00:00:48 | *** | Saving seen data "./dancer.seen" |
00:57:02 | | Quit massiveH (Quit: Leaving) |
01:00 |
01:35:33 | | Join tchan1 [0] (~tchan@c-98-206-141-238.hsd1.il.comcast.net) |
01:37:51 | | Quit tchan1 (Read error: Connection reset by peer) |
01:38:36 | | Join tchan1 [0] (~tchan@c-98-206-141-238.hsd1.il.comcast.net) |
01:39:10 | | Quit tchan (Ping timeout: 272 seconds) |
01:39:12 | | Quit tchan1 (Read error: Connection reset by peer) |
01:40:30 | | Join tchan1 [0] (~tchan@c-98-206-141-238.hsd1.il.comcast.net) |
01:45:50 | | Quit tchan1 (Ping timeout: 246 seconds) |
02:00 |
02:00:49 | *** | Saving seen data "./dancer.seen" |
02:13:33 | | Join tchan1 [0] (~tchan@c-98-206-141-238.hsd1.il.comcast.net) |
02:22:15 | _bilgus_ | I think QOI will be a viable way to store in bin assets, I might try using it as a way to store static assets loaded from disk as well 37% space savings might add up pretty nicely with all the buflib allocs for assets |
02:23:45 | _bilgus_ | lcd_update tempers the benchmark with QOI losing 17% but without lcd update its about 80% slower clipzip unboosted |
02:26:49 | | Quit tchan1 (Ping timeout: 260 seconds) |
02:28:43 | _bilgus_ | 11.4k v 18.4k bytes 97.8fps v 116.2 fps for the former and 501fps v 2449fps for the latter |
02:29:38 | _bilgus_ | without update is pretty much raw decode vs copying data between buffers |
03:00 |
03:20:35 | | Join alex39 [0] (~alex@p5b0fbe19.dip0.t-ipconnect.de) |
03:47:09 | | Quit alex39 (Quit: Connection closed) |
03:59:28 | | Join alex12 [0] (~alex@p5b0fbe19.dip0.t-ipconnect.de) |
04:00 |
04:00:53 | *** | Saving seen data "./dancer.seen" |
04:09:16 | | Quit alex12 (Ping timeout: 272 seconds) |
06:00 |
06:00:57 | *** | Saving seen data "./dancer.seen" |
06:32:38 | | Join tchan [0] (~tchan@c-98-206-141-238.hsd1.il.comcast.net) |
06:35:21 | | Quit tchan (Client Quit) |
06:44:46 | | Join tchan [0] (~tchan@c-98-206-141-238.hsd1.il.comcast.net) |
06:58:16 | | Quit tchan (Read error: Connection reset by peer) |
06:59:08 | | Join tchan [0] (~tchan@c-98-206-141-238.hsd1.il.comcast.net) |
08:00 |
08:01:01 | *** | Saving seen data "./dancer.seen" |
08:07:16 | | Quit CH23_M (Read error: Connection reset by peer) |
08:07:56 | | Join CH23_M [0] (~CH23@revspace/participant/ch23) |
08:09:11 | | Quit tchan (Read error: Connection reset by peer) |
08:09:55 | | Join tchan [0] (~tchan@c-98-206-141-238.hsd1.il.comcast.net) |
08:21:52 | | Join alex85 [0] (~alex@p5b0fbe19.dip0.t-ipconnect.de) |
08:38:14 | | Join speachy [0] (~speachy@tunnel658658-pt.tunnel.tserv12.mia1.ipv6.he.net) |
08:38:15 | | Quit speachy (Changing host) |
08:38:15 | | Join speachy [0] (~speachy@rockbox/developer/speachy) |
08:38:15 | Mode | "#rockbox +v speachy" by ChanServ (ChanServ@services.libera.chat) |
08:39:14 | | Quit tchan (Ping timeout: 246 seconds) |
08:48:59 | | Quit alex85 (Ping timeout: 246 seconds) |
09:00 |
09:07:00 | | Join tchan [0] (~tchan@c-98-206-141-238.hsd1.il.comcast.net) |
10:00 |
10:01:03 | *** | Saving seen data "./dancer.seen" |
10:18:07 | | Quit CH23_M (Ping timeout: 260 seconds) |
10:19:31 | | Join CH23_M [0] (~CH23@revspace/participant/ch23) |
10:50:33 | | Quit CH23_M (Read error: Connection reset by peer) |
10:51:00 | | Join CH23_M [0] (~CH23@revspace/participant/ch23) |
11:00 |
11:13:00 | | Quit speachy (Quit: WeeChat 3.6) |
11:27:54 | | Quit tchan (Ping timeout: 260 seconds) |
11:40:10 | | Join tchan [0] (~tchan@c-98-206-141-238.hsd1.il.comcast.net) |
11:52:29 | | Join dconrad [0] (~dconrad@152.117.104.235) |
12:00 |
12:01:05 | *** | Saving seen data "./dancer.seen" |
12:06:01 | dconrad | so I'm giving it another try to build the compiler on macos, and I'm hitting a vague roadblock - I get a "CC_FOR_BUILD doesn't seem to work", but I can't seem to figure out what about it that it doesn't like |
12:07:10 | dconrad | it appears to be using the native gcc rather than the homebrew gnu gcc which I imagine is what it's supposed to be using, but I'm not sure how to make it use that (probably symlink something somewhere into the path?) |
12:07:36 | dconrad | so any ideas on what to look at? |
12:07:55 | dconrad | logs at https://www.dropbox.com/scl/fo/0903omcadsglsn1bhy7ze/h?dl=0&rlkey=i6uweowglftm4qnti7py2v6wh |
12:21:55 | | Join amachronic [0] (~amachroni@user/amachronic) |
12:47:29 | rb-bluebot | Build Server message: New build round started. Revision 6e794c9a2d, 303 builds, 8 clients. |
13:00 |
13:12:37 | rb-bluebot | Build Server message: Build round completed after 1509 seconds. |
13:12:40 | rb-bluebot | Build Server message: Revision 6e794c9a2d result: All green |
13:28:59 | | Quit amachronic (Quit: amachronic) |
14:00 |
14:01:08 | *** | Saving seen data "./dancer.seen" |
14:42:35 | hactar|ant | dconrad i tried the same a while back and i seem to remember having a similar issue, i was never able to resolve it |
14:42:41 | hactar|ant | lmk if you figure it out |
14:50:37 | dconrad | yeah if I ever get it to work I'll definitely do a forum post at least |
14:52:55 | | Join amachronic [0] (~amachroni@user/amachronic) |
15:00 |
15:00:18 | amachronic | dconrad the error is actually coming from gmp not gcc itself |
15:01:05 | dconrad | oh I see |
15:02:05 | dconrad | I'll look at the gmp config.log then |
15:02:48 | amachronic | i see in config.log |
15:02:50 | amachronic | configure:4238: gcc −−version >&5 |
15:02:55 | amachronic | Apple clang version 12.0.0 (clang-1200.0.32.29) |
15:03:36 | amachronic | i guess the configure script isn't picking up the correct gcc? |
15:04:01 | dconrad | that was kind of the conclusion I was starting to get to |
15:04:48 | dconrad | I think if I export CC="gcc-12" I can get it to use the homebrew-installed version, that was going to be the next thing to try I think |
15:05:20 | amachronic | or maybe if you prepend the homebrew gcc bindir to your PATH |
15:05:44 | amachronic | assuming there's a "gcc" symlink there and not just versioned binaries |
15:06:00 | dconrad | that's exactly the issue I think I ran into trying to do that |
15:06:23 | dconrad | I made a symlink gcc -> gcc-12 but that introduced new errors trying to build binutils |
15:07:56 | amachronic | i think you might need to do the c++ compiler too |
15:08:05 | amachronic | at some point gcc started using c++ internally |
15:08:12 | dconrad | ah, maybe that's the key |
15:08:36 | dconrad | I do see some errors/warnings in the gmp config.log, though they don't seem much more helpful |
15:09:56 | dconrad | If you wouldn't mind taking a look, its gmp_config.log in that dropbox folder |
15:10:22 | dconrad | I'm going to try like you say and symlinking both and see what happens |
15:17:44 | amachronic | well judging by the gmp config log it's choking on this |
15:17:50 | amachronic | conftest.c:4:3: error: implicitly declaring library function 'exit' with type 'void (int) __attribute__((noreturn))' [-Werror,-Wimplicit-function-declaration] |
15:19:31 | dconrad | an incompatible version issue, I guess? to google! |
15:20:06 | amachronic | in gmp 6.2.1 that's lying around my system they have the same test but replaced the exit(0) call with return 0 |
15:21:22 | amachronic | maybe rbdev.sh should use a newer gmp version |
15:24:31 | dconrad | yeah binutils didn't like the gcc/g++/etc, binutils didn't even build |
15:25:22 | dconrad | gcc (Homebrew GCC 12.2.0) 12.2.0 |
15:26:27 | amachronic | logs? |
15:26:51 | dconrad | one sec |
15:28:16 | dconrad | https://www.dropbox.com/scl/fo/5jtzpsurv5tzmqcjhna59/h?dl=0&rlkey=0zs0weonwu1484mvr1mgcxooh |
15:28:21 | amachronic | i'm wondering if something else needs symlinking |
15:29:00 | dconrad | so apparently I do have gmp 6.2.1 |
15:29:04 | amachronic | nice now it's isl not finding gmp.h :D |
15:30:12 | dconrad | fix one thing, another breaks! |
15:30:39 | amachronic | well where is gmp.h on your system |
15:30:48 | amachronic | it seems to be looking for the system copy |
15:30:49 | dconrad | unless it's using the native gmp rather than the homebrew one |
15:30:53 | dconrad | yeah hmm |
15:32:28 | dconrad | /usr/local/include/gmp.h |
15:32:28 | dconrad | /usr/local/include/gmpxx.h |
15:32:29 | dconrad | /usr/local/lib/libgmp.10.dylib |
15:32:30 | DBUG | Enqueued KICK dconrad |
15:32:30 | dconrad | /usr/local/lib/libgmp.a |
15:32:32 | dconrad | /usr/local/lib/libgmp.dylib |
15:32:34 | dconrad | /usr/local/lib/libgmpxx.4.dylib |
15:32:36 | dconrad | /usr/local/lib/libgmpxx.a |
15:32:38 | dconrad | /usr/local/lib/libgmpxx.dylib |
15:32:40 | dconrad | /usr/local/lib/pkgconfig/gmp.pc |
15:32:42 | dconrad | /usr/local/lib/pkgconfig/gmpxx.pc |
15:32:44 | dconrad | /usr/local/share/info/gmp.info |
15:32:46 | dconrad | /usr/local/share/info/gmp.info-1 |
15:32:48 | dconrad | /usr/local/share/info/gmp.info-2 |
15:32:50 | dconrad | oh god I hoped that wouldn't do exactly that, sorry |
15:32:57 | dconrad | those are where it's linked, anyway |
15:32:59 | amachronic | lol |
15:33:02 | amachronic | anyhow |
15:33:07 | amachronic | https://stackoverflow.com/questions/4980819/what-are-the-gcc-default-include-directories |
15:33:37 | amachronic | so run "gcc -xc -E -v -" with homebrew gcc |
15:33:48 | amachronic | i'm guessing it's not searching /usr/local/include |
15:37:28 | dconrad | I'm still trying to parse the output, but here it is: https://pastebin.com/DrF9KsbK |
15:38:48 | amachronic | the paths are at the bottom |
15:39:05 | amachronic | no /usr/local/include |
15:39:17 | dconrad | the "/include"s? |
15:40:03 | amachronic | yeah after the "search starts here" part |
15:40:07 | dconrad | ok I see |
15:40:44 | amachronic | so i guess you need to add the correct include dir somehow |
15:40:53 | dconrad | so I need to point it to the gmp headers and such, yeah |
15:42:03 | amachronic | the rbdev script overrides CFLAGS= |
15:42:10 | amachronic | but not CC= or CXX= |
15:42:25 | amachronic | so maybe CC="/path/to/gcc -I/usr/local/include" |
15:42:33 | amachronic | ditto for g++ |
15:44:12 | dconrad | yeah I can try that, stackexchange was suggesting setting C_INCLUDE_PATH or LIBRARY_PATH too |
15:46:31 | dconrad | alright, giving it another go with -I/usr/local/include |
15:52:09 | dconrad | something's still going wonky, I get: |
15:52:17 | dconrad | configure:4376: /usr/local/opt/gcc/bin/gcc -I/usr/local/include -V >&5 |
15:52:17 | dconrad | gcc: error: unrecognized command-line option '-V' |
15:53:55 | amachronic | pretty sure that's harmless |
15:54:24 | dconrad | logs: https://www.dropbox.com/scl/fo/kgl89z57akqtng60ydlgo/h?dl=0&rlkey=fhyqqipl2y4cjcypyallceryn |
15:54:40 | dconrad | shoot, I thought it meant something haha |
15:55:23 | amachronic | the exit 0 at the end of config.log means configure succeeded (I think) |
15:55:51 | dconrad | makes sense |
15:55:56 | amachronic | checking for main in -lgmp... no |
15:56:00 | dconrad | it still can't seem to find gmp though |
15:56:00 | amachronic | configure: error: gmp library not found |
15:56:30 | amachronic | you'll have to add the lib search path as well |
15:57:49 | dconrad | -L? |
15:58:06 | amachronic | yep |
15:59:46 | amachronic | the part to look for in the build log is make[1]: *** [configure-isl] Error 1 |
16:00 |
16:01:11 | *** | No seen item changed, no save performed. |
16:04:11 | rb-bluebot | Build Server message: New build round started. Revision 5f2ca6718d, 303 builds, 8 clients. |
16:07:34 | dconrad | so I'm noticing two things: in the binutils config.log, it appears to be configuring with −−with-gmp=/usr/local/opt/gmp, and this new build does fail on different things: [isl_val_gmp.lo] Error 1 |
16:08:00 | dconrad | presumably −−with-gmp= needs to line up with -I and -L? |
16:08:48 | amachronic | huh, I'd have thought if it gets −−with-gmp it would find the includes and libs from that |
16:09:10 | dconrad | you would think so, and they're there... no idea what version is there though |
16:09:35 | dconrad | appears to be 6.2.1 |
16:11:13 | amachronic | whats the error now |
16:11:35 | amachronic | like the actual error message |
16:12:43 | dconrad | looks like ISL? |
16:12:45 | dconrad | https://pastebin.com/4Xm27pG4 |
16:13:04 | | Quit jacobk (Ping timeout: 252 seconds) |
16:13:37 | dconrad | let me get the logs up |
16:15:20 | dconrad | https://www.dropbox.com/scl/fo/xptvw681mv2b5s46ca7lq/h?dl=0&rlkey=ocfparbo473svfigjylw3go39 |
16:15:38 | dconrad | not sure if I need to look at the include/library paths for that as well |
16:16:08 | amachronic | ah it looks like it's a conflict between the local copy of isl and the system copy |
16:17:23 | amachronic | probably a side effect of putting -I/usr/local/include first |
16:18:34 | dconrad | that makes sense, I also have homebrew isl linked into /usr/local/include and /usr/local/lib |
16:18:57 | dconrad | I guess I probably need to get all the configure paths lined up to that as well |
16:19:15 | amachronic | it's building its own copy of isl |
16:19:21 | amachronic | lemme check if that's been put there by rockboxdev.sh |
16:19:23 | dconrad | oh I see |
16:20:27 | amachronic | ok open up rockboxdev.sh and try this |
16:20:36 | amachronic | at "kludge to avoid having to install GMP, MPFR, MPC and ISL" |
16:20:52 | amachronic | remove the whole if .. block |
16:21:09 | amachronic | so it doesn't try to add local copies to the build dir |
16:21:43 | dconrad | the whole "if test -n "$needs_libs"; then? |
16:21:51 | amachronic | yep |
16:21:58 | amachronic | the idea is to force it to use the system libs instead |
16:22:18 | dconrad | hmm, ok |
16:23:48 | dconrad | no way this wasn't going to open a can of worms, huh? |
16:24:16 | amachronic | i'm not sure what effect using the newer libs with old gcc & binutils might have, but hopefully nothing bad will happen :) |
16:24:31 | dconrad | well I'm certainly willing to give it a shot! |
16:25:20 | dconrad | I guess I don't understand why this would crop up only on macos/homebrew though, would think this would be the same on linux too |
16:25:42 | dconrad | though I guess homebrew is always going to do weird things |
16:26:27 | amachronic | probably because it has to spell out paths for everything |
16:26:28 | rb-bluebot | Build Server message: Build round completed after 1337 seconds. |
16:26:29 | rb-bluebot | Build Server message: Revision 5f2ca6718d result: All green |
16:26:35 | amachronic | on linux it's more likely to all be in /usr |
16:26:49 | dconrad | well it made it past binutils |
16:27:22 | dconrad | if this works I'll have to go and try to pare it down to see what all is actually necessary too |
16:30:28 | amachronic | i'm going to be afk for a bit |
16:30:46 | dconrad | hmm, made it past binutils but gcc didn't build |
16:33:51 | dconrad | link to logs: https://www.dropbox.com/scl/fo/eyg2gq5k571b9m9eqbpw2/h?dl=0&rlkey=ui54yoxedy6gg9nun6sx6l5oy |
16:34:17 | dconrad | (don't feel like you have to keep going on this btw, I appreciate all the help anyways!) |
16:35:01 | dconrad | I thought it would be a simple fix |
16:36:36 | | Quit amachronic (Ping timeout: 272 seconds) |
16:53:30 | dconrad | well I don't want to jinx it but it I cleared out $CC and it's sure getting a lot farther than it was before |
16:53:35 | dconrad | knock on wood |
16:55:32 | dconrad | eh spoke too soon |
17:00 |
17:16:47 | | Join amachronic [0] (~amachroni@user/amachronic) |
17:21:12 | amachronic | dconrad in that latest log it looks like gmp is still in the gcc source dir?? |
17:21:21 | amachronic | ../../../gcc-4.9.4/gmp/gmp-impl.h:3371:40: error: expected ')' before 'L' |
17:22:08 | dconrad | hmm, let me see if I can find that file |
17:22:35 | dconrad | oh wait I didn't delete the extracted gcc dir |
17:22:38 | dconrad | duh |
17:22:51 | dconrad | I didn't realize that's where the local libraries were put |
17:23:05 | dconrad | let me try it again |
17:24:31 | dconrad | (with specified $CC) |
17:32:12 | dconrad | ok, it can't seem to find libint1.h |
17:32:58 | amachronic | libintl? |
17:33:09 | amachronic | libint1 seems to be for quantum mechanics :D |
17:33:40 | dconrad | ¯\_(ツ)_/¯ |
17:33:59 | dconrad | hey, we all knew rockbox is ahead of its time! |
17:35:15 | amachronic | heh |
17:35:21 | dconrad | new versions should be in the same folder as before now |
17:35:34 | amachronic | i swear there's an option to disable libintl |
17:35:50 | dconrad | in rockbox or gcc? |
17:35:56 | amachronic | in gcc |
17:36:12 | amachronic | −−disable-nls |
17:36:13 | dconrad | something rockboxdev.sh should set |
17:36:33 | dconrad | well I can add that to $CC |
17:37:06 | amachronic | no it needs to be in GCC's ./configure options |
17:37:14 | dconrad | oh dang |
17:40:16 | amachronic | hopefully you're almost there :) |
17:40:52 | amachronic | i have to go again now, probably won't be on again tonight |
17:41:11 | amachronic | but i'll check tomorrow to see if you had success |
17:42:03 | | Quit amachronic (Quit: amachronic) |
17:42:29 | dconrad | alright, I really appreciate your help |
18:00 |
18:01:14 | *** | Saving seen data "./dancer.seen" |
18:45:47 | | Join massiveH [0] (~massiveH@2600:4040:a992:a300:80e:97c2:b3a9:582a) |
19:00 |
19:23:05 | dconrad | amachronic (logs): I'm still stumbling over libintl - gcc says it can't find libintl.h, but it's clearly symlinked in /usr/local/include |
19:23:31 | dconrad | wondering if I'm not setting the include/library paths correctly |
19:25:12 | dconrad | also weirdly, "find /usr -name libintl* -print 2>/dev/null" (from a stackexchange thread about missing libintl.h on macos, apparently a common problem) can't find them either... wonder if that's related |
20:00 |
20:01:15 | *** | No seen item changed, no save performed. |
20:42:36 | | Quit dconrad (Remote host closed the connection) |
20:43:09 | | Join dconrad [0] (~dconrad@152.117.104.235) |
20:46:06 | | Join dconrad_ [0] (~dconrad@152.117.104.235) |
20:46:06 | | Quit dconrad (Read error: Connection reset by peer) |
21:00 |
21:22:35 | | Join CH230 [0] (~CH23@revspace/participant/ch23) |
21:23:16 | | Quit CH23_M (Ping timeout: 252 seconds) |
21:24:46 | | Quit CH23 (Ping timeout: 272 seconds) |
21:30:38 | | Join CH23_M [0] (~CH23@revspace/participant/ch23) |
22:00 |
22:01:17 | *** | Saving seen data "./dancer.seen" |
22:13:11 | | Quit othello7 (Quit: othello7) |
22:51:47 | | Quit CH23_M (Read error: Connection reset by peer) |
22:52:07 | | Join CH23_M [0] (~CH23@revspace/participant/ch23) |
23:00 |
23:03:35 | | Quit dconrad_ (Read error: Connection reset by peer) |
23:04:07 | | Join dconrad [0] (~dconrad@152.117.104.235) |
23:29:00 | dconrad | amachronic (logs): after letting updates churn for several hours, I'm hitting up against some random package pygobject blocking me from doing a reinstall of gettext, which... might fix libintl.h? Anyway, I think I'm running into a consequence of not being on the latest macos version, so I might run that tonight |
23:30:52 | dconrad | I'm sure there's some simple, obvious reason it can't find libintl, but for now it's baffling :o |
23:37:11 | | Quit m01 (Quit: Konversation terminated.) |
23:39:25 | | Join m01 [0] (~quassel@vps-b172b88b.vps.ovh.net) |
23:46:50 | | Quit dconrad (Remote host closed the connection) |
23:49:27 | | Join dconrad [0] (~dconrad@152.117.104.235) |
23:51:12 | | Join jacobk [0] (~quassel@2600:1700:5410:dae0:cca:b077:4c2e:bfbd) |