| # 06:24:51 | mnt_real | quits : Quit: Leaving |
| # 11:02:49 | y_morin | joins #crosstool-ng |
| # 11:25:59 | dbtid | y_morin: thanks for ct-ng. |
| # 11:26:34 | y_morin | dbtid: Hello! Cheers! |
| # 11:27:35 | dbtid | which compiler builds linux? arm-linux-gnueabi? |
| # 11:29:08 | y_morin | dbtid: what do you mean? |
| # 11:29:19 | dbtid | well, you can't use arm-none-eabi- to build linux, right? |
| # 11:29:54 | y_morin | dbtid: indeed, that's not possible. |
| # 11:30:09 | dbtid | ok, so which tuples can i build linux with? |
| # 11:30:19 | y_morin | dbtid: in theory, it should be, because Linux does not rely on userland, so a bare-metal compiler should work. |
| # 11:30:44 | y_morin | dbtid: but the ACPI stuff in Linux ends up including headers from the C library... |
| # 11:31:49 | y_morin | dbtid: a compiler able to build Linux would have 'linux' in its tuple. |
| # 11:31:53 | dbtid | you know infinitely more about this than i do, for sure. so if my questions sound stupid, i apologize. i haven't found a good resource online that gives all of this information in one complete set. doesn't arm-none-eabi- also use headers from the C library? |
| # 11:31:59 | dbtid | y_morin: ok, so about any will do. |
| # 11:32:59 | y_morin | dbtid: probably. I don't really understand what the qestion is, though (I just woke up, so not every thing is running full-speed in my head yet!) |
| # 11:33:16 | dbtid | y_morin: i just woke up a bit ago myself. i'm in ohio, in the u.s. |
| # 11:33:33 | y_morin | dbtid: France! It's 12:30 here! ;-) |
| # 11:33:40 | dbtid | y_morin: you lazy bum! :) |
| # 11:33:44 | y_morin | :-) |
| # 11:33:46 | dbtid | i have codesourcery, arm-none-eabi- |
| # 11:33:54 | dbtid | it has a standard c library in it |
| # 11:33:56 | dbtid | several in fact |
| # 11:34:04 | dbtid | so why couldn't i build linux with that (if it wasn't on windows)? |
| # 11:35:20 | dbtid | uh-oh, i don't think my compiler went where i wanted it to go! |
| # 11:36:20 | y_morin | dbtid: I don't really know the CS toolchains. But if you build a toolchain with crosstool-NG (and this is the ct-ng support IRC chan, here!), then you'll get a toolchain properly setup! ;-) |
| # 11:36:39 | dbtid | y_morin: well, i will continue to work with it to get the tools i need |
| # 11:36:48 | dbtid | i DO appreciate all the work you put into making it work so well |
| # 11:36:54 | dbtid | i will let you know what i end up building :) |
| # 11:45:13 | dbtid | y_morin: if i specified an FPU in menuconfig, does that mean that libc and such will be compiled with VFP support? |
| # 11:47:04 | y_morin | dbtid: you mean, the 'Use specific FPU' option, or the 'hardware (FPU)' choice entry ? |
| # 11:47:09 | dbtid | yes |
| # 11:48:18 | y_morin | dbtid: yes what? The option, the choice entry, or both? |
| # 11:48:41 | dbtid | i did set hardware(FPU) and i set Use specified FPU to "neon" |
| # 11:49:04 | y_morin | dbtid: Neon id not 'VFP'. |
| # 11:49:40 | dbtid | y_morin: no, it's not, but i've been told that unless you tell gcc to use unsafe operations, -mfpu=neon generates VFPv3 instructions. |
| # 11:49:58 | dbtid | and here's another bit of weirdness. |
| # 11:50:07 | y_morin | dbtid: I think that this should be sufficient to have a hard-float toolchain, yes. And the C library will use FP instructions, yes. |
| # 11:50:13 | dbtid | y_morin: thanks. |
| # 11:50:27 | y_morin | dbtid: floating point on ARM is full of weirdness! ;-) |
| # 11:50:48 | dbtid | i'm rebuilding with the compiler i just built, and it complains "switch -mcpu=cortex-a8 conflicts with -march= switch" i'm using -march=armvy-a. |
| # 11:50:57 | dbtid | armv7-a, i mean. |
| # 11:51:10 | dbtid | but it works fine w/ the other compiler... why would it not work here? |
| # 11:51:24 | y_morin | dbtid: is it an error or a warning ? |
| # 11:51:34 | dbtid | error |
| # 11:51:51 | y_morin | dbtid: and what is complaining? gcc, glibc, other? |
| # 11:51:56 | dbtid | cc1 |
| # 11:52:05 | y_morin | But when building what? |
| # 11:52:09 | dbtid | a .c file |
| # 11:52:17 | y_morin | Your own? |
| # 11:52:20 | dbtid | yes |
| # 11:52:22 | y_morin | Ah. |
| # 11:52:29 | y_morin | Just a sec... |
| # 11:52:31 | dbtid | arm-unknown-eabi-gcc -c -g -Os -Wall -Werror -mlittle-endian -march=armv7-a -mcpu=cortex-a8 -Iinclude -Iam35xxGPIO -Iam35xxI2C -Iam35xxINTC -Iam35xxMMC -Iam35xxUART -Wp,-MMD,dep/.am35xxClocksInit.d,-MT,obj/am35xxClocks/am35xxClocksInit.o am35xxClocks/am35xxClocksInit.c -o obj/am35xxClocks/am35xxClocksInit.o |
| # 11:52:45 | dbtid | cc1: warnings being treated as errors |
| # 11:52:47 | dbtid | am35xxClocks/am35xxClocksInit.c:1:0: error: switch -mcpu=cortex-a8 conflicts with -march= switch |
| # 11:52:56 | dbtid | not sure why those would conflict |
| # 11:53:08 | dbtid | i'm going to rebuild the compiler anyway, because i didn't install it in the right palce. |
| # 11:53:11 | dbtid | *place |
| # 11:53:12 | dbtid | so no problem there |
| # 11:53:12 | dbtid | brb |
| # 11:53:14 | y_morin | dbtid: Ah! This is a warning, but it is treated as an error! ;-) |
| # 11:55:51 | y_morin | is now known as: y_morin|away |
| # 11:57:18 | dbtid | ok, i know i have -Werror |
| # 11:57:35 | dbtid | i just wanted to know how to rebuild the compiler so that it won't generate the error. |
| # 12:08:00 | dbtid | ok, lots of interesting linker errors. |
| # 12:08:30 | dbtid | this project builds and links just fine w/ CodeSourcery's arm-none-eabi- v4.5.2. so i just have to figure out what i haven't done yet. |
| # 13:17:01 | dbtid | y_morin|away: this REALLY is a great tool. |
| # 14:23:05 | y_morin|away | is now known as: y_morin |
| # 16:57:06 | y_morin | is now known as: y_morin|away |
| # 18:42:38 | y_morin|away | is now known as: y_morin |
| # 18:42:39 | dbtid | oh, rats, i missed you |
| # 18:42:41 | dbtid | oh you're back! |
| # 18:42:46 | dbtid | y_morin: do you have a moment? |
| # 18:43:12 | y_morin | dbtid: just a few minutes, my parents are due here any minutes now... |
| # 18:43:19 | dbtid | i found a web page that explains the root of a problem i'm having linking regarding VFP registers |
| # 18:43:24 | dbtid | from readelf -A |
| # 18:43:29 | dbtid | #36> readelf -A linpackc.o |
| # 18:43:31 | dbtid | (that's my file) |
| # 18:43:36 | dbtid | Tag_ABI_HardFP_use: SP and DP |
| # 18:43:43 | dbtid | from libc : |
| # 18:43:48 | dbtid | #37> readelf -A lib_a-reallocr.o |
| # 18:43:57 | dbtid | there is no such tag. |
| # 18:44:07 | dbtid | would you perchance know what i need to do with my compiler to fix this issue? |
| # 18:44:19 | dbtid | arm-none-eabi-gcc (crosstool-NG 1.12.3) 4.5.4 20110630 (prerelease) |
| # 18:44:27 | dbtid | let me see what i'm building |
| # 18:46:12 | dbtid | i'm building the experimental linaro-4.5.2011-07 |
| # 18:46:14 | y_morin | dbtid: sorry, no clue... :-/ |
| # 18:46:18 | dbtid | y_morin: ok, thanks |
| # 18:46:24 | dbtid | i will try and build 4.5.2 and see what happens |
| # 18:46:45 | y_morin | dbtid: can you upgrade to crosstool-ng-1.13.0 ? |
| # 18:46:51 | dbtid | sure |
| # 18:47:08 | y_morin | dbtid: also, could you pastebin your .config? |
| # 18:47:30 | dbtid | um, sure |
| # 18:47:51 | dbtid | where do i find that .config ??? |
| # 18:47:54 | dbtid | in my build directory? |
| # 18:48:24 | y_morin | dbtid: in the directory where you run ct-ng. |
| # 18:48:28 | dbtid | right |
| # 18:48:30 | dbtid | one sec please |
| # 18:48:57 | dbtid | i need a pastesite that lets me uplaod a file |
| # 18:48:58 | dbtid | grrr |
| # 18:49:25 | y_morin | pastebin.com ? |
| # 18:49:35 | y_morin | or send it to the list. |
| # 18:49:50 | dbtid | i will paste the information for you later |
| # 18:49:55 | dbtid | i can't do it in a short time :) |
| # 18:50:21 | y_morin | dbtid: OK, I'll have to go shortly any way... |
| # 18:50:37 | dbtid | what's the mailing list email?? |
| # 18:51:06 | y_morin | http://sourceware.org/ml/crossgcc |
| # 18:51:14 | y_morin | Got to go now, parents here! |
| # 18:51:17 | y_morin | is now known as: y_morin|away |
| # 18:51:18 | dbtid | bye |
| # 18:51:20 | dbtid | thanks |
| # 18:53:24 | dbtid | i need to take a break from this |
| # 21:09:55 | y_morin|away | is now known as: y_morin |
| # 22:00:37 | y_morin | is now known as: y_morin|away |
| # 22:00:40 | y_morin|away | is now known as: y_morin |
| # 22:17:50 | y_morin | is now known as: y_morin|away |
| # 22:49:23 | y_morin|away | is now known as: y_morin |
| # 22:49:46 | y_morin | quits : Quit: Night! |