--- Log opened Tue Jul 30 00:00:25 2019 01:52 -!- Delfina29Vanderv [~Delfina29@ns334669.ip-5-196-64.eu] has quit [Ping timeout: 276 seconds] 04:56 -!- jonatack [d598a155@213.152.161.85] has joined #bitcoin-builds 08:07 < dongcarl> fanquake: will check it ou 10:53 < dongcarl> fanquake: Done reviewing #16370 10:53 < dongcarl> Not sure I can contribute much to #16371 as I don't have a Mac to test on... 10:55 < dongcarl> EVERYONE: Just wrote up https://github.com/bitcoin/bitcoin/issues/16496, let me know if you want clarifications, I'd gladly provide! 11:49 -!- hebasto [~hebasto@95.164.65.194] has quit [Remote host closed the connection] 12:12 < cfields> dongcarl / fanquake: suggestions on where I'm needed? I can't keep up with everything flying around :( 12:12 * dongcarl looking now 12:14 < dongcarl> cfields: Just https://github.com/bitcoin/bitcoin/pull/16435 for me, I know you were previous not a fan of this approach 12:14 < dongcarl> cfields: I think fanquake updated https://github.com/bitcoin/bitcoin/pull/16392 last night 12:15 < cfields> dongcarl: -0. That's bound to flip/flop around over time, I'm not too concerned. 12:16 < cfields> Cool, will review that one. 12:17 < dongcarl> cfields: sorry not 100% sure what you mean by that? 12:20 < cfields> wait, no. That's not doing what I thought it was. 12:20 < cfields> I don't understand why -O0? It's not debuggable... 12:21 < dongcarl> cfields: my experience with gcc has been that `-Og` optimizes out various things, and only `-O0` doesn't, which is better for debugging 12:21 < dongcarl> s/gcc/gdb/ 12:21 < cfields> dongcarl: I've found -O0 generally impossible to step through due to lack of optimization. It also performs _nothing_ like real-world code. 12:22 < cfields> But again, I expect these flags to change over time. Maybe that's not the case with -O0 anymore. 12:23 < dongcarl> cfields: not sure why lack of optimization would make it impossible to step through? In my naive view, it makes for a better debugging experience 12:27 < cfields> dongcarl: unoptimized c++ binaries contain a _ton_ of copies that aren't yet elided, useless constructors, deconstructors, etc. The good stuff is buried. 12:28 < cfields> But I don't care that much. If the people who use it have a preference.. sure :) 12:29 * dongcarl looking at gcc docs 12:40 < dongcarl> cfields: I wanna understand what you mean though... My main use for gdb is just stepping through and printing things... In which case it doesn't matter that there are extra stuff, it only matters when the stuff I need isn't there... 12:44 < cfields> dongcarl: understood. But my argument (and the one made in the gcc docs :p) is that -g is sufficient most of the time for most debugging, and reasonably optimizes otherwise. 12:44 < cfields> I would argue that's what --enable-debug should mean. 12:44 < cfields> -Og, sorry. 12:50 < dongcarl> cfields: Hmmm, I see your point. It seems that users have a preference though, so I'm gunna go ahead with -O0, if there are usage problems we can always reenable individual optimizations 12:51 < cfields> Np 14:03 < fanquake> dongcarl: nice, shall read through 14:04 < fanquake> cfields: probably #16370 from me at the moment. --- Log closed Wed Jul 31 00:00:26 2019