# Format for this file: # Epoch Channel Nickname Action Message # # All separated with a single space, empty values are replaced with a dash '-' # except messages which are kept empty. # Lines starting with '#' are comments; comments are not allowed except as the only content of the line # 1330663327 - sh4rm4 quit Ping timeout: 276 seconds 1330664107 #crosstool-ng sh4rm4 join 1330678183 - daggs1 quit Read error: Operation timed out 1330679106 #crosstool-ng daggs1 join 1330679511 - daggs1 quit Ping timeout: 252 seconds 1330680318 #crosstool-ng daggs1 join 1330680786 - daggs1 quit Ping timeout: 272 seconds 1330681552 #crosstool-ng daggs1 join 1330683534 - daggs1 quit Ping timeout: 245 seconds 1330684369 #crosstool-ng daggs1 join 1330686015 - daggs1 quit Quit: Leaving 1330694690 #crosstool-ng smartin join 1330708688 #crosstool-ng daggs1 join 1330714751 #crosstool-ng bhundven join 1330714952 #crosstool-ng bhundven say got a mercurial question for anyone willing to help. I have a patch for the eglibc-2.15 issue I mentioned on the mailing list. I did 'hg qnew -e eglibc_remove_test_except_builtin' then put my change in the right spot in the patches/eglibc/2_15 directory. 'hg qrefresh', and the file is not in the queue i made. Anyone know how to add a new file to a queue, or does that only work for edits? 1330716304 - smartin quit 1330717669 #crosstool-ng bhundven say well, anyways. I've got an e500mc 32-bit mutlilib (hard-float/soft-float) toolchain working now. 1330718714 #crosstool-ng smartin join 1330721485 #crosstool-ng bhundven say GCC 4.7.0 (changes) Status: 2012-03-02 (frozen for release). 1330724720 #crosstool-ng y_morin join 1330724794 #crosstool-ng bhundven say hey y_morin 1330724808 #crosstool-ng y_morin say bhundven: hello there! :-) 1330724856 #crosstool-ng y_morin say bhundven: did you run 'hg add file-to-add' before qrefresh? 1330724865 #crosstool-ng bhundven say oh 1330724867 #crosstool-ng bhundven say hehe 1330724871 #crosstool-ng bhundven say no 1330724877 #crosstool-ng bhundven say thanks 1330724888 #crosstool-ng y_morin say bhundven: 'hg addremove' does all the 'adds' and 'removes' it can find, too. 1330724901 #crosstool-ng bhundven say ok. good tip 1330724988 #crosstool-ng bhundven say on x86_64, building multlib for x86_64/x86_32, in headers & start files for 32: 1330724991 #crosstool-ng bhundven say [CFG ] checking for x86-64 TLS support... no 1330724994 #crosstool-ng bhundven say [ERROR] configure: error: the assembler must support TLS 1330725023 #crosstool-ng y_morin say bhundven: yes, multilib failed on me, too. I did not have the incentive to investigate yet... 1330725089 #crosstool-ng bhundven say but, I did get a 32-bit multilib e500mc that outputs both hard and soft float. works great. 1330725114 #crosstool-ng y_morin say bhundven: I meant, multilib for x86 target. 1330725117 #crosstool-ng bhundven say right 1330725148 #crosstool-ng bhundven say I'm trying again with the binutils-multiarch package installed 1330725200 #crosstool-ng y_morin say bhundven: last I tried, I had to install the lib32 packages on my x86_64 host, otherwise, it can not find (something...). 1330725248 #crosstool-ng y_morin say It should not be required to install such host packages to be able to build a multilib toolchain. gcc should be able to build its own internal xgcc properly... 1330725251 #crosstool-ng y_morin say Sigh... 1330725289 #crosstool-ng bhundven say yea, I keep getting told about this perfect world... can't find it :D 1330725309 #crosstool-ng y_morin tell smiles... :-) 1330726063 #crosstool-ng bhundven say configure:24: x86_64-unknown-linux-gnu-gcc -m32 -c -U_FORTIFY_SOURCE -O2 conftest.s 1>&5 1330726066 #crosstool-ng bhundven say conftest.s: Assembler messages: 1330726069 #crosstool-ng bhundven say conftest.s:8: Error: bad register name `%rip)' 1330726075 #crosstool-ng bhundven say eh, ic now 1330726146 #crosstool-ng bhundven say eglibc issue 1330726246 #crosstool-ng y_morin say bhundven: if you do not know which to blame, start with gcc. If not gcc's fault, it's glibc's. If still not glibc's, it's binutils'. And so on.. until you bash me! ;-) 1330726286 #crosstool-ng bhundven say it's eglibc's fault. It's using 64-bit assmebly test to check for tls when building the 32-bit multilib part 1330726312 #crosstool-ng bhundven say well 1330726316 #crosstool-ng bhundven say (e)glibc 1330728720 - smartin quit 1330729124 #crosstool-ng bhundven say http://cross-lfs.org/view/svn/x86_64/cross-tools/variables.html 1330729126 #crosstool-ng bhundven say eh 1330729142 #crosstool-ng bhundven say lfs uses a different --target for 32-bit in multilib 1330729163 #crosstool-ng bhundven say and that would make sense 1330729445 #crosstool-ng y_morin say bhundven: does binutils 2.22 really require cloog, now ? 1330729454 #crosstool-ng y_morin say (and ppl?) 1330729462 #crosstool-ng bhundven say idk, it still checks for them 1330729575 #crosstool-ng bhundven say y_morin: so that was the problem. the --target needs to be i686-*-linux-* for the 32-bit part of the multilib components 1330729587 #crosstool-ng bhundven say or it runs the wrong autoconf tests 1330729643 #crosstool-ng y_morin say bhundven: OK, but that's just stupid. IIRC, there is a switch to gcc's or glibc's ./configure that specifies the 32-bit tuple... Lemme check... 1330729731 #crosstool-ng y_morin say bhundven: forget it, I was wrong. I was thinking about the --with-{arch,tune,cpu}-{32,64} flags to gcc's ./configure. Never mind... 1330729784 #crosstool-ng y_morin say bhundven: so, in case the multilib is not the same bitness as the default, we have to tweak the tuple, right? 1330729802 #crosstool-ng bhundven say right 1330729810 #crosstool-ng bhundven say I'm working on a patch for x86 1330729824 #crosstool-ng bhundven say I'll look at ppc later 1330729866 #crosstool-ng y_morin say bhundven: OK, thanks. I think the way to go would be to have each arch set both tuple, the default one, and the alternate multilib one if needed. 1330729878 #crosstool-ng bhundven say yes, if CT_MULTILIB 1330729890 #crosstool-ng y_morin say then components such as glibc could choose which to use (depending on... carefully crafted tests, I guess...) 1330729937 #crosstool-ng bhundven say hmm, this might be more tricky then I initially thought 1330729944 #crosstool-ng bhundven say hehe 1330729964 #crosstool-ng y_morin say yes, it's not gonna be easy... 1330730012 #crosstool-ng bhundven say for instance. on e500mc/e500v2, if you build multilib, you get soft or hard float. if you build e500mc64, you get 32 and 64, soft and hard float. 1330730023 #crosstool-ng y_morin say bhundven: BTW, do you know anything about how gcc behaves for ARM+EABI? I tried multilib on ARM EABI, but ./configure decides there was no multilib available.... 1330730039 #crosstool-ng bhundven say hmm 1330730045 #crosstool-ng bhundven say idk, I'd have to try it 1330730066 #crosstool-ng y_morin say bhundven: I think the test should check if -m{32,64} is set, and if it is and does not match the default bitness, switch to using the alternate tuple if set. 1330730080 #crosstool-ng bhundven say yea 1330730086 #crosstool-ng y_morin say bhundven: forget about ARM if you don;t know, I was just asking just in case... 1330730148 #crosstool-ng y_morin say (OK, just built a new ARM toolchain with latest components: gcc-4.6.2-linaro uclibc-0.9.33, binutils-2.22, kernel headers 3.2.6. Woohoo! :-) ) 1330730257 #crosstool-ng bhundven say hehe, nice 1330730291 #crosstool-ng bhundven say got gcc-4.6.3, eglibc-2.15, binutils-2.22, and linux-3.2.8 on my e500mc and x86_64 builds 1330730292 #crosstool-ng y_morin say With NPTL ! :-) 1330730305 #crosstool-ng bhundven say oh, yea, uclibc fixed that, huh. 1330730306 #crosstool-ng y_morin say bhundven: Nice! :-) 1330730332 #crosstool-ng bhundven say e500mc was much easier then e500v2 1330730347 #crosstool-ng bhundven say still has the -Os issue though 1330730782 #crosstool-ng bhundven say hmm 1330730816 #crosstool-ng bhundven say another problem is x86_64-multilib vs. i386-multilib 1330730881 #crosstool-ng bhundven say x86_64 being, default to x86_64, but able to output 32-bit with -m32, and i386-multilib being default 32-bit, but able to output x86_64 -m64 binaries 1330730929 #crosstool-ng y_morin say bhundven: yes, the test should be smthg like: 1330730951 #crosstool-ng bhundven say the variable ${multilib} will contain m32 1330730959 #crosstool-ng bhundven say on x86_64 multilib 1330730961 #crosstool-ng y_morin say - if (multilib args contain -m32) and (arch is 64), then use alternate tuple 1330730967 #crosstool-ng bhundven say yup 1330730972 #crosstool-ng bhundven say jinx 1330730986 #crosstool-ng y_morin say - and conversely for 64/32... 1330730996 #crosstool-ng bhundven say yup 1330731433 #crosstool-ng y_morin say bhundven: I think we could also do something like that: http://code.bulix.org/xpnt1b-81169 1330731466 #crosstool-ng y_morin say bhundven: for those arch that accept the --with-{arch,tune,cpu}-{32,64} configure flags of gcc. 1330731483 #crosstool-ng y_morin say It's just a brain-dump, of course! :-] 1330731500 #crosstool-ng bhundven say well, we get the multilibs variable when we get done with the core-cc-1 1330731515 #crosstool-ng bhundven say just before we start on headers & start files 1330731549 #crosstool-ng bhundven say probably better to get the alternate from gcc then some configuration that someone can screw up 1330731552 #crosstool-ng y_morin say bhundven: nope, I was refering to the default -mcpu -mtune and -march flags 1330731562 #crosstool-ng bhundven say oh 1330731576 #crosstool-ng bhundven tell looks again 1330731593 #crosstool-ng y_morin say bhundven: http://gcc.gnu.org/install/configure.html <- look for --with-arch-32 (and the likes) 1330731599 #crosstool-ng bhundven say yea 1330731639 #crosstool-ng y_morin tell runs one last build before heading to bed... 1330731651 #crosstool-ng bhundven say heh 1330731719 #crosstool-ng y_morin tell is trying to start working on a freedombox-like device... 1330731768 #crosstool-ng bhundven say hmm 1330731774 #crosstool-ng bhundven say never heard of that before 1330731780 #crosstool-ng bhundven say just found the website 1330731828 #crosstool-ng y_morin say bhundven: it's steered by Eben Moglen, of SFLC (https://www.softwarefreedom.org/) 1330732021 #crosstool-ng y_morin say ..../arm-unknown-linux-uclibcgnueabi/bin/ld: BFD (crosstool-NG hg+default-4fcedd2c14b2) 2.22 assertion fail ..../binutils-2.22/bfd/elf32-arm.c:12049 1330732031 #crosstool-ng y_morin say Not too good... :-( 1330732049 #crosstool-ng y_morin say OK, time for some rest. 1330732052 #crosstool-ng y_morin say See ya later! 1330732054 - y_morin quit Quit: Quitting... 1330732106 #crosstool-ng bhundven say g'night