00:04.27 | linuxjacques | yeah, happy birthday landley you youngster |
00:10.08 | landley | linuxjacques: thanks. |
00:41.38 | *** join/#uclibc xiangfu (~xiangfu@fidelio.qi-hardware.com) |
02:27.11 | *** join/#uclibc wagle (~wagle@static-50-43-56-158.bvtn.or.frontiernet.net) |
02:39.48 | *** join/#uclibc risca (~risca@wnpgmb0903w-ds01-177-34.dynamic.mtsallstream.net) |
02:48.48 | *** join/#uclibc antgreen (~user@64.134.168.217) |
03:48.51 | *** join/#uclibc bkuhn (~bkuhn@fsf/director/conservancy.president.bkuhn) |
03:51.25 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
04:22.32 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
05:12.17 | *** join/#uclibc risca (~risca@wnpgmb0903w-ds01-177-34.dynamic.mtsallstream.net) |
05:52.59 | *** join/#uclibc clio (~andrej@85.159.109.222) |
05:57.48 | *** join/#uclibc bkuhn (~bkuhn@fsf/director/conservancy.president.bkuhn) |
06:09.10 | *** join/#uclibc khem (~khem@99-57-141-118.lightspeed.sntcca.sbcglobal.net) |
07:18.18 | CIA-62 | 03rep.dot.nop 07refs/tags/v0.9.33 * r8c0b2c288647 10uClibc/Rules.mak: Release 0.9.33 |
07:49.14 | *** join/#uclibc ncopa (~ncopa@3.203.202.84.customer.cdi.no) |
08:03.53 | *** join/#uclibc War2 (~war2@unaffiliated/war2) |
08:13.57 | *** join/#uclibc Artemys (~quassel@stp25-2-82-234-232-91.fbx.proxad.net) |
08:30.44 | *** join/#uclibc War2_ (~war2@14.97.127.61) |
08:59.01 | *** join/#uclibc smartin (52e7e007@gateway/web/freenode/ip.82.231.224.7) |
09:06.12 | *** join/#uclibc acoul (~alex@212.70.193.241) |
09:08.32 | CIA-62 | 03jacmet 07master * r8ed392f3d649 10buildroot/toolchain/toolchain-buildroot/Config.in.2: toolchain-buildroot: enable / default to NPTL fr uClibc 0.9.33 as well |
09:08.36 | CIA-62 | 03jacmet 07master * r4461b3213d35 10buildroot/package/binutils/Config.in.host: binutils: default to 2.21.1 for uClibc 0.9.33 as well |
09:08.38 | CIA-62 | 03jacmet 07master * r4a4937957148 10buildroot/toolchain/uClibc/ (uClibc-0.9.33.config uClibc-snapshot.config): uClibc: tweak 0.9.33 / snapshot configs |
09:19.11 | *** part/#uclibc acoul (~alex@212.70.193.241) |
09:36.11 | *** join/#uclibc jcEcaSinters (~jc@fw-tatom-nantes.pacwan.net) |
09:38.39 | *** join/#uclibc War2 (~war2@unaffiliated/war2) |
09:41.10 | *** join/#uclibc shevy (~shevy@178-191-255-205.adsl.highway.telekom.at) |
10:30.09 | CIA-62 | 03vda.linux 07master * r9ec3cd400ad9 10busybox/libbb/procps.c: tweak comment, no code changes |
10:49.52 | *** join/#uclibc acoul (~alex@212.70.193.241) |
10:52.45 | *** part/#uclibc acoul (~alex@212.70.193.241) |
11:18.57 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
11:41.09 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
11:46.18 | CIA-62 | 03hartleys 07master * re8b320a7080c 10buildroot/package/fis/fis-003-list_output.patch: fis: make fis list more like RedBoot's output |
11:46.18 | CIA-62 | 03hartleys 07master * r5256c77b46fb 10buildroot/package/fis/fis-004-sorted_list.patch: fis: sort output of fis list |
12:10.44 | CIA-62 | 03will.newton 07master * rcee43cbcbd69 10buildroot/package/ (Config.in liboauth/Config.in liboauth/liboauth.mk): liboauth: Add new package |
12:10.45 | CIA-62 | 03will.newton 07master * ra2a9f6358d3c 10buildroot/package/pixman/ (pixman-0.10.0-no-tests.patch pixman.mk): pixman: Bump version to 0.24.2 |
12:12.07 | *** join/#uclibc blindvt (~brf@85-127-80-99.dynamic.xdsl-line.inode.at) |
12:13.51 | *** join/#uclibc cendres (~ashes@75.155.223.218) |
13:10.30 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
13:25.13 | *** join/#uclibc sh4rm4 (~sh4rm@gateway/tor-sasl/sh4rm4) |
13:28.30 | *** join/#uclibc curious (~curious@bwv190.internetdsl.tpnet.pl) |
13:42.53 | *** join/#uclibc avinashhm (~avinash-h@203.101.61.7) |
13:49.24 | *** join/#uclibc avinashhm (~avinash-h@203.101.61.7) |
14:07.35 | sh4rm4 | the busybox Makefile uses the gcc "-r" switch for building some stuff... it seems undocumented, so what does that do ? |
14:12.06 | jow_laptop | its passed through to ld where it means "Generate relocatable output" |
14:13.17 | *** join/#uclibc xiangfu (~xiangfu@fidelio.qi-hardware.com) |
14:19.19 | sh4rm4 | jow_laptop, thanks... how does that work ? i get link errors on that specific object file, saying that _start is already defined in crt.o |
14:20.25 | sh4rm4 | usually relocatable is used in the meaning of -fPIC, i suppose |
14:20.59 | sh4rm4 | the ld manual is rather vague on this switch |
14:27.14 | jow_laptop | I don't know the details either, I just inspected my gcc spec file to see how -r is handled, there I've seen its only part of link_command |
14:37.20 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
14:40.31 | *** join/#uclibc sh4rm4 (~sh4rm@gateway/tor-sasl/sh4rm4) |
14:44.21 | *** join/#uclibc drwhom (~drwhom@149-166-11-180.dhcp-in.iupui.edu) |
15:03.58 | CIA-62 | 03will.newton 07master * rdfa93b47f721 10buildroot/package/cairo/ (3 files): cairo: Bump version to 1.10.2 |
15:03.58 | CIA-62 | 03jacmet 07master * r84babf1c1a5f 10buildroot/boot/lpc32xxcdl/lpc32xxcdl.mk: lpc32xxcdl: fix patch step |
15:05.05 | *** join/#uclibc mnt_real (~mnt_real@bas1-montreal19-1177821094.dsl.bell.ca) |
15:07.17 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
15:07.51 | CIA-62 | 03jacmet 07master * r88404d6b80a5 10buildroot/boot/uboot/ (2 files in 2 dirs): u-boot: only apply 2011.12 at91 patch if building 2011.12 |
15:16.38 | CIA-62 | 03yegorslists 07master * rfaf266b460c3 10buildroot/package/python-serial/python-serial.mk: python-serial: bump to 2.6 |
15:17.52 | *** join/#uclibc antgreen (user@nat/redhat/x-nrzjnowshmgrukns) |
15:27.25 | *** join/#uclibc landley (~landley@140.242.26.2) |
15:28.30 | sh4rm4 | hi landley |
15:28.44 | landley | sh4rm4: good morning. |
15:28.49 | landley | Or at least "morning". |
15:29.02 | landley | is not a morning person. |
15:29.17 | sh4rm4 | i've trouble building busybox... one file gets compiled using the -r switch |
15:29.33 | sh4rm4 | which apparently means -relocatable to the linker |
15:29.43 | sh4rm4 | applets/built-in.o: In function `_start': |
15:29.44 | sh4rm4 | (.text+0x0): multiple definition of `_start' |
15:29.59 | landley | Huh. |
15:30.02 | landley | Which version? |
15:30.12 | sh4rm4 | any idea what's the purpose of that -r flag ? the manual is quite vague on it |
15:30.13 | landley | And which architecture? |
15:30.30 | sh4rm4 | i386/bb 1.19.2 |
15:30.39 | landley | Well, in theory busybox grew a "build libbb as a shared library" functionality we discussed on the list way back when. |
15:30.45 | landley | And shared libraries have to be relocatable. |
15:30.52 | landley | There's also the PIC/PIE stuff in the .config. |
15:30.56 | landley | General options I think. |
15:30.58 | sh4rm4 | musl-gcc -nostdlib -r -o applets/built-in.o applets/applets.o |
15:32.07 | sh4rm4 | hmm usually one would use -fPIC instead of -r, no ? |
15:33.30 | sh4rm4 | note that musl-gcc is a wrapper which uses -nostdlib itself |
15:38.25 | landley | sh4rm4: I've built 1.19.2 for i486 and it worked fine. |
15:38.29 | landley | I even put binaries up, I think. |
15:38.34 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
15:38.56 | landley | Eh last I bothered to post was 1.19.0 it seems: http://busybox.net/downloads/binaries/ |
15:39.24 | landley | but aboriginal linux was using 1.19.2 until a few days ago, now 1.19.3. And it has i486, i586, and i686 targets. |
15:39.40 | sh4rm4 | oh well, i'm trying to get it to compile as part of sabotage linux |
15:39.56 | landley | (I never bothered to do an actual i386 because nobody's made hardware that requires that for about 15 years.) |
15:40.19 | landley | Understood. Good luck with it. I'm just saying "it worked for me with uClibc and glibc". |
15:40.30 | landley | digs up build logs... |
15:40.44 | sh4rm4 | it used to work when i did a standard installation of gcc |
15:41.11 | sh4rm4 | now i tried building gcc/binutils/musl with -sysroot and put it in a different path, then symlinked everything into / |
15:41.17 | sh4rm4 | i.e all the files |
15:41.29 | landley | Alas, the logs were without V=1 so they just say "HOSTCC" and such. |
15:41.44 | sh4rm4 | so i must now use the musl-gcc wrapper to use /include and not /toolchain/include |
15:41.46 | landley | You changed your toolchain, busybox broke, so you're asking on the busybox list. Got it. |
15:42.08 | landley | -sysroot is creepy. |
15:42.08 | sh4rm4 | oh well, just trying to figure out stuff. |
15:42.21 | landley | Using a wrapper to make sysroot work sounds like par for the course to me, but other people have had better luck with it. |
15:42.28 | landley | Me, I use the wrapper and _not_ sysroot... |
15:42.47 | landley | The problem is you're not supposed to mix pic and non-pic code in the same binary. |
15:43.08 | landley | Is the -r there _before_ the wrapper runs, or after? |
15:43.18 | sh4rm4 | mhm, so i wonder why anybody would want to pass -r to that specific file |
15:43.22 | sh4rm4 | before |
15:44.49 | landley | Odd. |
15:45.37 | landley | I note that for i386 specifically busybox grew an insane arch/i386/Makefile that no other target has. |
15:45.39 | landley | I have no IDEA why. |
15:45.50 | landley | But I don't see a -r in there. |
15:46.08 | landley | Uh-huh. |
15:46.11 | landley | Top level Makefile: |
15:46.28 | landley | # For maximum performance (+ possibly random breakage, uncomment |
15:46.28 | landley | # the following) |
15:46.36 | landley | MAKEFLAGS += -rR |
15:46.46 | landley | And at least in my copy, it's not commented. |
15:46.52 | sh4rm4 | LDFLAGS_MODULE = -r |
15:46.53 | landley | Line 287. |
15:47.10 | landley | Busybox DOES NOT HAVE MODULES. |
15:47.20 | landley | Sigh. I have no idea why the katamari rolled up the kbuild infrastructure. |
15:47.22 | sh4rm4 | i'd guess MAKEFLAGS doesnt change gcc behaviour |
15:47.23 | landley | None at all. |
15:47.57 | mnemoc | unless MAKEFLAGS sets CFLAGS... |
15:47.59 | landley | Back in 1.2.2 this was one of the things I wanted to rip out, clean up, and simplify. |
15:48.02 | landley | Instead they made it bigger. |
15:48.04 | landley | Oh well. |
15:48.08 | landley | does toybox now... |
15:48.32 | landley | Working on getting a release out, but birthday social obligations have been interfering a bit. :) |
15:48.54 | sh4rm4 | toybox is not on par with busybox yet, isn't it ? |
15:56.34 | *** join/#uclibc risca (~risca@wi-secure-2685.cc.umanitoba.ca) |
16:02.20 | landley | sh4rm4: not remotely. :) |
16:02.24 | landley | It's about 1/3 of the way, I'd guess. |
16:03.26 | sh4rm4 | hmm. i'd imagine fixing the broken parts of busybox would be a lighter task. |
16:03.47 | landley | Back in 2006-2008 I solved the interesting infrastructure problems, and then mothballed the project (except for the patch command) because I didn't think "technically better" was good enough to displace a project with a 10 year headstart. |
16:04.17 | landley | I only revived it in November. Added a dozen commands since then, but I need over 100 total. |
16:04.55 | landley | Everybody kept linking to Tim's "hey, do we wanna fund this" wiki page, but the _interesting_ one was this one: |
16:04.56 | landley | http://www.elinux.org/Busybox_replacement |
16:05.15 | landley | That's the one I filled out with the "what do we actually _need_" info, including sources and methodology. |
16:05.20 | landley | That's my toybox roadmap, more or less. |
16:06.44 | landley | I need to update that to include xargs, now that's in... |
16:09.19 | landley | sh4rm4: I tried fixing busybox. It's not easy: http://landley.net/notes-2011.html#08-06-2011 |
16:10.26 | landley | I tried pushing the toybox infrastructure upstream, and they just didn't get it: http://landley.net/notes-2010.html#11-03-2010 |
16:10.45 | landley | I explained it in person to Denys at CELF 2010, and he merged... about half of it. Badly from my point of view. |
16:12.17 | landley | I was still trying to push cleanups in all along: http://lists.busybox.net/pipermail/busybox/2010-April/072320.html |
16:12.27 | landley | I just... couldn't keep up with the rate at which busybox got ugly. |
16:16.02 | sh4rm4 | oh well. how about "stealing" the missing applets at least partially from busybox ? |
16:16.20 | sh4rm4 | i'd guess some of them are reasonably clean |
16:16.26 | landley | sh4rm4: well the commands I wrote and which are entirely my code I intend to repurpose. Already grapped sort, for example. |
16:16.54 | landley | (I have to triage them to find the last version in source control nobody else touched, but eh.) |
16:17.04 | sh4rm4 | heh |
16:17.04 | landley | But really, writing new ones isn't that hard. |
16:17.15 | landley | I rewrote mount three times before I was happy with it. |
16:17.19 | landley | (And then they broke it.) |
16:17.41 | landley | I'm a bit disappointed that even after I "left" I kept having to fix things in busybox at a design level, ala http://lists.busybox.net/pipermail/busybox/2008-October/067372.html |
16:18.38 | landley | I'd much rather put that work into toybox, it just seemed pointless for a while there until Tim pointed out Android as a target for it, and offered to help get it working there. |
16:19.14 | landley | (I still don't have a bionic build enviornment set up, and I just _know_ I'm going to have to write my own regex implementation on there. Last time I did that was under OS/2 in 1997. Hopefully it'll be easier this time. :) |
16:19.41 | landley | And I am glad my option parsing infrastructure doesn't use getopt() because I doubt bionic's is particularly feature-complete, assuming it even exists... |
16:19.51 | sh4rm4 | hmm i think musl is a viable alternative for bionic |
16:20.00 | landley | I thought that for a while too. |
16:21.02 | landley | sh4rm4: http://landley.net/notes-2012.html#18-01-2012 |
16:21.25 | landley | http://source.android.com/source/licenses.html |
16:21.42 | landley | Their "no GPL in userspace" specifically excludes LGPL too. |
16:22.24 | landley | So using musl on Android is about as viable as using uClibc or glibc: google's never shipping it, android vendors who follow Google's policies are never shipping it. |
16:22.32 | landley | And if there's one thing I learned working on OS/2: preinstalls matter. |
16:32.10 | *** join/#uclibc Fallenou (~fallen@yozora-irc.net) |
16:39.21 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
16:52.44 | *** join/#uclibc x-fak (~WinGuru@ANancy-554-1-136-73.w109-217.abo.wanadoo.fr) |
16:52.44 | *** join/#uclibc x-fak (~WinGuru@reactos/tester/x-fak) |
17:07.50 | *** join/#uclibc NIN101 (~NIN@2001:530::216:3cff:fe71:5e1e) |
17:16.16 | khem | landley: why dont you use git for your projects. Not that mercurial is bad |
17:25.31 | *** join/#uclibc NIN102 (~NIN@2001:530::216:3cff:fe71:5e1e) |
17:52.20 | *** join/#uclibc luneff (~yury@85.26.183.99) |
18:06.40 | *** join/#uclibc cendres (~ashes@75.155.223.218) |
18:07.20 | *** join/#uclibc bkuhn (~bkuhn@fsf/director/conservancy.president.bkuhn) |
18:08.49 | bkuhn | Hey, does anyone know if Denys is regularly on IRC? |
18:08.56 | bkuhn | I've never seen him here, but maybe I just don't know his username? |
18:12.20 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
18:28.10 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
18:33.45 | *** join/#uclibc y_morin (~ymorin@ARennes-256-1-67-147.w90-32.abo.wanadoo.fr) |
18:35.32 | Jacmet | bkuhn: I don't think he uses irc |
18:45.12 | bkuhn | Jacmet: Yeah, I didn't think so; I've never seen him here. I think he tried to phone me -- I got a missed call from the Czech Republic (which doesn't happen often) and odds are it was him. I emailed him to let him know I'll call him back if it was him. I figured I'd try to see if he was maybe here. |
19:03.29 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
19:05.03 | landley | bkuhn: I've never seen denys on irc either. |
19:05.14 | landley | khem: I've been meaning to set up a git mirror. |
19:05.35 | landley | I prefer mercurial because both git's user interface and internal implementation are a giant pile of unintelligible corner cases. |
19:06.50 | landley | Basically OLS 2006 I tracked down both matt mackall and junio hamano and asked them both about their respective thingies, and went with mercurial as the better tool for my purposes. |
19:10.10 | landley | khem: http://landley.net/notes-2006.html#23-07-2006 |
19:15.46 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
19:28.50 | *** join/#uclibc AnywhereIs (~edK@h62-133-168-70.static.bashtel.ru) |
20:15.19 | CIA-62 | 03thomas.petazzoni 07master * rd06e802223b2 10buildroot/ (Config.in package/Config.in.host): Add basic config infrastructure for host utilities |
20:15.20 | CIA-62 | 03thomas.petazzoni 07master * r6b04b2dda6bf 10buildroot/package/ (Config.in.host uboot-tools/Config.in.host): uboot-tools: expose host package in menuconfig |
20:15.20 | CIA-62 | 03thomas.petazzoni 07master * r66ac654d8ebe 10buildroot/package/libusb/libusb.mk: libusb: add host variant |
20:15.20 | CIA-62 | 03thomas.petazzoni 07master * ree673bae7702 10buildroot/package/libusb-compat/libusb-compat.mk: libusb-compat: add host variant |
20:15.21 | CIA-62 | 03thomas.petazzoni 07master * r76cd6b18e3b6 10buildroot/package/libftdi/libftdi.mk: libftdi: add host variant |
20:15.21 | CIA-62 | 03thomas.petazzoni 07master * r375d68d110a0 10buildroot/package/ (Config.in.host openocd/Config.in.host openocd/openocd.mk): openocd: add host variant |
20:15.21 | CIA-62 | 03jacmet 07master * r2b81a500f923 10buildroot/package/openocd/openocd.mk: openocd: fix libftdi dependency if _FT2XXX option is selected |
20:15.21 | CIA-62 | 03thomas.petazzoni 07master * ra23c0e596846 10buildroot/package/ (Config.in.host sam-ba/Config.in.host sam-ba/sam-ba.mk): sam-ba: new package with host variant only |
20:46.29 | *** join/#uclibc luneff (~yury@84.51.201.99) |
21:17.53 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
21:25.32 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
21:28.53 | *** join/#uclibc cendres (~ashes@75.155.223.218) |
21:30.19 | *** join/#uclibc WildPikachu (~nkukard@about/linux/staff/wildpikachu) |
22:04.02 | CIA-62 | 03arnout 07master * r5fe4b2c3998b 10buildroot/package/ (Makefile.autotools.in Makefile.package.in): pkg-infra: improve reliability of foo-install |
22:04.02 | CIA-62 | 03yegorslists 07master * r8c3e2cbb53b0 10buildroot/fs/skeleton/run: Introduce /run directory |
22:30.07 | *** join/#uclibc ashes (~ashes@75.155.223.218) |
22:33.08 | *** join/#uclibc cendres (~ashes@75.155.223.218) |
22:56.16 | sh4rm4 | landley, i've got an android mobile in my hands recently and it had a page about software licenses... a big part of them were using GPL/LGPL |
22:56.31 | sh4rm4 | although no single GPL 3 sighted |
22:57.43 | landley | sh4rm4: you're not going to find any for-profit corporation using GPLv3 that I'm aware of. |
22:58.06 | landley | Your mobile vendor may have been ok adding GPL or LGPL code to their base system. The ones I've worked with aren't. |
22:58.19 | landley | Which page? |
22:58.22 | sh4rm4 | it was a samsung galaxy tablet |
22:58.37 | sh4rm4 | under settings, it has a page called "open source licenses" |
22:59.06 | sh4rm4 | basically just a huge page with concatenated text files |
23:00.00 | sh4rm4 | iirc it had about 2/3 apache and BSD licenses, and about one third GPL |
23:41.03 | *** join/#uclibc War2 (~war2@unaffiliated/war2) |