# 00:40:56 |
nandub |
quits : Quit: WeeChat 0.4.2 |
# 00:41:13 |
nandub |
joins #crosstool-ng |
# 01:10:23 |
dFence |
joins #crosstool-ng |
# 03:19:53 |
dFence |
quits : Quit: dFence |
# 03:20:19 |
dFence |
joins #crosstool-ng |
# 03:20:24 |
dFence |
will do, thanks guys |
# 03:35:15 |
nandub |
quits : Quit: WeeChat 0.4.2 |
# 03:35:33 |
nandub |
joins #crosstool-ng |
# 04:07:08 |
perr |
joins #crosstool-ng |
# 04:14:40 |
perr |
quits : Quit: Leaving |
# 04:47:24 |
nandub |
quits : Quit: WeeChat 0.4.2 |
# 04:48:48 |
nandub |
joins #crosstool-ng |
# 04:52:11 |
nandub |
quits : Client Quit |
# 04:52:27 |
nandub |
joins #crosstool-ng |
# 05:01:09 |
smartin |
quits : Ping timeout: 256 seconds |
# 05:01:16 |
nandub |
quits : Quit: WeeChat 0.4.2 |
# 05:01:36 |
nandub |
joins #crosstool-ng |
# 05:01:44 |
nandub |
quits : Client Quit |
# 05:02:51 |
nandub |
joins #crosstool-ng |
# 05:04:02 |
smartin |
joins #crosstool-ng |
# 05:19:19 |
nandub |
quits : Quit: WeeChat 0.4.2 |
# 05:24:16 |
daggs1-work |
joins #crosstool-ng |
# 05:27:21 |
alan_o |
quits : Remote host closed the connection |
# 05:34:47 |
smartin |
quits : Ping timeout: 255 seconds |
# 05:37:57 |
smartin |
joins #crosstool-ng |
# 06:23:53 |
sh4rm4 |
quits : Ping timeout: 256 seconds |
# 06:29:13 |
daggs1-work |
bhundven, patches 0001-add_missing_signal_limits.patch and 0002-add_missing_ISbit_defs.patch should exist for musl-1.1.4 too |
# 06:32:54 |
daggs1-work |
bhundven, I think the right place is in musl build |
# 06:38:41 |
smartin_ |
joins #crosstool-ng |
# 06:48:19 |
voltagex |
quits : Quit: WeeChat 0.4.3 |
# 06:53:54 |
bhundven |
daggs1-work: I'm currently working on getting the musl gcc patches upstream. If you have patches for musl, the best place for them is in #musl. |
# 06:54:30 |
bhundven |
I really don't want to add more patches to crosstool-ng. |
# 06:57:30 |
daggs1-work |
bhundven, I understand, already asked there on the missing defines, I'm waiting for a reply |
# 06:57:49 |
bhundven |
Cool! |
# 06:58:27 |
bhundven |
I'll have updates tomorrow evening. Today was my "day off" |
# 06:58:54 |
daggs1-work |
enjoy it :) |
# 07:17:40 |
perr |
joins #crosstool-ng |
# 07:17:40 |
perr |
quits : Changing host |
# 07:17:40 |
perr |
joins #crosstool-ng |
# 07:18:42 |
y_morin |
joins #crosstool-ng |
# 07:27:50 |
perr |
quits : Quit: Leaving |
# 07:28:10 |
perr |
joins #crosstool-ng |
# 07:41:00 |
voltagex |
joins #crosstool-ng |
# 07:44:04 |
voltagex |
parts #crosstool-ng |
# 07:44:37 |
voltagex |
joins #crosstool-ng |
# 07:45:00 |
voltagex |
quits : Changing host |
# 07:45:01 |
voltagex |
joins #crosstool-ng |
# 07:50:41 |
perr |
quits : Ping timeout: 255 seconds |
# 07:51:52 |
perr |
joins #crosstool-ng |
# 07:59:26 |
perr |
quits : Quit: Leaving |
# 08:04:36 |
daggs1-work |
bhundven, sorry for bugging you, latest head doesn't get build on x64 with musl, I know you don't like to patch the musl lib bu that is the only way I can see, the patches doesn't get applied properly and it fails compilation |
# 08:10:26 |
daggs123_ |
joins #crosstool-ng |
# 08:12:25 |
daggs1-work |
quits : Ping timeout: 255 seconds |
# 08:20:05 |
daggs123_ |
is now known as: daggs-work |
# 08:20:12 |
daggs-work |
is now known as: daggs1-work |
# 09:37:35 |
daggs1-work |
quits : Quit: Leaving |
# 10:11:30 |
maxime_ |
is now known as: maxime |
# 10:11:34 |
maxime |
quits : Changing host |
# 10:11:35 |
maxime |
joins #crosstool-ng |
# 10:13:29 |
daggs1-work |
joins #crosstool-ng |
# 10:32:52 |
doc2 |
joins #crosstool-ng |
# 11:25:50 |
dFence |
gmornin |
# 11:33:20 |
sh4rm4 |
joins #crosstool-ng |
# 12:25:09 |
dFence |
can I manually trigger ct-ng build steps? |
# 12:25:54 |
dFence |
Iâm trying to patch binutils to compile with gold on OSX and that requires some patching, so I donât wanna have to go through the entire build chain over and over again |
# 12:43:06 |
sh4rm4 |
gold is just a toy |
# 12:43:41 |
sh4rm4 |
https://sourceware.org/bugzilla/show_bug.cgi?id=16728 |
# 13:33:28 |
perr |
joins #crosstool-ng |
# 13:54:35 |
perr |
quits : Quit: Leaving |
# 14:00:07 |
diorcety |
joins #crosstool-ng |
# 14:14:17 |
dFence |
muahaha okay, thought as much ;D |
# 14:14:37 |
dFence |
alright - got a couple of small patches for ct-ng to make sure it compiles on OSX |
# 14:14:49 |
dFence |
whatâs the preferred method of submitting a patch for ct-ng? |
# 14:42:54 |
smartin_ |
quits : Quit: brb |
# 14:43:01 |
sh4rm4 |
mailing list ,i'd guess |
# 14:48:15 |
y_morin |
dFence: 'git send-email' to the mailing list, indeed. |
# 14:55:39 |
sh4rm4 |
what if git send-email does not work ? it seems to be a fragile perl script depending on some fragile perl libs (git,svn,ssl,smtp,...) |
# 14:56:04 |
sh4rm4 |
in my case, it doesn't work due to this |
# 14:56:40 |
y_morin |
sh4rm4: git format-patch; and then send the mails. |
# 14:56:59 |
y_morin |
sh4rm4: But if git-send-email is broken, you should report upstream so it gets fixed |
# 14:57:55 |
y_morin |
sh4rm4: Besides, I am a bit surprised it is broken, given the huge number of people using it in various situations. If it was, then I know some Linus T. would have alrady barked at it... |
# 14:58:59 |
sh4rm4 |
i'm not surprised at all, but that may be due to the fact that i spent 2 hours to get the hideous svn/apr build tools generate the svn perl binding with swig |
# 14:59:23 |
dFence |
y_morin: crossgcc@sourceware.org? |
# 14:59:33 |
y_morin |
dFence: Yes. And Cc me. |
# 15:00:36 |
y_morin |
dFence: http://crosstool-ng.org/#contacts |
# 15:08:00 |
dFence |
done |
# 15:08:21 |
dFence |
hope it came through alright - itâs been a while since I submitted patches ;D |
# 15:08:54 |
y_morin |
dFence: I got three mails: a cover-letter and two patches. Is that it? |
# 15:08:59 |
dFence |
yup |
# 15:09:05 |
y_morin |
dFence: Then all good! ;-) |
# 15:09:09 |
dFence |
wohoo |
# 15:09:30 |
y_morin |
dFence: I'll look at it by the evening (GMT+2 here) |
# 15:09:47 |
dFence |
thx |
# 15:17:51 |
sh4rm4 |
the gperf patch looks legit |
# 15:18:01 |
sh4rm4 |
but the other one not so much |
# 15:18:24 |
sh4rm4 |
if OSX features libintl support in the libc, it should also ship the header, no ? |
# 15:18:58 |
sh4rm4 |
btw it might be a good idea to get the gperf patch into glibc upstream |
# 16:08:24 |
dFence |
sh4rm4: OSX from what I can see, libintl.h is only required for some typedefs which are in other headers on OSX |
# 16:09:43 |
sh4rm4 |
the post you linked to doesnt remove the inclusion tho |
# 16:10:23 |
sh4rm4 |
instead, it adds mac's -lintl to LDFLAGS |
# 16:10:28 |
dFence |
nope, it works together. The post is to skip the redefinition and my patch is to get rid of the âcouldnât find libintl" |
# 16:10:29 |
sh4rm4 |
which seems dubious |
# 16:11:03 |
sh4rm4 |
ah, so the bug is that mac headers pollute the namespace |
# 16:11:10 |
dFence |
yup |
# 16:11:13 |
sh4rm4 |
or maybe that host headers are used at all |
# 16:11:16 |
dFence |
⦠I think? ;D |
# 16:11:46 |
sh4rm4 |
i'd guess as soon as host headers are mixed in, the glibc build will be spoiled |
# 16:12:02 |
sh4rm4 |
so whatever step is done there should use -nostdinc |
# 16:13:43 |
dFence |
hard to tell â itâs one of those âworks for meâ¢â kinda situations |
# 16:14:08 |
sh4rm4 |
when you undo that patch, what's the exact error you're getting ? |
# 16:15:12 |
sh4rm4 |
having #ifdef __APPLE__ in glibc files seems silly.. because it should *never* touch the mac's headers |
# 16:16:29 |
dFence |
I had to remove the build-folder, ran out of space on the image. Iâll give it another spin tonight/tomorrow, then can give you the exact error message |
# 16:16:37 |
sh4rm4 |
nice |
# 16:16:47 |
sh4rm4 |
and pls save the full log for context |
# 16:17:04 |
alan_o |
joins #crosstool-ng |
# 16:17:06 |
dFence |
Basically, it started of with âlibintl.h not foundâ and then clang just lost its shit b/c of the typedefs (something __u_char64 methinks and a bunch of others) |
# 16:18:04 |
dFence |
so yeah, maybe only the 2 defines should do the trick, but Iâve been trying to get this to work for 3 days now, so I didnât want to take any chances ;) |
# 16:18:51 |
sh4rm4 |
if i'm correct and your host headers have been used, you glibc is broken anyway |
# 16:19:25 |
sh4rm4 |
s/you/your/ |
# 16:19:32 |
sh4rm4 |
lol |
# 16:19:44 |
sh4rm4 |
s/you /your / |
# 16:20:30 |
sh4rm4 |
s/lol/if i'm correct and your host headers have been used, your glibc is broken anyway |
# 16:20:45 |
sh4rm4 |
sigh |
# 16:21:03 |
sh4rm4 |
s/sigh/if i'm correct and your host headers have been used, your glibc is broken anyway/ |
# 16:22:12 |
sh4rm4 |
for example, imagine that you got the mac's struct stat definition instead of glibc's |
# 16:22:48 |
sh4rm4 |
so if it's smaller the kernel may just overwrite your process' RAM with random junk |
# 16:23:49 |
dFence |
good point. hooray. |
# 16:25:23 |
dFence |
so youâre saying -nostdinc is the way to go? |
# 16:25:54 |
sh4rm4 |
probably so |
# 16:26:18 |
sh4rm4 |
but i'd like to know the context - what it was trying to build, and which commandline it used |
# 16:27:04 |
sh4rm4 |
and ultimately, if the command line looks proper, the output of $CC -v with the exact same command (so one can see which include dirs are used) |
# 16:27:41 |
sh4rm4 |
(built-in includedirs) |
# 16:32:36 |
dFence |
gotcha |
# 16:33:10 |
dFence |
Iâll get first hand results in a couple minutes, trying to boot the first ccâd kernel. If the tablet blows up, Iâll know I fucked up. |
# 16:33:41 |
sh4rm4 |
yes, if you're lucky it blows up immediately |
# 16:34:11 |
sh4rm4 |
memory corruption is not always leading to straight crashes tho... |
# 16:35:25 |
sh4rm4 |
the kernel itself will work |
# 16:35:31 |
sh4rm4 |
it doesnt use glibc |
# 17:32:02 |
mingwandroid |
joins #crosstool-ng |
# 17:44:15 |
dFence |
buildroot is impossible to get it to work on OSX, using my current initrd from the debian machine - should be fine⦠I hope |
# 17:50:14 |
daggs1-work |
y_morin, there? |
# 17:54:07 |
y_morin |
daggs1-work: Yep, but not for long (almost dinner here) |
# 17:55:35 |
maxime_ |
joins #crosstool-ng |
# 17:57:33 |
maxime |
quits : Ping timeout: 240 seconds |
# 17:58:44 |
daggs1-work |
y_morin, I need a another eye on this, can you take a look at this file? patches/experimental/gcc/4.6.0/0001-musl.patch in musl tree |
# 17:58:50 |
daggs1-work |
does it seems correct to you? |
# 17:59:16 |
y_morin |
daggs1-work: Full url, please? |
# 17:59:25 |
y_morin |
(can't recall...) |
# 18:00:00 |
daggs1-work |
y_morin, tried but I have to idea how to access the cgit in that page |
# 18:00:24 |
djerome |
joins #crosstool-ng |
# 18:00:43 |
y_morin |
daggs1-work: On bhundven's tree? |
# 18:00:59 |
y_morin |
on bitbucket? |
# 18:01:00 |
daggs1-work |
yes |
# 18:01:35 |
daggs1-work |
no idea how to navigate bitbucket |
# 18:02:21 |
y_morin |
daggs1-work: What do you want me to look at in this patch? |
# 18:03:17 |
daggs1-work |
it seems that the patch is a patch of a patch, not a patch of a file |
# 18:03:57 |
daggs1-work |
I've took a look on the gcc 4.8.3 one an it is patch of files. |
# 18:04:08 |
y_morin |
daggs1-work: Ah, yes... |
# 18:04:18 |
y_morin |
daggs1-work: Sorry, I got to go now (I was called for dinner!) |
# 18:04:25 |
y_morin |
daggs1-work: I'll be back later tonight... |
# 18:04:29 |
daggs1-work |
enjoy |
# 18:04:36 |
y_morin |
sh4rm4: Thanks! |
# 18:05:39 |
sh4rm4 |
hm? |
# 18:06:19 |
maxime |
joins #crosstool-ng |
# 18:09:35 |
maxime_ |
quits : Ping timeout: 264 seconds |
# 18:25:36 |
smartin_ |
joins #crosstool-ng |
# 18:36:48 |
sh4rm4 |
quits : Remote host closed the connection |
# 18:37:55 |
sh4rm4 |
joins #crosstool-ng |
# 18:50:24 |
y_morin |
sh4rm4: orry, that was for daggs1-work... |
# 18:50:27 |
y_morin |
daggs1-work: Thanks! |
# 18:51:30 |
daggs1-work |
np |
# 19:50:36 |
dFence |
diorcety: can you send me a link? |
# 20:03:40 |
daggs1-work |
bhundven, this patch fixes errors, after that the cc is created. http://pastebin.com/qJmmaa7D |
# 20:05:58 |
sh4rm4 |
quits : Remote host closed the connection |
# 20:20:39 |
sh4rm4 |
joins #crosstool-ng |
# 20:20:41 |
doc2 |
quits : Remote host closed the connection |
# 20:29:16 |
sh4rm4 |
quits : Remote host closed the connection |
# 20:40:04 |
sh4rm4 |
joins #crosstool-ng |
# 20:42:26 |
sh4rm4 |
quits : Remote host closed the connection |
# 20:47:10 |
sh4rm4 |
joins #crosstool-ng |
# 21:09:01 |
sh[4]rm4 |
joins #crosstool-ng |
# 21:11:17 |
sh4rm4 |
quits : Ping timeout: 256 seconds |
# 21:11:43 |
sh[4]rm4 |
is now known as: sh4rm4 |
# 21:12:38 |
sh4rm4 |
quits : Remote host closed the connection |
# 21:17:23 |
sh4rm4 |
joins #crosstool-ng |
# 21:20:08 |
djerome |
quits : Remote host closed the connection |
# 21:41:46 |
sh4rm4 |
quits : Remote host closed the connection |
# 21:42:13 |
sh4rm4 |
joins #crosstool-ng |
# 21:44:39 |
sh4rm4 |
quits : Remote host closed the connection |
# 21:46:04 |
sh4rm4 |
joins #crosstool-ng |
# 21:52:20 |
smartin_ |
quits : Quit: good night |
# 21:52:43 |
maxime_ |
joins #crosstool-ng |
# 21:55:28 |
maxime |
quits : Ping timeout: 255 seconds |
# 22:26:41 |
sh4rm4 |
quits : Remote host closed the connection |
# 22:27:12 |
sh4rm4 |
joins #crosstool-ng |
# 22:46:19 |
y_morin |
quits : Quit: Nighty Night! |
# 23:14:15 |
nandub |
joins #crosstool-ng |
# 23:36:08 |
smartin |
quits : Ping timeout: 244 seconds |
# 23:38:57 |
smartin |
joins #crosstool-ng |