Re: [gentoo-dev] LANG=en_GB.UTF-8 by default
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Mr. Aaron W. Swenson schrieb: P.S. would be nice to have a wd_WD.UTF-8 with WD standing for world, just a country is so 1900 wd_WD.UTF-8 is certainly a no go. WD doesn't match any ISO country code. To support it, we'd have to create the necessary supporting files and that would lead to a lot of work and headaches trying to determine what should be where in what order, et cetera. C or POSIX does not match any country code either. FWIW, Debian has patched their glibc fork to remove the charset restriction on the C locale, and added C.UTF-8. It has the advantage of not messing with transliteration as LC_CTYPE=en_US.UTF-8 would. http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=609306 Best regards, Chí-Thanh Christopher Nguyễn -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.17 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk9CFw8ACgkQ+gvH2voEPRBz1ACdG7XqIJ21D9hBA6e+bpKPGiXq AY8An0osz/G2PnzKnAGOLw2q9UzW7ChW =kuR0 -END PGP SIGNATURE-
[gentoo-dev] Re: LANG=en_GB.UTF-8 by default
On 20/02/2012 07:47, Fabian Groffen wrote: On 20-02-2012 03:07:33 +, Kerin Millar wrote: I know that adding LANG=POSIX doesn't do anything in this case but I have a feeling that its presence would be instructive to new users. If a user is asked to configure something which isn't present, it often generates questions which might otherwise be avoided. I've changed en_US.UTF-8 to en_US.utf8 there for similar reasons. I don't understand. UTF-8 is the codeset, that utf8 is recognised as the same thing is IMO a GNUism. glibc understands UTF-8 perfectly fine these days, so it should preferably be used instead. (Even the man-page, utf8(7), suggests that.) Most users don't read man pages. The rationale was that the user can copy-paste exactly what they see from locale -a, which might diminish the number of questions asked about it via mainstream support channels, as well as simplifying the instructions in the sample comment. It was just a thought; no big deal. --Kerin
[gentoo-dev] Re: [gentoo-commits] gentoo-x86 commit in dev-db/libiodbc: libiodbc-3.52.7-r1.ebuild ChangeLog
On 02/20/2012 05:57 PM, Tomas Chvatal (scarabeus) wrote:
scarabeus12/02/20 15:57:31
Modified: libiodbc-3.52.7-r1.ebuild ChangeLog
Log:
Use single thread build as it is broken in paralel on some machines wrt
bug#405029.
(Portage version: 2.2.0_alpha87/cvs/Linux x86_64)
Revision ChangesPath
1.2 dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild
file :
http://sources.gentoo.org/viewvc.cgi/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild?rev=1.2view=markup
plain:
http://sources.gentoo.org/viewvc.cgi/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild?rev=1.2content-type=text/plain
diff :
http://sources.gentoo.org/viewvc.cgi/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild?r1=1.1r2=1.2
Index: libiodbc-3.52.7-r1.ebuild
===
RCS file: /var/cvsroot/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild,v
retrieving revision 1.1
retrieving revision 1.2
diff -u -r1.1 -r1.2
--- libiodbc-3.52.7-r1.ebuild 20 Feb 2012 10:21:59 - 1.1
+++ libiodbc-3.52.7-r1.ebuild 20 Feb 2012 15:57:31 - 1.2
@@ -1,6 +1,6 @@
# Copyright 1999-2012 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
-# $Header: /var/cvsroot/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild,v
1.1 2012/02/20 10:21:59 scarabeus Exp $
+# $Header: /var/cvsroot/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild,v
1.2 2012/02/20 15:57:31 scarabeus Exp $
EAPI=4
@@ -22,6 +22,8 @@
DOCS=AUTHORS ChangeLog NEWS README
+MAKEOPTS=-j1
+
Why not use MAKEOPTS=${MAKEOPTS} -j1 to allow other make flags get
respected?
[gentoo-dev] Gentoo Janitor scripts
Hi,
Since I plan to use the remote remote-id tag for euscan, and I already
use SRC_URI but I'd like all ebuild to use mirrors, I've wrote to
scripts to cleanup your ebuilds and metadata.
There are available here: https://github.com/iksaif/portage-janitor
Here is what you can do with them:
python remoteids.py --diff pycuda Test-Tester Alien-SDL ostinato
--- a/dev-python/pycuda/metadata.xml
+++ b/dev-python/pycuda/metadata.xml
@@ -4,4 +4,7 @@
maintainer
emailsp...@gentoo.org/email
/maintainer
+upstream
+remote-id type=pypipycuda/remote-id
+/upstream
/pkgmetadata
--- a/dev-perl/Alien-SDL/metadata.xml
+++ b/dev-perl/Alien-SDL/metadata.xml
@@ -7,4 +7,7 @@
emailssuomi...@gentoo.org/email
nameSamuli Suominen/name
/maintainer
+ upstream
+remote-id type=cpanAlien-SDL/remote-id
+ /upstream
/pkgmetadata
--- a/net-analyzer/ostinato/metadata.xml
+++ b/net-analyzer/ostinato/metadata.xml
@@ -7,5 +7,7 @@
/maintainer
longdescription lang=en
/longdescription
+upstream
+remote-id type=google-codeostinato/remote-id
+/upstream
/pkgmetadata
$ eix -C dev-python --only-names | python mirrors.py --diff
--- a/dev-python/asciitable/asciitable-0.8.0.ebuild
+++ b/dev-python/asciitable/asciitable-0.8.0.ebuild
@@ -11,7 +11,7 @@
DESCRIPTION=An extensible ASCII table reader
HOMEPAGE=http://pypi.python.org/pypi/asciitable
http://cxc.harvard.edu/contrib/asciitable;
-SRC_URI=http://pypi.python.org/packages/source/a/${PN}/${P}.tar.gz;
+SRC_URI=mirror://pypi/a/${PN}/${P}.tar.gz
LICENSE=GPL-2
SLOT=0
--- a/dev-python/cosmolopy/cosmolopy-0.1.102.ebuild
+++ b/dev-python/cosmolopy/cosmolopy-0.1.102.ebuild
@@ -15,7 +15,7 @@
DESCRIPTION=Cosmology routines built on NumPy/SciPy
HOMEPAGE=http://roban.github.com/CosmoloPy/
http://pypi.python.org/pypi/CosmoloPy;
-SRC_URI=http://pypi.python.org/packages/source/C/${MY_PN}/${MY_P}.tar.gz;
+SRC_URI=mirror://pypi/C/${MY_PN}/${MY_P}.tar.gz
LICENSE=MIT
SLOT=0
Feel free to test them, and if they are broken I'll gladly accept a patch :).
Maybe some bits could be integrated to repoman...
Thanks,
--
Corentin Chary
http://xf.iksaif.net
Re: [gentoo-dev] Help on getting media-libs/svgalib fixed
Maybe it's time to just punt svgalib? There are only 46 ebuilds that use it (some, optionally). On Sun, Feb 19, 2012 at 6:30 AM, Pacho Ramos pa...@gentoo.org wrote: Hello You can see current opened bugs for svgalib here: https://bugs.gentoo.org/buglist.cgi?quicksearch=media-libs% 2Fsvgalib;list_id=812773 Most of them already contain a patch that is supposed to fix each bug report, the problem is that svgalib doesn't build at all on amd64 and, then, would be interesting if anybody with a x86 system could check if patches fix the problems and commit them. Thanks a lot :-)
Re: [gentoo-dev] Re: [gentoo-commits] gentoo-x86 commit in dev-db/libiodbc: libiodbc-3.52.7-r1.ebuild ChangeLog
El lun, 20-02-2012 a las 17:57 +0200, Samuli Suominen escribió:
On 02/20/2012 05:57 PM, Tomas Chvatal (scarabeus) wrote:
scarabeus12/02/20 15:57:31
Modified: libiodbc-3.52.7-r1.ebuild ChangeLog
Log:
Use single thread build as it is broken in paralel on some machines wrt
bug#405029.
(Portage version: 2.2.0_alpha87/cvs/Linux x86_64)
Revision ChangesPath
1.2 dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild
file :
http://sources.gentoo.org/viewvc.cgi/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild?rev=1.2view=markup
plain:
http://sources.gentoo.org/viewvc.cgi/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild?rev=1.2content-type=text/plain
diff :
http://sources.gentoo.org/viewvc.cgi/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild?r1=1.1r2=1.2
Index: libiodbc-3.52.7-r1.ebuild
===
RCS file:
/var/cvsroot/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild,v
retrieving revision 1.1
retrieving revision 1.2
diff -u -r1.1 -r1.2
--- libiodbc-3.52.7-r1.ebuild 20 Feb 2012 10:21:59 - 1.1
+++ libiodbc-3.52.7-r1.ebuild 20 Feb 2012 15:57:31 - 1.2
@@ -1,6 +1,6 @@
# Copyright 1999-2012 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
-# $Header:
/var/cvsroot/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild,v 1.1
2012/02/20 10:21:59 scarabeus Exp $
+# $Header:
/var/cvsroot/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild,v 1.2
2012/02/20 15:57:31 scarabeus Exp $
EAPI=4
@@ -22,6 +22,8 @@
DOCS=AUTHORS ChangeLog NEWS README
+MAKEOPTS=-j1
+
Why not use MAKEOPTS=${MAKEOPTS} -j1 to allow other make flags get
respected?
About this, I would also like to know if we are allowed to use MAKEOPTS
+=... (and the same for other variables in ebuild) instead or it should
be avoided for some reason.
Thanks for the info :)
signature.asc
Description: This is a digitally signed message part
Re: [gentoo-dev] Help on getting media-libs/svgalib fixed
El lun, 20-02-2012 a las 13:09 -0500, Michael Sterrett escribió: Maybe it's time to just punt svgalib? There are only 46 ebuilds that use it (some, optionally). On Sun, Feb 19, 2012 at 6:30 AM, Pacho Ramos pa...@gentoo.org wrote: Hello You can see current opened bugs for svgalib here: https://bugs.gentoo.org/buglist.cgi?quicksearch=media-libs% 2Fsvgalib;list_id=812773 Most of them already contain a patch that is supposed to fix each bug report, the problem is that svgalib doesn't build at all on amd64 and, then, would be interesting if anybody with a x86 system could check if patches fix the problems and commit them. Thanks a lot :-) The problem is that users CCed on their bug reports have provided patches and fixes for them and would probably get angry if we punt them without even applying the patches to the tree (but I don't want to commit them as I cannot even test them at build time due being x86 specific). Also, looks like upstream is dead, but some distributions are still providing it :-/ signature.asc Description: This is a digitally signed message part
Re: [gentoo-dev] Re: [gentoo-commits] gentoo-x86 commit in dev-db/libiodbc: libiodbc-3.52.7-r1.ebuild ChangeLog
On 02/20/2012 09:36 PM, Pacho Ramos wrote:
El lun, 20-02-2012 a las 17:57 +0200, Samuli Suominen escribió:
On 02/20/2012 05:57 PM, Tomas Chvatal (scarabeus) wrote:
scarabeus12/02/20 15:57:31
Modified: libiodbc-3.52.7-r1.ebuild ChangeLog
Log:
Use single thread build as it is broken in paralel on some machines wrt
bug#405029.
(Portage version: 2.2.0_alpha87/cvs/Linux x86_64)
Revision ChangesPath
1.2 dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild
file :
http://sources.gentoo.org/viewvc.cgi/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild?rev=1.2view=markup
plain:
http://sources.gentoo.org/viewvc.cgi/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild?rev=1.2content-type=text/plain
diff :
http://sources.gentoo.org/viewvc.cgi/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild?r1=1.1r2=1.2
Index: libiodbc-3.52.7-r1.ebuild
===
RCS file: /var/cvsroot/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild,v
retrieving revision 1.1
retrieving revision 1.2
diff -u -r1.1 -r1.2
--- libiodbc-3.52.7-r1.ebuild 20 Feb 2012 10:21:59 - 1.1
+++ libiodbc-3.52.7-r1.ebuild 20 Feb 2012 15:57:31 - 1.2
@@ -1,6 +1,6 @@
# Copyright 1999-2012 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
-# $Header: /var/cvsroot/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild,v
1.1 2012/02/20 10:21:59 scarabeus Exp $
+# $Header: /var/cvsroot/gentoo-x86/dev-db/libiodbc/libiodbc-3.52.7-r1.ebuild,v
1.2 2012/02/20 15:57:31 scarabeus Exp $
EAPI=4
@@ -22,6 +22,8 @@
DOCS=AUTHORS ChangeLog NEWS README
+MAKEOPTS=-j1
+
Why not use MAKEOPTS=${MAKEOPTS} -j1 to allow other make flags get
respected?
About this, I would also like to know if we are allowed to use MAKEOPTS
+=... (and the same for other variables in ebuild) instead or it should
be avoided for some reason.
Thanks for the info :)
+= wasn't allowed because it isn't compatible with bash-3.1 (and
earlier) if I remember correctly
but that ship has long sailed and += syntax is used all over the tree
(and by the package manager itself!)
so feel free :)
- Samuli
[gentoo-dev] Re: Packages/build systems not honoring LINGUAS and a sane solution
On Fri, 17 Feb 2012 22:58:06 +0100 Piotr Szymaniak szar...@grubelek.pl wrote: localepurge will be removed from portage [1]. As I was (/am) heavy user of it I found it funny that linguas takes care of the proper locale installation [2]. Maybe it should, but there's some major failure in lots of packages. I don't think we should drop it unless there's a reason to. It works fine and people obviously use it (me too!). So I filled few dozen bugs about this, but it's not always an easy task to make a package respect LINGUAS. As we talked about it on #gentoo-bugs there were few suggestions to solve this issue. Not directly related to solving, but will help (by ssuominen): make a tracker bug for offending packages Yes, please. Once these get fixed then we can drop localepurge. And please ignore hyperactive bug wranglers who think they get to decide when you've filed enough bugs. We want these reports. -- fonts, gcc-porting toolchain, wxwidgets @ gentoo.org signature.asc Description: PGP signature
Re: [gentoo-dev] Help on getting media-libs/svgalib fixed
On Mon, Feb 20, 2012 at 2:38 PM, Pacho Ramos pa...@gentoo.org wrote: The problem is that users CCed on their bug reports have provided patches and fixes for them and would probably get angry if we punt them without even applying the patches to the tree (but I don't want to commit them as I cannot even test them at build time due being x86 specific). It seems to me that the package still needs a maintainer. That could be a developer, or a proxy-maintainer if one of those users wants to commit to tending it. If nobody wants to step up, and the package is buggy, then treecleaning is the only recourse. Rich
[gentoo-dev] About gcc-4.6 unmasking
I don't know if this has been discussed before but, what issues are preventing us from unmasking gcc-4.6 (and think on a near stabilization)? I have read hardmask message but it simply explains that it's masked for testing purposes :-/ Thanks a lot for the info signature.asc Description: This is a digitally signed message part
Re: [gentoo-dev] About gcc-4.6 unmasking
On 20.02.2012 21:34, Pacho Ramos wrote: I don't know if this has been discussed before but, what issues are preventing us from unmasking gcc-4.6 (and think on a near stabilization)? I have read hardmask message but it simply explains that it's masked for testing purposes :-/ Thanks a lot for the info Here was the last one http://archives.gentoo.org/gentoo-dev/msg_b6db68b41a4b318ea2122fb982c10dfb.xml For me it worked fine for months now. justin
Re: [gentoo-dev] About gcc-4.6 unmasking
El lun, 20-02-2012 a las 21:41 +0100, Justin escribió: On 20.02.2012 21:34, Pacho Ramos wrote: I don't know if this has been discussed before but, what issues are preventing us from unmasking gcc-4.6 (and think on a near stabilization)? I have read hardmask message but it simply explains that it's masked for testing purposes :-/ Thanks a lot for the info Here was the last one http://archives.gentoo.org/gentoo-dev/msg_b6db68b41a4b318ea2122fb982c10dfb.xml For me it worked fine for months now. justin Bleh, looks like grub is blocking this :(, will need to wait then (or maybe move to grub2 ;)) signature.asc Description: This is a digitally signed message part
Re: [gentoo-dev] Re: Packages/build systems not honoring LINGUAS and a sane solution
On Mon, 20 Feb 2012 14:00:31 -0600 Ryan Hill dirtye...@gentoo.org wrote: Yes, please. Once these get fixed then we can drop localepurge. That's a lot of bugs to fix, and the way LINGUAS now works, we could at least do with a QA check that helps figure out which packages break when new translations are added to new releases, or when old translations get dropped. And please ignore hyperactive bug wranglers who think they get to decide when you've filed enough bugs. We want these reports. Give the guy bug assignment privileges and move on. And/or get the automated check in place, and make sure package maintainers see them. jer
Re: [gentoo-dev] About gcc-4.6 unmasking
We had a chat about this in #gentoo-dev the other night. I might come up with a solution as part of the ZFS stuff that I am doing, but it won't happen for at least a month. With that said, it doesn't look like GRUB is the only blocker: https://bugs.gentoo.org/show_bug.cgi?id=gcc-4.6 On Mon, Feb 20, 2012 at 3:49 PM, Pacho Ramos pa...@gentoo.org wrote: El lun, 20-02-2012 a las 21:41 +0100, Justin escribió: On 20.02.2012 21:34, Pacho Ramos wrote: I don't know if this has been discussed before but, what issues are preventing us from unmasking gcc-4.6 (and think on a near stabilization)? I have read hardmask message but it simply explains that it's masked for testing purposes :-/ Thanks a lot for the info Here was the last one http://archives.gentoo.org/gentoo-dev/msg_b6db68b41a4b318ea2122fb982c10dfb.xml For me it worked fine for months now. justin Bleh, looks like grub is blocking this :(, will need to wait then (or maybe move to grub2 ;))
Re: [gentoo-dev] Re: [gentoo-commits] gentoo-x86 commit in dev-db/libiodbc: libiodbc-3.52.7-r1.ebuild ChangeLog
On Mon, 20 Feb 2012, Samuli Suominen wrote: About this, I would also like to know if we are allowed to use MAKEOPTS +=... (and the same for other variables in ebuild) instead or it should be avoided for some reason. += wasn't allowed because it isn't compatible with bash-3.1 (and earlier) if I remember correctly This is almost right: The += operator was introduced with bash 3.1, so previously it couldn't be used in the tree because compatibility with 3.0 was required. but that ship has long sailed and += syntax is used all over the tree (and by the package manager itself!) so feel free :) It's officially allowed since more than two years ago: http://www.gentoo.org/proj/en/council/meeting-logs/20091109-summary.txt Ulrich
Re: [gentoo-dev] Re: Packages/build systems not honoring LINGUAS and a sane solution
On Mon, Feb 20, 2012 at 1:26 PM, Jeroen Roovers j...@gentoo.org wrote: On Mon, 20 Feb 2012 14:00:31 -0600 Ryan Hill dirtye...@gentoo.org wrote: Yes, please. Once these get fixed then we can drop localepurge. That's a lot of bugs to fix, and the way LINGUAS now works, we could at least do with a QA check that helps figure out which packages break when new translations are added to new releases, or when old translations get dropped. And please ignore hyperactive bug wranglers who think they get to decide when you've filed enough bugs. We want these reports. Give the guy bug assignment privileges and move on. And/or get the automated check in place, and make sure package maintainers see them. I gave him editbugs last week. -A jer
Re: Re: [gentoo-dev] About gcc-4.6 unmasking
Bleh, looks like grub is blocking this :(, will need to wait then (or maybe move to grub2 ;)) Yeah... anyone helping to debug this tricky thingy [*] is likely welcome. Would like to help, but cant do much atm because of real-life work load... [*] https://bugs.gentoo.org/show_bug.cgi?id=360513 -- Andreas K. Huettel Gentoo Linux developer kde, sci, arm, tex, printing signature.asc Description: This is a digitally signed message part.
[gentoo-dev] Re: About gcc-4.6 unmasking
On Mon, 20 Feb 2012 21:34:14 +0100 Pacho Ramos pa...@gentoo.org wrote: I don't know if this has been discussed before but, what issues are preventing us from unmasking gcc-4.6 (and think on a near stabilization)? I have read hardmask message but it simply explains that it's masked for testing purposes :-/ Grub is the only blocker. I don't want to unmask something that makes people's systems unbootable. I'm also out of ideas and open to suggestions. -- fonts, gcc-porting toolchain, wxwidgets @ gentoo.org signature.asc Description: PGP signature
Re: [gentoo-dev] Re: About gcc-4.6 unmasking
On Mon, Feb 20, 2012 at 8:03 PM, Ryan Hill dirtye...@gentoo.org wrote: On Mon, 20 Feb 2012 21:34:14 +0100 Pacho Ramos pa...@gentoo.org wrote: I don't know if this has been discussed before but, what issues are preventing us from unmasking gcc-4.6 (and think on a near stabilization)? I have read hardmask message but it simply explains that it's masked for testing purposes :-/ Grub is the only blocker. I don't want to unmask something that makes people's systems unbootable. I'm also out of ideas and open to suggestions. Is it a bad idea to go ahead and unmask it on architectures that don't use grub?
Re: [gentoo-dev] Re: About gcc-4.6 unmasking
On 02/20/2012 05:03 PM, Ryan Hill wrote: On Mon, 20 Feb 2012 21:34:14 +0100 Pacho Ramos pa...@gentoo.org wrote: I don't know if this has been discussed before but, what issues are preventing us from unmasking gcc-4.6 (and think on a near stabilization)? I have read hardmask message but it simply explains that it's masked for testing purposes :-/ Grub is the only blocker. I don't want to unmask something that makes people's systems unbootable. I'm also out of ideas and open to suggestions. Stabilize grub-1.99, and modify the grub-0.9x ebuilds to die if they can't find a supported compiler. -- Thanks, Zac
Re: [gentoo-dev] Re: About gcc-4.6 unmasking
On Mon, 2012-02-20 at 19:03 -0600, Ryan Hill wrote: On Mon, 20 Feb 2012 21:34:14 +0100 Pacho Ramos pa...@gentoo.org wrote: I don't know if this has been discussed before but, what issues are preventing us from unmasking gcc-4.6 (and think on a near stabilization)? I have read hardmask message but it simply explains that it's masked for testing purposes :-/ Grub is the only blocker. I don't want to unmask something that makes people's systems unbootable. I'm also out of ideas and open to suggestions. gcc is slotted. Is there any reason why we can't simply make grub depend on a working slot of gcc and set CC appropriately in the ebuild? -Alexandre
Re: [gentoo-dev] Re: About gcc-4.6 unmasking
Ryan, I took a look at the problem cited in your bug report. I suggest compiling sys-boot/grub with CFLAGS=-O0 -ggdb3, attaching gdb to grub-install and then watching what happens in the debugger. If you compare runs with a GCC 4.5.3 built stage2 and a GCC 4.6.2 built stage2, you should be able to find the bug. Yours truly, Richard Yao On Mon, Feb 20, 2012 at 8:03 PM, Ryan Hill dirtye...@gentoo.org wrote: On Mon, 20 Feb 2012 21:34:14 +0100 Pacho Ramos pa...@gentoo.org wrote: I don't know if this has been discussed before but, what issues are preventing us from unmasking gcc-4.6 (and think on a near stabilization)? I have read hardmask message but it simply explains that it's masked for testing purposes :-/ Grub is the only blocker. I don't want to unmask something that makes people's systems unbootable. I'm also out of ideas and open to suggestions. -- fonts, gcc-porting toolchain, wxwidgets @ gentoo.org
Re: [gentoo-dev] Re: About gcc-4.6 unmasking
I took a look at the problem cited in your bug report. I suggest compiling sys-boot/grub with CFLAGS=-O0 -ggdb3, attaching gdb to grub-install and then watching what happens in the debugger. If you compare runs with a GCC 4.5.3 built stage2 and a GCC 4.6.2 built stage2, you should be able to find the bug. I should add that I was able to use this technique to fix a bug that I encountered during my initial attempt to port Illumos GRUB a month ago. The code the introduced the Illumos GRUB bug is not present in sys-boot/grub, but I imagine that the same technique should work here. Also, for anyone interested in what happened to the sys-boot/grub-illumos port I mention, there are issues with the generated stage2 binary, grub-install is broken (Solaris uses a separate install-grub tool) and I would prefer to rework the Sun Microsystems code into a patch for sys-boot/grub, but the diff between Illumos GRUB and GRUB 0.97 is a few megabytes in size, so that won't happen this month.
[gentoo-dev] Re: About gcc-4.6 unmasking
On Mon, 20 Feb 2012 20:30:40 -0500 Alexandre Rostovtsev tetrom...@gentoo.org wrote: On Mon, 2012-02-20 at 19:03 -0600, Ryan Hill wrote: Grub is the only blocker. I don't want to unmask something that makes people's systems unbootable. I'm also out of ideas and open to suggestions. gcc is slotted. Is there any reason why we can't simply make grub depend on a working slot of gcc and set CC appropriately in the ebuild? We have no way of forcing an ebuild to be built with a particular version of GCC. This is on purpose, and there are both technical and sociological reasons for it. What we can do is take some kind of action if the compiler is 4.6, such as die with a message to use grub-static instead. -- fonts, gcc-porting toolchain, wxwidgets @ gentoo.org signature.asc Description: PGP signature
[gentoo-dev] Re: About gcc-4.6 unmasking
On Mon, 20 Feb 2012 17:17:30 -0800 Zac Medico zmed...@gentoo.org wrote: On 02/20/2012 05:03 PM, Ryan Hill wrote: On Mon, 20 Feb 2012 21:34:14 +0100 Pacho Ramos pa...@gentoo.org wrote: I don't know if this has been discussed before but, what issues are preventing us from unmasking gcc-4.6 (and think on a near stabilization)? I have read hardmask message but it simply explains that it's masked for testing purposes :-/ Grub is the only blocker. I don't want to unmask something that makes people's systems unbootable. I'm also out of ideas and open to suggestions. Stabilize grub-1.99, and modify the grub-0.9x ebuilds to die if they can't find a supported compiler. What's the state of 1.99? I know someone was working on it recently. We'd also have to update the handbooks. I think it could be several months of work to get it ready, and I'd like to unmask 4.6 last September. -- fonts, gcc-porting toolchain, wxwidgets @ gentoo.org signature.asc Description: PGP signature
[gentoo-dev] Re: About gcc-4.6 unmasking
On Mon, 20 Feb 2012 20:37:39 -0500 Richard Yao r...@cs.stonybrook.edu wrote: Ryan, I took a look at the problem cited in your bug report. I suggest compiling sys-boot/grub with CFLAGS=-O0 -ggdb3, attaching gdb to grub-install and then watching what happens in the debugger. If you compare runs with a GCC 4.5.3 built stage2 and a GCC 4.6.2 built stage2, you should be able to find the bug. Sorry, the bug report is confusing. It's actually two bugs, the first being a miscompiled stage2 causing an error when running grub-install and making the system unbootable. I fixed that back in Sept. The second bug is a continuous boot loop that only seems to manifest on certain machines or configurations. This is the one I'm having trouble with. I should have opened a new report for it, but at the time I thought it was due to fallout from the first patch. The biggest problem is that I can't reproduce it on either of my systems, so I have no way of narrowing it down. FWIW, I did a comparison of /boot/grub/* from a broken system and my own and they are byte-for-byte identical. -- fonts, gcc-porting toolchain, wxwidgets @ gentoo.org signature.asc Description: PGP signature
