| # 02:35:55 | mnt_real_ | joins #crosstool-ng |
| # 03:55:39 | mnt_real_ | quits : Quit: Leaving |
| # 04:20:39 | mnt_real | quits : Quit: Ex-Chat |
| # 07:48:54 | smartin | joins #crosstool-ng |
| # 08:01:01 | y_morin | joins #crosstool-ng |
| # 08:49:36 | sh4rm4 | quits : Ping timeout: 255 seconds |
| # 08:50:55 | sh4rm4 | joins #crosstool-ng |
| # 11:34:36 | sh4rm4 | quits : *.net *.split |
| # 12:31:00 | mnt_real | joins #crosstool-ng |
| # 12:48:17 | y_morin | quits |
| # 14:00:03 | sh4rm4 | joins #crosstool-ng |
| # 14:00:14 | sh4rm4 | quits : Changing host |
| # 14:00:14 | sh4rm4 | joins #crosstool-ng |
| # 15:24:16 | y_morin | joins #crosstool-ng |
| # 15:35:19 | y_morin | is now known as: y_morin|away |
| # 15:39:54 | y_morin|away | is now known as: y_morin |
| # 16:27:26 | y_morin | is now known as: y_morin|away |
| # 16:28:30 | y_morin|away | is now known as: y_morin |
| # 17:51:49 | sh4rm4 | mmh, musl supports ARM since yesterday |
| # 17:52:14 | sh4rm4 | http://git.etalabs.net/cgi-bin/gitweb.cgi?p=musl;a=commitdiff;h=d960d4f2cbf18ff3476c7ac03698ec253885dd8e |
| # 17:54:16 | y_morin | sh4rm4: hey! But it's a bit crippeld so far: hard-coded kernel entry points for threads helpers, LE defines... |
| # 17:56:38 | sh4rm4 | it seems the hardcoded entry point is needed for atomic compare and swap |
| # 17:57:08 | sh4rm4 | since the CPU doesnt offer it |
| # 17:57:27 | sh4rm4 | and it is required to implement pthreads without race conditions |
| # 17:59:49 | y_morin | sh4rm4: the proper way to handle that would be to query for the entry point, either at build-time if possible, of at run-time. |
| # 18:03:33 | sh4rm4 | forwarded |
| # 18:03:47 | sh4rm4 | but anyway, i think its pretty sweet |
| # 18:03:55 | sh4rm4 | going to test it asap |
| # 18:04:27 | sh4rm4 | btw |
| # 18:04:44 | sh4rm4 | you might be interested in the build/package manager i wrote for sabotage linux |
| # 18:05:11 | sh4rm4 | it features multithreaded download and builds of non-interdependant packages |
| # 18:05:52 | sh4rm4 | https://github.com/rofl0r/sabotage/blob/master/pkg/gcc4 |
| # 18:05:55 | sh4rm4 | ^ a package |
| # 18:07:10 | sh4rm4 | https://github.com/rofl0r/butch/blob/master/butch.c <- da codez |
| # 18:07:33 | y_morin | sh4rm4: interesting. But it does not handle multi-step builds. |
| # 18:07:53 | sh4rm4 | multi-step ? |
| # 18:08:17 | sh4rm4 | you can add a metapackage which builds the packages it depends on first, if you mean that |
| # 18:08:18 | y_morin | sh4rm4: for the toolchain, we need to build gcc up to three times: core-pass-1, core-pass-2 and final. |
| # 18:08:43 | y_morin | sh4rm4: and for the C library, we need to install headers+start files, and later build the full library. |
| # 18:08:44 | sh4rm4 | then make a package gcc_stage0 gcc_stage1 gcc_stage2 |
| # 18:09:00 | y_morin | sh4rm4: yes, but it's a hack. |
| # 18:09:03 | y_morin | ;-) |
| # 18:09:41 | y_morin | sh4rm4: but yes, I'm looming over switching to a dependency-based build, rather than a linear scripting. |
| # 18:10:13 | sh4rm4 | this is the only build tool i'm aware of, which can handle multiple downloads in parallel |
| # 18:10:53 | sh4rm4 | and either sequential builds of dependant stuff, or using all available cores for stuff that doesnt depend on each other |
| # 18:11:15 | sh4rm4 | also it should start with the build as soon as all required things are downloaded/built |
| # 18:11:23 | y_morin | sh4rm4: well, in my experience, paralle downloads are not that interesting. For example, at work, the network is clogged not by the remote server, but by the borderline routers of the company. |
| # 18:11:54 | sh4rm4 | sabotage linux currently downloads the tarballs from all kind of different servers |
| # 18:12:17 | sh4rm4 | i experienced things like waiting for a 40 MB package to download with 30 KB/s |
| # 18:12:52 | y_morin | sh4rm4: what I mean is, if I download, say a kernel and a gcc, the downloads do not really add up., It just increases the overall download time... |
| # 18:13:05 | sh4rm4 | in that time, all other packages could have been downloaded as well, and most even built... |
| # 18:13:49 | y_morin | is now known as: y_morin|away |
| # 18:25:10 | smartin | quits |
| # 18:41:54 | sh4rm4 | 2011.08.19 18:36:34 downloading psmisc(/src/dl_psmisc.sh) -> /src/logs/dl_psmisc.log |
| # 18:41:54 | sh4rm4 | got error 1 from download script of expat, retrying |
| # 18:41:54 | sh4rm4 | 2011.08.19 18:36:36 downloading expat(/src/dl_expat.sh) -> /src/logs/dl_expat.log |
| # 18:41:54 | sh4rm4 | 2011.08.19 18:36:41 building bison(/src/build_bison.sh) -> /src/logs/build_bison.log |
| # 20:12:05 | y_morin|away | is now known as: y_morin |
| # 20:29:20 | smartin | joins #crosstool-ng |
| # 22:39:00 | smartin | quits |
| # 23:49:50 | y_morin | quits : Quit: Bye! |