00:40 <+bridge> [ddnet] @deen where do you get that SDL2.framework? 00:40 <+bridge> [ddnet] it seems to have a wrong path for hidapi in it 00:53 <+bridge> [ddnet] not our fault, sdl2.0.12 seems to be linking weirdly 00:54 <+bridge> [ddnet] idk how @rpath is supposed to work exactly but the path it links to for hidapi feels like it should be "@rpath/Frameworks/hidapi.framework" 07:21 <+bridge> [ddnet] @Learath2 from the official libsdl.org site 07:28 <+bridge> [ddnet] Same issue here I guess: https://github.com/bvschaik/julius/issues/386 08:50 <+bridge> [ddnet] @onby @Learath2 can you try again on mac osx? 08:50 <+bridge> [ddnet] (i fixed the 13.0.1 release on osx) 08:50 <+bridge> [ddnet] ok ill try 08:50 <+bridge> [ddnet] Did you upload the dmg? 08:52 <+bridge> [ddnet] Identifier: org.DDNetClient.app 08:52 <+bridge> [ddnet] Version: 13.0.1 08:52 <+bridge> [ddnet] Code Type: X86-64 (Native) 08:52 <+bridge> [ddnet] ... 08:52 <+bridge> [ddnet] Crashed Thread: 0 08:52 <+bridge> [ddnet] 08:52 <+bridge> [ddnet] Exception Type: EXC_CRASH (SIGABRT) 08:52 <+bridge> [ddnet] Exception Codes: 0x0000000000000000, 0x0000000000000000 08:52 <+bridge> [ddnet] Exception Note: EXC_CORPSE_NOTIFY 08:52 <+bridge> [ddnet] 08:52 <+bridge> [ddnet] Termination Reason: DYLD, [0x1] Library missing 08:52 <+bridge> [ddnet] 08:52 <+bridge> [ddnet] Application Specific Information: 08:52 <+bridge> [ddnet] dyld: launch, loading dependent libraries 08:52 <+bridge> [ddnet] 08:53 <+bridge> [ddnet] Dyld Error Message: 08:53 <+bridge> [ddnet] Library not loaded: @rpath/hidapi.framework/Versions/A/hidapi 08:53 <+bridge> [ddnet] Referenced from: /Users/nobody1/Desktop/DDNet.app/Contents/Frameworks/SDL2.framework/SDL2 08:53 <+bridge> [ddnet] Reason: image not found 08:54 <+bridge> [ddnet] Binary Images: 08:54 <+bridge> [ddnet] 0x10ff4a000 - 0x110287fff +org.DDNetClient.app (13.0.1) /Users/USER/Desktop/DDNet.app/Contents/MacOS/DDNet 08:54 <+bridge> [ddnet] 0x110541000 - 0x11062efff +libfreetype.6.dylib (24.1) /Users/USER/Desktop/DDNet.app/Contents/Frameworks/libfreetype.6.dylib 08:54 <+bridge> [ddnet] 0x11064d000 - 0x11078efff +org.libsdl.SDL2 (2.0.12 - 2.0.12) <3E6D5145-6DE8-3734-95A7-AE31ED09A1A9> /Users/USER/Desktop/DDNet.app/Contents/Frameworks/SDL2.framework/SDL2 08:54 <+bridge> [ddnet] 0x11747e000 - 0x1174c8acf dyld (551.4) <8A72DE9C-A136-3506-AA02-4BA2B82DCAF3> /usr/lib/dyld 08:55 <+bridge> [ddnet] @deen didn't work for me either 08:58 <+bridge> [ddnet] The hidapi rpath is wrong in SDL2 itself, not the client 08:58 <+bridge> [ddnet] last argument of install_name_tool should be sdl2 08:58 <@deen> ah, too bad... 08:59 <@deen> so we have to overwrite the rpath in SDL2 directly 08:59 <+bridge> [ddnet] Yeah, they should be the ones doing it tbh, but we can for the time being 09:03 <+bridge> [ddnet] Should I merge and try another build or can someone test locally? ^ 09:03 <+bridge> [ddnet] Wait for a second for CI to build and I can test that 09:05 <+bridge> [ddnet] I think in this case executable path would be SDL2 not DDNet but ill try 09:06 <+bridge> [ddnet] oh 09:06 <+bridge> [ddnet] I'm not completely sure, let me try first 09:07 <+bridge> [ddnet] Mach-O binaries are weird sometimes 09:11 <+bridge> [ddnet] ```dyld: Library not loaded: @executable_path/../Frameworks/SDL2.framework/SDL2/Versions/A/Frameworks/hidapi.framework/Versions/A/hidapi 09:11 <+bridge> [ddnet] Referenced from: /Volumes/DDNet-13.0.1-osx/DDNet.app/Contents/Frameworks/SDL2.framework/SDL2 09:11 <+bridge> [ddnet] Reason: no suitable image found. Did find: 09:11 <+bridge> [ddnet] /Volumes/DDNet-13.0.1-osx/DDNet.app/Contents/MacOS/../Frameworks/SDL2.framework/SDL2/Versions/A/Frameworks/hidapi.framework/Versions/A/hidapi: stat() failed with errno=20 09:11 <+bridge> [ddnet] /Volumes/DDNet-13.0.1-osx/DDNet.app/Contents/MacOS/../Frameworks/SDL2.framework/SDL2/Versions/A/Frameworks/hidapi.framework/Versions/A/hidapi: stat() failed with errno=20 09:11 <+bridge> [ddnet] Abort trap: 6``` 09:17 <+bridge> [ddnet] oops, I also had a wrong path in there, pushed again 09:17 <+bridge> [ddnet] so we don't know where executable_path points to yet? 09:19 <+bridge> [ddnet] Not sure :/ 09:19 < Test123> Hi 09:19 <+bridge> [ddnet] It should be tho, how else could you link a library to a library? 09:21 <@deen> I think executable_path will be the path to DDNet.app 09:21 <+bridge> [ddnet] Then this should work 09:25 <+bridge> [ddnet] I wonder if we can mount and run the dmg on the CI 09:25 <@deen> would be great 09:26 <@deen> can you try the version I pushed again? 09:26 <+bridge> [ddnet] yep, it works 09:26 <@deen> ok, thanks 09:27 <+bridge> [ddnet] @deen there is a patch I got into sdl, but it wasnt released on a version yet, you think we should apply that to the macOS ddnet sdl? 09:27 <@deen> I'm not building my own sdl2 for mac 09:27 <@deen> do we need it? 09:27 <+bridge> [ddnet] Oooooh, maybe you should be compiling with the 10.14 or 13 sdk then 09:27 <+bridge> [ddnet] I forgot about that 09:27 <@deen> too late! 09:28 <+bridge> [ddnet] on binaries linked on 10.15 there is a new check enabled in AppKit which crashes on launch 09:28 <+bridge> [ddnet] if you compile on older AppKit compatibility disables the asserts 09:28 <@deen> so we're crashing? 09:29 <+bridge> [ddnet] Yeah, at launch it hits an assert 09:31 <@deen> so I should downgrade to 10.13 sdk and we're fine? 09:32 <+bridge> [ddnet] Yeah I think so 09:32 <+bridge> [ddnet] Or you apply my sdl patch 09:32 <+bridge> [ddnet] But compiling sdl on our own for mac too is just a pain 09:32 <+bridge> [ddnet] exactly, so I'll downgrade sdl 09:33 <+bridge> [ddnet] We can upgrade sdk when the patch is in a stable sdl 09:33 <+bridge> [ddnet] good to know 09:33 <+bridge> [ddnet] Its broken on old sdl too, you need old sdk 09:34 <+bridge> [ddnet] do i need to recompile libs? 09:35 <+bridge> [ddnet] I never tried, mach-o binaries should be forward compatible though 09:56 <+bridge> [ddnet] @deen lmk when you have a build, I can test 10:19 <+bridge> [ddnet] @Learath2 available now 10:24 <+bridge> [ddnet] I'm sad to report that that didn't work 10:25 <+bridge> [ddnet] I mean it did do what's intended but now we get some lazy loading issue 10:25 <+bridge> [ddnet] what was your `MACOSX_DEPLOYMENT_TARGET`? 10:27 <+bridge> [ddnet] ```dyld: lazy symbol binding failed: Symbol not found: ___isOSVersionAtLeast 10:27 <+bridge> [ddnet] Referenced from: /Volumes/DDNet-13.0.1-osx/DDNet.app/Contents/MacOS/DDNet 10:27 <+bridge> [ddnet] Expected in: flat namespace 10:27 <+bridge> [ddnet] 10:27 <+bridge> [ddnet] dyld: Symbol not found: ___isOSVersionAtLeast 10:27 <+bridge> [ddnet] Referenced from: /Volumes/DDNet-13.0.1-osx/DDNet.app/Contents/MacOS/DDNet 10:27 <+bridge> [ddnet] Expected in: flat namespace 10:27 <+bridge> [ddnet] 10:27 <+bridge> [ddnet] Abort trap: 6 10:27 <+bridge> [ddnet] logout 10:27 <+bridge> [ddnet] Saving session...completed.``` 10:30 <+bridge> [ddnet] I don't set one for the DDNet build, for the libs -mmacosx-version-min=10.9 10:30 <+bridge> [ddnet] should I try setting -mmacosx-version-min=10.9 on ddnet as well? 10:31 <+bridge> [ddnet] ah, CMakeLists.txt sets 10.7 10:31 <+bridge> [ddnet] Also, we have something in CMakeLists.txt just for that failure you see 10:32 <+bridge> [ddnet] target_link_libraries(${target} -Wl,-undefined -Wl,dynamic_lookup) 10:32 <+bridge> [ddnet] yeah noticed that, but that enables lazy binding, idk why it won't bind though 10:32 <+bridge> [ddnet] let's try setting all to 10.9? 10:34 <+bridge> [ddnet] yeah sure, let's try 10:35 <+bridge> [ddnet] I don't think you can cross compile if you disable the lazy binding @deen 10:35 <+bridge> [ddnet] you probably don't have libclang_rt.osx.a as it's part of the toolchain not the sdk 10:37 <+bridge> [ddnet] ok 10:37 <+bridge> [ddnet] this release is really annoying 10:41 <+bridge> [ddnet] apparently osxcross has a way to build compiler_rt 10:41 <+bridge> [ddnet] I'll get a linux vm 10:42 <+bridge> [ddnet] I guess I'll have to recompile the libs with 10.7 10:42 <+bridge> [ddnet] 10.9 also fails in central thing 10:43 <+bridge> [ddnet] @Learath2 you can try 13.0.1 again on macosx? I rebuilt with 10.9 10:43 <+bridge> [ddnet] I'm not sure what's the issue, lazy binding should resolve that symbol just fine 10:43 <+bridge> [ddnet] @deen is it on ddnet.tw? 10:44 <+bridge> [ddnet] yes 10:45 <+bridge> [ddnet] nope still failing 10:46 <+bridge> [ddnet] then I'll rebuild the libs 10:51 <+bridge> [ddnet] I don't get this, how was this not a problem when you compiled with the 10.15 sdk? 10:52 <+bridge> [ddnet] maybe it was but we never got this far in booting up? 10:52 <+bridge> [ddnet] *starting up 10:52 <+bridge> [ddnet] since we always crashed before 10:52 <+bridge> [ddnet] ah hidapi wasn't lazily linked, yeah it probably was a problem 10:53 <+bridge> [ddnet] curl doesn't build with 10.7 10:53 <+bridge> [ddnet] vtls/sectransp.c:1313:6: error: 'SSLSetProtocolVersionMax' is only available on macOS 10.8 or newer [-Werror,-Wunguarded-availability] 10:57 <+bridge> [ddnet] weren't you going to recompile everything with 10.9 anyway? 10:57 <+bridge> [ddnet] no, i had everything with 10.9 and you said it doesn't work 10:57 <+bridge> [ddnet] so now I'm going everything 10.7 10:57 <+bridge> [ddnet] ah, that was with all 10.9 10:59 <+bridge> [ddnet] maybe the sdk change caused the problem? should i try rebuilding all? 10:59 <+bridge> [ddnet] with 10.9 again 11:00 <+bridge> [ddnet] i'm kind of searching in the dark here :/ 11:00 <+bridge> [ddnet] and before I didn't notice this curl error since ssl was accidentally disabled 😄 11:01 <+bridge> [ddnet] I'm setting up osxcross on a vm 11:02 <+bridge> [ddnet] if you are bored of this I can finish it up 11:02 <+bridge> [ddnet] I'll give you another dmg file to test soon 11:03 <+bridge> [ddnet] @deen are you building with --without-ssl btw? 11:03 <+bridge> [ddnet] no 11:03 <+bridge> [ddnet] try --without-ssl --with-secure-transport 11:18 <+bridge> [ddnet] uhm when I'm cross building cmake can't find sdl framework 11:24 <+bridge> [ddnet] @deen can you find the bundled SDL2 properly? 11:24 <+bridge> [ddnet] cmake on linux even when cross compiling doesn't seem to want to find frameworks 11:32 <+bridge> [ddnet] what do you mean by find? 11:33 <+bridge> [ddnet] I don't have SDL2 on my VM, but it should use the bundled lib anyway 11:33 <+bridge> [ddnet] it says SDL2 not found 11:34 <+bridge> [ddnet] -- * SDL2 not found (using bundled version) 11:34 <+bridge> [ddnet] for me 11:34 <+bridge> [ddnet] did you clone ddnet-libs? 11:34 <+bridge> [ddnet] yep 11:34 <+bridge> [ddnet] it finds curl just fine 11:34 <+bridge> [ddnet] what version of cmake do you use? 11:34 <+bridge> [ddnet] I also set -DPREFER_BUNDLED_LIBS=ON -DCMAKE_TOOLCHAIN_FILE=cmake/toolchains/darwin.toolchain 11:34 <+bridge> [ddnet] cmake version 3.17.0 11:35 <+bridge> [ddnet] let me upgrade, maybe that'll help, but I set the exact same thing you set 11:43 <+bridge> [ddnet] why is libqt a dependency of cmake? 11:43 <+bridge> [ddnet] it pulled in emacs even 11:43 <+bridge> [ddnet] probably they have a graphical interface 11:50 <+bridge> [ddnet] @Learath2 please try dmg from ddnet.tw again, now rebuilt all with all 10.7 11:52 <+bridge> [ddnet] it works 11:52 <+bridge> [ddnet] party time 11:52 <+bridge> [ddnet] :D 11:53 <+bridge> [ddnet] wierd graphical thing goin on here 11:53 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/697383490169340004/Screen_Shot_2020-04-08_at_4.52.50_AM.png 11:53 <+bridge> [ddnet] but worth it for new ddnet with good fps 11:53 <+bridge> [ddnet] @deen wait, onby is on an old osx 11:53 <+bridge> [ddnet] let me check just to be sure 11:53 <+bridge> [ddnet] the lines inbetween? they look like opengl3.3 11:55 <+bridge> [ddnet] it works, but for a different reason 11:56 <+bridge> [ddnet] ? 11:56 <+bridge> [ddnet] clang optimizes out the version check 11:56 <+bridge> [ddnet] 😄 11:56 <+bridge> [ddnet] i guess it's fine for now 11:56 <+bridge> [ddnet] yeah 11:56 <+bridge> [ddnet] if I ever get cmake working with osxcross I'll fix it 11:56 <+bridge> [ddnet] it's much easier when you have access to a mac 11:56 <+bridge> [ddnet] yes 11:57 <+bridge> [ddnet] i now have good fps and no fan spinning on ddnet 13 :D 11:57 <+bridge> [ddnet] cool 11:57 <+bridge> [ddnet] I accidentally killed my terminal, which killed the cmake build I was running.... 11:58 <+bridge> [ddnet] Anyone here running archlinux? 11:58 <+bridge> [ddnet] yes 11:58 <+bridge> [ddnet] Do you know why pacman might be stuck not upgrading things? 11:58 <+bridge> [ddnet] cmake is 3.17 in extra and it seems to be 3.17 in all mirrors 11:58 <+bridge> [ddnet] IgnorePkg/IgnoreGroup in /etc/pacman.conf? 11:59 <+bridge> [ddnet] ah let me check 12:00 <+bridge> [ddnet] nope 12:00 <+bridge> [ddnet] pacman -S cmake? 12:01 <+bridge> [ddnet] Packages (1) cmake-3.16.4-1 12:03 <@deen> pacman -Syu cmake? 12:03 <+bridge> [ddnet] ```:: Synchronizing package databases... 12:03 <+bridge> [ddnet] core is up to date 12:03 <+bridge> [ddnet] extra is up to date 12:04 <+bridge> [ddnet] community is up to date 12:04 <+bridge> [ddnet] :: Starting full system upgrade... 12:04 <+bridge> [ddnet] resolving dependencies... 12:04 <+bridge> [ddnet] looking for conflicting packages... 12:04 <+bridge> [ddnet] 12:04 <+bridge> [ddnet] Packages (1) cmake-3.16.4-1 12:04 <+bridge> [ddnet] 12:04 <+bridge> [ddnet] Total Download Size: 6.77 MiB 12:04 <+bridge> [ddnet] Total Installed Size: 35.53 MiB``` 12:04 <+bridge> [ddnet] oh rip irc 12:04 <+bridge> [ddnet] weird, maybe you use an out-of-date mirror 12:05 <+bridge> [ddnet] oh the top mirror is indeed out of date 12:05 <+bridge> [ddnet] who would do such a thing 12:09 <+bridge> [ddnet] and mirror status says that mirror is up to date 12:10 <+bridge> [ddnet] oh nvm, they do list it as out of sync 12:10 <+bridge> [ddnet] imagine hosting a widely used mirror and not actually mirroring 12:11 <+bridge> [ddnet] do u only have 1 mirror enabled? 12:11 <+bridge> [ddnet] shouldnt it check several mirrors or smth 12:12 <+bridge> [ddnet] no, the first that works is always used 12:12 <+bridge> [ddnet] it's the top mirror and it connected fine so it never checks another 12:12 <+bridge> [ddnet] they should really add a way to check sync, like a last sync timestamp at the repo root 12:13 <+bridge> [ddnet] pacman could then display a `WARNING: selected mirror could be out of date (last sync %d)` 12:13 <+bridge> [ddnet] tbh I never had an out-of-date mirror 12:14 <+bridge> [ddnet] btw kernel 5.6 made a thing which theorically makes "make" way faster when parallel 12:14 <+bridge> [ddnet] well if it happens it's quite hard to diagnose 12:14 <+bridge> [ddnet] https://www.phoronix.com/scan.php?page=news_item&px=Linux-Pipe-Parallel-Job-Opt 12:14 <+bridge> [ddnet] we are up to 5.6 now? 12:14 <+bridge> [ddnet] @Learath2 i see updates every day, if there are none something is wrong with the mirror 12:14 <+bridge> [ddnet] "In a simplified test case written by Linus Torvalds, this patch caused the number of context switches on the test program to drop from 11 million to just 1.2 million. " 12:14 <+bridge> [ddnet] :poggers: 12:15 <+bridge> [ddnet] oh wow, that's an entire order of magnitude better 12:15 <+bridge> [ddnet] yep 12:16 <+bridge> [ddnet] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=0ddad21d3e99c743a3aa473121dc5561679e26bb 12:16 <+bridge> [ddnet] here its explained in detail xd 12:16 <+bridge> [ddnet] context switches are pretty cheap tbf, but when you have millions happening it does stack up 12:16 <+bridge> [ddnet] uname -a 12:17 <+bridge> [ddnet] Linux arch 5.6.2-arch1-2 #1 SMP PREEMPT Sun, 05 Apr 2020 05:13:14 +0000 x86_64 GNU/Linux 12:17 <+bridge> [ddnet] 5.6.2 12:18 <+bridge> [ddnet] @Learath2 It really depends on your CPU and the mitigations on Intel made it much slower, see bottom here: https://www.phoronix.com/scan.php?page=article&item=3900x-9900k-400&num=7 12:19 <+bridge> [ddnet] https://github.com/google/llvm-propeller 12:19 <+bridge> [ddnet] oh the meltdown/spectre mitigations? I can see them raising the cost of a context switch 12:19 <+bridge> [ddnet] can u optimize ddnet with this? 12:19 <+bridge> [ddnet] :poggers: 12:20 <+bridge> [ddnet] anyway even cmake 3.17 doesn't find SDL2.framework 12:20 <+bridge> [ddnet] yes, you can use PGO, can also try LTO 12:20 <+bridge> [ddnet] but in my experience you don't win much with PGO 12:21 <+bridge> [ddnet] i see 12:21 <+bridge> [ddnet] "2% to 6%", sounds realistic 12:22 <+bridge> [ddnet] using a profiler and improving the slow spots selectively should help more 12:23 <+bridge> [ddnet] what profiler you recommend 12:23 <+bridge> [ddnet] VTune is what we usually use at work, but I think it costs money 12:23 <+bridge> [ddnet] is gprof good 12:23 <+bridge> [ddnet] gprof is a sampling profiler, keeping in mind the limitations of sampling profilers it's decent 12:24 <+bridge> [ddnet] and valgrind --tool=callgrind is what I use privately, but that's just to get a rough estimate since it changes runtimes hugely 12:25 <+bridge> [ddnet] yeah callgrind actually logs all the calls but at the cost of runtime performance 12:25 <+bridge> [ddnet] you are kinda hoping that everything is affected pretty much equally 12:25 <+bridge> [ddnet] yes, but it's not true 12:26 <+bridge> [ddnet] I wonder if there is an architecture that allows a hardware profiler of sorts 12:27 <+bridge> [ddnet] if you have cache misses you definitely won't find those wil callgrind. I've seen perf used to find cache misses 12:27 <+bridge> [ddnet] what about "perf· 12:27 <+bridge> [ddnet] @Learath2 yes, that's what vtune does. intel cpus have specific registers for profiling 12:27 <+bridge> [ddnet] perf 12:27 <+bridge> [ddnet] and perf uses them as well 12:28 <+bridge> [ddnet] but aren't they all just counters? you wouldn't get any callstack information from those 12:28 <+bridge> [ddnet] yeah, right 12:28 <+bridge> [ddnet] it's a little beyond my ability to go from counter information to optimization 12:30 <+bridge> [ddnet] perf report :monkaS: 12:30 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/697393045347237888/unknown.png 15:58 <+bridge> [ddnet] i didnt test on windows 15:58 <+bridge> [ddnet] @Ryozuki did you test it? (open the client, go somewhere?) 15:59 <+bridge> [ddnet] i tested on linux works fine 15:59 <+bridge> [ddnet] okay 16:00 <+bridge> [ddnet] @heinrich5991 macos failed xd 16:00 <+bridge> [ddnet] seen it 16:00 <+bridge> [ddnet] Could NOT find OpenSSL, try to set the path to OpenSSL root folder in the system variable OPENSSL_ROOT_DIR (missing: OPENSSL_INCLUDE_DIR) 16:00 <+bridge> [ddnet] -- Could NOT find Crypto (missing: CRYPTO_LIBRARY CRYPTO_INCLUDEDIR) 16:00 <+bridge> [ddnet] i think its not related? 16:00 <+bridge> [ddnet] that's not the error I think 16:01 <+bridge> [ddnet] it has so many warnings 16:01 <+bridge> [ddnet] https://github.com/ddnet/ddnet/runs/570919188#step:6:355 16:01 <+bridge> [ddnet] that's the reason the build failed 16:01 <+bridge> [ddnet] that warning 16:01 <+bridge> [ddnet] not sure why it appeared now though 16:02 <+bridge> [ddnet] @heinrich5991 note: before #include "SDL_opengl.h" was only on non macos 16:02 <+bridge> [ddnet] but theorically 16:02 <+bridge> [ddnet] sdl_opengl handles cross platform stuff 16:02 <+bridge> [ddnet] it should include opengl fine on macos 16:03 <+bridge> [ddnet] in my change its included on every os 16:03 <+bridge> [ddnet] like in tw 16:03 <+bridge> [ddnet] note: insert '_Nullable' if the pointer may be null 16:03 <+bridge> [ddnet] never seen this 16:03 <+bridge> [ddnet] maybe some compiler flag on macos? 16:04 <+bridge> [ddnet] @heinrich5991 this nullable stuff is objective-c 16:04 <+bridge> [ddnet] https://stackoverflow.com/questions/32539285/pointer-is-missing-a-nullability-type-specifier 16:04 <+bridge> [ddnet] it's also C I think 16:05 <+bridge> [ddnet] @heinrich5991 i think -Wno-nullability-completeness 16:05 <+bridge> [ddnet] disables this 16:05 <+bridge> [ddnet] or -Wno-nonnull 16:05 <+bridge> [ddnet] idk 16:08 <+bridge> [ddnet] why does it start happening with this commit though? 16:09 <+bridge> [ddnet] no idea 16:09 <+bridge> [ddnet] maybe its cuz SDL_opengl.h 16:10 <+bridge> [ddnet] maybe also something on the CI was updated 16:10 <+bridge> [ddnet] I can't re-run master CI unfortunately 16:10 <+bridge> [ddnet] fix a typo somewhere 16:10 <+bridge> [ddnet] xd 16:21 <+bridge> [ddnet] @Learath2 ^? 16:52 <+bridge> [ddnet] i ran valgrind on ddnet client, maybe someone finds something 16:52 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/697458959409479750/valgrind-out.txt 17:55 <+bridge> [ddnet] @heinrich5991 `CUIRect MenuBar, CModeBar, ToolBar, StatusBar, ExtraEditor, UndoList, ToolBox;` declaring a value like this initializes them explicitly right? 17:55 <+bridge> [ddnet] or maybe not 17:56 <+bridge> [ddnet] ah nevermind this isnt the issue i think 17:58 <+bridge> [ddnet] it only initializes them if the constructor does something or the constructors of the struct members do something 18:09 <+bridge> [ddnet] sometimes it feels like valgrind tells u nothing :lol: 18:11 <+bridge> [ddnet] ``` 18:11 <+bridge> [ddnet] Uninitialised value was created by a stack allocation 18:11 <+bridge> [ddnet] ==84015== at 0x1F998E: CDemoRecorder::Start(IStorage*, IConsole*, char const*, char const*, char const*, SHA256_DIGEST*, unsigned int, char const*, unsigned int, unsigned char*, IOINTERNAL*, bool (*)(void const*, int, void*), void*) (demo.cpp:45) 18:11 <+bridge> [ddnet] ``` 18:11 <+bridge> [ddnet] sure but why u dont tell me exactly what stack allocated variable it is 😤 18:12 <+bridge> [ddnet] rip IRC, but it says here: 18:12 <+bridge> [ddnet] ah, message too long, I'm not allowed to rip IRC 18:12 <+bridge> [ddnet] ? 18:12 <+bridge> [ddnet] make a pastebin or smth 18:14 <+bridge> [ddnet] it says in the lines above that it comes from this function: 18:14 <+bridge> [ddnet] ``` 18:14 <+bridge> [ddnet] ==84015== by 0x1FA13A: CDemoRecorder::Start(IStorage*, IConsole*, char const*, char const*, char const*, SHA256_DIGEST*, unsigned int, char const*, unsigned int, unsigned char*, IOINTERNAL*, bool (*)(void const*, int, void*), void*) (demo.cpp:167) 18:14 <+bridge> [ddnet] ``` 18:14 <+bridge> [ddnet] check line 167 and you see that it's `aChunk` 18:14 <+bridge> [ddnet] oh true 18:14 <+bridge> [ddnet] it isnt initialized 18:15 <+bridge> [ddnet] I don't really see how it cannot be uninitialized though 18:15 <+bridge> [ddnet] I don't really see how it cannot be initialized though 18:15 <+bridge> [ddnet] hmm 18:15 <+bridge> [ddnet] maybe i should memzero it? 18:15 <+bridge> [ddnet] I don't really see how it can not be initialized though 18:15 <+bridge> [ddnet] no 18:15 <+bridge> [ddnet] > @heinrich5991 i think -Wno-nullability-completeness 18:15 <+bridge> [ddnet] @Ryozuki I tried that and there were even more other failures: https://github.com/ddnet/ddnet/pull/2124/files 18:16 <+bridge> [ddnet] @heinrich5991 then i suppose this should be ignored right? 18:16 <+bridge> [ddnet] @deen :o 18:18 <+bridge> [ddnet] @Ryozuki I don't think valgrind has false positives, so I'm kinda suprised here 18:18 <+bridge> [ddnet] well valgrind says it isnt perfect in their web xD 18:18 <+bridge> [ddnet] ok 18:18 <+bridge> [ddnet] link? 18:19 <+bridge> [ddnet] https://valgrind.org/docs/manual/quick-start.html 18:19 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/697480735266701392/unknown.png 18:19 <+bridge> [ddnet] in the caveeats section 18:19 <+bridge> [ddnet] broken encoding, nice 18:19 <+bridge> [ddnet] xd 18:20 <+bridge> [ddnet] maybe its not aChunk? 18:20 <+bridge> [ddnet] but only aChunk makes sense 18:21 <+bridge> [ddnet] well its definitly aChunk 18:21 <+bridge> [ddnet] cuz the top error is Syscall param write(buf) points to uninitialised byte(s 18:21 <+bridge> [ddnet] and io_write calls fwrite 18:22 <+bridge> [ddnet] @heinrich5991 why shouldnt we memzero it? does fwrite handle it? 18:22 <+bridge> [ddnet] fread should initialize all the bytes up to the number returned 18:23 <+bridge> [ddnet] fwrite should only read as far as the parameter goes 18:23 <+bridge> [ddnet] so there shouldn't™ be uninitialized bytes there 18:23 <+bridge> [ddnet] true 18:23 <+bridge> [ddnet] `mem_zero` would probably silence that warning, but it would do unnecessary work 18:23 <+bridge> [ddnet] yea 18:24 <+bridge> [ddnet] You could create a suppression file if we have actual false positives 18:24 <+bridge> [ddnet] then we don't have to keep looking into the same stuff again and again 18:24 <+bridge> [ddnet] and add it to the repository I mean 18:24 <+bridge> [ddnet] ok 18:24 <+bridge> [ddnet] i guess i have to rerun with -gen-suppressions=all 18:25 <+bridge> [ddnet] you could write the suppression file manually, but probably easier that way 18:44 <+bridge> [ddnet] it complains aBuf is uninitialized but i dont see how, m_pTooltip is initialized to 0 in the ctor 18:44 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/697486983370506300/unknown.png 18:44 <+bridge> [ddnet] is valgrind not that good? 18:46 <+bridge> [ddnet] @Ryozuki can you point me to the line in the valgrind-out.txt file? 18:47 <+bridge> [ddnet] notification from github: @Ryozuki pushed 0 commits. 18:47 <+bridge> [ddnet] i removed a commit i made 18:47 <+bridge> [ddnet] ah 18:47 <+bridge> [ddnet] cuz it was already initialized 18:47 <+bridge> [ddnet] how to print current line in less? 18:47 <+bridge> [ddnet] xD 18:48 <+bridge> [ddnet] idk 18:48 <+bridge> [ddnet] https://stackoverflow.com/a/831707 18:48 <+bridge> [ddnet] ah -N 18:48 <+bridge> [ddnet] > You can also toggle line numbers without quitting less by typing `-N`. 18:49 <+bridge> [ddnet] > You can also press `=` while `less` is open to just display (at the bottom of the screen) information about the current screen, including line numbers, with format: 18:50 <+bridge> [ddnet] 556 18:50 <+bridge> [ddnet] use the new file i uploaded with gen suppressions btw 18:51 <+bridge> [ddnet] valgrind maybe doesnt know what str_format does 18:51 <+bridge> [ddnet] it shouldn't care what `str_format` does, afaik it tracks which bytes have been initialized and which have not 18:54 <+bridge> [ddnet] well its either str_format or str_copy 18:54 <+bridge> [ddnet] or `str_utf8_decode` 18:55 <+bridge> [ddnet] or even `TextEx`? 18:56 <+bridge> [ddnet] all those functions just use ptext 18:56 <+bridge> [ddnet] which is aBuf 18:56 <+bridge> [ddnet] iirc 18:56 <+bridge> [ddnet] well, but if they read past the 0 byte, then they read uninit data 18:57 <+bridge> [ddnet] (I'm looking into it right now) 18:57 <+bridge> [ddnet] if this leaks it would be big right? its called every render call xD 18:57 <+bridge> [ddnet] this isn't mem leaks, but reading of uninit data 18:57 <+bridge> [ddnet] ah yeah 18:57 <+bridge> [ddnet] which could lead to ub? 18:58 <+bridge> [ddnet] which is UB 18:59 <+bridge> [ddnet] unlikely to cause problems that deep in the call stack, should still be fixed 18:59 <+bridge> [ddnet] well ill take a break now, ill come later at night maybe 18:59 <+bridge> [ddnet] I'll highlight you if I find something 19:00 <+bridge> [ddnet] ok 19:00 <+bridge> [ddnet] @Ryozuki ah, found it 19:00 <+bridge> [ddnet] line 920 in TextEx reads two characters ahead 19:01 <+bridge> [ddnet] i.e. one past the null byte 19:03 <+bridge> [ddnet] Ok send remind me it on git, i wont be at pc for a whilw 21:23 <+bridge> [ddnet] @heinrich5991 i guess this fixes it? 21:23 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/697527153633263666/unknown.png 21:24 <+bridge> [ddnet] hm wait 21:25 <+bridge> [ddnet] i guess this 21:25 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/697527566109769779/unknown.png 21:28 <+bridge> [ddnet] i think a check like this is also good 21:28 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/697528249521274910/unknown.png 21:29 <+bridge> [ddnet] https://github.com/ddnet/ddnet/pull/2128/commits/8f1cc4c83d3e88c3db33dd46ba3ecbbdc18b6932 21:33 <+bridge> [ddnet]