# 01:08:32 |
devcoder |
quits : Quit: devcoder |
# 01:28:33 |
sh4rm4 |
quits : Ping timeout: 276 seconds |
# 01:41:26 |
sh4rm4 |
joins #crosstool-ng |
# 04:53:39 |
sfan5|OFF |
is now known as: sfan5 |
# 04:55:25 |
sfan5 |
is now known as: sfan5|OFF |
# 06:36:05 |
sspiff |
joins #crosstool-ng |
# 07:27:21 |
sh4rm4 |
quits : Ping timeout: 276 seconds |
# 07:28:49 |
sh4rm4 |
joins #crosstool-ng |
# 07:58:41 |
smartin |
joins #crosstool-ng |
# 08:10:08 |
Net147 |
joins #crosstool-ng |
# 08:11:40 |
smartin |
quits : Quit: leaving |
# 08:13:36 |
smartin |
joins #crosstool-ng |
# 08:27:05 |
bhundven |
quits : Ping timeout: 268 seconds |
# 11:45:49 |
jf1976 |
quits : Quit: Leaving |
# 12:45:31 |
Net147 |
quits : Quit: HydraIRC -> http://www.hydrairc.com <- Would you like to know more? |
# 13:30:49 |
ccole |
quits : Read error: Connection reset by peer |
# 13:56:57 |
sfan5|OFF |
is now known as: sfan5 |
# 14:02:24 |
ccole |
joins #crosstool-ng |
# 14:06:01 |
alan_o |
joins #crosstool-ng |
# 14:56:44 |
smartin |
quits : Ping timeout: 252 seconds |
# 14:58:40 |
smartin |
joins #crosstool-ng |
# 15:30:04 |
codyps |
quits : Ping timeout: 260 seconds |
# 15:52:26 |
Kasreyn |
quits : Read error: Connection reset by peer |
# 15:56:24 |
codyps |
joins #crosstool-ng |
# 16:08:30 |
smartin |
quits : *.net *.split |
# 16:08:30 |
gotit |
quits : *.net *.split |
# 16:08:38 |
gotit |
joins #crosstool-ng |
# 16:09:57 |
smartin |
joins #crosstool-ng |
# 16:43:15 |
y_morin |
joins #crosstool-ng |
# 17:44:39 |
abique|work |
quits : Quit: Leaving |
# 17:56:49 |
Dyblast |
joins #crosstool-ng |
# 17:56:59 |
Dyblast |
Hi, |
# 17:57:15 |
Dyblast |
quits : Quit: Leaving. |
# 17:57:28 |
diorcety |
joins #crosstool-ng |
# 17:57:35 |
diorcety |
Hello, ! |
# 18:08:54 |
codyps |
quits : Ping timeout: 276 seconds |
# 18:25:22 |
codyps |
joins #crosstool-ng |
# 18:51:28 |
y_morin |
diorcety: Hello! :-) |
# 18:51:48 |
y_morin |
diorcety: I'm just here for a few minutes, I'll be back ~10:00pm GMT+2 |
# 18:51:58 |
EsbenH |
Yann overflow.... |
# 18:52:09 |
y_morin |
diorcety: sounds like we're going for a very nice compromise! :) |
# 18:52:14 |
y_morin |
EsbenH: Hehe! :-) |
# 18:52:36 |
y_morin |
EsbenH: BTW, where will you end up staying in Barcelona? |
# 18:53:08 |
EsbenH |
I found a mucho cheapo hotel 1 km from the conference |
# 18:53:28 |
EsbenH |
El Abrevadero |
# 18:54:01 |
y_morin |
EsbenH: then I can forget about you inviting us in your suite to drin beers on the terasse, I guess! :-)) |
# 18:54:50 |
EsbenH |
yes, that part of the plan drowned in corporate requirements for minimizing spending :-D |
# 18:55:20 |
y_morin |
got to go for a little while... |
# 18:56:07 |
diorcety |
EsbenH: Yann domination ! |
# 18:56:59 |
EsbenH |
hehe |
# 18:57:38 |
EsbenH |
diorcety: are you coming to ELC-E also? |
# 18:59:47 |
diorcety |
no :D .. i don't even know what is it :D |
# 19:01:13 |
EsbenH |
http://events.linuxfoundation.org/events/embedded-linux-conference-europe |
# 19:01:25 |
EsbenH |
Barcelona, 5-7 nov. |
# 19:02:05 |
EsbenH |
I think it is not to late to register. |
# 19:02:15 |
diorcety |
and i don't work in embedded |
# 19:05:50 |
codyps |
quits : Quit: Leaving. |
# 19:06:23 |
kos_tom |
EsbenH: hello! I'll be at ELCE as well. |
# 19:11:35 |
codyps |
joins #crosstool-ng |
# 19:29:12 |
devshark |
joins #crosstool-ng |
# 19:29:25 |
devshark |
looks like i'm in luck |
# 19:29:29 |
devshark |
yann's here. |
# 19:29:34 |
sfan5 |
is now known as: sfan5|OFF |
# 19:29:59 |
devshark |
or is everyone AFK again? |
# 19:30:07 |
diorcety |
devshark: he will come back at 10:00 GMT+2 |
# 19:30:17 |
devshark |
as in 30 mins? |
# 19:30:21 |
diorcety |
yes |
# 19:30:24 |
devshark |
oh good. |
# 19:30:25 |
devshark |
:D |
# 19:30:36 |
devshark |
maybe you can help me out until then? |
# 19:31:18 |
EsbenH |
kos_tom: where are you staying in Barcelona? |
# 19:32:11 |
kos_tom |
Fira Palace |
# 19:32:20 |
kos_tom |
but sharing a room with a colleague, to reduce costs |
# 19:33:08 |
EsbenH |
lucky you |
# 19:33:19 |
EsbenH |
it was fully booked long ago |
# 19:35:17 |
kos_tom |
I booked long ago |
# 19:45:38 |
devshark |
so... now that its quiet again |
# 19:45:49 |
devshark |
can i de-rail back to a more serious issue? |
# 19:46:23 |
EsbenH |
you can try ;-) |
# 19:47:47 |
y_morin |
devshark: speaking about ELC-E *is* a serious subject! ;-) |
# 19:48:29 |
y_morin |
devshark: which Yan do you want? diorcety, or y_morin? ;-) |
# 19:49:22 |
devshark |
the one from the mailing list. |
# 19:49:25 |
devshark |
i'm guessing you :) |
# 19:49:33 |
devshark |
because diorcety wasn't very helpful earlier. |
# 19:49:49 |
devshark |
so... what's ELC-E |
# 19:49:49 |
devshark |
? |
# 19:49:57 |
y_morin |
devshark: we all do what we can, given knowledge and time. Tiem especially. |
# 19:50:13 |
y_morin |
devshark: http://events.linuxfoundation.org/events/embedded-linux-conference-europe |
# 19:50:20 |
devshark |
oh. |
# 19:50:29 |
devshark |
hmmm.... intredasting. |
# 19:50:56 |
devshark |
well, i'd really love to re-ask my question |
# 19:51:02 |
devshark |
or ask the question i came here for. |
# 19:51:03 |
devshark |
but.... |
# 19:51:17 |
devshark |
it's still building. maybe i resolved it. |
# 19:51:25 |
devshark |
heh :) |
# 19:51:51 |
y_morin |
devshark: what issue? |
# 19:52:34 |
devshark |
i'm trying to build an arm CC |
# 19:52:46 |
devshark |
and PLL has been on my ass for the past 4-5 hours. |
# 19:52:58 |
devshark |
to sum it up shortly. |
# 19:53:15 |
y_morin |
devshark: You're Aleksandar? |
# 19:53:22 |
devshark |
yes :) |
# 19:53:23 |
codyps |
quits : Quit: Leaving. |
# 19:53:24 |
devshark |
omg [ERROR] >> [ERROR] >> Build failed in step 'Installing final compiler' [ERROR] >> called in step '(top-level)' |
# 19:53:42 |
devshark |
do you think a sudo will fix this? |
# 19:54:37 |
devshark |
hm no, it's a libiberty issue it seems. |
# 19:54:39 |
y_morin |
devshark: do *not* run as root. |
# 19:55:05 |
devshark |
oh ok. so then i just drop libiberty from the build and it might work? |
# 19:55:42 |
y_morin |
devshark: no, you need to try to _understand_ why it fails. |
# 19:55:58 |
devshark |
i understand why, it has an undefined reference. |
# 19:56:10 |
devshark |
what i do not understand (yet) is why i need libiberty on my target. |
# 19:56:15 |
devshark |
and why i even turned that option on. |
# 19:56:21 |
y_morin |
devshark: no, the real problem is something else. |
# 19:56:31 |
devshark |
i was afraid of that.... :) |
# 19:57:06 |
devshark |
[ERROR] /home/milenkovic/crosstool-ng/.build/src/gcc-4.4.6/libiberty/fibheap.c:258: error: 'LONG_MIN' undeclared (first use in this function) [ERROR] /home/milenkovic/crosstool-ng/.build/src/gcc-4.4.6/libiberty/fibheap.c:258: error: (Each undeclared identifier is reported only once [ERROR] /home/milenkovic/crosstool-ng/.build/src/gcc-4.4.6/libiberty/fibheap.c:258: error: for each function it appears in.) |
# 19:57:08 |
y_morin |
devshark: first, you do not provide your complete build.log. Without that, and the brief "it takes too long" expanation are not enough to understand the issue, much les to solve it. |
# 19:57:24 |
devshark |
no that "takes too long" issue has been resolved yesterday |
# 19:57:29 |
devshark |
my issues keep evolving. |
# 19:57:41 |
y_morin |
devshark: put your _complete_ build.log on some pastebin or such, so we can actualy _see_ what the problem is. |
# 19:58:03 |
y_morin |
devshark: Also, provide your .config file, so I can run it here. |
# 19:59:21 |
y_morin |
devshark: and do not post the build.log to the list, it's too big. |
# 20:00:31 |
devshark |
ok, where do i find the build.log? it's not in ctng root |
# 20:00:58 |
y_morin |
diorcety: OK, back to mingw: so, mingw64 already supports both 64- and 32-bit runtimes? |
# 20:01:11 |
y_morin |
devshark: the build.log *is* in the directory where you run ct-ng. |
# 20:01:33 |
EsbenH |
y_morin: yes, this is called mingw-w32 (as opposed to mingw32) |
# 20:02:07 |
y_morin |
EsbenH: Oh, that's very good! So, definitely, we *want* to get rid of our mingw32, and replace it with mingw. |
# 20:02:59 |
diorcety |
EsbenH: i686-w64-mingw32 no? |
# 20:02:59 |
y_morin |
has to admit that Windows stuff is definitely not his cup of tea. |
# 20:03:00 |
EsbenH |
mingw-w64 looks like a live fork of a mostly dead dead mingw32 project. |
# 20:03:14 |
diorcety |
y_morin: it supports multilib --enable-targets=x86_64-w64-mingw32,i686-w64-mingw32 |
# 20:03:24 |
EsbenH |
I don't know why the "w64" sneak in as vendor... |
# 20:03:48 |
diorcety |
As of 2009-05-31, x86_64-pc-mingw32 has been replaced by x86_64-w64-mingw32. |
# 20:03:59 |
EsbenH |
I prefer (or rather, have standardized on) using the vendor part of the arch tuple for cpu specific details. |
# 20:04:02 |
diorcety |
The change allows for mingw-w64 to have a host/target type independent of mingw.org's MinGW. |
# 20:04:12 |
y_morin |
diorcety: not even speaking about multilib. But using the same source, we can build either a i686-pc-mingw32 *or* a x86_64-pc-mingw32 toolchains? |
# 20:04:28 |
diorcety |
y_morin: i work on it |
# 20:04:48 |
y_morin |
diorcety: Nice! :-) |
# 20:04:54 |
y_morin |
diorcety: do you mean that the vendor part of the tuple *is* meaningful? |
# 20:05:33 |
diorcety |
y_morin: the target (mingw32) is important. I have try with mingw64 ... the configure fails |
# 20:05:55 |
diorcety |
i will try with another vendor |
# 20:06:25 |
y_morin |
diorcety: OK, that sounds weird, but that's windows stuff. The vendor part is the second part between dashes, eg. 'pc' in 'i686-pc-mingw32'. |
# 20:06:29 |
y_morin |
OK |
# 20:08:07 |
EsbenH |
current (newest) config.guess/config.sub does have some *-pc-mingw64 support there |
# 20:09:01 |
diorcety |
i have start the compilation with unknown vendor ... |
# 20:09:34 |
codyps |
joins #crosstool-ng |
# 20:11:03 |
devshark |
bah, You have exceeded the maximum file size of 500 kilobytes per paste. PRO users don't have this limit! |
# 20:11:29 |
devshark |
but you were right |
# 20:11:45 |
devshark |
this one is really serious.... internal compiler segmentation fault. |
# 20:12:27 |
devshark |
would something trivial liek make-cleaning do something... ? |
# 20:12:32 |
y_morin |
diorcety, EsbenH: http://code.bulix.org/7uunla-82357 |
# 20:15:35 |
diorcety |
y_morin: by the way, how use ~/src directory |
# 20:15:40 |
diorcety |
? |
# 20:15:58 |
y_morin |
diorcety: you mean ~/src has all your tarballs? |
# 20:16:00 |
EsbenH |
y_morin: yes, that is what I see in config.sub/config.guess also |
# 20:17:19 |
diorcety |
y_morin: yes, it's painfull to redownload all src when i make a distclean |
# 20:18:00 |
y_morin |
diorcety: Go in menuconfig, in "Paths and misc options" there's "Local tarball dir" (or smthg like that) |
# 20:18:22 |
y_morin |
diorcety: Enter (with no quotes): ${HOME}/src |
# 20:18:38 |
y_morin |
Then you can also say 'y' to "save to local ytarball dir". |
# 20:19:42 |
y_morin |
devshark: at least, paste your .config, so I can give it a psin here to see if I can at least reproduce the issue. |
# 20:19:52 |
y_morin |
psin -> spin |
# 20:20:11 |
y_morin |
GTG, BBL... |
# 20:23:59 |
devshark |
jno wait |
# 20:24:07 |
devshark |
i'm pasting the buildlog. |
# 20:24:10 |
devshark |
also the config too. |
# 20:24:30 |
devshark |
i'm back to the long_min libiberty issue. |
# 20:24:35 |
devshark |
so thats comforting. |
# 20:28:45 |
devshark |
and hey - the pastebin is ready: http://paste.ubuntu.com/1303525/ |
# 20:32:30 |
y_morin |
Sigh... paste.ubuntu.com requires a launchpad login to just downld the stuff as raw text... :-( Tsss.... |
# 20:34:26 |
y_morin |
devshark: I don't see any error in that build.log. It seems it's not complete... |
# 20:35:11 |
y_morin |
Wonderfull. pate.ubuntu.com freezed my browser, now... :-( |
# 20:46:46 |
devshark_ |
joins #crosstool-ng |
# 20:47:33 |
devshark |
quits : Ping timeout: 245 seconds |
# 20:50:10 |
devshark_ |
@ymorin: here's the latest build.log http://paste.ubuntu.com/1303575/ ... also I didn't really expect a response on the same day, though i saw it coming this way. |
# 20:52:16 |
devshark_ |
and here's the config file - http://paste.ubuntu.com/1303581/ |
# 20:53:48 |
y_morin |
devshark_: we don't need the config.log, but the .config file (hidden by default, to see it: ls -A ) |
# 20:54:42 |
devshark_ |
oh pasted the wrong one i guess. |
# 20:54:51 |
y_morin |
;-) |
# 20:55:13 |
devshark_ |
i'm still in the office, i'm sure you'll find it in your heart to forgive me.... :) |
# 20:55:57 |
devshark_ |
http://paste.ubuntu.com/1303592/ there we go |
# 20:56:16 |
devshark_ |
also, i just 'confirmed' that what i'm seeing is somewhat of a known issue. |
# 20:57:22 |
devshark_ |
getpwd.c:75: error: storage size of 'dotstat' isn't known - this is clearly libc |
# 20:58:21 |
devshark_ |
maybe I can try with newer gcc? |
# 20:58:35 |
devshark_ |
this config is using gcc 4.4.6 |
# 20:58:38 |
y_morin |
devshark_: dud you try to first build one of the bundled samples? |
# 20:59:09 |
devshark_ |
no, because by the time I noticed them I was already knees deep strugling with this one. |
# 20:59:30 |
devshark_ |
and then i just got stubborn. |
# 20:59:49 |
devshark_ |
but i guess it'll prove if i can build anything or not. |
# 21:00:40 |
devshark_ |
so i:ll try to build arm-unknown-linux-gnueabi |
# 21:00:42 |
devshark_ |
from samples. |
# 21:00:54 |
devshark_ |
which may as well be what i was trying to build all day. |
# 21:00:59 |
y_morin |
devshark_: yes. I know that the samples do build (at least on my machine!), so it is a good start. |
# 21:02:09 |
devshark_ |
[ERROR] Missing: 'x86_64-unknown-linux-gnu-gcj' or 'x86_64-unknown-linux-gnu-gcj' or 'gcj' : either needed! |
# 21:02:11 |
devshark_ |
bummer. |
# 21:02:56 |
y_morin |
devshark_: just disable building java for a starter. Or isntall a native Java compiler. |
# 21:03:34 |
devshark_ |
oh i know what you mean, menuconfig -> languages |
# 21:04:13 |
devshark_ |
the default one had fortran and java in it as well. |
# 21:06:57 |
diorcety |
y_morin: seems to compile with another vendor |
# 21:07:14 |
y_morin |
diorcety: OK, that's good :-) |
# 21:07:21 |
diorcety |
y_morin: directx for mingw32 (actual) is a joke no? |
# 21:08:17 |
y_morin |
diorcety: Hmmm. I don't know. I do not use it. I was against adding it, but people said that the directx (and others) were really part of the 'system' (whatever that be on winblows). |
# 21:08:37 |
y_morin |
diorcety: I would not be too sad if it were to disapear. |
# 21:08:52 |
y_morin |
diorcety: especially if we can get rid of the libc_finish step |
# 21:09:02 |
diorcety |
its very imcomplete |
# 21:09:03 |
diorcety |
5 headers |
# 21:09:08 |
diorcety |
no standard names |
# 21:09:29 |
diorcety |
same for opengl |
# 21:09:41 |
diorcety |
2 headers |
# 21:10:58 |
y_morin |
diorcety: OK, so, just send a patch that nukes them. |
# 21:11:12 |
y_morin |
;-) |
# 21:11:28 |
diorcety |
it compiles but i can't compile my directx test application ... i succeed before ... i have to check if it isn't coming from this vendor stuff |
# 21:11:40 |
diorcety |
another question |
# 21:12:10 |
diorcety |
we can use CT_CC_STATIC_LIBSTDCXX which means that the target lib std c++ is static, right ? |
# 21:12:50 |
y_morin |
diorcety: nope, it statically links the compiler against the host libstdc++ |
# 21:13:30 |
devshark_ |
you really did a number on this script though. gj man. |
# 21:13:45 |
y_morin |
devshark_: Thanks! :-) |
# 21:14:16 |
devshark_ |
i give credit where credit is due... |
# 21:14:17 |
diorcety |
y_morin: hum so the toolchain compiler have libstdc++ integrated? |
# 21:14:24 |
diorcety |
has |
# 21:16:06 |
y_morin |
diorcety: not really. If you link your compiler binary statically against libstdc++, in a way it is embedded, yes, as it's statically linked. But the shared *host* lib is not. |
# 21:17:12 |
diorcety |
the *build* so ? |
# 21:19:27 |
diorcety |
for make clear (to me) does it impact the target system ? |
# 21:19:42 |
devshark_ |
omg omg [INFO ] Build completed at 20121024.231819 [INFO ] (elapsed: 14:17.65) how do i use it now? :D |
# 21:23:53 |
y_morin |
diorcety: no impact on the target system. |
# 21:24:12 |
y_morin |
devshark_: look in the docs/ sub-dir, there a chapter on how to use the toolchain. |
# 21:24:34 |
devshark_ |
awwww... i tried compiling a printf*"hello world"); and got this. |
# 21:24:45 |
devshark_ |
[milenkovic@mcs02 bin]$ arm-linux-gnueabi-gcc -o test test.c test.c:1:19: fatal error: stdio.h: No such file or directory compilation terminated. |
# 21:25:03 |
devshark_ |
is that bad or really bad? |
# 21:26:32 |
devshark_ |
no wait, that s the wrong one. |
# 21:26:36 |
devshark_ |
there might still be hope |
# 21:27:00 |
y_morin |
devshark_: you mean, you have another arm compiler in your PATH? |
# 21:27:11 |
diorcety |
y_morin: thanks ... so EsbenH has misunderstood this stuff :) |
# 21:28:14 |
y_morin |
diorcety: host machine: the machine on which the compilers _runs_. If you select CT_CC_STATIC_LIBSTDCXX, the compiler will be statically linked against libstdc++. |
# 21:28:34 |
y_morin |
It does not impact what _target_ libraries are being built, nor how they are built. |
# 21:32:51 |
devshark_ |
yea, i navigated to tools/bin and invoked the wrong one, the arm-unknown-linux-gnueabi made a binary which i'm about to test . |
# 21:33:20 |
devshark_ |
7 hours of fucking with this shit got me all excited. it'd be even better if it ends up working :) |
# 21:34:07 |
y_morin |
devshark_: "Heaven is waiting, and waiting is Hell!" ;-) |
# 21:34:45 |
diorcety |
y_morin: ok |
# 21:34:51 |
diorcety |
y_morin: another vendor works :) |
# 21:53:04 |
devshark_ |
it works. |
# 21:53:56 |
devshark_ |
thanks for all the help and effort, i really appreciate it. |
# 21:54:04 |
devshark_ |
i'm gonna go drink one in your name. |
# 21:54:09 |
devshark_ |
cheers :) |
# 21:54:22 |
y_morin |
devshark_: Good! That gives you a known-good starting point, if you want to tweak the toolchain |
# 21:54:27 |
y_morin |
devshark_: Cheers! ;-) |
# 21:54:33 |
devshark_ |
i probably will tomorrow. |
# 21:54:41 |
devshark_ |
but even this gave decent results. |
# 21:54:53 |
devshark_ |
well preliminary results, who knows if softfp works and all that jazz |
# 21:54:59 |
devshark_ |
but i'll get it running |
# 21:55:45 |
devshark_ |
and i learned how to deal with bullshit that comes your way while using this to build toolchains. now that i actually know how to use it properly after two days it's really a great tool. |
# 22:15:06 |
diorcety |
y_morin: still here? |
# 22:15:22 |
y_morin |
diorcety: yes (but not for too long, I think) |
# 22:17:30 |
diorcety |
it almost works ... it's fail to final cc with "Build GCC with plugin support requires a host that supports" |
# 22:18:47 |
diorcety |
strange stuff ... it works with x86-64 ... |
# 22:20:21 |
y_morin |
diorcety: Hmm. Too late here to think straight. We can live with gcc not having plugins on mingw-w32, I guess. |
# 22:22:17 |
diorcety |
x86_64-unknown-mingw32 works, i686-unknown-mingw32 not (the only change is that, same code) |
# 22:22:24 |
diorcety |
i will investigate |
# 22:22:38 |
y_morin |
diorcety: thanks! |
# 22:22:53 |
diorcety |
objdump: conftest: File format not recognized |
# 22:22:57 |
diorcety |
hum ... |
# 22:25:10 |
smartin |
quits : Quit: heading home |
# 22:38:57 |
y_morin |
diorcety: I'm off to bed in a few minutes. Do you think you'll have a patch ready before the end of the week-end? |
# 22:40:27 |
diorcety |
yes |
# 22:40:42 |
diorcety |
y_morin: it seems a bug in gcc 4.5.1 |
# 22:40:46 |
diorcety |
and before |
# 22:40:50 |
diorcety |
fixed in gcc 4.5.2 |
# 22:40:51 |
diorcety |
i will try |
# 22:41:19 |
diorcety |
normally a clean patch tomorrow ... |
# 22:41:23 |
diorcety |
heu today :D |
# 22:43:04 |
y_morin |
diorcety: You're in France, too? |
# 22:43:50 |
y_morin |
diorcety: I don't mind if it's a bug in gcc. Just be sure your sample uses a compiler that works. |
# 22:44:22 |
y_morin |
diorcety: OK, I'm off. See^WRead you tomorow! :-) |
# 22:44:38 |
y_morin |
quits : Quit: Nighty Night! |
# 23:14:17 |
bhundven |
joins #crosstool-ng |