to introduce bugs
2572020-10-10T17:27:15 <sipa> if the code is perfect, all thread annotations/checks are wasted effort
2582020-10-10T17:27:56 <sipa> we could also delete all tests ;)
2592020-10-10T17:28:11 <hebasto> no, please :)
2602020-10-10T17:28:51 <sipa> it isn't the case that only one thread has access - it's that (i assume!) in the current code, only one thread actually accesses it
2612020-10-10T17:29:53 <hebasto> re "i think net_processing should move to its own locks, instead of cs_main" agree
2622020-10-10T17:32:26 *** Talkless has joined #bitcoin-core-dev
2632020-10-10T17:48:04 *** Victorsueca has quit IRC
2642020-10-10T17:59:27 *** DeanGuss has joined #bitcoin-core-dev
2652020-10-10T18:00:02 *** Guest982 has quit IRC
2662020-10-10T18:22:04 *** litenull has joined #bitcoin-core-dev
2672020-10-10T18:38:25 *** roconnor has quit IRC
2682020-10-10T19:02:52 *** [\\\] has quit IRC
2692020-10-10T19:03:08 *** tripleslash has joined #bitcoin-core-dev
2702020-10-10T19:06:46 *** belcher_ has joined #bitcoin-core-dev
2712020-10-10T19:09:09 *** belcher has quit IRC
2722020-10-10T19:16:38 *** belcher_ is now known as belcher
2732020-10-10T19:17:05 *** mdunnio has joined #bitcoin-core-dev
2742020-10-10T19:18:24 *** ZAQfx has quit IRC
2752020-10-10T19:22:24 *** mdunnio has quit IRC
2762020-10-10T19:29:25 *** EagleTM has joined #bitcoin-core-dev
2772020-10-10T19:40:46 *** rc_423_ has quit IRC
2782020-10-10T20:12:30 <stevenroose> is there any protocol rule limiting future env type identifiers to 32 bytes?
2792020-10-10T20:12:42 <stevenroose> or can it in principle be any length?
2802020-10-10T20:13:19 <sipa> "env type identifiers" ?
2812020-10-10T20:14:56 <stevenroose> inv*
2822020-10-10T20:15:04 <stevenroose> from the inventory messages
2832020-10-10T20:15:12 <stevenroose> txid, wtixd, block hash, ...
2842020-10-10T20:15:17 <stevenroose> sipa: ^
2852020-10-10T20:15:48 <stevenroose> all current ones are some variant of sha256 hash, but I'm not sure if that's a hard rule
2862020-10-10T20:17:13 <sipa> no, they're just 32-bit numbers
2872020-10-10T20:17:22 <stevenroose> I mean CInv is currently hash256, so Core would refuse any future inv variants that are > 32 bytes. perhaps that's enough to make sure it won't every be any other length
2882020-10-10T20:17:50 <sipa> the current protocol defines a CInv as 32-bit type + 256-bit value
2892020-10-10T20:17:59 <stevenroose> key thanks
2902020-10-10T20:18:29 <sipa> there isn't even a way to convey anything else
2912020-10-10T20:18:38 <sipa> it's not like the "CInv value length" is sent anywhere
2922020-10-10T20:19:07 <sipa> a negotiated protocol change could modify that of course
2932020-10-10T20:21:53 <stevenroose> ah good point there is no length indicator, hadn't thought about that
2942020-10-10T20:23:40 *** S3RK has joined #bitcoin-core-dev
2952020-10-10T20:28:18 *** S3RK has quit IRC
2962020-10-10T20:37:01 <jonatack> stevenroose: until very recently, CInv type was implemented as an int and was changed to uint32_t in #19818
2972020-10-10T20:37:03 <gribble> https://github.com/bitcoin/bitcoin/issues/19818 | p2p: change `CInv::type` from `int` to `uint32_t`, fix UBSan warning by jonatack · Pull Request #19818 · bitcoin/bitcoin · GitHub
2982020-10-10T20:37:29 <jonatack> though it was documented as uint32_t in https://en.bitcoin.it/wiki/Protocol_documentation#Inventory_Vectors
2992020-10-10T20:37:36 <jonatack> and in https://btcinformation.org/en/developer-reference#data-messages
3002020-10-10T20:39:39 <stevenroose> jonatack: awesome :) yeah in rust-bitcoin it's been a u32 always AFAIK, but I was confused, my question didn't make much sense because as Pieter pointed out, the inv message can only have 32-byte hashes as there is no way to convey items of other lengths
3012020-10-10T20:39:50 <jonatack> (by MagicalTux and harding, respectively)
3022020-10-10T20:50:31 <luke-jr> jonatack: on all supported platforms, int is 32-bit
3032020-10-10T20:59:17 <jonatack> luke-jr: right (https://en.cppreference.com/w/cpp/language/types), recent change was signedness
3042020-10-10T21:00:02 *** litenull has quit IRC
3052020-10-10T21:02:56 *** awesome_doge has joined #bitcoin-core-dev
3062020-10-10T21:04:37 *** awesome_doge has quit IRC
3072020-10-10T21:06:26 <luke-jr> not sure it matters :P
3082020-10-10T21:19:43 *** Guyver2 has quit IRC
3092020-10-10T21:21:53 *** gchaix has joined #bitcoin-core-dev
3102020-10-10T21:24:47 *** awesome_doge has joined #bitcoin-core-dev
3112020-10-10T21:26:49 *** awesome_doge has quit IRC
3122020-10-10T21:34:48 *** awesome_doge has joined #bitcoin-core-dev
3132020-10-10T21:37:07 *** belcher_ has joined #bitcoin-core-dev
3142020-10-10T21:39:49 *** belcher has quit IRC
3152020-10-10T21:50:49 *** mrostecki has joined #bitcoin-core-dev
3162020-10-10T22:03:14 *** Talkless has quit IRC
3172020-10-10T22:06:59 *** shesek has quit IRC
3182020-10-10T22:07:31 *** S3RK has joined #bitcoin-core-dev
3192020-10-10T22:10:43 *** vasild has quit IRC
3202020-10-10T22:12:25 *** S3RK has quit IRC
3212020-10-10T22:12:59 *** vasild has joined #bitcoin-core-dev
3222020-10-10T22:49:13 *** owowo has quit IRC
3232020-10-10T22:53:56 *** owowo has joined #bitcoin-core-dev
3242020-10-10T23:04:03 *** EagleTM has quit IRC
3252020-10-10T23:12:43 *** promag has joined #bitcoin-core-dev
3262020-10-10T23:25:28 *** mdunnio has joined #bitcoin-core-dev
3272020-10-10T23:30:34 *** morcos has quit IRC
3282020-10-10T23:31:19 *** morcos has joined #bitcoin-core-dev
3292020-10-10T23:33:16 *** AaronvanW has quit IRC
3302020-10-10T23:35:46 *** mdunnio has quit IRC
3312020-10-10T23:38:23 *** davterra has quit IRC