GCC 4.6.3 is just release with a number of important fixes. Please bump new ebuild. Thanks!
I was bored last night and made gcc-4.6.3 build successfully. The following tarballs can be renamed from 4.6.2 to 4.6.3 without modification: - gcc-4.6.2-piepatches-v0.5.0.tar.bz2 - gcc-4.6.2-uclibc-patches-1.0.tar.bz2 gcc-4.6.2-patches-1.4.tar.bz2 needs to be unpacked, the following patches removed: - 75_all_arm_4.7.0_strict-volatile-bitfields.patch - 77_all_pr51106_4.6.3_ice-move_insn.patch Those seem to have been committed upstream. These packages didn't apply without porting: - 74_all_gcc46_cloog-dl.patch - 76_all_4.7.0_c-family-headers.patch (kdiff3 called this one malformed in 1.4) I will upload my work so you can replace those two patches and then re-create the tar.bz2 package, I renamed it to gcc-4.6.3-patches-1.0.tar.bz2 The 4.6.2 ebuild works just fine, rename it to 4.6.3 and don't forget to set PATCH_VER="1.0" if you did rename the gentoo patchset too.
Created attachment 303993 [details, diff] 74_all_gcc46_cloog-dl.patch
Created attachment 303995 [details, diff] 76_all_4.7.0_c-family-headers.patch
Works like a charm (But I didnt test the graphite-cloog) Very thanx.
Raw output of a ebuild test attached. Contrib/test_results does not see output ofmake -k check.
Created attachment 304211 [details] FUATUERS=test USE=test ebuild test output
strangly I am not able to compile gcc 4.6.3 as the simple renamong of the ebuild does not help. I stumbled upon strange error gcc # ebuild gcc-4.6.3.ebuild digest Error(s) in metadata for 'sys-devel/gcc-4.6.3': PDEPEND: USE flag 'go' referenced in conditional 'go?' is not in IUSE I bypassed it by adding IUSE="${IUSE} go" in the ebuild but now I get gcc # emerge gcc Calculating dependencies... done! >>> Verifying ebuild manifests >>> Emerging (1 of 1) sys-devel/gcc-4.6.3 from bigvo * gcc-4.6.3.tar.bz2 RMD160 SHA1 SHA256 size ;-) ...[ ok ] * gcc-4.6.3-uclibc-patches-1.0.tar.bz2 RMD160 SHA1 SHA256 size ;-) ...[ ok ] * gcc-4.6.3-patches-1.0.tar.bz2 RMD160 SHA1 SHA256 size ;-) ...[ ok ] * gcc-4.6.3-piepatches-v0.5.0.tar.bz2 RMD160 SHA1 SHA256 size ;-) ...[ ok ] * gcc-4.4.3-specs-0.2.0.tar.bz2 RMD160 SHA1 SHA256 size ;-) ...[ ok ] * ERROR: sys-devel/gcc-4.6.3 failed (setup phase): * Your ebuild needs to set the ETYPE variable * * Call stack: * ebuild.sh, line 85: Called pkg_setup * gcc-4.6.3.ebuild, line 59: Called toolchain_pkg_setup * toolchain.eclass, line 23: Called gcc_pkg_setup * toolchain.eclass, line 783: Called die * The specific snippet of code: * [[ -z ${ETYPE} ]] && die "Your ebuild needs to set the ETYPE variable" What might be a cause of that? In number of years with Gentoo I have not stumbeled upon this error myself nor I have found it in the forums... please advice
Is there some mixup with an old toolchain.eclass going on in your local overlay?
yes that was it. I forgot I had such a file when I copied it from toolchain overlay back in the day. btw ebuild works perfectly and gcc works fine when build using instructions from Andreas Sturmlechner Thnka You both
For some reason the patch 74_all_gcc46_cloog-dl.patch is failing on my system. I don't know if the second patch works yet since it fails before it get to 76_all_4.7.0_c-family-headers.patch. gcc-4.4.3-specs-0.2.0.tar.bz2 RMD160 SHA1 SHA256 size ;-) ... [ ok ] * ecj-4.5.jar RMD160 SHA1 SHA256 size ;-) ... [ ok ] * * LTO support is still experimental and unstable. * Any bugs resulting from the use of LTO will not be fixed. * >>> Unpacking source... >>> Unpacking gcc-4.6.3.tar.bz2 to /var/tmp/portage/sys-devel/gcc-4.6.3/work >>> Unpacking gcc-4.6.3-patches-1.0.tar.bz2 to /var/tmp/portage/sys-devel/gcc-4.6.3/work >>> Unpacking gcc-4.6.3-uclibc-patches-1.0.tar.bz2 to /var/tmp/portage/sys-devel/gcc-4.6.3/work >>> Unpacking gcc-4.6.3-piepatches-v0.5.0.tar.bz2 to /var/tmp/portage/sys-devel/gcc-4.6.3/work >>> Unpacking gcc-4.4.3-specs-0.2.0.tar.bz2 to /var/tmp/portage/sys-devel/gcc-4.6.3/work * Applying Gentoo patches ... * 01_all_joined-cpp-defs.patch ... [ ok ] * 03_all_java-nomulti.patch ... [ ok ] * 08_all_cross-compile.patch ... [ ok ] * 10_all_default-fortify-source.patch ... [ ok ] * 11_all_default-warn-format-security.patch ... [ ok ] * 12_all_default-warn-trampolines.patch ... [ ok ] * 15_all_libgomp-Werror.patch ... [ ok ] * 25_all_alpha-mieee-default.patch ... [ ok ] * 26_all_alpha-asm-mcpu.patch ... [ ok ] * 29_all_arm_armv4t-default.patch ... [ ok ] * 34_all_ia64_note.GNU-stack.patch ... [ ok ] * 38_all_sh_pr24836_all-archs.patch ... [ ok ] * 42_all_superh_default-multilib.patch ... [ ok ] * 50_all_libiberty-asprintf.patch ... [ ok ] * 51_all_libiberty-pic.patch ... [ ok ] * 52_all_netbsd-Bsymbolic.patch ... [ ok ] * 74_all_gcc46_cloog-dl.patch ... * Failed Patch: 74_all_gcc46_cloog-dl.patch ! * ( /var/tmp/portage/sys-devel/gcc-4.6.3/work/patch/74_all_gcc46_cloog-dl.patch ) * * Include in your bugreport the contents of: * * /var/tmp/portage/sys-devel/gcc-4.6.3/temp/74_all_gcc46_cloog-dl.patch.out * ERROR: sys-devel/gcc-4.6.3 failed (unpack phase): * Failed Patch: 74_all_gcc46_cloog-dl.patch! * * Call stack: * ebuild.sh, line 84: Called src_unpack * environment, line 4495: Called toolchain_src_unpack * environment, line 5058: Called gcc_src_unpack * environment, line 2816: Called epatch '/var/tmp/portage/sys-devel/gcc-4.6.3/work/patch' * environment, line 1773: Called die * The specific snippet of code: * die "Failed Patch: ${patchname}!"; * * If you need support, post the output of 'emerge --info =sys-devel/gcc-4.6.3', * the complete build log and the output of 'emerge -pqv =sys-devel/gcc-4.6.3'. * This ebuild is from an overlay named 'local': '/var/lib/layman/local/' * The complete build log is located at '/var/tmp/portage/sys-devel/gcc-4.6.3/temp/build.log'. * The ebuild environment file is located at '/var/tmp/portage/sys-devel/gcc-4.6.3/temp/environment'. * S: '/var/tmp/portage/sys-devel/gcc-4.6.3/work/gcc-4.6.3' >>> Failed to emerge sys-devel/gcc-4.6.3, Log file: >>> '/var/tmp/portage/sys-devel/gcc-4.6.3/temp/build.log'
Created attachment 305513 [details] /var/tmp/portage/sys-devel/gcc-4.6.3/temp/74_all_gcc46_cloog-dl.patch.out
If you guys can wait a couple days you won't have to worry about it.
Sorry about the delay. The patch for PR51106 (77_all_pr51106_4.6.3_ice-move_insn.patch) is causing an ICE elsewhere. We'll wait for an updated patch to go in upstream.
Magnus, I'm including piepatches 0.5.1 in this bump, is that correct?
I fail to understand where the actual problem is. As far as I can see from the referenced GCC bug (http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51106), this issue affects several GCC versions (including GCC 4.6.2).
Any updates on when this may be hitting the tree? I've tried the method Andreas detailed in comment 1 in my local overlay, but I'm having issues with the patches not applying correctly, so emerge fails. I've had few serious problems with this branch (the exception being grub), but have experienced significant speed improvements over the stable 4.5.x branch on my core2 duo laptop. It used to take me over 32+ hours to recompile the 1100+ packages in my setup. Now with 4.6.2 it takes around 26. That's a tremendous improvement IMHO. I love to see if there's further speed improvement with 4.6.3. I'd really like to see this update incorporated soon. Even if it's hard masked like previous versions of gcc-4.6.x. TIA
As soon as I get a day off work. It should be Saturday unless something else goes wrong.
Actually, screw it, I'll do it now.
(In reply to comment #18) > Actually, screw it, I'll do it now. I don't think you added it to CVS... http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/sys-devel/gcc/
The commit failed but I wandered away without noticing. Fixed now.
(In reply to comment #20) > The commit failed but I wandered away without noticing. Fixed now. Thanks a lot!
Thanks, your rock!!!
(In reply to comment #22) > Thanks, your rock!!! I meant you rock!!! LOL!
