' from ./bin/gbuild:304:in `each' from ./bin/gbuild:304:in `block in

' [11:44:37] from ./bin/gbuild:302:in `each' from ./bin/gbuild:302:in `

' anyone have any idea why? [11:45:49] ahh just saw @mxaddict posts about docker [11:47:19] i'll try a docker build instead [11:54:31] Yep sorry, we switched to docker. Should be easier to get running than the old vm setup [12:17:39] Also, if you need to run gitian from OSX [12:17:46] @prole has a fork that works on OSX [12:17:53] I have only tested on linux [12:17:58] @mc290 [13:15:27] @aguycalled @prole we are using bionic right now, but seems we will need to switch to different distro in the future [13:15:36] 20.04 ubuntu will not have 32bit support anymore [13:15:56] I guess we wait for what BTC does, then we copy? 😄 [13:16:14] But for now bionic is good, just thinking about the future [13:27:16] i'll try set up in linux [13:27:18] @prole about the RC releases [13:27:23] BTC puts RCs in https://github.com/bitcoin/bitcoin/releases [13:28:00] naming convension is v0.0.0rc0 replace 0 with the correct version numbers [13:28:24] IE if we release RC1 of v4.7.0 the release tag would be v4.7.0rc1 [13:29:24] And it looks like if the RC is successfull, they then retag the same tag as final version [13:29:51] IE if v4.7.0rc1 passes checking, then we re-tag same commmit/hash as v4.7.0 [13:30:04] That is what I can understand from the releases that BTC is doing [13:30:24] We could follow same method and have RCs on github to streamline the release process [13:48:14] hmmm not sure if it's just me but I'm getting an error when I run navcoin-qt RC binary on linux [13:48:30] navcoind binary seems to be running fine so far [13:48:50] /navcoin-4.7.0/bin$ ./navcoin-qt navcoin-qt: compat/glibc_compat.cpp:113: int __poll_chk(pollfd*, nfds_t, int, size_t): Assertion `(fdslen / sizeof(*fds)) < nfds' failed. Aborted (core dumped) [13:49:21] i'll try again tomorrow. sleep time [14:08:51] @mc290 64bit linux? [14:08:55] I'll see if I can replicate [15:57:37] @mc290 I was able to replicate the issue [15:57:44] I'm creating a PR now to fix it. [15:58:29] @prole @aguycalled [15:58:41] @mc290 found a bug with the glibc compat code [15:58:48] I'm creating a new PR to fix it. [16:15:23] I was a bit naive and thought an assert would stop the execution of any bad poll calls [16:15:50] I've changed it to match glibc implementation which uses chk_fail [16:17:08] that pr included more additions to glibc_compat [16:17:11] which are not in upstream [16:17:40] did you get them from somewhere else or are all of your own? [16:24:32] Those were to make libunbound pass the glibc symbols check [16:30:24] yes, but i mean if you wrote the new functions or if you took them from somewhere [16:52:14] glob_old, wrap_glob and __wrap_glob64 I coded based on the __wrap_log2f code that was already in the file. __poll_chk I based on glibc implementation, but I used assert (I did not know __chk_fail existed) : https://sourceware.org/git/?p=glibc.git;a=blob;f=debug/poll_chk.c;h=34596de9777b3fe9f05e8a57840dd75076c51672;hb=HEAD __explicit_bzero_chk I based on monero implementation: [16:52:15] https://github.com/monero-project/monero/blob/master/src/common/compat/glibc_compat.cpp reallocarray from : libunbound in our ./depends dir getentropy I think I coded myself. [16:53:18] @aguycalled [17:00:12] Did we already upload RC binaries? [17:01:11] Windows and OSX should be unaffected by this bug, but linux/riscv64/arm/aarch will all have been affected [17:03:32] I think __poll_chk was only called by libcurl [17:03:47] So navcoind on the linux builds should be unaffected as well [17:03:58] Only the QT wallets [17:04:09] i haven't upload anything. my gitian is still going [17:04:15] I see [17:35:18] I wish we could get more users to test our RC, we can find and squash so many more issues in RC if we had more guys compiling and testing 😢 [17:35:40] We are so out numbered right now (Out numbered by work/issues) 🤣 [17:35:53] I feel like the guys from sparta: 300 [18:04:19] @mc290 can you test my PR and see if you still see the crash with: navcoin-qt: compat/glibc_compat.cpp:113: int __poll_chk(pollfd*, nfds_t, int, size_t): Assertion(fdslen / sizeof(*fds)) < nfds' failed.` [18:16:43] im testing it right now too [18:18:24] Thanks [18:18:35] I updated the pr to also remove the assert header [18:41:50] I'll can try testing the RC when I'm back from vacation [18:45:27] I'll pledge my time to guide anyone interested in building [18:45:47] Just ping me here