# 00:01:06 |
imMute^ |
is now known as: imMute |
# 00:26:34 |
mingwandroid |
bhundven: hey |
# 00:30:41 |
mnt_real |
quits : Ping timeout: 245 seconds |
# 00:41:18 |
mingwandroid |
night all. |
# 00:41:25 |
mingwandroid |
quits : Remote host closed the connection |
# 01:12:36 |
mingwandroid |
joins #crosstool-ng |
# 01:14:57 |
mingwandroid |
bhundven: ping. |
# 01:15:03 |
bhundven |
hey, sorry |
# 01:15:26 |
mingwandroid |
bhundven: np, I was going to go to bed but figured I'd check on my build first. |
# 01:16:26 |
mingwandroid |
ok, really going this time! |
# 01:16:33 |
bhundven |
:) |
# 01:16:46 |
bhundven |
I'll talk with you later, lets catch up tomorrow morning. |
# 01:17:34 |
mingwandroid |
aye! |
# 01:17:39 |
mingwandroid |
quits : Remote host closed the connection |
# 02:29:23 |
sh4rm4 |
quits : Ping timeout: 252 seconds |
# 03:54:51 |
perr |
joins #crosstool-ng |
# 07:49:56 |
diorcety |
quits : Quit: Leaving. |
# 08:07:55 |
aalv |
joins #crosstool-ng |
# 09:45:55 |
mingwandroid |
joins #crosstool-ng |
# 09:48:54 |
mingwandroid |
bhundven: ping |
# 09:59:18 |
Net147 |
joins #crosstool-ng |
# 10:09:31 |
bhundven |
mingwandroid: hey |
# 10:09:43 |
mingwandroid |
bhundven: hey |
# 10:10:18 |
mingwandroid |
bhundven: last night I had a crack at finishing GNU make 3.81 'stuff'. |
# 10:10:59 |
mingwandroid |
i.e. if detect you don't have that exact version then build it as a companion tool. |
# 10:12:50 |
mingwandroid |
bhundven: can you eyeball the patch please? http://paste.kde.org/pmv4o9rzb |
# 10:13:43 |
mingwandroid |
urgh, better: http://paste.kde.org/plgjzpe4k |
# 10:15:57 |
bhundven |
well |
# 10:16:04 |
bhundven |
one problem is |
# 10:16:17 |
bhundven |
"GNU Make 3.81" |
# 10:16:23 |
bhundven |
you can change the version string |
# 10:16:28 |
bhundven |
(some distro's do) |
# 10:16:40 |
bhundven |
I would just check the substring for 3.81 |
# 10:17:14 |
bhundven |
some sources that are older won't handle 3.82 |
# 10:17:23 |
bhundven |
but newer sources will be able to handle it |
# 10:17:37 |
bhundven |
most current source can't handle 4.0 |
# 10:17:48 |
bhundven |
but I'll assume that they will in the future |
# 10:17:58 |
bhundven |
with that, I would suggest |
# 10:18:08 |
mingwandroid |
I'm piggybacking the existing check for: |
# 10:18:13 |
mingwandroid |
[[MAKE_ver=$($ac_path_MAKE --version 2>&1 \ |
# 10:18:13 |
mingwandroid |
|$EGREP '^GNU Make (3.[89][[:digit:]]|[4-9])') |
# 10:19:27 |
bhundven |
btw, that makes me happy: http://crosstool-ng.org/hg/crosstool-ng/rev/bddc642fc8eb |
# 10:20:02 |
bhundven |
fixes some static toolchain issues |
# 10:20:14 |
mingwandroid |
Yeah, I forgot about that one until I saw the email! |
# 10:20:15 |
bhundven |
or, should, I need to test more |
# 10:20:52 |
bhundven |
there was some problems with ppc and target binutils |
# 10:21:34 |
mingwandroid |
FWIW, I've only found so far that {e}glibc don't like gnumake 4. But Yann wanted "trusty 3.81" to avoid weird issues (target order is randomised in 4.0!) |
# 10:21:47 |
bhundven |
sure. |
# 10:21:54 |
bhundven |
I think what I'm saying is |
# 10:22:15 |
bhundven |
tools should require make or greater or less then |
# 10:22:28 |
bhundven |
instead of forcing a version |
# 10:22:40 |
bhundven |
s/require/depend on/ |
# 10:23:42 |
bhundven |
so usually what yann tells me is, make an intermediate/hidden kconfig option |
# 10:23:53 |
bhundven |
that tools depend on |
# 10:24:03 |
mingwandroid |
yeah, is mine not hidden btw? |
# 10:24:31 |
bhundven |
no, it is |
# 10:25:23 |
bhundven |
I'm just bad at explaining |
# 10:25:39 |
mingwandroid |
thing is it's not just tools that depend on gnumake 3.81, in fact not even tools. It's anything we build, the thing that goes wrong beling libc builds .. |
# 10:25:56 |
bhundven |
for now |
# 10:25:56 |
mingwandroid |
.. you could see it as a min-max of all the acceptable ranges for every tool and library you build, but that might just be overkill for the sake of it? |
# 10:26:12 |
bhundven |
not every tool will have that requirement |
# 10:26:20 |
bhundven |
to depend on a version of make |
# 10:26:31 |
bhundven |
libc is probably more prone to different versions |
# 10:26:33 |
bhundven |
and for now |
# 10:26:40 |
bhundven |
yes, depending on make 3.81 is right |
# 10:26:47 |
bhundven |
but eventually, libc will work with 4.0 |
# 10:27:24 |
mingwandroid |
it'll also work with 3.81 (until it doesn't) |
# 10:27:25 |
bhundven |
I'm just saying that versions up until then, depend on that option |
# 10:27:47 |
bhundven |
so that it's not forced unless a tool depends on it. |
# 10:28:04 |
bhundven |
that's all |
# 10:28:17 |
bhundven |
otherwise, looks good to me! |
# 10:28:55 |
bhundven |
PST: 2:28am... |
# 10:29:01 |
bhundven |
is looking for bed |
# 10:29:01 |
mingwandroid |
yeah, I get what you are saying. I guess I think that I managed to do it with few lines added, and Yann just wanted trusty 3.81 .. |
# 10:29:07 |
mingwandroid |
you mind if I add it to our queue? |
# 10:29:24 |
bhundven |
sure! |
# 10:29:25 |
mingwandroid |
there's another patch to make tools not depend on experimental, that's simple |
# 10:29:31 |
bhundven |
:) |
# 10:29:36 |
bhundven |
that would be great too |
# 10:29:38 |
mingwandroid |
ok great. |
# 10:29:45 |
bhundven |
there are a lot of things experimental |
# 10:29:49 |
bhundven |
that don't need to be |
# 10:29:55 |
mingwandroid |
yeah. |
# 10:29:58 |
bhundven |
they were just new at the time we did it |
# 10:30:22 |
bhundven |
I usually just enable experimental and ctng debugging |
# 10:30:27 |
bhundven |
by default |
# 10:30:42 |
bhundven |
but, I'm not a 'normal user' |
# 10:30:44 |
bhundven |
;) |
# 10:30:57 |
mingwandroid |
I added an option to build the compilers as debuggable btw .. |
# 10:31:06 |
bhundven |
nice! |
# 10:31:07 |
mingwandroid |
so with "-O0 -ggdb" and unstripped. |
# 10:31:21 |
bhundven |
now I can do something with those ICE errors I find |
# 10:31:31 |
mingwandroid |
I will try to get that submitted to the ML. |
# 10:31:56 |
mingwandroid |
I think I'll start posting random bits and pieces like that to it soon. |
# 10:32:10 |
bhundven |
I want to branch out the wip repo |
# 10:32:22 |
mingwandroid |
you waiting for git ;-) |
# 10:32:47 |
mingwandroid |
thing is, looks like MQs are quite cheap on the server, so setup a load of different ones on bitbucket for now? |
# 10:33:09 |
bhundven |
yea |
# 10:33:16 |
mingwandroid |
so wip -> multilib, new-versions, gnumake-381, etc. |
# 10:33:21 |
bhundven |
maybe that's better then this stupid hg bookmark crap |
# 10:33:37 |
bhundven |
not a real branch |
# 10:34:02 |
mingwandroid |
is it just a lot of different "series" files or somesuch? |
# 10:34:07 |
bhundven |
I never noticed that, kinda funny how one feature makes me dislike a whole version control system |
# 10:34:50 |
bhundven |
mingwandroid: kinda |
# 10:35:03 |
mingwandroid |
heh. well, branching needs to be cheap and fast these days as we've all got attention deficit disorder. |
# 10:35:44 |
bhundven |
:) |
# 10:36:00 |
bhundven |
oo, look, shinny! |
# 10:36:13 |
bhundven |
mq is nice for wip |
# 10:36:51 |
bhundven |
I just need to stop my dependency on it |
# 10:36:56 |
mingwandroid |
mq fits my mental model quite well for a single set of patches yeah. |
# 10:37:02 |
bhundven |
s/dependency/addiction/ |
# 10:37:07 |
bhundven |
:) |
# 10:37:46 |
bhundven |
the work flow for branching, commiting and pushing makes much more sense to me. |
# 10:38:04 |
bhundven |
on the other side, reviewing and merging. |
# 10:38:30 |
bhundven |
patchworks really takes that pain out too |
# 10:39:04 |
mingwandroid |
yeah, mq sort-of allows for rebase-like operations after having pushed .. |
# 10:39:15 |
mingwandroid |
whereas you shouldn't do that with git of course |
# 10:39:21 |
bhundven |
anyways, it'd be nice to have the updates separate from the mq stuff, and push that soon to yann before he cuts the next release |
# 10:39:29 |
bhundven |
he doesn't want the ml stuff till it's ready |
# 10:39:50 |
mingwandroid |
yeah, I'd like GNU make 3.81 in there for now. |
# 10:40:00 |
mingwandroid |
we'll see. |
# 10:40:02 |
bhundven |
kernel updates |
# 10:40:18 |
mingwandroid |
how about GCC 4.9 snapshots? |
# 10:40:25 |
mingwandroid |
I'd be very keen on seeing them land. |
# 10:40:27 |
bhundven |
config-sub-config-guess update |
# 10:40:52 |
bhundven |
nah, if you want snapshots, you can either use the custom tarball/directory or do what I did. |
# 10:40:58 |
bhundven |
yann won't add that |
# 10:41:19 |
mingwandroid |
"do what I did."? |
# 10:41:58 |
bhundven |
yea, add it yourself in your patch queue so you can play with snapshots while you test new features (to see if they will work with the next version) |
# 10:42:05 |
mingwandroid |
ok |
# 10:42:18 |
bhundven |
sooner or later, preferably sooner |
# 10:42:29 |
bhundven |
I'd like to get into this refactoring stuff |
# 10:42:38 |
bhundven |
and clean up a bunch of annoying crap |
# 10:42:51 |
bhundven |
one being how versions and sources are collected |
# 10:43:01 |
bhundven |
getting a tarball |
# 10:43:09 |
bhundven |
and checking out a git tree at a tag is the same thing |
# 10:43:23 |
mingwandroid |
they should be stamped with revision of sha1 I think. |
# 10:43:44 |
mingwandroid |
s/of/or/ |
# 10:44:24 |
bhundven |
well any reference |
# 10:44:27 |
bhundven |
branch |
# 10:44:28 |
bhundven |
tag |
# 10:44:31 |
bhundven |
revision |
# 10:44:40 |
bhundven |
branch + revision |
# 10:44:46 |
bhundven |
ref-spec |
# 11:03:12 |
perr |
quits : Quit: Leaving |
# 11:08:22 |
sh4rm4 |
joins #crosstool-ng |
# 11:15:11 |
sh4rm4 |
quits : Ping timeout: 252 seconds |
# 11:45:57 |
sh4rm4 |
joins #crosstool-ng |
# 12:01:32 |
sh4rm4 |
quits : Remote host closed the connection |
# 12:14:32 |
sh4rm4 |
joins #crosstool-ng |
# 12:18:08 |
sh4rm4 |
quits : Remote host closed the connection |
# 12:20:21 |
sh4rm4 |
joins #crosstool-ng |
# 12:22:52 |
sh4rm4 |
quits : Remote host closed the connection |
# 12:23:22 |
sh4rm4 |
joins #crosstool-ng |
# 12:25:18 |
sh4rm4 |
quits : Remote host closed the connection |
# 12:26:57 |
sh4rm4 |
joins #crosstool-ng |
# 12:30:37 |
sh4rm4 |
quits : Remote host closed the connection |
# 12:36:32 |
sh4rm4 |
joins #crosstool-ng |
# 12:38:03 |
sh4rm4 |
quits : Remote host closed the connection |
# 12:45:40 |
Net147 |
quits : Quit: HydraIRC -> http://www.hydrairc.com <- Wibbly Wobbly IRC |
# 12:58:47 |
Net147 |
joins #crosstool-ng |
# 13:13:54 |
aalv |
quits : Quit: Leaving. |
# 13:14:29 |
aalv |
joins #crosstool-ng |
# 13:23:14 |
aalv1 |
joins #crosstool-ng |
# 13:24:46 |
aalv |
quits : Ping timeout: 245 seconds |
# 13:41:50 |
aalv1 |
quits : Quit: Leaving. |
# 13:42:14 |
aalv |
joins #crosstool-ng |
# 13:53:05 |
ctngbot |
quits : Ping timeout: 246 seconds |
# 14:01:10 |
ctngbot_ |
joins #crosstool-ng |
# 14:37:23 |
mingwandroid |
joins #crosstool-ng |
# 14:38:20 |
mingwandroid_ |
quits : *.net *.split |
# 14:40:32 |
Net147 |
quits : Quit: HydraIRC -> http://www.hydrairc.com <- Organize your IRC |
# 22:30:22 |
ctngbot |
joins #crosstool-ng |
# 22:33:06 |
hurdman |
quits : *.net *.split |
# 22:33:07 |
codyps |
quits : *.net *.split |
# 22:36:25 |
codyps |
joins #crosstool-ng |
# 22:36:25 |
hurdman |
joins #crosstool-ng |
# 22:46:08 |
hurdman |
quits : *.net *.split |
# 22:46:08 |
codyps |
quits : *.net *.split |
# 22:46:59 |
codyps |
joins #crosstool-ng |
# 22:46:59 |
hurdman |
joins #crosstool-ng |
# 23:03:44 |
smartin_ |
quits : Quit: leaving |
# 23:18:26 |
sh4rm4 |
quits : Ping timeout: 252 seconds |
# 23:27:36 |
sh4rm4 |
joins #crosstool-ng |
# 23:47:57 |
mingwandroid |
quits : Remote host closed the connection |