Re: Bug#632452: *ERROR*: Could not open file: /dev/ttyp0
On Sun, Jul 03, 2011 at 12:16:08PM +0200, Robert Millan wrote: > 2011/7/3 Axel Beckert : > > Hi Christoph? > > > > Christoph Egger wrote: > >> I'm getting a "*ERROR*: Could not open file: /dev/ttyp0" e.g. when > >> trying to start an emacsclient here after today's updates (sid). Anyone > >> else seeing those as well? Any idea where to report that bug? > > > > Could that be another symptom of http://bugs.debian.org/632452? > > Seems a lot like it, adding it to CC. > > Aurelien, I notice you don't consider this bug release-critical. Do > you plan to let this migrate to wheezy? If there's no fix readily I am seriously annoyed how you very often abuse the bug severities. This is one of the reason why I downgrade this bug, but also as the symptoms you reported are clearly not a grave bug: critical makes unrelated software on the system (or the whole system) break, or causes serious data loss, or introduces a security hole on systems where you install the package. Now with the new facts reported by Christoph, we can probably consider this bug grave, but still not critical. > available, please consider reverting r3435 so that wheezy users are > not affected. IIRC the previous state only caused a harmless warning > with sshd. > It is my plan if I don't see new commits from Petr. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110703170214.gb7...@hall.aurel32.net
Re: [kfreebsd] massive report for uninstallable FUSE packages
On Sat, Jul 16, 2011 at 09:38:35PM +0200, Robert Millan wrote: > After receiving some feedback which allowed me to improve the > template, this is the message I intend to use for the bug reports: > > Package: %package% > Severity: important > User: debian-bsd@lists.debian.org > Usertags: kfreebsd > > Hi > > This package is not installable on kfreebsd-i386 or kfreebsd-amd64 because it > depends unconditionally on fuse-utils. > > If %package% depends on fuse-utils only to ensure that FUSE support is > enabled, > please consider adjusting the dependency to something like: > > Depends: fuse [linux-any] | fuse4bsd [kfreebsd-any] > > (example uses "fuse" because fuse-utils is now a transitional package that > depends on fuse) I think it should be ',' instead of '|' here, there is no need for alternative, the architecture conditionals already do the job. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110720081738.gb26...@volta.aurel32.net
kfreebsd-headers-8.2-1-amd64 provides kernel-headers (was asdfasdf package installation please)
On Fri, Aug 12, 2011 at 01:56:59AM +0200, Axel Beckert wrote:
> Hi,
>
> Samuel Thibault wrote:
> > Please install in the sid chroot of asdfasdf
> >
> > kfreebsd-headers-8-amd64
>
> Discussed the issue with Samuel on IRC:
>
> It's more or less uninstallable. I would have to remove libc0.1-dev:
> libc0.1-dev depends on kfreebsd-kernel-headers (>= 0.11) which
> conflicts with kfreebsd-headers which is provided by
> kfreebsd-headers-8.2-1-amd64 which is a dependency of
> kfreebsd-headers-8-amd64.
kfreebsd-headers used to be the old name of kfreebsd-kernel-headers,
hence the conflicts there. Up to a few weeks ago, it was possible to
co-install the packages, but revision 3592 added the Provides:
| Index: debian/changelog
| ===
| --- debian/changelog(révision 3591)
| +++ debian/changelog(révision 3592)
| @@ -3,8 +3,10 @@
|* Replace msk backport patch with version from 8-STABLE (instead of
| 9-CURRRENT).
|* Fix another FTBFS on hurd-i386 (bogus use of linux macro in aicasm).
| + * Provide "kfreebsd-headers" virtual package in kfreebsd-headers-
| +@version@-@abiname@-@flavor@.
|
| - -- Robert Millan Mon, 11 Jul 2011 18:15:44 +0200
| + -- Robert Millan Fri, 15 Jul 2011 17:45:50 +0200
|
| kfreebsd-8 (8.2-6) unstable; urgency=low
|
| Index: debian/control.flavor.in
| ===
| --- debian/control.flavor.in(révision 3591)
| +++ debian/control.flavor.in(révision 3592)
| @@ -19,7 +19,7 @@
|
| Package: kfreebsd-headers-@version@-@abiname@-@flavor@
| Architecture: any-@arch@
| -Provides: kfreebsd-headers-@major@
| +Provides: kfreebsd-headers-@major@, kfreebsd-headers
| Depends: kfreebsd-headers-@version@-@abiname@ (= ${binary:Version}),
${misc:Depends}
| Description: header files for kernel of FreeBSD @version@
| This package provides the header files for kernel of FreeBSD @version@,
generally
I don't know the reason of this provides, so it's difficult to comment
on it right now, but not being able to install the kernel headers looks
a major issue to me.
> The installed kfreebsd-kernel-headers doesn't seem to suffice. He
> needs /usr/include/sys/ipmi.h which is no package at all according to
> "apt-file search /usr/include/sys/ipmi.h".
>
It looks like to me this file should be added to
kfreebsd-kernel-headers.
Cheers,
Aurelien
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurel...@aurel32.net http://www.aurel32.net
--
To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20110812074913.gk5...@hall.aurel32.net
Bug#570805: [kfreebsd-*] please rebuild elfutils/sid, ignoring the 2 known testsuite failures (Re: transition: liblzma 5)
On Sat, Oct 29, 2011 at 11:38:18AM +0200, Kurt Roeckx wrote: > On Sat, Oct 29, 2011 at 11:07:42AM +0200, Julien Cristau wrote: > > On Sat, Oct 29, 2011 at 02:12:25 -0500, Jonathan Nieder wrote: > > > > > Julien Cristau wrote: > > > > On Wed, Oct 26, 2011 at 12:01:41 +0200, Julien Cristau wrote: > > > > > > >>> libdw1 (DWARF parser for elfutils) > > > >> > > > >> FTBFS on kfreebsd, needs a bug filed. > > > > > > > > Apparently that's #570805 (thanks, Jakub). > > > > > > That's a pair of testsuite failures due to a kernel bug and not a > > > regression in elfutils as far as I can tell. > > > > > > kfreebsd buildd admins, is it possible to schedule a rebuild with > > > DEB_BUILD_OPTIONS=nocheck, or would it be better to request that > > > directly in debian/rules as a temporary workaround? (Please forgive > > > my ignorance.) > > > > > I think this should be worked around in the elfutils package. > > How do you suggest I work around this? I really don't like to > ignore those errors. They're the one that test the platform > specific parts. > > Can I instead suggest someone looks at the kernel and fixes it? The kernel part is not trivial to solve. > It used to work, it works on the porter machines, it just fails > on the buidds. It now fails because of the multiple bind mounts needed by schroot. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111029094523.gc31...@hall.aurel32.net
Bug#646957: "mount -o remount,rw" has no effect with kFreeBSD 9.0
On Sat, Oct 29, 2011 at 02:27:53PM -0400, Benjamin Kaduk wrote:
> On Sat, 29 Oct 2011, Robert Millan wrote:
>
> >2011/10/29 Aurelien Jarno :
> >>Which seems to be kfreebsd-9 related, not busybox related.
> >>
> >>This is very likely to be the same bug than the one fixed in kfreebsd-8
> >>by 107_mount_update.diff. This patch has been disabled in kfreebsd-9
> >>(probably because this part has been changed upstream), so this is where
> >>to look first.
> >
> >Uhm... ISTR it's me who removed that patch during the upgrade. In any
> >case, I re-checked, and it seems to me that the code that made iovec
> >arguments a requirement is no longer present.
>
> Err, I don't see any iovec arguments in the 107_mount_update.diff
> that google finds for me as [Glibc-bsd-commits] r3165.
>
> >
> >I don't have time to look deeper right now though :-(
> >
> >>Until we are sure it's not a kfreebsd-9 bug, it's better to keep the bug
> >>assigned to kfreebsd-9.
>
> Since stock freebsd is documented as requiring -u ("update") when
> changing the mount options of an already-mounted filesystem, it
> seems very likely to be a kfreebsd bug.
>
Requiring -u is a requirement from the userland mount utility. Here we
are talking about the kernel interface. In this case the MNT_UPDATE flag
passed to the nmount() function is ignored if the filesystem was mounted
ro and the update is to mount it rw.
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurel...@aurel32.net http://www.aurel32.net
--
To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20111029233526.gd31...@hall.aurel32.net
Re: (forw) Debian installer failed builds
On Mon, Oct 31, 2011 at 08:02:24AM +0530, Christian PERRIER wrote: > Hello, > > I'm sorry for sending this only now, but it seems that kfreebsd-amd64 > builds of D-I are failing for about 2 weeks with the following > message. I suspect this might be related to changes in kernel in > kfreebsd-amd64 (udebs not rebuilt?) but this is only wild guess. > The problem is that udebs for kernel 8.2 have been uploaded, but the configuration in debian-installer hasn't been changed accordingly. I have just committed the necessary changes, it should build in the next run. Thanks for the notice. Aurelien -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111031100914.gj31...@hall.aurel32.net
Re: Bug#637236: gcc-4.6/gcj-4.6 build failures on kfreebsd-amd64 are back again
On Sun, Oct 30, 2011 at 10:43:03PM +0100, Ludovic Brenta wrote: > Robert Millan writes: > >> Matthias tried to disable parallel builds on kfreebsd-*: > >> > >> ifneq (,$(filter $(DEB_TARGET_ARCH), kfreebsd-amd64 kfreebsd-i386)) > >> USE_NJOBS = no > >> endif > >> > >> but it looks like the DEB_BUILD_OPTIONS on fano overrode him: > >> > >> # Support parallel= in DEB_BUILD_OPTIONS (see #209008) > >> ifneq (,$(filter parallel=%,$(subst $(COMMA), ,$(DEB_BUILD_OPTIONS > >> NJOBS := -j $(subst parallel=,,$(filter parallel=%,$(subst $(COMMA), > >> ,$(DEB_BUILD_OPTIONS > >> endif > >> > >> The package FTBFS on fano (the kfreebsd-amd64 buildd). > >> > >> As Matthias is travelling, I'll try to upload 4.6.1-3 this weekend. > > > > Cool, thanks! > > The build on fano failed again despite using only one job :( It was expected, but at least it allows to compare the build logs more easily. > At this point I'll have to leave it to kFreeBSD porters to find the real > cause. There must be a difference between the buildds and > asdfasdf.debian.net. It's worse than that, gcc-4.6 builds fine in the build daemon if the build is done directly in the chroot or started directly by calling sbuild. The problem only appears when the buildd process calls sbuild, which makes things difficult to debug. The next step is probably to do the next source uploads with another architecture than amd64, to see if the problem is reproducible on the amd64 build daemon. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111031113340.gk31...@hall.aurel32.net
Re: (forw) Debian installer failed builds
On Mon, Oct 31, 2011 at 05:05:36PM +0100, Robert Millan wrote: > 2011/10/31 Aurelien Jarno : > > The problem is that udebs for kernel 8.2 have been uploaded, but the > > configuration in debian-installer hasn't been changed accordingly. > > > > I have just committed the necessary changes, it should build in the next > > run. > > Hi Aurelien, > > It seems that kfreebsd-kernel-di-i386 is still outdated. Do you plan > to handle this one? > Oh, I didn't realized you only uploaded kfreebsd-kernel-di-amd64. I am going to work on it soon. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111031171230.ge13...@hall.aurel32.net
Re: (forw) Debian installer failed builds
Le 02/11/2011 12:19, Robert Millan a écrit : > 2011/10/31 Aurelien Jarno : >> Oh, I didn't realized you only uploaded kfreebsd-kernel-di-amd64. I am >> going to work on it soon. > > If you update kfreebsd-kernel-di-i386, beware not to use 8.2-12 (still > not uploaded), it'd break a few places in D-I (I'm currently looking > at those). > I already uploaded it two days ago, so I guess it's fine. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4eb13443.9030...@aurel32.net
Bug#658185: (no subject)
On Thu, Feb 02, 2012 at 06:28:52PM +, Robert Millan wrote: > El 2 de febrer de 2012 18:27, Robert Millan ha escrit: > > I believe that a rebuild of smartmontools with libcam-dev 9.0 (just > > accepted into unstable) would solve the problem. > > > > Could someone test? > > We need the rebuild for ABI transition anyway, so it's no harm to > queue it right away. > Does it means that after that it won't work with kernel 8.x anymore? -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120202185319.gf2...@hall.aurel32.net
Bug#658185: freebsd-9: freebsd-9 kernel + smartmontools: "error sending CAMIOCOMMAND ioctl: Inappropriate ioctl for device"
On Thu, Feb 02, 2012 at 06:28:52PM +, Robert Millan wrote: > El 2 de febrer de 2012 18:27, Robert Millan ha escrit: > > I believe that a rebuild of smartmontools with libcam-dev 9.0 (just > > accepted into unstable) would solve the problem. > > > > Could someone test? > > We need the rebuild for ABI transition anyway, so it's no harm to > queue it right away. > Does it mean that after a rebuild, it won't work anymore with a kernel 8.x? For binNMUs, please ask the release team who can handle the transition. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120202185429.gg2...@hall.aurel32.net
Re: Bug#658185: freebsd-9: freebsd-9 kernel + smartmontools: "error sending CAMIOCOMMAND ioctl: Inappropriate ioctl for device"
On Thu, Feb 02, 2012 at 09:41:46PM +, Robert Millan wrote: > El 2 de febrer de 2012 18:54, Aurelien Jarno ha escrit: > >> We need the rebuild for ABI transition anyway, so it's no harm to > >> queue it right away. > >> > > > > Does it mean that after a rebuild, it won't work anymore with a kernel > > 8.x? > > Probably. After several hours I've given up trying to make userland > compatible with both kernels. Specifically with CAM, the required > patch would be very intrusive (a struct is modified carelessly, which > would require duplicating the struct, all structs that include it, and > versioning in all the functions that rely on any of that). If someone > volunteers feel free to supply a patch, but IMHO with manpower being > so scarce this is a dead end. > The problem there is that given that all 9.x packages have been already pushed to the archive with ABI changes and so on, we *must* switch the default kernel for wheezy to 9.x. I don't say it's a bad decision, but I would have prefer to have some discussion about, it including the possible consequences of such a choice, instead of getting to the point of not having any other choice. Now we have no choice than making a real plan for switching to 9.x kernel: - We have to make sure users are using wheezy/sid with a 9.x kernel. - We have to provide an upgrade path for users, including the best moment to switch from one kernel to another in the release notes. - The build daemons are going to stay with the 8.1 kernel up to the release of wheezy. Will it work? Sometimes after they are going to switch to a 9.x kernel, but they should still be able to build squeeze packages. Will it work? Who wants to work on addressing these points? -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120205105029.gc7...@ohm.aurel32.net
Re: [kfreebsd-*] please rebuild elfutils/sid, ignoring the 2 known, testsuite failures (Re: transition: liblzma 5)
On Fri, Feb 03, 2012 at 12:33:33AM +, peter green wrote: > Kurt Roeckx wrote: > >Can I instead suggest someone looks at the kernel and fixes it? > >It used to work, it works on the porter machines, it just fails > >on the buidds. > > Aurelien Jarno wrote: > >The kernel part is not trivial to solve. > > >It now fails because of the multiple bind mounts needed by schroot. > > This issue was first run into nearly a year ago, and the last post on it > was arround 2 months ago is there any progress on solving the root > cause? No, it hasn't changed. It's due to multiple bind mounting in schroot, and therefore is not a usual case, so it's not on top priority of people TODO list. > If not and the maintainers don't feel comfortable disabling the > tests would it be possible to manually build and upload the binnmus > in an environment that doesn't suffer the issue so that the packages > are installable on kfreebsd again and transitions and builds of > dependent packages can proceed? > I can do manual builds and uploads if needed. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120205110653.gd7...@ohm.aurel32.net
Re: Directories created in /tmp are owned by group root
On Sat, Mar 03, 2012 at 12:47:10PM +, Javi Merino wrote: > Hi, > > In kfreebsd-* buildds and porterboxes, directories created in /tmp are > owned by group root. In linux-* buildds and porterboxes they are > owned by the "Debian" group. > > vicho@asdfasdf:/tmp$ mkdir test > vicho@asdfasdf:/tmp$ ls -ld test > drwxr-xr-x 2 vicho root 512 mar 3 13:44 test > vicho@asdfasdf:/tmp$ > > mercurial's testsuite fails because of this and, though I've fixed it > to change the group of the directory after creating it, I was > wondering why is this? > This is the default behaviour in FreeBSD: a newly created file or directory has the same group as the parent directory. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120303144105.gg4...@hall.aurel32.net
Re: new patch. eglibc upload urgently required!
On Sun, Apr 22, 2012 at 03:11:35PM +0200, Robert Millan wrote: > Hi, Hi, > I notice that my initial patch didn't handle the library dependency > part. I'm attaching a new patch that fixes this issue. This makes new > packages built against the patched libc0.1-dev depend on libc0.1 >= > 2.13-31. > > VERY IMPORTANT NOTICE: if this patch is introduced in a later > version (rather than 2.13-31), you MUST change the version in > debian/libc0.1.symbols.common to the first version that provides > pthread_condattr_*etclock on GNU/kFreeBSD. > > Please could you give an estimate on when you'll be able to include it > in glibc? This issue is currently blocking libsoup2.4 builds (see > #663056). It does also have a major impact on other programs that use > glib (see #667686). > > Also, the other fixes already committed to pkg-glibc are also quite > important, as they are also blocking RC bugs in other packages. > > If you're all busy and can't prepare an upload right now, please say > it so the package be NMUed. > I have been mostly away from Debian stuff in the last weeks due to moving, but i'll try to catchup with glibc stuff during the next days and schedule an upload for the middle of the week. There are other important bugs not related to kfreebsd to handle. Aurelien -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120422150828.ga22...@hall.aurel32.net
Re: Bug#666510: please make udev|makedev dependency Linux-only
unmerge 666510 thanks Hi, On Sat, Mar 31, 2012 at 01:43:24PM +0200, Robert Millan wrote: > Hi, > > El 31 de març de 2012 12:26, Aurelien Jarno ha escrit: > > Given lm-sensors is completely useless on non Linux-kernel, I don't > > really see the point. The package is kept on non-Linux architectures > > just to avoid removing the build-deps in dozens of packages, to add it > > later when either support for other kernels is added to lm-sensors or > > when the non-Linux kernels start to provide access to sensors via their > > /sys compatibility layer. > > I'm afraid I don't understand your point. You're basically saying > that either [1] or [2] may happen, and that if [1] happens, then the > dependency needs to be removed, if [2] happens, then the dependency > needs to be removed, and until either of [1] or [2] happen, the > dependency needs to be removed anyway. I.e. you're saying the > dependency needs to be removed no matter if/when [1] or [2] happen. > > Additionally, there's [3] which is not something that can happen, but > a decision that has been discarded. If you apply solution [3], then > of course the situation is different, but you didn't say that you're > taking [3] into consideration. > > I can't make any sense of it. But maybe I'm just obtuse today. Is > there someone else who understands this reasoning? > Sorry I read the bug report to fast, and thought the dependency was on lm-sensors and not libsensors4. I am going to fix that in the next upload. Aurelien -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120423170254.ga22...@volta.aurel32.net
Re: new patch. eglibc upload urgently required!
On Sun, Apr 29, 2012 at 08:24:30PM +0200, Robert Millan wrote: > 2012/4/22 Robert Millan : > > I notice that my initial patch didn't handle the library dependency > > part. I'm attaching a new patch that fixes this issue. This makes new > > packages built against the patched libc0.1-dev depend on libc0.1 >= > > 2.13-31. > > > > VERY IMPORTANT NOTICE: if this patch is introduced in a later > > version (rather than 2.13-31), you MUST change the version in > > debian/libc0.1.symbols.common to the first version that provides > > pthread_condattr_*etclock on GNU/kFreeBSD. > > Uhm, I notice you didn't include this part. Now packages that build > with libc0.1-dev 2.13-31 won't have a versioned dependency on libc0.1 > >= 2.13-31. > > Are you sure this is correct? It seems to me that this can be a big > problem if those packages migrate to testing before eglibc does. > Not it's a mistake, I don't know how I managed that. I am going to fix that in an upload today. Aurelien -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120501103310.gk22...@hall.aurel32.net
Re: Possible pthread_cond_timedwait() regression on kfreebsd-i386
On Sat, Jun 02, 2012 at 08:31:20PM +0100, Steven Chamberlain wrote: > On 02/06/12 20:16, Stefan Fritsch wrote: > > since recently, I get some test failure in apr on kfreebsd-i386 that > > seem to indicate a bug in the kernel or libc. The failing test is part > > of testlock.c and basically does a pthread_cond_timedwait() ... > > Hi Stefan, > > I think you might be seeing #673711 which was already fixed in eglibc > SVN, but we have been waiting for a new release... > > http://bugs.debian.org/673711 > > This bug caused #672152 which is a failure of a similar test in Perl. > > There was some correspondence with eglibc maintainers off-list: > > On Sat, May 26, 2012 at 04:29:08PM +0200, Petr Salinger wrote: > > Please do an eglibc upload, with urgency at least medium. > > On 26/05/12 23:47, Aurelien Jarno wrote: > > My plan is to do an eglibc upload on Monday > > I guess Aurelien didn't have time to do the release Monday but I've > included him in Cc: in case he forgot. > Yes, didn't found time on Monday, but I am working on it since the beginning of the afternoon. Aurelien -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120602200635.gc4...@hall.aurel32.net
Re: Please upgrade eglibc and give back ruby1.8
On Mon, Jun 04, 2012 at 02:59:28AM +0100, Steven Chamberlain wrote: > Hi Christoph, > > eglibc 2.13-33 has been uploaded and built; please could you upgrade > the buildds to use this as soon as possible and then give back ruby1.8. Done for kfreebsd-i386. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120604045416.gg4...@hall.aurel32.net
Re: Please upgrade eglibc and give back ruby1.8
On Mon, Jun 04, 2012 at 06:54:16AM +0200, Aurelien Jarno wrote: > On Mon, Jun 04, 2012 at 02:59:28AM +0100, Steven Chamberlain wrote: > > Hi Christoph, > > > > eglibc 2.13-33 has been uploaded and built; please could you upgrade > > the buildds to use this as soon as possible and then give back ruby1.8. > > Done for kfreebsd-i386. > And now done for kfreebsd-amd64. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120604064316.gh4...@hall.aurel32.net
Re: Update buildds to gcc-4.7 4.7.0-13, and givebacks
On Wed, Jun 13, 2012 at 11:08:29PM +0100, Steven Chamberlain wrote: > Hi Christoph, > > Please could you update all the kfreebsd-* buildds with gcc-4.7 > 4.7.0-13. That fixes GCC PR c++/53602. > > Then please could you arrange givebacks for magics++ on kfreebsd-*, and > digicam just on kfreebsd-i386, as I think that was the reason for their > recent build failures, and they seem to build okay for me now in an > up-to-date sid chroot. > The chroots were already updated, so I have just requeued magics++. For digicam, I have requeued digikam instead, I guess it was a typo. Aurelien -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120614121236.gi30...@hall.aurel32.net
Re: packages stuck on buildd fils
On Thu, Jul 19, 2012 at 06:47:21PM +0100, Steven Chamberlain wrote: > Hi, > > Just a reminder that some packages are still stuck in 'Built' state on > fils for over 10 days without reaching 'Uploaded' or 'Installed' state. > I would guess this was some temporary problem such as lack of diskspace > or connection problem. > > This is keeping some of those packages from migrating despite having > freeze exceptions. > > Maybe givebacks would be the easiest way to fix this (the problem seems > to have gone away, and the buildds are basically idle now anyways). > > fritzing (18d 23h 12m, fils) > shoes (18d 23h 5m, fils) > git-annex (18d 22h 57m, fils) > dico (18d 22h 52m, fils) > file (18d 22h 45m, fils) > I have requeued all of them, and they have been built successfully. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120721120940.gg31...@hall.aurel32.net
Re: Bug#679762: dolfin: FTBFS: Armadillo could not be found. Be sure to set ARMADILLO_DIR. (missing: ARMADILLO_TEST_RUNS)
On Fri, Aug 03, 2012 at 01:27:34PM +0100, Steven Chamberlain wrote: > Hi Christoph, > > Please could you give back dolfin after updating kfreebsd-* buildds to > gcc-4.7 4.7.1-6, fixing PR c++/53549. > > fasch/fano/fils/finzi appear to have 4.7.1-5 at the moment. > Done, let's see if it builds. -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120803174752.gd3...@ohm.aurel32.net
Re: d-i failing to build on fils
On Fri, Aug 31, 2012 at 12:45:24AM +0100, Steven Chamberlain wrote: > Hi, > > These build failures of d-i on fils suggest a problem with > mirror-ubc.debian.org; I notice that other buildds use that mirror, > whereas fils refers to debian.netcologne.de instead, could that be > causing the problem here? Or if not, any idea what is wrong? > This was a configuration issue, /etc/apt/sources.list was using the ubece mirror while the correct mirrors where listed in /etc/apt/sources.list.d/*. It seems debian-installer is only looking up in the former file, not in the laters. Aurelien -- Aurelien Jarno GPG: 1024D/F1BCDB73 aurel...@aurel32.net http://www.aurel32.net -- To UNSUBSCRIBE, email to debian-bsd-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120831055351.gl17...@hall.aurel32.net
Bug#690021: libc0.1: getaddrinfo() might fail with IPv6 DNS record and no IPv6 connection
Package: libc0.1 Version: 2.13-35 Severity: grave Tags: d-i upstream patch ipv6 Justification: renders package unusable On a GNU/kFreeBSD machine without IPv6 connection, but with a DNS server returning IPv6 addresses, getaddrinfo() might fail to return only IPv4 addresses, causing some applications to fail to connect to some host using an host name. This happens when getaddrinfo() is called with service = 0 (aka port), which is perfectly valid. getaddrinfo() determine the source address the kernel would use for a given destination address by opening a SOCK_DGRAM connection with ŧhe same port as the service. A Linux kernel accepts such a connection with the port 0, while the FreeBSD kernel doesn't. A lot of applications are calling getaddrinfo() with the port they are going to use later, but a few others do not. Among them busybox is affected, causing wget on IPv6 addresses for non-IPv6 connected machine to fail, in turn causing the installation to fail: | $ busybox wget http://ftp.debian.org | Connecting to ftp.debian.org ([2001:610:1908:b000::148:12]:80) | wget: can't connect to remote host: No route to host As more and more Debian mirrors (including the default ftp.debian.org) have an IPv6 address, it makes GNU/kFreeBSD difficultly installable on such machines. The fix is to use port = 1 in such case, as it is done in the FreeBSD libc. This is what the attached patch does. I'll do an upload of eglibc in the next days, including this patch. -- System Information: Debian Release: wheezy/sid APT prefers unstable APT policy: (500, 'unstable') Architecture: kfreebsd-amd64 (x86_64) Kernel: kFreeBSD 8.3-1-amd64 Locale: LANG=fr_FR.UTF-8, LC_CTYPE=fr_FR.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages libc6 depends on: ii libc-bin 2.13-35 ii libgcc1 1:4.7.1-7 libc6 recommends no packages. Versions of packages libc6 suggests: ii debconf [debconf-2.0] 1.5.46 ii glibc-doc 2.13-35 ii locales2.13-35 ii locales-all [locales] 2.13-35 -- debconf information excluded --- a/sysdeps/posix/getaddrinfo.c +++ b/sysdeps/posix/getaddrinfo.c @@ -2267,8 +2267,15 @@ } socklen_t sl = sizeof (results[i].source_addr); + struct sockaddr sa = *q->ai_addr; +#ifdef __FreeBSD_kernel__ + /* The FreeBSD kernel doesn't allow connections on port 0. Use + port 1 instead, as on the FreeBSD libc. */ + if (((struct sockaddr_in *)&sa)->sin_port == htons(0)) + ((struct sockaddr_in *)&sa)->sin_port = htons(1); +#endif if (fd != -1 - && __connect (fd, q->ai_addr, q->ai_addrlen) == 0 + && __connect (fd, &sa, q->ai_addrlen) == 0 && __getsockname (fd, (struct sockaddr *) &results[i].source_addr, &sl) == 0)
Re: t1lib for Debian BSD
Artur R. Czechowski a écrit : > Hello, > A year ago Robert Millan asked me to update libtool stuff in t1lib[1]. > Since this time upstream released a new version with newer libtool stuff. > I prepared the package[2] and it is now available in experimental. > Could you, please, confirm, that t1lib builds correctly on k*BSD port > with original upstream libtool? I have just built it on my machine, and I can confirm you every thing is ok. So you can now upload it to unstable ;) Bye, Aurelien -- .''`. Aurelien Jarno GPG: 1024D/F1BCDB73 : :' : Debian GNU/Linux developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bug#312524: kfreebsd5-source: suggested by Linux kernel patches
Hi ! Braun Gabor a écrit : > Package: kfreebsd5-source > Version: 5.3-13 > > I do not have the package installed, I was just browsing the packages with > aptitude. I have found that many kernel-patch-* packages and > kernel-package > suggests kernel-source, which is provided by kfreebsd5-source. > I think that since the above packages deal only with Linux kernels, they > mustn't suggest a FreeBSD kernel. I agree with that. > I think the problem is that many packages assume that kernel-source is > only > provided by Linux kernel source packages. That's why I am not sure the bug is totally on the kfreebsd5-source side. I would suggest something like linux-kernel-source and freebsd-kernel-source, but I doubt the kernel team (the one dealing with the linux kernel would accept that). I'll change the "Provides: kernel-source" line by "Provides: freebsd-kernel-source" in the next upload. Anyway no kfreebsd package currently depends on kernel-source so I am not going to break any packages. Bye, Aurelien -- .''`. Aurelien Jarno GPG: 1024D/F1BCDB73 : :' : Debian GNU/Linux developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#295125: freebsd-buildutils: broken freebsd-mkdep
On Sun, Feb 13, 2005 at 09:56:23PM +0100, [EMAIL PROTECTED] wrote: > Package: freebsd-buildutils > Version: 5.2.1+20041028 freebsd-buildutils has been replaced by freebsd5-buildutils, anyway, the files in the package are the same. > It seems like there is a missing link for freebsd-mkdep that > should point to mkdep. adding a link fixed the problem: > > ~$ which freebsd-mkdep > /usr/bin/freebsd-mkdep > ~$ file /usr/bin/freebsd-mkdep > /usr/bin/freebsd-mkdep: symbolic link to `../lib/freebsd/mkdep' Could you explain a bit more the problem? /usr/lib/freebsd/mkdep exists, and running freebsd-mkdep works well on my system. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian GNU/Linux developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[Fwd: GCC 4.0 as the default GCC / C++ ABI change]
Hi all,
As you can see the C++ ABI transition has begun. We are not totally
ready on GNU/kFreeBSD, so please don't build and upload any C++ package
until we are ready.
Thinks that need to be done:
- Build an up to date gcc-4.0 (build currently running on my machine)
- Build an up to date gcc-defaults (done, but not uploaded)
- Build an up to date build-essential (FTBFS, but I haven't time to
investigate the problem now, I'll do that during the night).
I hope I be able to upload all of these packages before the dinstall at
04:00 UTC. I'll send you another mail when we are ready.
Bye,
Aurelien
--
.''`. Aurelien Jarno | GPG: 1024D/F1BCDB73
: :' : Debian GNU/Linux developer | Electrical Engineer
`. `' [EMAIL PROTECTED] | [EMAIL PROTECTED]
`-people.debian.org/~aurel32 | www.aurel32.net
--- Begin Message ---
-BEGIN PGP SIGNED MESSAGE-
With today's dinstall run, new gcc/g++ packages are entering the
archives and GCC 4.0 is the default gcc/g++. Starting from now, please
DON'T upload any C++ code, which build-depends on a library written in
C++ that is not yet converted to the new C++ ABI. Details for the C++
ABI change are at the end of the message. How do we go on?
- - The buildd admins make sure, that build-essential (>= 11), gcc (>=
4:4.0) and g++ (>= 4:4.0) are installed on the buildds.
- - Package maintainers must update all their build environments to
gcc (>= 4:4.0) and g++ (>= 4:4.0).
Each of the three following points can be done in parallel:
- - Rebuild C++ applications, which do not depend on any other C++
library besides libstdc++.
- - Rename and rebuild C++ libraries, which do not depend on any
other C++ library besides libstdc++. See below for the details.
- - All other applications and libraries have to wait, until the C++
related build dependencies are available for the new ABI. It's
important to adjust the build dependencies and the dependencies of
the -dev packages to the first version of a package, which is built
for the new ABI. See below for the details.
For the time until all C++ libraries are converted, we use the
following NMU policy for uploads related to the C++ ABI change:
- - 0-day NMU's allowed for all C++ library packages, which are uploaded
after the g++ default change, and are completely ignoring the C++
ABI change.
- - 2-day NMU's allowed for all C++ library packages, which are uploaded
after the g++ default change, with serious bugs in the packaging
(i.e. wrong package name in shlibs file, missing conflicts/replaces,
library package without a library, etc).
- - 5-day NMU for all C++ library packages, which can be converted, but
are left alone.
i.e. if libfoo1++ depends on libbar1++, libfoo1++ can be NMU'ed 5 days
after libbar1++ is uploaded.
Besides the C++ ABI change, all m68k and hppa packages depending on
libgcc1 have to be rebuilt to use libgcc2.
Have fun
Matthias
PS: Some of you know, that Ubuntu did the C++ ABI change at the start
of it's release cycle. For most if the libraries, patches are
available. These have to be adjusted, at least for the version
numbers. These patches may be outdated, some think they are too simple,
and it doesn't make sense to submit those to the Debian BTS. All
others may have a look at http://wiki.ubuntu.com/CxxLibraryList
- --
C++ ABI Transition
==
Debian already had a C++ transition in 2002. Still remember? We are
proposing the same schema for the forthcoming ABI transition. The
following text is derived from this plan.
- - Why do we need one?
Because GCC 3.4/4.0 changed the C++ ABI. You can't mix a C++ library
compiled with GCC 3.4/4.0 and a C++ application compiled with an
earlier version, or vice versa.
Transitions are painful. This will be no exception. The rules here
are designed to make it as smooth as possible, but it's still going
to be unpleasant. We have to do it, we can't stay with GCC 3.3 for
ever.
Other distributions did already switch to 3.4 or 4.0, and most of
our ports will have much better toolchains with the newer compiler.
- - How is it called?
The C++ ABI has many names (no, actually only two). G++ 3.2/3.3 have
the Version 1, 3.4/4.0 have the version 2. To get the ABI version:
g++ -E -dM - < /dev/null | awk '/GXX_ABI/ {print $3}'
Compilers with ABI version 1 print '102', those with version 2 print
1002. During the last C++ ABI transition package names were renamed
from libfoo to libfooc102. So rename them to libfooc1002 this time?
No,
- if your package is called libfoo1, add the string ''c2'' to the
package name (see below).
- if your package is called libfoo1c102, then replace c102 with c2
or drop the
GCC 4.0 as the default GCC / C++ ABI change on GNU/kFreeBSD
Hi all, I have successfully built gcc-4.0, gcc-defaults and build-essential on kfreebsd-i386. They are now installed on gnuab, so please upgrade your system before uploading C++ packages. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian GNU/Linux developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#323154: header protection in
Robert Millan a écrit : Package: kfreebsd-5 Severity: wishlist Tags: patch Hi, This patch adds header protection to kfreebsd's . Note that this is a dynamicaly-generated file. My patch fixes _both_ the generator script and the header so that the fix is also applied when building header packages. After this is applied, it will be suitable to replace glibc's (which is just an old version from kfreebsd) with up-to-date version in kfreebsd-kernel-headers. Well, I am not sure it is the way to do that. If the glibc already provides it, we should update it in the glibc instead of deleting the file in debian/rules of glibc and providing it in kfreebsd-kernel-headers. I'm forwarding it to upstream, too. Ok, nice. However I don't think it is necessary to fix that in our kernel if the bug is fixed in the glibc. After all, the kernel currently build fine. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian GNU/Linux developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Invitation to the Oldenburg Linux Developers Meeting
Robert Millan a écrit : Uhm.. perhaps we should request the listmasters to do something with debian-ports, so that we don't recieve off-topic messages like this one. - Create debian-cpu-ports and debian-kernel-ports, then deprecate debian-ports? - Create debian-kernel-ports, then re-define debian-ports as only for cpu ports? What do you people think? Well, we are receiving very few off-topic messages like that one on this list. And I am not even sure it is that much off-topic, as I thought to go there to have a first contact with the Debian installer-team and try to see how we could work to have d-i working on GNU/kFreeBSD. Unfortunately I won't be able to go there. So I think we could just ignore such messages coming there until they get very annoying. Cheers, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian GNU/Linux developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#330193: kfreebsd-5: should not be arch any
Blars Blarson a écrit : Package: kfreebsd-5 Severity: important Justification: fails to build from source kfreebsd-5 attempts to build and fails on all buildds. Please change the arch on the package source so it will not waste buildd resources. The source package is declared as "arch: any" by dpkg-dev, because it builds an "arch: all" and an "arch: kfreebsd-i386". The solution would be to make the source package being "arch: all kfreebsd-i386", but I don't know how to do. Do you have a solution for that? Thanks, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#334357: unneeded hunks for GENERIC in debian/patches
Robert Millan a écrit : Package: kfreebsd-5 Severity: minor The following patches contain hunks for sys/i386/conf/GENERIC which aren't really used (since the config file is now in config/*). Perhaps they should be removed: This GENERIC configuration file is still provided in kfreebsd-source-5.4, so I think it is a good idea to provide it. What do you think? Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian GNU/Linux developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#334357: unneeded hunks for GENERIC in debian/patches
Robert Millan a écrit : On Tue, Oct 18, 2005 at 11:58:58AM +0200, Aurelien Jarno wrote: Robert Millan a ?crit : Package: kfreebsd-5 Severity: minor The following patches contain hunks for sys/i386/conf/GENERIC which aren't really used (since the config file is now in config/*). Perhaps they should be removed: This GENERIC configuration file is still provided in kfreebsd-source-5.4, so I think it is a good idea to provide it. What do you think? I'm not sure. I think it's confusing to provide complete files in config/, but still provide the patches. That can lead to inconsistencies easily. How do Linux packages do this? The Linux packages make some modifications on the config files, just as we do. However, I have looked at the patch, and it is not necessary to use it to get a working FreeBSD kernel. I'll remove it in the next upload. Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#335898: bogus "all rights reserved" message
severity 335898 wishlist thanks Robert Millan a écrit : Package: kfreebsd-5 Severity: normal The following lines are printed by kFreeBSD when boot starts: "Copyright (c) 1992-2005 The FreeBSD Project. Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 The Regents of the University of California. All rights reserved." I think there two problems with that: [...] snip - These lines were added to advertise BSD 4.4 and FreeBSD, but our system is much different, and contains code copyrighted by a lot other contributors (FSF, SPI, X, etc). In this context, I think advertising UCB doesn't make Exactly as FreeBSD (except SPI). any sense. As for FreeBSD, I'm not so sure. Perhaps we should keep it, but still indicate that this copyright doesn't refer to the whole system as it did on FreeBSD. On FreeBSD, it also don't apply to the whole system. I think it applies to the kernel, even on FreeBSD. Well I don't really know what we should do, but that's clearly not a problem nor a bug, that's why I am decreasing the severity to wishlist. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Please don't upload packages built with gcc-4.0 4.0.2-3
Hi all, gcc-4.0 4.0.2-3 seems to be horribly fucked. It breaks the C++ ABI, and there is also some big regressions discovered on mips (floating point exception) and on ppc (wrong assembly instructions). It may be possible that some more regression will be found also on i386. The problem also affects Debian ports, and we have to wait to see how it will be addressed. Meanwhile, I suggest that everybody who upload packages to gnuab downgrade gcc-4.0 to version 4.0.2-2. It could be found on: http://ftp.gnuab.org/debian/obsolete/2005-10-25/gcc-4.0/ Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#336496: ldd
Robert Millan a écrit : I observed this some time ago, but I didn't give it any importance untill seeing this bug. It is the same bug, right? On GNU/Linux: [EMAIL PROTECTED]:~$ ldd /bin/ls | grep /lib/ld /lib/ld-linux.so.2 (0xb7f48000) On GNU/kFreeBSD: [EMAIL PROTECTED]:~$ ldd /bin/ls | grep /lib/ld /lib/ld.so.1 => /lib/ld.so.1 (0x) It maybe be due to the ugly hack we have in the glibc to make ld.so.1 both executable and a library. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#338378: kfreebsd-5: please enable altq support
tag 338378 + pending thanks On Wed, Nov 09, 2005 at 10:59:30PM +, Brian M. Carlson wrote: > Package: kfreebsd-5 > Severity: wishlist > > Please enable altq support, as I'd like to be able to use the features > it provides wrt pf, such as limiting the bandwidth to Windows boxes. ;-) > Fixed in SVN. I plan to release new kernel image this week-end, I am not sure though if I'll upload them directly or if I will ask for test first because there will be a lot of changes in the /boot directory to allow more than one kernel to be installed at a time. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#338378: kfreebsd-5: please enable altq support
Robert Millan a écrit : On Thu, Nov 10, 2005 at 12:46:25AM +0100, Aurelien Jarno wrote: tag 338378 + pending thanks On Wed, Nov 09, 2005 at 10:59:30PM +, Brian M. Carlson wrote: Package: kfreebsd-5 Severity: wishlist Please enable altq support, as I'd like to be able to use the features it provides wrt pf, such as limiting the bandwidth to Windows boxes. ;-) Fixed in SVN. I plan to release new kernel image this week-end, I am not sure though if I'll upload them directly or if I will ask for test first because there will be a lot of changes in the /boot directory to allow more than one kernel to be installed at a time. How about `experimental'? currently gnuab does not support experimental. Maybe support could be added? -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: knetbsd maintained? want glibc patched for it?
On Sun, Nov 13, 2005 at 02:55:04PM -0600, Drew Scott Daniels wrote: > Hi, > Is Debian knetbsd maintained? I know it's nowhere near where Debian > kfreebsd is. I ran across the following in bug 213503 that may be of > interest to those interested in Debian knetbsd. I haven't checked at all > to see if glibc has been patched for knetbsd, but if it has, then the > bug could use an update. Well, as far as I know, Debian knetbsd is unmaintained for a while. All the efforts are now ported on Debian GNU/kFreeBSD. We will have patch to merge into glibc for it, but we have to write them first :) Currently, we are still using an old glibc 2.3. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#338182: kfreebsd-5: signals for linuxthreads internal operation
Petr Salinger a écrit :
Package: kfreebsd-5
Severity: wishlist
Tags: patch
Hi,
with attached patch for kernel and clone.c it should be possible leave
SIGUSR1 and SIGUSR2 free for user code.
Current libc will work with current and new kernel,
new libc will require new kernel.
Thanks for the patch, a new kernel has just been uploaded. A new glibc
will follow soon, probably tomorrow. I have already a working one on my
disk, but I need to make a clean patch first.
--- clone.c.OLD 2005-11-08 17:25:02.0 +0100
+++ clone.c 2005-11-08 17:32:18.0 +0100
@@ -64,13 +64,13 @@
if ((flags & CSIGNAL) != 0 && (flags & CSIGNAL) != SIGCHLD)
{
- /* This implementation of clone() supports only the SIGUSR1 signal. */
- if ((flags & CSIGNAL) != SIGUSR1)
+ /* This implementation of clone() supports only the 128 signals. */
+ if ((flags & CSIGNAL) & RFTHPNMASK)
This line has to be
+ if ((flags & CSIGNAL) & ~RFTHPNMASK)
{
__set_errno (EINVAL);
return -1;
}
- rfork_flags |= RFLINUXTHPN;
+ rfork_flags |= (RFLINUXTHPN | ((flags & CSIGNAL) << RFTHPNSHIFT));
}
if (flags & CLONE_VM)
Bye,
Aurelien
--
.''`. Aurelien Jarno | GPG: 1024D/F1BCDB73
: :' : Debian developer | Electrical Engineer
`. `' [EMAIL PROTECTED] | [EMAIL PROTECTED]
`-people.debian.org/~aurel32 | www.aurel32.net
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#338182: kfreebsd-5: signals for linuxthreads internal operation
Petr Salinger a écrit : I have another one: --- sysdeps/unix/bsd/bsd4.4/kfreebsd/bits/sigset.h~ +++ sysdeps/unix/bsd/bsd4.4/kfreebsd/bits/sigset.h @@ -42,9 +42,9 @@ # ifndef _EXTERN_INLINE # define _EXTERN_INLINE extern __inline # endif /* Return a mask that includes SIG only. */ -# define __sigmask(sig)((unsigned int) 1 << ((sig) - 1)) +# define __sigmask(sig)((unsigned int) 1 << ((sig) - 1) % 32) /* Return the word index for SIG. */ # define __sigword(sig)(((sig) - 1) >> 5) Fixed, thanks. However, I don't know which code use this definition and so may be buggy. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#264118: type-handling man page is missing
On Wed, Nov 30, 2005 at 07:59:13PM +0100, Peter Eisentraut wrote: > Could the maintainers comment on this patch? As long as type-handling > exists, it should have a man page, I think. Yes, but you should understand that there is a lot to do in GNU/kFreeBSD, and adding a manpage for a package that will disappear has a low priority. I will try to upload a fixed version this evening. Thanks for the patch. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1061 - trunk/glibc-2.3-head
Author: aurel32 Date: 2006-01-18 00:29:18 + (Wed, 18 Jan 2006) New Revision: 1061 Modified: trunk/glibc-2.3-head/NOTES Log: All the packages affected by the ucontext ABI changed have been binNMUed Modified: trunk/glibc-2.3-head/NOTES === --- trunk/glibc-2.3-head/NOTES 2006-01-17 20:15:41 UTC (rev 1060) +++ trunk/glibc-2.3-head/NOTES 2006-01-18 00:29:18 UTC (rev 1061) @@ -96,18 +96,18 @@ sigreturn The affected source packages are only: - gstreamer0.8 - lwp - pdns - pth - ruby1.8 - ruby1.9 + gstreamer0.8 [done] + lwp [done] + pdns [done] + pth [done] + ruby1.8 [done] + ruby1.9 [done] Transition should be as follows: - 1) enlarge struct declarations in glibc [done for i386] - 2) upload glibc - 3) binNMU affected packages + 1) enlarge struct declarations in glibc [done] + 2) upload glibc [done] + 3) binNMU affected packages [done] 4) change implementation of ucontext_t related calls to access added mmx fields in structs, test 5) upload glibc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1063 - trunk/kfreebsd-kernel-headers/debian
Author: aurel32 Date: 2006-01-18 10:37:51 + (Wed, 18 Jan 2006) New Revision: 1063 Modified: trunk/kfreebsd-kernel-headers/debian/changelog trunk/kfreebsd-kernel-headers/debian/includes.list Log: * Remove sys/dirent.h Modified: trunk/kfreebsd-kernel-headers/debian/changelog === --- trunk/kfreebsd-kernel-headers/debian/changelog 2006-01-18 10:26:44 UTC (rev 1062) +++ trunk/kfreebsd-kernel-headers/debian/changelog 2006-01-18 10:37:51 UTC (rev 1063) @@ -1,3 +1,9 @@ +kfreebsd-kernel-headers (0.14) unreleased; urgency=low + + * Remove sys/dirent.h + + -- Aurelien Jarno <[EMAIL PROTECTED]> Wed, 18 Jan 2006 11:36:20 +0100 + kfreebsd-kernel-headers (0.13) unreleased; urgency=low * Remove a lot of file from netinet6/in6.h. Modified: trunk/kfreebsd-kernel-headers/debian/includes.list === --- trunk/kfreebsd-kernel-headers/debian/includes.list 2006-01-18 10:26:44 UTC (rev 1062) +++ trunk/kfreebsd-kernel-headers/debian/includes.list 2006-01-18 10:37:51 UTC (rev 1063) @@ -128,7 +128,6 @@ sys/dataacq.h sys/device_port.h sys/devicestat.h -sys/dirent.h sys/disk.h sys/disklabel.h sys/dkstat.h -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1071 - trunk/type-handling/debian
Author: aurel32
Date: 2006-01-18 19:53:08 + (Wed, 18 Jan 2006)
New Revision: 1071
Modified:
trunk/type-handling/debian/changelog
trunk/type-handling/debian/rules
Log:
* Provides and filters out both the Debian and GNU type/CPU. Should fix
the problem with type-handling providing not+hurd on GNU/Hurd.
* Does not provide hurd on GNU/Hurd, as type-handling does not provide
the hurd kernel.
Modified: trunk/type-handling/debian/changelog
===
--- trunk/type-handling/debian/changelog2006-01-18 19:27:45 UTC (rev
1070)
+++ trunk/type-handling/debian/changelog2006-01-18 19:53:08 UTC (rev
1071)
@@ -1,3 +1,12 @@
+type-handling (0.2.18) unstable; urgency=low
+
+ * Provides and filters out both the Debian and GNU type/CPU. Should fix
+the problem with type-handling providing not+hurd on GNU/Hurd.
+ * Does not provide hurd on GNU/Hurd, as type-handling does not provide
+the hurd kernel.
+
+ -- Aurelien Jarno <[EMAIL PROTECTED]> Wed, 18 Jan 2006 20:29:09 +0100
+
type-handling (0.2.17) unstable; urgency=low
[ Peter Eisentraut ]
Modified: trunk/type-handling/debian/rules
===
--- trunk/type-handling/debian/rules2006-01-18 19:27:45 UTC (rev 1070)
+++ trunk/type-handling/debian/rules2006-01-18 19:53:08 UTC (rev 1071)
@@ -6,21 +6,26 @@
#export DH_VERBOSE=1
DEB_BUILD_ARCH_CPU ?= $(shell dpkg-architecture -qDEB_BUILD_ARCH_CPU)
+DEB_BUILD_ARCH_OS ?= $(shell dpkg-architecture -qDEB_BUILD_ARCH_OS)
DEB_BUILD_GNU_SYSTEM ?= $(shell dpkg-architecture -qDEB_BUILD_GNU_SYSTEM)
-DEB_BUILD_GNU_TYPE = $(DEB_BUILD_ARCH_CPU)-$(DEB_BUILD_GNU_SYSTEM)
+DEB_BUILD_GNU_CPU ?= $(shell dpkg-architecture -qDEB_BUILD_GNU_CPU)
+DEB_BUILD_GNU_TYPE ?= $(shell dpkg-architecture -qDEB_BUILD_GNU_TYPE)
-cpus := $(shell grep -v ^\# /usr/share/dpkg/cputable | (while read debian gnu
regex ; do echo $${debian} ; done))
-systems=$(shell grep -v ^\# /usr/share/dpkg/ostable | (while read debian gnu
regex ; do echo $${debian} ; done))
-systems := $(shell echo $(systems) linux-gnu | tr " " "\n" | grep -vx linux |
sort | uniq)
+cpus := $(shell grep -v ^\# /usr/share/dpkg/cputable | (while read debian gnu
regex ; do echo $${debian} ; echo $${gnu}; done) | sort -u)
+systems := $(shell grep -v ^\# /usr/share/dpkg/ostable | (while read debian
gnu regex ; do echo $${debian} ; echo $${gnu}; done) | sort -u)
-not_cpus := $(shell echo $(cpus) | tr " " "\n" | grep -vx
$(DEB_BUILD_ARCH_CPU))
-not_systems := $(shell echo $(systems) | tr " " "\n" | grep -vx
$(DEB_BUILD_GNU_SYSTEM))
+not_cpus := $(shell echo $(cpus) | tr " " "\n" | grep -vx
$(DEB_BUILD_ARCH_CPU) | grep -vx $(DEB_BUILD_GNU_CPU))
+not_systems := $(shell echo $(systems) | tr " " "\n" | grep -vx
$(DEB_BUILD_ARCH_OS) | grep -vx $(DEB_BUILD_GNU_SYSTEM))
not_cpus := $(shell for cpu in $(not_cpus) ; do echo -n , not+$$cpu ; done)
not_systems := $(shell for system in $(not_systems) ; do echo -n ,
not+$$system ; done)
-provides = $(DEB_BUILD_ARCH_CPU), $(DEB_BUILD_GNU_SYSTEM),
$(DEB_BUILD_GNU_TYPE)$(not_cpus)$(not_systems)
+ifneq (hurd,$(DEB_HOST_ARCH_OS))
+provides = $(DEB_HOST_ARCH_OS),
+endif
+provides += $(DEB_BUILD_ARCH_CPU), $(DEB_BUILD_GNU_SYSTEM),
$(DEB_BUILD_GNU_TYPE), $(not_cpus), $(not_systems)
+
clean:
dh_testdir
dh_testroot
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: r1073 - trunk/glibc-2.3-head/sysdeps/kfreebsd
On Wed, Jan 18, 2006 at 11:27:24PM +, Robert Millan wrote: > Author: rmh > Date: 2006-01-18 23:27:24 + (Wed, 18 Jan 2006) > New Revision: 1073 > > Modified: >trunk/glibc-2.3-head/sysdeps/kfreebsd/uname.c > Log: > Rewrite uname.c to use SYS_uname. Runs 1.54 faster. > This is the comment in the FreeBSD kernel, file sys/kern/kern_xxx.c /* * This is the FreeBSD-1.1 compatable uname(2) interface. These * days it is done in libc as a wrapper around a bunch of sysctl's. * This must maintain the old 1.1 binary ABI. */ Then if you look at the code corresponding to the syscall, you will see that the syscall correspond to a call to the corresponding sysctl, but truncated: ... name[0] = CTL_KERN; name[1] = KERN_OSTYPE; len = sizeof (uap->name->sysname); mtx_lock(&Giant); error = userland_sysctl(td, name, 2, uap->name->sysname, &len, 1, 0, 0, 0); if (error) goto done2; subyte( uap->name->sysname + sizeof(uap->name->sysname) - 1, 0); ... So, in short the current implementation in our glibc is the correct one. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1084 - trunk/glibc-2.3-head/sysdeps/kfreebsd
Author: aurel32
Date: 2006-01-20 23:13:52 + (Fri, 20 Jan 2006)
New Revision: 1084
Removed:
trunk/glibc-2.3-head/sysdeps/kfreebsd/getdomain.c
trunk/glibc-2.3-head/sysdeps/kfreebsd/setdomain.c
Log:
Remve setdomain.c and getdomain.c, as we now have the corresponding syscalls.
Deleted: trunk/glibc-2.3-head/sysdeps/kfreebsd/getdomain.c
===
--- trunk/glibc-2.3-head/sysdeps/kfreebsd/getdomain.c 2006-01-20 16:54:08 UTC
(rev 1083)
+++ trunk/glibc-2.3-head/sysdeps/kfreebsd/getdomain.c 2006-01-20 23:13:52 UTC
(rev 1084)
@@ -1,103 +0,0 @@
-/* Copyright (C) 2002 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
- Contributed by Bruno Haible <[EMAIL PROTECTED]>, 2002.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-
-#include
-#include
-#include
-#include
-#include
-
-/* Put the name of the current NIS domain in no more than LEN bytes of NAME.
- The result is null-terminated if LEN is large enough for the full
- name and the terminator. */
-
-int
-getdomainname (char *name, size_t len)
-{
- /* Fetch the "kern.domainname" sysctl value. */
- int request[2] = { CTL_KERN, KERN_NISDOMAINNAME };
-#if 1
- size_t result_len = len;
-
- if (__sysctl (request, 2, name, &result_len, NULL, 0) < 0)
-{
- if (errno == ENOMEM)
- __set_errno (ENAMETOOLONG);
- return -1;
-}
-
- if (result_len == len)
-{
- __set_errno (ENAMETOOLONG);
- return -1;
-}
-
- name[result_len] = '\0';
- return 0;
-#else
- char buf[MAXHOSTNAMELEN + 1];
- char *result;
- size_t result_len;
- char *bufend;
- size_t buflen;
-
- if (len >= MAXHOSTNAMELEN)
-{
- result = name;
- result_len = len - 1;
-}
- else
-{
- /* Use a temporary buffer, so that we can detect the ENAMETOOLONG
-condition. (Well, we could also rely on the ENOMEM error code.) */
- result = buf;
- result_len = MAXHOSTNAMELEN;
-}
-
- if (__sysctl (request, 2, result, &result_len, NULL, 0) < 0)
-return -1;
-
- /* If we used no temporary buffer, we are done. */
- if (result == name)
-{
- result[resultlen] = '\0';
- return 0;
-}
-
- /* See if the result fits in the caller's buffer. */
- bufend = memchr (buf, '\0', result_len);
- if (bufend == NULL)
-{
- bufend = buf + result_len;
- *bufend = '\0';
-}
- buflen = bufend - buf + 1;
-
- /* Copy into the caller's buffer. */
- memcpy (name, buf, len < buflen ? len : buflen);
-
- if (len < buflen)
-{
- __set_errno (ENAMETOOLONG);
- return -1;
-}
- return 0;
-#endif
-}
-libc_hidden_def (getdomainname)
Deleted: trunk/glibc-2.3-head/sysdeps/kfreebsd/setdomain.c
===
--- trunk/glibc-2.3-head/sysdeps/kfreebsd/setdomain.c 2006-01-20 16:54:08 UTC
(rev 1083)
+++ trunk/glibc-2.3-head/sysdeps/kfreebsd/setdomain.c 2006-01-20 23:13:52 UTC
(rev 1084)
@@ -1,37 +0,0 @@
-/* Copyright (C) 2002 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
- Contributed by Bruno Haible <[EMAIL PROTECTED]>, 2002.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-
-#include
-#include
-
-/* Set the name of the current NIS domain to NAME, which is LEN bytes long
- (excluding a possible trailing NUL byte). This call is restricted to
- the super-user. */
-
-int
-setdomainname (const char *name, size_t len)
-{
- /* Set the "kern.domainname" sysctl value. */
- int request[2] = { CTL_KERN
r1085 - trunk/glibc-2.3-head
Author: aurel32 Date: 2006-01-21 08:49:48 + (Sat, 21 Jan 2006) New Revision: 1085 Modified: trunk/glibc-2.3-head/NOTES Log: glibc revision r1084 upload. It uses the new context structs, so removing the corresponding paragraph from NOTES. Modified: trunk/glibc-2.3-head/NOTES === --- trunk/glibc-2.3-head/NOTES 2006-01-20 23:13:52 UTC (rev 1084) +++ trunk/glibc-2.3-head/NOTES 2006-01-21 08:49:48 UTC (rev 1085) @@ -73,21 +73,6 @@ problems: = -* update machine context structs (ABI change) - - FreeBSD 5 changed extended context structs (on i386 added support for mmx), - we still using FreeBSD 4 structs - - Transition should be as follows: - - 1) enlarge struct declarations in glibc [done] - 2) upload glibc [done] - 3) binNMU affected packages [done] - 4) change implementation of ucontext_t related calls - to access added mmx fields in structs, test [done] - 5) upload glibc - - * thread cancelation cancelable functions are not properly marked in syscalls.list, ... informal list of such functions is include in linuxthreads/tst-cancel4.c, nptl/tst-cancel4.c -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: r1073 - trunk/glibc-2.3-head/sysdeps/kfreebsd
Robert Millan a écrit : On Sat, Jan 21, 2006 at 12:11:52AM +0100, Aurelien Jarno wrote: This is the comment in the FreeBSD kernel, file sys/kern/kern_xxx.c /* * This is the FreeBSD-1.1 compatable uname(2) interface. These * days it is done in libc as a wrapper around a bunch of sysctl's. * This must maintain the old 1.1 binary ABI. */ Then if you look at the code corresponding to the syscall, you will see that the syscall correspond to a call to the corresponding sysctl, but truncated: ... name[0] = CTL_KERN; name[1] = KERN_OSTYPE; len = sizeof (uap->name->sysname); mtx_lock(&Giant); error = userland_sysctl(td, name, 2, uap->name->sysname, &len, 1, 0, 0, 0); if (error) goto done2; subyte( uap->name->sysname + sizeof(uap->name->sysname) - 1, 0); ... So, in short the current implementation in our glibc is the correct one. Weird. How about disabling SYS_uname macro, then? I don't see why we should disable it here. This syscall does its job correclty, ie it returns the requested data, but in a format different from our userland structure (32 bytes field instead of 256 bytes). There is a lot of unused syscalls in sys/syscalls.h. Maybe we could decide not to export this file, but I don't think it is a good idea to remove some syscalls just because they are not suitable for the corresponding glibc functions. Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1091 - trunk/glibc-2.3-head/sysdeps/kfreebsd
Author: aurel32
Date: 2006-01-22 18:48:06 + (Sun, 22 Jan 2006)
New Revision: 1091
Modified:
trunk/glibc-2.3-head/sysdeps/kfreebsd/uname.c
Log:
uname.c: also fill the domainname field. It is a GNU extension, but as
we can easily have the information using getdomainname(), we can support it.
Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/uname.c
===
--- trunk/glibc-2.3-head/sysdeps/kfreebsd/uname.c 2006-01-22 08:20:14 UTC
(rev 1090)
+++ trunk/glibc-2.3-head/sysdeps/kfreebsd/uname.c 2006-01-22 18:48:06 UTC
(rev 1091)
@@ -21,6 +21,7 @@
#include
#include
#include
+#include /* For getdomainname() */
#define SYSNAME"GNU/kFreeBSD"
#define SYSNAME_LEN13
@@ -94,6 +95,13 @@
name->machine[len] = '\0';
}
+ /* Fill domainname. Use getdomainname(). */
+ if (getdomainname(name->domainname, sizeof(name->domainname)) == -1)
+{
+ if (errno != EINVAL)
+strncpy (name->domainname, UNAME_DOMAINNAME, sizeof
(name->domainname));
+}
+
return 0;
}
libc_hidden_def (__uname)
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1092 - trunk/glibc-2.3-head
Author: aurel32 Date: 2006-01-22 21:12:08 + (Sun, 22 Jan 2006) New Revision: 1092 Modified: trunk/glibc-2.3-head/debian.sh Log: Split out i386 specific parts of debian/sysdeps/kfreebsd.mk into debian/sysdeps/kfreebsd-i386.mk Modified: trunk/glibc-2.3-head/debian.sh === --- trunk/glibc-2.3-head/debian.sh 2006-01-22 18:48:06 UTC (rev 1091) +++ trunk/glibc-2.3-head/debian.sh 2006-01-22 21:12:08 UTC (rev 1092) @@ -138,14 +138,11 @@ cat debian/control.in/amd64 >> [EMAIL PROTECTED] --- glibc-2.3.5.orig/debian/sysdeps/kfreebsd.mk +++ glibc-2.3.5/debian/sysdeps/kfreebsd.mk -@@ -0,0 +1,73 @@ +@@ -0,0 +1,61 @@ +GLIBC_OVERLAYS ?= $(shell ls glibc-linuxthreads* glibc-ports* glibc-libidn*) +MIN_KERNEL_SUPPORTED := 5.4.0 +libc = libc0.1 + -+GLIBC_PASSES += i686 -+DEB_ARCH_REGULAR_PACKAGES += libc0.1-i686 -+ +# Support multiple makes at once based on number of processors +# Common wisdom says parallel make can be up to 2n+1. +# Should we do that to get faster builds? @@ -163,15 +160,6 @@ +libc_add-ons = linuxthreads $(add-ons) +libc_extra_config_options = $(extra_config_options) --with-tls --with-__thread --disable-compatible-utmp + -+# We use -march=i686 and glibc's i686 routines use cmov, so require it. -+# A Debian-local glibc patch adds cmov to the search path. -+i686_add-ons = linuxthreads $(add-ons) -+i686_configure_target=i686-kfreebsd -+i686_configure_build=i686-kfreebsd -+i686_extra_cflags = -march=i686 -mtune=i686 -g1 -O3 -+i686_LIBDIR = /i686/cmov -+i686_extra_config_options = $(extra_config_options) --disable-profile --with-tls --with-__thread --disable-compatible-utmp -+ +ifndef KFREEBSD_SOURCE + KFREEBSD_HEADERS := /usr/include +else @@ -212,6 +200,20 @@ +fi) +endef + +--- glibc-2.3.5.orig/debian/sysdeps/kfreebsd-i386.mk glibc-2.3.5/debian/sysdeps/kfreebsd-i386.mk +@@ -0,0 +1,11 @@ ++GLIBC_PASSES += i686 ++DEB_ARCH_REGULAR_PACKAGES += libc0.1-i686 ++ ++# We use -march=i686 and glibc's i686 routines use cmov, so require it. ++# A Debian-local glibc patch adds cmov to the search path. ++i686_add-ons = linuxthreads $(add-ons) ++i686_configure_target=i686-kfreebsd ++i686_configure_build=i686-kfreebsd ++i686_extra_cflags = -march=i686 -mtune=i686 -g1 -O3 ++i686_LIBDIR = /i686/cmov ++i686_extra_config_options = $(extra_config_options) --disable-profile --with-tls --with-__thread --disable-compatible-utmp --- glibc-2.3.5.old/debian/sysdeps/depflags.pl 2005-12-23 00:40:20.0 +0100 +++ glibc-2.3.5/debian/sysdeps/depflags.pl 2005-12-23 00:44:38.0 +0100 @@ -33,6 +33,13 @@ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1093 - trunk/glibc-2.3-head/sysdeps/kfreebsd/bits
Author: aurel32
Date: 2006-01-23 00:44:49 + (Mon, 23 Jan 2006)
New Revision: 1093
Modified:
trunk/glibc-2.3-head/sysdeps/kfreebsd/bits/resource.h
Log:
Update bits/resource.h from the Linux version of the file, and define RLIMIT_AS
as we have it.
Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/bits/resource.h
===
--- trunk/glibc-2.3-head/sysdeps/kfreebsd/bits/resource.h 2006-01-22
21:12:08 UTC (rev 1092)
+++ trunk/glibc-2.3-head/sysdeps/kfreebsd/bits/resource.h 2006-01-23
00:44:49 UTC (rev 1093)
@@ -23,50 +23,70 @@
#include
-/* These are the values for FreeBSD. Earlier BSD systems have a subset of
+/* Transmute defines to enumerations. The macro re-definitions are
+ necessary because some programs want to test for operating system
+ features with #ifdef RUSAGE_SELF. In ISO C the reflexive
+ definition is a no-op.
+
+ These are the values for FreeBSD. Earlier BSD systems have a subset of
these kinds of resource limit. */
/* Kinds of resource limit. */
enum __rlimit_resource
- {
-/* Per-process CPU limit, in seconds. */
-RLIMIT_CPU,
-#defineRLIMIT_CPU RLIMIT_CPU
-/* Largest file that can be created, in bytes. */
-RLIMIT_FSIZE,
-#defineRLIMIT_FSIZERLIMIT_FSIZE
-/* Maximum size of data segment, in bytes. */
-RLIMIT_DATA,
-#defineRLIMIT_DATA RLIMIT_DATA
-/* Maximum size of stack segment, in bytes. */
-RLIMIT_STACK,
-#defineRLIMIT_STACKRLIMIT_STACK
-/* Largest core file that can be created, in bytes. */
-RLIMIT_CORE,
-#defineRLIMIT_CORE RLIMIT_CORE
-/* Largest resident set size, in bytes.
- This affects swapping; processes that are exceeding their
- resident set size will be more likely to have physical memory
- taken from them. */
-RLIMIT_RSS,
-#defineRLIMIT_RSS RLIMIT_RSS
-/* Locked-in-memory address space. */
-RLIMIT_MEMLOCK,
-#defineRLIMIT_MEMLOCK RLIMIT_MEMLOCK
-/* Number of processes. */
-RLIMIT_NPROC,
-#defineRLIMIT_NPROCRLIMIT_NPROC
-/* Number of open files. */
-RLIMIT_NOFILE,
-#defineRLIMIT_NOFILE RLIMIT_NOFILE
-/* Maximum size of all socket buffers. */
-RLIMIT_SBSIZE,
-#define RLIMIT_SBSIZE RLIMIT_SBSIZE
+{
+ /* Per-process CPU limit, in seconds. */
+ RLIMIT_CPU = 0,
+#defineRLIMIT_CPU RLIMIT_CPU
-RLIMIT_NLIMITS,/* Number of limit flavors. */
-RLIM_NLIMITS = RLIMIT_NLIMITS /* Traditional name for same. */
- };
+ /* Largest file that can be created, in bytes. */
+ RLIMIT_FSIZE = 1,
+#defineRLIMIT_FSIZE RLIMIT_FSIZE
+ /* Maximum size of data segment, in bytes. */
+ RLIMIT_DATA = 2,
+#defineRLIMIT_DATA RLIMIT_DATA
+
+ /* Maximum size of stack segment, in bytes. */
+ RLIMIT_STACK = 3,
+#defineRLIMIT_STACK RLIMIT_STACK
+
+ /* Largest core file that can be created, in bytes. */
+ RLIMIT_CORE = 4,
+#defineRLIMIT_CORE RLIMIT_CORE
+
+ /* Largest resident set size, in bytes.
+ This affects swapping; processes that are exceeding their
+ resident set size will be more likely to have physical memory
+ taken from them. */
+ __RLIMIT_RSS = 5,
+#defineRLIMIT_RSS __RLIMIT_RSS
+
+ /* Locked-in-memory address space. */
+ __RLIMIT_MEMLOCK = 6,
+#defineRLIMIT_MEMLOCK __RLIMIT_MEMLOCK
+
+ /* Number of processes. */
+ __RLIMIT_NPROC = 7,
+#defineRLIMIT_NPROC __RLIMIT_NPROC
+
+ /* Number of open files. */
+ RLIMIT_NOFILE = 8,
+#defineRLIMIT_NOFILE RLIMIT_NOFILE
+
+ /* Maximum size of all socket buffers. */
+ __RLIMIT_SBSIZE = 9,
+#define RLIMIT_SBSIZE __RLIMIT_SBSIZE
+
+ /* Address space limit. */
+ RLIMIT_AS = 10,
+#define RLIMIT_AS RLIMIT_AS
+
+ __RLIMIT_NLIMITS = 11,
+ __RLIM_NLIMITS = __RLIMIT_NLIMITS
+#define RLIMIT_NLIMITS __RLIMIT_NLIMITS
+#define RLIM_NLIMITS __RLIM_NLIMITS
+};
+
/* Value to indicate that there is no limit. */
#define RLIM_INFINITY 0x7fffLL
#ifdef __USE_LARGEFILE64
@@ -100,20 +120,18 @@
/* Whose usage statistics do you want? */
enum __rusage_who
-/* The macro definitions are necessary because some programs want
- to test for operating system features with #ifdef RUSAGE_SELF.
- In ISO C the reflexive definition is a no-op. */
- {
-/* The calling process. */
-RUSAGE_SELF = 0,
-#define RUSAGE_SELF RUSAGE_SELF
-/* All of its terminated child processes. */
-RUSAGE_CHILDREN = -1
+{
+ /* The calling process. */
+ RUSAGE_SELF = 0,
+#define RUSAGE_SELF RUSAGE_SELF
+
+ /* All of its terminated child processes. */
+ RUSAGE_CHILDREN = -1
#define RUSAGE_CHILDREN RUSAGE_CHILDREN
- };
+};
#define __need_timeval
-#include/* For `struct timeval'. */
+#include /* For `struct timeval'. */
/* Structure which says how much of each resource has been used. */
struct ru
r1096 - trunk/type-handling/debian
Author: aurel32
Date: 2006-01-23 11:19:28 + (Mon, 23 Jan 2006)
New Revision: 1096
Modified:
trunk/type-handling/debian/changelog
trunk/type-handling/debian/rules
Log:
type-handling (0.2.19) unstable; urgency=low
* Don't provide anything containing x86_64, as the underscore symbol is
not allowed in packages name.
-- Aurelien Jarno <[EMAIL PROTECTED]> Mon, 23 Jan 2006 12:16:24 +0100
Modified: trunk/type-handling/debian/changelog
===
--- trunk/type-handling/debian/changelog2006-01-23 10:05:11 UTC (rev
1095)
+++ trunk/type-handling/debian/changelog2006-01-23 11:19:28 UTC (rev
1096)
@@ -1,3 +1,10 @@
+type-handling (0.2.19) unstable; urgency=low
+
+ * Don't provide anything containing x86_64, as the underscore symbol is
+not allowed in packages name.
+
+ -- Aurelien Jarno <[EMAIL PROTECTED]> Mon, 23 Jan 2006 12:16:24 +0100
+
type-handling (0.2.18) unstable; urgency=low
* Provides and filters out both the Debian and GNU type/CPU. Should fix
Modified: trunk/type-handling/debian/rules
===
--- trunk/type-handling/debian/rules2006-01-23 10:05:11 UTC (rev 1095)
+++ trunk/type-handling/debian/rules2006-01-23 11:19:28 UTC (rev 1096)
@@ -11,7 +11,7 @@
DEB_BUILD_GNU_CPU ?= $(shell dpkg-architecture -qDEB_BUILD_GNU_CPU)
DEB_BUILD_GNU_TYPE ?= $(shell dpkg-architecture -qDEB_BUILD_GNU_TYPE)
-cpus := $(shell grep -v ^\# /usr/share/dpkg/cputable | (while read debian gnu
regex ; do echo $${debian} ; echo $${gnu}; done) | sort -u)
+cpus := $(shell grep -v ^\# /usr/share/dpkg/cputable | (while read debian gnu
regex ; do echo $${debian} ; echo $${gnu}; done) | grep -vx x86_64 | sort -u)
systems := $(shell grep -v ^\# /usr/share/dpkg/ostable | (while read debian
gnu regex ; do echo $${debian} ; echo $${gnu}; done) | sort -u)
not_cpus := $(shell echo $(cpus) | tr " " "\n" | grep -vx
$(DEB_BUILD_ARCH_CPU) | grep -vx $(DEB_BUILD_GNU_CPU))
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: r1097 - in trunk/glibc-2.3-head/sysdeps/kfreebsd: . bits
Petr Salinger a écrit : Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/bits/socket.h === --- trunk/glibc-2.3-head/sysdeps/kfreebsd/bits/socket.h 2006-01-23 11:19:28 UTC (rev 1096) +++ trunk/glibc-2.3-head/sysdeps/kfreebsd/bits/socket.h 2006-01-23 13:35:42 UTC (rev 1097) @@ -189,8 +189,10 @@ #define MSG_DONTWAIT MSG_DONTWAIT MSG_EOF= 0x100, /* Data completes connection. */ #define MSG_EOF MSG_EOF -MSG_COMPAT = 0x8000 /* Used in sendit(). */ +MSG_COMPAT = 0x8000, /* Used in sendit(). */ #define MSG_COMPAT MSG_COMPAT + MSG_NOSIGNAL = 0x2 /* do not generate SIGPIPE on EOF */ +#define MSG_NOSIGNAL MSG_NOSIGNAL }; Does our kernel actually support this flag? Looking at the linux compat layer code, this flag is currently ignored. If it is really the case, it should be better to remove it from the .h file, as some configure scripts try to detect if they could use this flag or not. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: r1097 - in trunk/glibc-2.3-head/sysdeps/kfreebsd: . bits
On Mon, Jan 23, 2006 at 05:11:45PM +0100, Petr Salinger wrote: > >Does our kernel actually support this flag? Looking at the linux compat > >layer code, this flag is currently ignored. > > You are right, it became available with FreeBSD 6.0. Ok, in such case we may keep it, and speed up the switch FreeBSD 6.0 as the default kernel. Maybe we have to wait for the rate of Security Advisories to go lower first. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1101 - trunk/glibc-2.3-head/sysdeps/kfreebsd
Author: aurel32
Date: 2006-01-25 12:49:41 + (Wed, 25 Jan 2006)
New Revision: 1101
Modified:
trunk/glibc-2.3-head/sysdeps/kfreebsd/uname.c
Log:
Revert patch r1091
Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/uname.c
===
--- trunk/glibc-2.3-head/sysdeps/kfreebsd/uname.c 2006-01-23 16:18:58 UTC
(rev 1100)
+++ trunk/glibc-2.3-head/sysdeps/kfreebsd/uname.c 2006-01-25 12:49:41 UTC
(rev 1101)
@@ -21,7 +21,6 @@
#include
#include
#include
-#include /* For getdomainname() */
#define SYSNAME"GNU/kFreeBSD"
#define SYSNAME_LEN13
@@ -95,13 +94,6 @@
name->machine[len] = '\0';
}
- /* Fill domainname. Use getdomainname(). */
- if (getdomainname(name->domainname, sizeof(name->domainname)) == -1)
-{
- if (errno != EINVAL)
-strncpy (name->domainname, UNAME_DOMAINNAME, sizeof
(name->domainname));
-}
-
return 0;
}
libc_hidden_def (__uname)
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1113 - trunk/glibc-2.3-head
Author: aurel32
Date: 2006-01-27 05:00:47 + (Fri, 27 Jan 2006)
New Revision: 1113
Modified:
trunk/glibc-2.3-head/debian.sh
Log:
Ignore .disabled files when creating debian patches
Modified: trunk/glibc-2.3-head/debian.sh
===
--- trunk/glibc-2.3-head/debian.sh 2006-01-26 21:10:49 UTC (rev 1112)
+++ trunk/glibc-2.3-head/debian.sh 2006-01-27 05:00:47 UTC (rev 1113)
@@ -12,8 +12,8 @@
cp -a $1/sysdeps/kfreebsd ${tmp}/sysdeps/unix/bsd/bsd4.4/
cp -a $1/linuxthreads/kfreebsd ${tmp}/linuxthreads/sysdeps/unix/bsd/bsd4.4/
echo "kfreebsd-sysdeps.diff -p0" >> debian/patches/series
-(cd ${tmp} && diff -x .svn -Nurd null sysdeps/ ) >
debian/patches/kfreebsd-sysdeps.diff
-(cd ${tmp} && diff -x .svn -Nurd null linuxthreads/ ) >>
debian/patches/kfreebsd-sysdeps.diff
+(cd ${tmp} && diff -x .svn -x *.disabled -Nurd null sysdeps/ ) >
debian/patches/kfreebsd-sysdeps.diff
+(cd ${tmp} && diff -x .svn -x *.disabled -Nurd null linuxthreads/ ) >>
debian/patches/kfreebsd-sysdeps.diff
rm -rf ${tmp}
for i in `ls $1/patches` ; do
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1114 - trunk/glibc-2.3-head
Author: aurel32
Date: 2006-01-27 05:06:36 + (Fri, 27 Jan 2006)
New Revision: 1114
Modified:
trunk/glibc-2.3-head/debian.sh
Log:
Fix to my previous patch, seems I was not really awaken...
Modified: trunk/glibc-2.3-head/debian.sh
===
--- trunk/glibc-2.3-head/debian.sh 2006-01-27 05:00:47 UTC (rev 1113)
+++ trunk/glibc-2.3-head/debian.sh 2006-01-27 05:06:36 UTC (rev 1114)
@@ -12,12 +12,12 @@
cp -a $1/sysdeps/kfreebsd ${tmp}/sysdeps/unix/bsd/bsd4.4/
cp -a $1/linuxthreads/kfreebsd ${tmp}/linuxthreads/sysdeps/unix/bsd/bsd4.4/
echo "kfreebsd-sysdeps.diff -p0" >> debian/patches/series
-(cd ${tmp} && diff -x .svn -x *.disabled -Nurd null sysdeps/ ) >
debian/patches/kfreebsd-sysdeps.diff
-(cd ${tmp} && diff -x .svn -x *.disabled -Nurd null linuxthreads/ ) >>
debian/patches/kfreebsd-sysdeps.diff
+(cd ${tmp} && diff -x .svn -x -Nurd null sysdeps/ ) >
debian/patches/kfreebsd-sysdeps.diff
+(cd ${tmp} && diff -x .svn -x -Nurd null linuxthreads/ ) >>
debian/patches/kfreebsd-sysdeps.diff
rm -rf ${tmp}
for i in `ls $1/patches` ; do
- cat $1/patches/$i/* > debian/patches/kfreebsd-$i.diff
+ cat $1/patches/$i/*.patch > debian/patches/kfreebsd-$i.diff
echo "kfreebsd-$i.diff -p0" >> debian/patches/series
done
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1115 - trunk/glibc-2.3-head
Author: aurel32
Date: 2006-01-27 05:37:21 + (Fri, 27 Jan 2006)
New Revision: 1115
Modified:
trunk/glibc-2.3-head/debian.sh
Log:
Really revert revision 1113
Modified: trunk/glibc-2.3-head/debian.sh
===
--- trunk/glibc-2.3-head/debian.sh 2006-01-27 05:06:36 UTC (rev 1114)
+++ trunk/glibc-2.3-head/debian.sh 2006-01-27 05:37:21 UTC (rev 1115)
@@ -12,8 +12,8 @@
cp -a $1/sysdeps/kfreebsd ${tmp}/sysdeps/unix/bsd/bsd4.4/
cp -a $1/linuxthreads/kfreebsd ${tmp}/linuxthreads/sysdeps/unix/bsd/bsd4.4/
echo "kfreebsd-sysdeps.diff -p0" >> debian/patches/series
-(cd ${tmp} && diff -x .svn -x -Nurd null sysdeps/ ) >
debian/patches/kfreebsd-sysdeps.diff
-(cd ${tmp} && diff -x .svn -x -Nurd null linuxthreads/ ) >>
debian/patches/kfreebsd-sysdeps.diff
+(cd ${tmp} && diff -x .svn -Nurd null sysdeps/ ) >
debian/patches/kfreebsd-sysdeps.diff
+(cd ${tmp} && diff -x .svn -Nurd null linuxthreads/ ) >>
debian/patches/kfreebsd-sysdeps.diff
rm -rf ${tmp}
for i in `ls $1/patches` ; do
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1122 - in trunk/kfreebsd-5/debian: . patches
Author: aurel32
Date: 2006-02-03 17:11:18 + (Fri, 03 Feb 2006)
New Revision: 1122
Added:
trunk/kfreebsd-5/debian/patches/000_sack.diff
Modified:
trunk/kfreebsd-5/debian/changelog
Log:
* Fix an infinite loop in SACK handling (FreeBSD-SA-06:08.sack /
CVE-2006-0433).
Modified: trunk/kfreebsd-5/debian/changelog
===
--- trunk/kfreebsd-5/debian/changelog 2006-02-02 12:59:35 UTC (rev 1121)
+++ trunk/kfreebsd-5/debian/changelog 2006-02-03 17:11:18 UTC (rev 1122)
@@ -1,3 +1,11 @@
+kfreebsd-5 (5.4-13) unstable; urgency=high
+
+ * Urgency set to high as this fixes a security bug.
+ * Fix an infinite loop in SACK handling (FreeBSD-SA-06:08.sack /
+CVE-2006-0433).
+
+ -- Aurelien Jarno <[EMAIL PROTECTED]> Fri, 3 Feb 2006 17:50:38 +0100
+
kfreebsd-5 (5.4-12) unstable; urgency=low
* Recommends libc0.1-i686 in kfreebsd-image*, not kfreebsd-headers*
Added: trunk/kfreebsd-5/debian/patches/000_sack.diff
===
--- trunk/kfreebsd-5/debian/patches/000_sack.diff 2006-02-02 12:59:35 UTC
(rev 1121)
+++ trunk/kfreebsd-5/debian/patches/000_sack.diff 2006-02-03 17:11:18 UTC
(rev 1122)
@@ -0,0 +1,24 @@
+Index: sys/netinet/tcp_sack.c
+===
+RCS file: /home/ncvs/src/sys/netinet/tcp_sack.c,v
+retrieving revision 1.3
+diff -u -p -I__FBSDID -r1.3 tcp_sack.c
+--- sys/netinet/tcp_sack.c 17 Aug 2004 22:05:54 - 1.3
sys/netinet/tcp_sack.c 26 Jan 2006 15:18:05 -
+@@ -301,6 +301,7 @@ tcp_sack_option(struct tcpcb *tp, struct
+ tp->snd_numholes = 0;
+ if (tp->t_maxseg == 0)
+ panic("tcp_sack_option"); /* Should never happen */
++next_block:
+ while (tmp_olen > 0) {
+ struct sackblk sack;
+
+@@ -390,7 +391,7 @@ tcp_sack_option(struct tcpcb *tp, struct
+ temp = (struct sackhole *)
+ uma_zalloc(sack_hole_zone,M_NOWAIT);
+ if (temp == NULL)
+- continue; /* ENOBUFS */
++ goto next_block; /* ENOBUFS */
+ temp->next = cur->next;
+ temp->start = sack.end;
+ temp->end = cur->end;
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1128 - in trunk/glibc-2.3-head: patches/fixes sysdeps/kfreebsd
Author: aurel32
Date: 2006-02-04 18:32:19 + (Sat, 04 Feb 2006)
New Revision: 1128
Removed:
trunk/glibc-2.3-head/patches/fixes/elfosabi.patch
Modified:
trunk/glibc-2.3-head/sysdeps/kfreebsd/dl-machine.h
Log:
Change the OSABI in dl-machine.h, without touching to elf/dl-load.c, as
suggested by Roland.
Deleted: trunk/glibc-2.3-head/patches/fixes/elfosabi.patch
===
--- trunk/glibc-2.3-head/patches/fixes/elfosabi.patch 2006-02-04 13:45:24 UTC
(rev 1127)
+++ trunk/glibc-2.3-head/patches/fixes/elfosabi.patch 2006-02-04 18:32:19 UTC
(rev 1128)
@@ -1,29 +0,0 @@
elf/dl-load.c.orig 2005-04-06 04:50:10.0 +0200
-+++ elf/dl-load.c 2006-01-13 20:50:51.0 +0100
-@@ -1478,12 +1478,16 @@
- static int
- open_verify (const char *name, struct filebuf *fbp)
- {
-+ /* This is the expected ELF OSABI */
-+#ifndef ELF_OSABI
-+# define ELF_OSABI ELFOSABI_SYSV
-+#endif
- /* This is the expected ELF header. */
- #define ELF32_CLASS ELFCLASS32
- #define ELF64_CLASS ELFCLASS64
- #ifndef VALID_ELF_HEADER
- # define VALID_ELF_HEADER(hdr,exp,size) (memcmp (hdr, exp, size) == 0)
--# define VALID_ELF_OSABI(osabi) (osabi == ELFOSABI_SYSV)
-+# define VALID_ELF_OSABI(osabi) (osabi == ELF_OSABI)
- # define VALID_ELF_ABIVERSION(ver)(ver == 0)
- #endif
- static const unsigned char expected[EI_PAD] =
-@@ -1495,7 +1499,7 @@
- [EI_CLASS] = ELFW(CLASS),
- [EI_DATA] = byteorder,
- [EI_VERSION] = EV_CURRENT,
--[EI_OSABI] = ELFOSABI_SYSV,
-+[EI_OSABI] = ELF_OSABI,
- [EI_ABIVERSION] = 0
- };
- static const struct
Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/dl-machine.h
===
--- trunk/glibc-2.3-head/sysdeps/kfreebsd/dl-machine.h 2006-02-04 13:45:24 UTC
(rev 1127)
+++ trunk/glibc-2.3-head/sysdeps/kfreebsd/dl-machine.h 2006-02-04 18:32:19 UTC
(rev 1128)
@@ -21,6 +21,11 @@
/* Contrary to most kernels which use ELFOSABI_SYSV aka ELFOSABI_NONE,
FreeBSD uses ELFOSABI_FREEBSD for the OSABI field. */
-#define ELF_OSABI ELFOSABI_FREEBSD
+# define VALID_ELF_OSABI(osabi)(osabi == ELFOSABI_SYSV)
+# define VALID_ELF_ABIVERSION(ver) (ver == 0)
+# define VALID_ELF_HEADER(hdr,exp,size) \
+ memcmp (hdr,exp,size-2) == 0 \
+ && VALID_ELF_OSABI (hdr[EI_OSABI]) \
+ && VALID_ELF_ABIVERSION (hdr[EI_ABIVERSION])
#include_next "dl-machine.h"
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1129 - trunk/glibc-2.3-head/sysdeps/kfreebsd
Author: aurel32 Date: 2006-02-04 18:33:15 + (Sat, 04 Feb 2006) New Revision: 1129 Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/dl-machine.h Log: Oops, we are using ELFOSABI_FREEBSD, not ELFOSABI_SYSV Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/dl-machine.h === --- trunk/glibc-2.3-head/sysdeps/kfreebsd/dl-machine.h 2006-02-04 18:32:19 UTC (rev 1128) +++ trunk/glibc-2.3-head/sysdeps/kfreebsd/dl-machine.h 2006-02-04 18:33:15 UTC (rev 1129) @@ -21,7 +21,7 @@ /* Contrary to most kernels which use ELFOSABI_SYSV aka ELFOSABI_NONE, FreeBSD uses ELFOSABI_FREEBSD for the OSABI field. */ -# define VALID_ELF_OSABI(osabi)(osabi == ELFOSABI_SYSV) +# define VALID_ELF_OSABI(osabi)(osabi == ELFOSABI_FREEBSD) # define VALID_ELF_ABIVERSION(ver) (ver == 0) # define VALID_ELF_HEADER(hdr,exp,size) \ memcmp (hdr,exp,size-2) == 0 \ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1148 - in trunk/kfreebsd-5/debian: . patches
Author: aurel32
Date: 2006-02-06 23:07:37 + (Mon, 06 Feb 2006)
New Revision: 1148
Added:
trunk/kfreebsd-5/debian/patches/000_pf.diff
Modified:
trunk/kfreebsd-5/debian/changelog
Log:
* Fix IP fragment handling panic in pf (FreeBSD-SA-06:07.pf / CVE-2006-0381).
Modified: trunk/kfreebsd-5/debian/changelog
===
--- trunk/kfreebsd-5/debian/changelog 2006-02-06 17:25:27 UTC (rev 1147)
+++ trunk/kfreebsd-5/debian/changelog 2006-02-06 23:07:37 UTC (rev 1148)
@@ -1,3 +1,11 @@
+kfreebsd-5 (5.4-14) unstable; urgency=high
+
+ * Urgency set to high as this fixes a security bug.
+ * Fix IP fragment handling panic in pf (FreeBSD-SA-06:07.pf /
+CVE-2006-0381).
+
+ -- Aurelien Jarno <[EMAIL PROTECTED]> Tue, 7 Feb 2006 00:07:16 +0100
+
kfreebsd-5 (5.4-13) unstable; urgency=high
* Urgency set to high as this fixes a security bug.
Added: trunk/kfreebsd-5/debian/patches/000_pf.diff
===
--- trunk/kfreebsd-5/debian/patches/000_pf.diff 2006-02-06 17:25:27 UTC (rev
1147)
+++ trunk/kfreebsd-5/debian/patches/000_pf.diff 2006-02-06 23:07:37 UTC (rev
1148)
@@ -0,0 +1,16 @@
+Index: sys/contrib/pf/net/pf_norm.c
+===
+RCS file: /home/ncvs/src/sys/contrib/pf/net/pf_norm.c,v
+retrieving revision 1.11.2.2
+diff -u -p -I__FBSDID -r1.11.2.2 pf_norm.c
+--- sys/contrib/pf/net/pf_norm.c 17 Jan 2006 13:05:32 - 1.11.2.2
sys/contrib/pf/net/pf_norm.c 22 Jan 2006 16:38:31 -
+@@ -818,7 +818,7 @@ pf_fragcache(struct mbuf **m0, struct ip
+ } else {
+ hosed++;
+ }
+- } else {
++ } else if (frp == NULL) {
+ /* There is a gap between fragments */
+ DPFPRINTF(("fragcache[%d]: gap %d %d-%d (%d-%d)\n",
+ h->ip_id, -aftercut, off, max, fra->fr_off,
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1150 - trunk/web/patches
Author: aurel32
Date: 2006-02-08 14:09:23 + (Wed, 08 Feb 2006)
New Revision: 1150
Modified:
trunk/web/patches/cdparanoia_3a9.8-11.diff
Log:
* Use cdda_device_name of struct cdrom_drive to expose the path of the
device node. Previously this was only present in the Linux version and
is needed for KDE 3.5.1.
Modified: trunk/web/patches/cdparanoia_3a9.8-11.diff
===
--- trunk/web/patches/cdparanoia_3a9.8-11.diff 2006-02-08 12:14:49 UTC (rev
1149)
+++ trunk/web/patches/cdparanoia_3a9.8-11.diff 2006-02-08 14:09:23 UTC (rev
1150)
@@ -1,3 +1,6 @@
+Author: FreeBSD + aurel32
+Status: in BTS
+
diff -u cdparanoia-3a9.8/Makefile.in cdparanoia-3a9.8/Makefile.in
--- cdparanoia-3a9.8/Makefile.in
+++ cdparanoia-3a9.8/Makefile.in
@@ -147,7 +150,7 @@
cdrom_drive *cdda_identify_cooked(const char *dev, int messagedest,
char **messages){
-@@ -275,6 +299,60 @@
+@@ -275,6 +299,61 @@
return(d);
}
@@ -184,6 +187,7 @@
+ goto cdda_identify_cooked_fail;
+ }
+
++ d->cdda_device_name = copystring(dev);
+ d->drive_model = copystring("Generic cooked ioctl CDROM");
+ d->interface = COOKED_IOCTL;
+ d->bigendianp = -1;
@@ -208,7 +212,7 @@
struct sg_id {
longl1; /* target | lun << 8 | channel << 16 | low_ino << 24 */
longl2; /* Unique id */
-@@ -390,6 +468,7 @@
+@@ -390,6 +469,7 @@
if(dev!=-1)close(dev);
return(NULL);
}
@@ -216,7 +220,7 @@
void strscat(char *a,char *b,int n){
int i;
-@@ -401,6 +480,7 @@
+@@ -401,6 +481,7 @@
strcat(a," ");
}
@@ -224,7 +228,7 @@
/* At this point, we're going to punt compatability before SG2, and
allow only SG2 and SG3 */
static int verify_SG_version(cdrom_drive *d,int messagedest,
-@@ -654,6 +734,88 @@
+@@ -654,6 +735,89 @@
if(g_fd!=-1)close(g_fd);
return(NULL);
}
@@ -276,6 +280,7 @@
+ goto cdda_identify_scsi_fail;
+ }
+
++ d->cdda_device_name = copystring(devname);
+ d->ioctl_fd = -1;
+ d->bigendianp = -1;
+ d->nsectors = -1;
@@ -317,9 +322,9 @@
--- cdparanoia-3a9.8/interface/utils.h
+++ cdparanoia-3a9.8/interface/utils.h
@@ -1,4 +1,9 @@
-+#ifdef LINUX
++#if defined (__linux__) || defined (__GLIBC__)
#include
-+#elif defined(__FreeBSD__) || defined(__FreeBSD_kernel__)
++#elif defined(__FreeBSD__)
+#include
+#endif
+
@@ -329,7 +334,15 @@
diff -u cdparanoia-3a9.8/debian/changelog cdparanoia-3a9.8/debian/changelog
--- cdparanoia-3a9.8/debian/changelog
+++ cdparanoia-3a9.8/debian/changelog
-@@ -1,3 +1,9 @@
+@@ -1,3 +1,17 @@
++cdparanoia (3a9.8-11+kbsd.1) unreleased; urgency=low
++
++ * Use cdda_device_name of struct cdrom_drive to expose the path of the
++device node. Previously this was only present in the Linux version and
++is needed for KDE 3.5.1.
++
++ -- Aurelien Jarno <[EMAIL PROTECTED]> Wed, 8 Feb 2006 14:43:21 +0100
++
+cdparanoia (3a9.8-11+kbsd) unreleased; urgency=low
+
+ * Added support for GNU/kFreeBSD
@@ -370,9 +383,9 @@
+++ cdparanoia-3a9.8/utils.h
@@ -1,5 +1,9 @@
#include
-+#if defined (__linux__)
++#if defined (__linux__) || defined(__GLIBC__)
#include
-+#elif defined(__FreeBSD__) || defined(__FreeBSD_kernel__)
++#elif defined(__FreeBSD__)
+#include
+#endif
#include
@@ -424,12 +437,11 @@
#define MAXTRK 100
typedef struct TOC { /* structure of table of contents */
-@@ -46,14 +51,20 @@
-
+@@ -47,13 +52,19 @@
int opened; /* This struct may just represent a candidate for opening */
+ char *cdda_device_name;
+#if defined(__linux__)
- char *cdda_device_name;
char *ioctl_device_name;
int cdda_fd;
@@ -706,12 +718,11 @@
d->set_speed = cooked_setspeed;
--- cdparanoia-3a9.8.orig/interface/interface.c
+++ cdparanoia-3a9.8/interface/interface.c
-@@ -29,12 +29,25 @@
- d->enable_cdda(d,0);
+@@ -30,11 +30,24 @@
_clean_messages(d);
+ if(d->cdda_device_name)free(d->cdda_device_name);
+#if defined(__linux__)
- if(d->cdda_device_name)free(d->cdda_device_name);
if(d->ioctl_device_name)free(d->ioctl_device_name);
if(d->drive_model)free(d->drive_model);
if(d->cdda_fd!=-1)close(d->cdda_fd);
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1151 - in trunk/glibc-2.3-head/patches: . fixes upstream
Author: aurel32
Date: 2006-02-08 15:29:24 + (Wed, 08 Feb 2006)
New Revision: 1151
Added:
trunk/glibc-2.3-head/patches/upstream/
trunk/glibc-2.3-head/patches/upstream/errlist.patch
trunk/glibc-2.3-head/patches/upstream/no_nfs_h.patch
Removed:
trunk/glibc-2.3-head/patches/fixes/errlist.patch
trunk/glibc-2.3-head/patches/fixes/no_nfs_h.patch
Log:
Move patches merged upstream to a separate dir
Deleted: trunk/glibc-2.3-head/patches/fixes/errlist.patch
===
--- trunk/glibc-2.3-head/patches/fixes/errlist.patch2006-02-08 14:09:23 UTC
(rev 1150)
+++ trunk/glibc-2.3-head/patches/fixes/errlist.patch2006-02-08 15:29:24 UTC
(rev 1151)
@@ -1,14 +0,0 @@
-
-Status: in upstream CVS.
-
sysdeps/gnu/errlist-compat.awk.orig2005-11-18 07:48:29.0
+0100
-+++ sysdeps/gnu/errlist-compat.awk 2005-11-18 07:48:41.0 +0100
-@@ -35,7 +35,7 @@
- # Don't process any further Versions files
- ARGC = ARGIND + 1;
- cnt = $2 + 0;
-- if (cnt < 100) {
-+ if (cnt < 80) {
- print "*** this line seems bogus:", $0 > "/dev/stderr";
- exit 1;
- }
Deleted: trunk/glibc-2.3-head/patches/fixes/no_nfs_h.patch
===
--- trunk/glibc-2.3-head/patches/fixes/no_nfs_h.patch 2006-02-08 14:09:23 UTC
(rev 1150)
+++ trunk/glibc-2.3-head/patches/fixes/no_nfs_h.patch 2006-02-08 15:29:24 UTC
(rev 1151)
@@ -1,17 +0,0 @@
-
-Author: rmh
-Status: Fixed in CVS HEAD as of 20060117 (by adding sysdeps/generic/nfs/nfs.h)
-
sunrpc/rpcsvc/bootparam_prot.x~1999-04-18 19:16:14.0 +0200
-+++ sunrpc/rpcsvc/bootparam_prot.x 2004-11-27 22:27:20.0 +0100
-@@ -44,8 +44,10 @@
- %#include
- %#include
- %#include
-+#ifdef __linux__
- %#include
- #endif
-+#endif
-
- const MAX_MACHINE_NAME = 255;
- const MAX_PATH_LEN= 1024;
Copied: trunk/glibc-2.3-head/patches/upstream/errlist.patch (from rev 1150,
trunk/glibc-2.3-head/patches/fixes/errlist.patch)
Copied: trunk/glibc-2.3-head/patches/upstream/no_nfs_h.patch (from rev 1150,
trunk/glibc-2.3-head/patches/fixes/no_nfs_h.patch)
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1173 - trunk/glibc-2.3-head
Author: aurel32 Date: 2006-02-12 01:42:21 + (Sun, 12 Feb 2006) New Revision: 1173 Modified: trunk/glibc-2.3-head/debian.sh Log: Support for glibc 2.3.6 Modified: trunk/glibc-2.3-head/debian.sh === --- trunk/glibc-2.3-head/debian.sh 2006-02-11 17:34:21 UTC (rev 1172) +++ trunk/glibc-2.3-head/debian.sh 2006-02-12 01:42:21 UTC (rev 1173) @@ -37,7 +37,7 @@ Priority: required -Build-Depends: gettext (>= 0.10.37-1), make (>= 3.80-1), dpkg-dev (>= 1.13.5), debianutils (>= 1.13.1), tar (>= 1.13.11), bzip2, texinfo (>= 4.0), linux-kernel-headers (>= 2.6.13+0rc3-2) [!hurd-i386], mig (>= 1.3-2) [hurd-i386], hurd-dev (>= 20020608-1) [hurd-i386], gnumach-dev [hurd-i386], texi2html, file, gcc-4.0 [!powerpc !m68k !hppa !hurd-i386], gcc-3.4 (>= 3.4.4-6) [powerpc], gcc-3.4 [m68k hppa], gcc-3.3 [hurd-i386], autoconf, binutils (>= 2.16.1cvs20051109-1), sed (>= 4.0.5-4), gawk, debhelper (>= 4.1.76), libc6-dev-amd64 [i386], libc6-dev-ppc64 [powerpc], quilt +Build-Depends: gettext (>= 0.10.37-1), make (>= 3.80-1), dpkg-dev (>= 1.13.5), debianutils (>= 1.13.1), tar (>= 1.13.11), bzip2, texinfo (>= 4.0), linux-kernel-headers (>= 2.6.13+0rc3-2) [!hurd-i386 !kfreebsd-i386], mig (>= 1.3-2) [hurd-i386], hurd-dev (>= 20020608-1) [hurd-i386], gnumach-dev [hurd-i386], kfreebsd-kernel-headers (>= 0.01) [kfreebsd-i386], texi2html, file, gcc-4.0 [!powerpc !m68k !hppa !hurd-i386], gcc-3.4 (>= 3.4.4-6) [powerpc], gcc-3.4 [m68k hppa], gcc-3.3 [hurd-i386], autoconf, binutils (>= 2.16.1cvs20051109-1), sed (>= 4.0.5-4), gawk, debhelper (>= 4.1.76), libc6-dev-amd64 [i386], libc6-dev-ppc64 [powerpc], quilt - Build-Depends-Indep: perl, po-debconf, po4a + Build-Depends-Indep: perl, po-debconf Maintainer: GNU Libc Maintainers Uploaders: Ben Collins <[EMAIL PROTECTED]>, GOTO Masanori <[EMAIL PROTECTED]>, Philip Blundell <[EMAIL PROTECTED]>, Jeff Bailey <[EMAIL PROTECTED]>, Daniel Jacobowitz <[EMAIL PROTECTED]>, Clint Adams <[EMAIL PROTECTED]>, Denis Barbier <[EMAIL PROTECTED]> --- glibc-2.3.5/debian/sysdeps/kfreebsd-gnu.mk -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1208 - trunk/freebsd-libs/debian
Author: aurel32 Date: 2006-02-15 21:02:53 + (Wed, 15 Feb 2006) New Revision: 1208 Removed: trunk/freebsd-libs/debian/libfreebsd-dev.install trunk/freebsd-libs/debian/libfreebsd0.install Log: Remove libfreebsd0.install libfreebsd-dev.install, this library no has separate source Deleted: trunk/freebsd-libs/debian/libfreebsd-dev.install === --- trunk/freebsd-libs/debian/libfreebsd-dev.install2006-02-15 20:41:06 UTC (rev 1207) +++ trunk/freebsd-libs/debian/libfreebsd-dev.install2006-02-15 21:02:53 UTC (rev 1208) @@ -1,5 +0,0 @@ -usr/lib/libfreebsd*.so -usr/lib/libfreebsd*.a -usr/include/nlist.h -usr/share/man/man3/nlist.3.gz - Deleted: trunk/freebsd-libs/debian/libfreebsd0.install === --- trunk/freebsd-libs/debian/libfreebsd0.install 2006-02-15 20:41:06 UTC (rev 1207) +++ trunk/freebsd-libs/debian/libfreebsd0.install 2006-02-15 21:02:53 UTC (rev 1208) @@ -1 +0,0 @@ -usr/lib/libfreebsd*.so.* -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1210 - trunk/freebsd-libs/debian
Author: aurel32
Date: 2006-02-15 21:13:28 + (Wed, 15 Feb 2006)
New Revision: 1210
Modified:
trunk/freebsd-libs/debian/control
trunk/freebsd-libs/debian/rules
Log:
Compile fix for libkvm
libcam0 don't need to depend on libfreebsd0
Modified: trunk/freebsd-libs/debian/control
===
--- trunk/freebsd-libs/debian/control 2006-02-15 21:04:09 UTC (rev 1209)
+++ trunk/freebsd-libs/debian/control 2006-02-15 21:13:28 UTC (rev 1210)
@@ -24,7 +24,7 @@
Package: libcam0
Architecture: kfreebsd-i386
-Depends: ${shlibs:Depends}, libfreebsd0 (>= ${Source-Version})
+Depends: ${shlibs:Depends}
Description: FreeBSD CAM (Common Access Method) library
The CAM (Common Access Method) library provides a way for userland
SCSI applications to construct a SCSI command.
Modified: trunk/freebsd-libs/debian/rules
===
--- trunk/freebsd-libs/debian/rules 2006-02-15 21:04:09 UTC (rev 1209)
+++ trunk/freebsd-libs/debian/rules 2006-02-15 21:13:28 UTC (rev 1210)
@@ -9,7 +9,7 @@
# This has to be exported to make some magic below work.
export DH_OPTIONS
-CFLAGS = -Wall -g -pipe -fPIC -D__FBSDID\(s\)=
+CFLAGS = -Wall -g -pipe -fPIC -D__FBSDID\(s\)= -I.
ifneq (,$(findstring noopt,$(DEB_BUILD_OPTIONS)))
CFLAGS += -O0
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: mailing lists
Robert Millan a écrit : Hi, I spotted a few problems with glibc-bsd-devel mailing list: - It needs maintanance work (Santiago kindly offered to help on this, but this is still a PITA). - Message posts lag a lot (I've seen messages arriving the list ~40h after they were sent). - Users are confused and go to debian-bsd to ask questions instead. Which is not strange, since: - debian-bsd is the official list, and is marked as a general Debian *BSD list in http://lists.debian.org/ - glibc-bsd-devel has -devel in its name. I think it would be good to find a solution, but I'm not sure which one. Is everyone happy with the current situation? How about switching everything to debian-bsd ? -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1225 - trunk/kfreebsd-5/debian/arch/i386
Author: aurel32 Date: 2006-02-17 09:30:00 + (Fri, 17 Feb 2006) New Revision: 1225 Modified: trunk/kfreebsd-5/debian/arch/i386/586-smp.config Log: Also define I686_CPU for 586-smp flavour for kfreebsd-5 Modified: trunk/kfreebsd-5/debian/arch/i386/586-smp.config === --- trunk/kfreebsd-5/debian/arch/i386/586-smp.config2006-02-17 09:25:37 UTC (rev 1224) +++ trunk/kfreebsd-5/debian/arch/i386/586-smp.config2006-02-17 09:30:00 UTC (rev 1225) @@ -3,6 +3,7 @@ machinei386 cpuI586_CPU +cpuI686_CPU makeoptions COPTFLAGS="-O -pipe -march=i586" ident Debian-i586-SMP -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: GCC 4.1 in experimental / GCC for etch
Hi! Matthias Klose a écrit : The GCC (GNU compiler collection) 4.1 release candidate 1 can be found in experimental. "Porters", please make sure that the package is built and uploaded (if it's not built by the experimental buildd). Please check that the symbols exported in the 4.1 libraries are a superset of those exported in the 4.0 libraries (these are libgcc1 (except m68k and hppa), libstdc++6, libffi4, libobjc1, libmudflap0). The proposed GCC plan for etch consists of: - uploading GCC 4.0.3 to unstable (this release is expected shortly after the GCC 4.1.0 release), let 4.0.3 migrate to testing. The GCC website seems to be down currently, could you please tell us when the release are expected? - uploading GCC 4.1 to unstable for those architectures which do not have ABI problems (these should be all, but should be validated). Nice! - Once the 4.1 packages are migrated to testing, make 4.1 the default compiler for i386, amd64, powerpc. These are the architectures, which are considred primary (linux) architectures by GCC upstream. For the other Debian architectures, the GCC port maintainers and the Debian port maintainers should make the call, if and when the default GCC is changed. I think you could also make 4.1 the default one for kfreebsd-i386. They are very few differences with the linux version, and the testsuite shows good results. - Make gij-4.1/gcj-4.1 the default for all architectures. - Stop building compiler packages from the GCC 3.3 source; the only packages built will be libstdc++5 (and libgcc1 on hppa/m68k). AFAIK the 2.4 kernels in Debian are built with gcc-3.3. What are the plans wrt to them? - Stop building compilers from GCC 3.4.x, namely gobjc-3.4, gnat-3.4 and g++-3.4 (it looks like we can go without g++-3.4 for the etch release). Still build gpc-3.4, g77-3.4 and gcc-3.4, as g77 cannot be found anymore in 4.x releases. Currently three architectures are still using gcc 3.4 to build the glibc, namely m68k, hppa and powerpc. For powerpc, it seems we can make the switch to at least gcc 4.0. There is currently a problem of locales, but not sure it is related to gcc-4.0. For hppa, the glibc builds well with gcc 4.0, but create problem with python/perl. It still has to be investigated. For m68k, the glibc does not build, gcc 4.0 segfault on system.c. It looks like gcc 4.1 fixes the problem, but the resulting glibc has not been tested. I think compilers are critical for the glibc, so we will have to coordinate the changes. Aurélien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1245 - trunk/glibc-2.3-head
Author: aurel32
Date: 2006-02-23 19:01:20 + (Thu, 23 Feb 2006)
New Revision: 1245
Modified:
trunk/glibc-2.3-head/debian.sh
Log:
Remove the patch to the debian/ directory, now merged.
Modified: trunk/glibc-2.3-head/debian.sh
===
--- trunk/glibc-2.3-head/debian.sh 2006-02-22 08:22:09 UTC (rev 1244)
+++ trunk/glibc-2.3-head/debian.sh 2006-02-23 19:01:20 UTC (rev 1245)
@@ -21,222 +21,5 @@
echo "kfreebsd-$i.diff -p0" >> debian/patches/series
done
-patch -p1 < $0
-
-# re-generate debian/control
-debian/rules debian/control
-
exit 0
-diff -u glibc-2.3.5/debian/control.in/main glibc-2.3.5/debian/control.in/main
glibc-2.3.5/debian/control.in/main
-+++ glibc-2.3.5/debian/control.in/main
-@@ -1,7 +1,7 @@
- Source: @glibc@
- Section: libs
- Priority: required
--Build-Depends: gettext (>= 0.10.37-1), make (>= 3.80-1), dpkg-dev (>=
1.13.5), debianutils (>= 1.13.1), tar (>= 1.13.11), bzip2, texinfo (>= 4.0),
linux-kernel-headers (>= 2.6.13+0rc3-2) [!hurd-i386], mig (>= 1.3-2)
[hurd-i386], hurd-dev (>= 20020608-1) [hurd-i386], gnumach-dev [hurd-i386],
texi2html, file, gcc-4.0 [!powerpc !m68k !hppa !hurd-i386], gcc-3.4 (>=
3.4.4-6) [powerpc], gcc-3.4 [m68k hppa], gcc-3.3 [hurd-i386], autoconf,
binutils (>= 2.16.1cvs20051109-1), sed (>= 4.0.5-4), gawk, debhelper (>=
4.1.76), libc6-dev-amd64 [i386], libc6-dev-ppc64 [powerpc], quilt
-+Build-Depends: gettext (>= 0.10.37-1), make (>= 3.80-1), dpkg-dev (>=
1.13.5), debianutils (>= 1.13.1), tar (>= 1.13.11), bzip2, texinfo (>= 4.0),
linux-kernel-headers (>= 2.6.13+0rc3-2) [!hurd-i386 !kfreebsd-i386], mig (>=
1.3-2) [hurd-i386], hurd-dev (>= 20020608-1) [hurd-i386], gnumach-dev
[hurd-i386], kfreebsd-kernel-headers (>= 0.01) [kfreebsd-i386], texi2html,
file, gcc-4.0 [!powerpc !m68k !hppa !hurd-i386], gcc-3.4 (>= 3.4.4-6)
[powerpc], gcc-3.4 [m68k hppa], gcc-3.3 [hurd-i386], autoconf, binutils (>=
2.16.1cvs20051109-1), sed (>= 4.0.5-4), gawk, debhelper (>= 4.1.76),
libc6-dev-amd64 [i386], libc6-dev-ppc64 [powerpc], quilt
- Build-Depends-Indep: perl, po-debconf
- Maintainer: GNU Libc Maintainers
- Uploaders: Ben Collins <[EMAIL PROTECTED]>, GOTO Masanori <[EMAIL
PROTECTED]>, Philip Blundell <[EMAIL PROTECTED]>, Jeff Bailey <[EMAIL
PROTECTED]>, Daniel Jacobowitz <[EMAIL PROTECTED]>, Clint Adams <[EMAIL
PROTECTED]>, Denis Barbier <[EMAIL PROTECTED]>
glibc-2.3.5/debian/sysdeps/kfreebsd-gnu.mk
-+++ glibc-2.3.5.orig/debian/sysdeps/kfreebsd-gnu.mk
-@@ -1,11 +0,0 @@
--# This is for a Glibc-using FreeBSD system.
--
--GLIBC_OVERLAYS ?= $(shell ls glibc-linuxthreads* glibc-ports* glibc-libidn*)
--
--libc = libc1
--
--# Linuxthreads Config
--threads = yes
--libc_add-ons = linuxthreads $(add-ons)
--
--extra_config_options = $(extra_config_options) --disable-compatible-utmp
--enable-kernel-include=4.6
-diff -u glibc-2.3.5/debian/sysdeps/linux.mk glibc-2.3.5/debian/sysdeps/linux.mk
glibc-2.3.5/debian/sysdeps/linux.mk
-+++ glibc-2.3.5/debian/sysdeps/linux.mk
-@@ -39,7 +39,7 @@
- nptl_MIN_KERNEL_SUPPORTED = 2.6.0
- nptl_LIBDIR = /tls
-
--LINUX_HEADER_DIR = $(stamp)mkincludedir
-+KERNEL_HEADER_DIR = $(stamp)mkincludedir
- $(stamp)mkincludedir:
- rm -rf debian/include
- mkdir debian/include
-diff -u glibc-2.3.5/debian/rules.d/build.mk glibc-2.3.5/debian/rules.d/build.mk
glibc-2.3.5/debian/rules.d/build.mk
-+++ glibc-2.3.5/debian/rules.d/build.mk
-@@ -13,7 +13,7 @@
-
-
- $(patsubst %,mkbuilddir_%,$(GLIBC_PASSES)) :: mkbuilddir_% :
$(stamp)mkbuilddir_%
--$(stamp)mkbuilddir_%: $(stamp)patch-stamp $(LINUX_HEADER_DIR)
-+$(stamp)mkbuilddir_%: $(stamp)patch-stamp $(KERNEL_HEADER_DIR)
- @echo Making builddir for $(curpass)
- test -d $(DEB_BUILDDIR) || mkdir $(DEB_BUILDDIR)
- touch $@
glibc-2.3.5/debian/control.in/opt~ 2006-01-12 14:48:52.0 +0100
-+++ glibc-2.3.5/debian/control.in/opt 2006-01-12 14:49:46.0 +0100
-@@ -56,3 +56,23 @@
- Most notably, IBM's JDK. If you experience problems with such
- applications, you will need to remove this package.
-
-+Package: libc0.1-i686
-+Architecture: kfreebsd-i386
-+Section: libs
-+Priority: extra
-+Pre-Depends: @libc@ (= ${Source-Version})
-+Description: GNU C Library: Shared libraries [i686 optimized]
-+ Contains the standard libraries that are used by nearly all programs on
-+ the system. This package includes shared versions of the standard C
-+ library and the standard math library, as well as many others.
-+ .
-+ This set of libraries is optimized for i686 machines, and will only be
-+ used on an i686 class CPU (check the output of `uname -m'). This includes
-+ Pentium Pro, Pentium II/III/IV, Celeron CPU's and similar class CPU's
-+ (including clones such as AMD Athlon/Opteron, VIA C3 Nehemiah, but not VIA
-+ C3 Ezla).
-+ .
-+ WARNING: Some third-party binaries may not work well with these libraries.
-+ Most notably, IBM's JDK. If you experience problems with such
-+ appl
r1273 - trunk/glibc-2.3-head/sysdeps/kfreebsd
Author: aurel32 Date: 2006-02-28 10:24:03 + (Tue, 28 Feb 2006) New Revision: 1273 Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in Log: Also set arch_minimum_kernel for x86_64 Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in === --- trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in 2006-02-27 22:29:15 UTC (rev 1272) +++ trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in 2006-02-28 10:24:03 UTC (rev 1273) @@ -73,7 +73,7 @@ # if appropriate too. test -n "$arch_minimum_kernel" || case "$machine" in - i386*) + i386*|x86_64*) libc_cv_gcc_unwind_find_fde=yes arch_minimum_kernel=5.4.0 ;; -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1276 - trunk/glibc-2.3-head/sysdeps/kfreebsd
Author: aurel32 Date: 2006-02-28 15:05:21 + (Tue, 28 Feb 2006) New Revision: 1276 Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/configure trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in Log: Actually separate the definition of arch_minimum_kernel for each arches, as on Linux. Also don't set libc_cv_gcc_unwind_find_fde on x86_64, as on Linux. Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/configure === --- trunk/glibc-2.3-head/sysdeps/kfreebsd/configure 2006-02-28 12:41:24 UTC (rev 1275) +++ trunk/glibc-2.3-head/sysdeps/kfreebsd/configure 2006-02-28 15:05:21 UTC (rev 1276) @@ -116,10 +116,16 @@ # if appropriate too. test -n "$arch_minimum_kernel" || case "$machine" in - i386*|x86_64*) + i386*) libc_cv_gcc_unwind_find_fde=yes arch_minimum_kernel=5.4.0 ;; + x86_64*) +arch_minimum_kernel=5.4.0 +;; + *) +arch_minimum_kernel=5.4.0 +;; esac if test -n "$minimum_kernel"; then Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in === --- trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in 2006-02-28 12:41:24 UTC (rev 1275) +++ trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in 2006-02-28 15:05:21 UTC (rev 1276) @@ -73,10 +73,16 @@ # if appropriate too. test -n "$arch_minimum_kernel" || case "$machine" in - i386*|x86_64*) + i386*) libc_cv_gcc_unwind_find_fde=yes arch_minimum_kernel=5.4.0 ;; + x86_64*) +arch_minimum_kernel=5.4.0 +;; + *) +arch_minimum_kernel=5.4.0 +;; esac if test -n "$minimum_kernel"; then changequote(,) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1282 - trunk/kfreebsd-5/debian
Author: aurel32 Date: 2006-03-01 19:16:44 + (Wed, 01 Mar 2006) New Revision: 1282 Modified: trunk/kfreebsd-5/debian/changelog Log: * Fix a remote denial of service in NFS server (FreeBSD-SA-06:10.nfs / CVE-2006-0900). Modified: trunk/kfreebsd-5/debian/changelog === --- trunk/kfreebsd-5/debian/changelog 2006-03-01 16:05:15 UTC (rev 1281) +++ trunk/kfreebsd-5/debian/changelog 2006-03-01 19:16:44 UTC (rev 1282) @@ -1,3 +1,11 @@ +kfreebsd-5 (5.4-15) unstable; urgency=high + + * Urgency set to high as this fixes a security bug. + * Fix a remote denial of service in NFS server (FreeBSD-SA-06:10.nfs / +CVE-2006-0900). + + -- Aurelien Jarno <[EMAIL PROTECTED]> Wed, 1 Mar 2006 20:16:05 +0100 + kfreebsd-5 (5.4-14) unstable; urgency=high * Urgency set to high as this fixes a security bug. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1283 - trunk/kfreebsd-5/debian/patches
Author: aurel32
Date: 2006-03-01 19:18:46 + (Wed, 01 Mar 2006)
New Revision: 1283
Added:
trunk/kfreebsd-5/debian/patches/000_nfsserver.diff
Log:
... and the patch.
Added: trunk/kfreebsd-5/debian/patches/000_nfsserver.diff
===
--- trunk/kfreebsd-5/debian/patches/000_nfsserver.diff 2006-03-01 19:16:44 UTC
(rev 1282)
+++ trunk/kfreebsd-5/debian/patches/000_nfsserver.diff 2006-03-01 19:18:46 UTC
(rev 1283)
@@ -0,0 +1,11 @@
+--- sys/nfsserver/nfs_srvsock.c28 Jan 2006 19:24:40 - 1.96
sys/nfsserver/nfs_srvsock.c25 Feb 2006 21:27:10 -
+@@ -592,7 +592,7 @@
+ slp->ns_flag |= SLP_LASTFRAG;
+ else
+ slp->ns_flag &= ~SLP_LASTFRAG;
+- if (slp->ns_reclen > NFS_MAXPACKET) {
++ if (slp->ns_reclen > NFS_MAXPACKET || slp->ns_reclen <= 0) {
+ slp->ns_flag &= ~SLP_GETSTREAM;
+ return (EPERM);
+ }
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1284 - in trunk/kfreebsd-6/debian: . patches
Author: aurel32
Date: 2006-03-01 19:31:19 + (Wed, 01 Mar 2006)
New Revision: 1284
Added:
trunk/kfreebsd-6/debian/patches/000_nfsserver.diff
Modified:
trunk/kfreebsd-6/debian/changelog
Log:
* Fix a remote denial of service in NFS server (FreeBSD-SA-06:10.nfs /
CVE-2006-0900).
Modified: trunk/kfreebsd-6/debian/changelog
===
--- trunk/kfreebsd-6/debian/changelog 2006-03-01 19:18:46 UTC (rev 1283)
+++ trunk/kfreebsd-6/debian/changelog 2006-03-01 19:31:19 UTC (rev 1284)
@@ -40,6 +40,8 @@
- arch/amd64/em64t-p4.config: New. P4 config.
- patches/907_cpu_class.diff: Set hw.machine to "x86_64". This improves
compatibility with old config.sub scripts.
+ [ Aurelien Jarno ]
+ * Fix a remote denial of service in NFS server (FreeBSD-SA-06:10.nfs /
CVE-2006-0900).
-- Robert Millan <[EMAIL PROTECTED]> Tue, 28 Feb 2006 12:42:18 +0100
Added: trunk/kfreebsd-6/debian/patches/000_nfsserver.diff
===
--- trunk/kfreebsd-6/debian/patches/000_nfsserver.diff 2006-03-01 19:18:46 UTC
(rev 1283)
+++ trunk/kfreebsd-6/debian/patches/000_nfsserver.diff 2006-03-01 19:31:19 UTC
(rev 1284)
@@ -0,0 +1,11 @@
+--- sys/nfsserver/nfs_srvsock.c28 Jan 2006 19:24:40 - 1.96
sys/nfsserver/nfs_srvsock.c25 Feb 2006 21:27:10 -
+@@ -592,7 +592,7 @@
+ slp->ns_flag |= SLP_LASTFRAG;
+ else
+ slp->ns_flag &= ~SLP_LASTFRAG;
+- if (slp->ns_reclen > NFS_MAXPACKET) {
++ if (slp->ns_reclen > NFS_MAXPACKET || slp->ns_reclen <= 0) {
+ slp->ns_flag &= ~SLP_GETSTREAM;
+ return (EPERM);
+ }
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1286 - trunk/glibc-2.3-head/sysdeps/kfreebsd
Author: aurel32
Date: 2006-03-02 09:24:02 + (Thu, 02 Mar 2006)
New Revision: 1286
Modified:
trunk/glibc-2.3-head/sysdeps/kfreebsd/configure
trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in
Log:
Put the libraries in /lib64 and (/usr)/lib64 on x86_64, this is where the
upstream
put 64-bit version of libraries for architectures that can do biarch.
Consequences:
- the amd64 glibc package on i386 will go directly in /lib64 and (/usr)/lib64
- the path has to be fixed for the native amd64 port. This is already the case
in the
Debian package (glibc 2.3.6-3).
Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/configure
===
--- trunk/glibc-2.3-head/sysdeps/kfreebsd/configure 2006-03-02 08:31:54 UTC
(rev 1285)
+++ trunk/glibc-2.3-head/sysdeps/kfreebsd/configure 2006-03-02 09:24:02 UTC
(rev 1286)
@@ -201,7 +201,7 @@
/usr | /usr/)
# 64-bit libraries on bi-arch platforms go in /lib64 instead of /lib
case $machine in
- sparc/sparc64 | powerpc/powerpc64 | s390/s390-64 | \
+ sparc/sparc64 | x86_64 | powerpc/powerpc64 | s390/s390-64 | \
mips/mips64/n64/* )
libc_cv_slibdir="/lib64"
if test "$libdir" = '${exec_prefix}/lib'; then
Modified: trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in
===
--- trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in 2006-03-02 08:31:54 UTC
(rev 1285)
+++ trunk/glibc-2.3-head/sysdeps/kfreebsd/configure.in 2006-03-02 09:24:02 UTC
(rev 1286)
@@ -133,7 +133,7 @@
/usr | /usr/)
# 64-bit libraries on bi-arch platforms go in /lib64 instead of /lib
case $machine in
- sparc/sparc64 | powerpc/powerpc64 | s390/s390-64 | \
+ sparc/sparc64 | x86_64 | powerpc/powerpc64 | s390/s390-64 | \
mips/mips64/n64/* )
libc_cv_slibdir="/lib64"
if test "$libdir" = '${exec_prefix}/lib'; then
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
r1304 - trunk/glibc-2.3-head
Author: aurel32 Date: 2006-03-03 15:31:53 + (Fri, 03 Mar 2006) New Revision: 1304 Modified: trunk/glibc-2.3-head/AMD64 Log: Bug fixed in debian SVN Modified: trunk/glibc-2.3-head/AMD64 === --- trunk/glibc-2.3-head/AMD64 2006-03-03 14:19:24 UTC (rev 1303) +++ trunk/glibc-2.3-head/AMD64 2006-03-03 15:31:53 UTC (rev 1304) @@ -7,6 +7,7 @@ [stdlib/tst-setcontext.out] Error 255 actually bug in testsuite, not in our code, reported upstream, fixed in HEAD, see http://sourceware.org/bugzilla/show_bug.cgi?id=2414 +fixed in debian SVN fails also on kfreebsd-i386: -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
New list for commits (was backporting changes from 6.x to 5.x)
On Fri, Mar 03, 2006 at 04:17:44PM +0100, Robert Millan wrote: > > BTW, what to do with mailing lists usage > > (http://lists.debian.org/debian-bsd/2006/02/msg00113.html) > > Swap them, leave it as is ? > > Aurelien said (via IRC) he'd move the logs to a separate list this weekend. > Done, the new list for commits has been created: [EMAIL PROTECTED] If you want to subscribe, just go to: http://lists.alioth.debian.org/mailman/listinfo/glibc-bsd-commits Then I don't really what to do with the mailing lists. I think we should keep in mind: 1) mails to [EMAIL PROTECTED] are delivered to all ports mailing list, including [EMAIL PROTECTED] 2) the address used for the maintainer field in the packages is [EMAIL PROTECTED] It could be changed, but it is work, and would take time. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: New list for commits (was backporting changes from 6.x to 5.x)
Petr Salinger a écrit : Now, when logs from SVN are elsewhere, I second Aurelien's "switch everything to debian-bsd" (http://lists.debian.org/debian-bsd/2006/02/msg00114.html). Then we supress glibc-bsd-devel ? Yes, hibernate glibc-bsd-devel. Or maybe we want to keep it for development issues? (and let debian-bsd be for user questions and general stuff) For now discuss on debian-bsd both user and development issues. I guess it will not bother users to know what's going on with their port. Agreed. This is already true for other ports, maybe not for amd64, but for other lists I am subscribed like -sparc, -mips or -hppa where there is as much development questions as users questions. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian GNU/Linux developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[kfreebsd-amd64] Problem building glib2.0
Hi all, While trying to get xorg built without hack on kfreebsd-amd64, I face a problem with glib2.0: x86_64-kfreebsd-gnu-gcc -shared .libs/garray.o .libs/gasyncqueue.o .libs/gatomic.o .libs/gbacktrace .o .libs/gcache.o .libs/gcompletion.o .libs/gconvert.o .libs/gdataset.o .libs/gdate.o .libs/gdir.o . libs/gerror.o .libs/gfileutils.o .libs/ghash.o .libs/ghook.o .libs/giochannel.o .libs/gkeyfile.o .li bs/glist.o .libs/gmain.o .libs/gmappedfile.o .libs/gmarkup.o .libs/gmem.o .libs/gmessages.o .libs/gn ode.o .libs/goption.o .libs/gpattern.o .libs/gprimes.o .libs/gqsort.o .libs/gqueue.o .libs/grel.o .l ibs/grand.o .libs/gscanner.o .libs/gshell.o .libs/gslist.o .libs/gstdio.o .libs/gstrfuncs.o .libs/gs tring.o .libs/gthread.o .libs/gthreadpool.o .libs/gtimer.o .libs/gtree.o .libs/guniprop.o .libs/gutf 8.o .libs/gunibreak.o .libs/gunicollate.o .libs/gunidecomp.o .libs/gutils.o .libs/gprintf.o .libs/gi ounix.o .libs/gspawn.o -Wl,--whole-archive libcharset/.libs/libcharset.a -Wl,--no-whole-archive -W l,-z -Wl,defs -Wl,-O1 -Wl,-soname -Wl,libglib-2.0.so.0 -Wl,-retain-symbols-file -Wl,.libs/libglib-2. 0.exp -o .libs/libglib-2.0.so.0.800.6 /usr/bin/ld: .libs/gconvert.o: relocation R_X86_64_PC32 against `_g_charset_get_aliases' can not be used when making a shared object; recompile with -fPIC /usr/bin/ld: final link failed: Bad value collect2: ld returned 1 exit status All the object files are built with -fPIC, so I think we have a problem with the toolchain. Any hints? Thanks, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [kfreebsd-amd64] Problem building glib2.0
Petr Salinger a écrit :
Hi.
/usr/bin/ld: .libs/gconvert.o: relocation R_X86_64_PC32 against
`_g_charset_get_aliases' can not be
used when making a shared object; recompile with -fPIC
All the object files are built with -fPIC, so I think we have a problem
with the toolchain. Any hints?
It might be due to non-working visibility.
It is not sometimes fully supported in cross compiling.
Probably simply (native) rebuild of binutils and/or gcc-4.0 should suffice.
I will see with Robert Millan if they have been rebuilt at least once
natively.
Prototype is "_g_charset_get_aliases () G_GNUC_INTERNAL;"
Expansion is either
#define G_GNUC_INTERNAL __attribute__((visibility("hidden")))
or
#define G_GNUC_INTERNAL
Could you, please, check configure log for checking for GNUC visibility
attribute... yes
There is a "yes" in the config log, so this seems to be supported.
Thanks,
Aurelien
--
.''`. Aurelien Jarno | GPG: 1024D/F1BCDB73
: :' : Debian developer | Electrical Engineer
`. `' [EMAIL PROTECTED] | [EMAIL PROTECTED]
`-people.debian.org/~aurel32 | www.aurel32.net
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [kfreebsd-amd64] Problem building glib2.0
Petr Salinger a écrit : It might be due to non-working visibility. It is not sometimes fully supported in cross compiling. Probably simply (native) rebuild of binutils and/or gcc-4.0 should suffice. I will see with Robert Millan if they have been rebuilt at least once natively. According to timestamps in http://ftp.gnuab.org/debian/pool-kfreebsd-amd64/main/b/binutils http://ftp.gnuab.org/debian/pool-kfreebsd-amd64/main/g/gcc-4.0/ gcc-4.0 is built against older (may be non-native) binutils. Ok, then I will rebuild it. Could you, please, check configure log for "checking for GNUC visibility attribute... yes" There is a "yes" in the config log, so this seems to be supported. Can you try to force it to "no", and consequently #define G_GNUC_INTERNAL as empty ? Will it work ? That works thanks for you help. I will rebuild gcc-4.0 and then give another try without modifying glibconfig.h. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [kfreebsd-amd64] Problem building glib2.0
Aurelien Jarno a écrit : Petr Salinger a écrit : It might be due to non-working visibility. It is not sometimes fully supported in cross compiling. Probably simply (native) rebuild of binutils and/or gcc-4.0 should suffice. I will see with Robert Millan if they have been rebuilt at least once natively. According to timestamps in http://ftp.gnuab.org/debian/pool-kfreebsd-amd64/main/b/binutils http://ftp.gnuab.org/debian/pool-kfreebsd-amd64/main/g/gcc-4.0/ gcc-4.0 is built against older (may be non-native) binutils. Ok, then I will rebuild it. Could you, please, check configure log for "checking for GNUC visibility attribute... yes" There is a "yes" in the config log, so this seems to be supported. Can you try to force it to "no", and consequently #define G_GNUC_INTERNAL as empty ? Will it work ? That works thanks for you help. I will rebuild gcc-4.0 and then give another try without modifying glibconfig.h. I still have the problem with a gcc-4.0 rebuilt from source. I have uploaded it to the archive, as it is a newer version. Moreover the problem does not seem to come from binutils, as using the .o files from GNU/Linux fixes the problem. On the contrary I see the same problem using the .o files from GNU/kFreeBSD on GNU/Linux. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [kfreebsd-amd64] Problem building glib2.0
On Tue, Mar 07, 2006 at 09:19:06AM +0100, Petr Salinger wrote: > >I still have the problem with a gcc-4.0 rebuilt from source. I have > >uploaded it to the archive, as it is a newer version. > > > >Moreover the problem does not seem to come from binutils, as using the .o > >files from GNU/Linux fixes the problem. On the contrary I see the same > >problem using the .o files from GNU/kFreeBSD on GNU/Linux. > > It still can be bug in binutils - in gas. > > Please could you compile attached lib.c by commads bellow > on both linux-amd64 and kfreebsd-amd64 > and compare .s and .o, especially "readelf -s" ? > > gcc -fPIC -O2 -Wall -S lib.c -o lib1.s -DFIRST > gcc -fPIC -O2 -Wall -S lib.c -o lib2.s > gcc -fPIC -O2 -Wall -c lib1.s > gcc -fPIC -O2 -Wall -c lib2.s > gcc -shared -o liblib.so lib1.o lib2.o I have done you test, and I compared the resulting files. The .s files are identical, but the .o and .so files are different. readelf -s show the same sections (and a few difference than can be explained), but the assembly code is a bit different. Sorry I can tell you more, however I have attached the files to the mail. > BTW, glibc 2.4 in Debian might be closer than we think > http://sourceware.org/ml/libc-alpha/2006-03/msg00050.html Well, we have started a glibc 2.4 branch that will be targeted for experimental, however, we still don't know if it will be enough stable on all architectures to be considered for etch (the freeze of the toolchain will occur in July). Thanks for you help, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net visibility_amd64.tar.gz Description: Binary data
Re: [kfreebsd-amd64] Problem building glib2.0 (solved)
Aurelien Jarno a écrit : On Tue, Mar 07, 2006 at 09:19:06AM +0100, Petr Salinger wrote: I still have the problem with a gcc-4.0 rebuilt from source. I have uploaded it to the archive, as it is a newer version. Moreover the problem does not seem to come from binutils, as using the .o files from GNU/Linux fixes the problem. On the contrary I see the same problem using the .o files from GNU/kFreeBSD on GNU/Linux. It still can be bug in binutils - in gas. I have actually found the problem, it was a problem with our version of libtool. I have sent a patch under bug#355789 in the Debian BTS. Thanks Petr for the time you spent on this problem. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [Glibc-bsd-commits] r1335 - trunk/glibc-2.3-head
Robert Millan a écrit : +--- glibc-2.3.6/debian/sysdeps/kfreebsd-amd64.mk~ 2006-03-15 21:53:59.0 +0100 glibc-2.3.6/debian/sysdeps/kfreebsd-amd64.mk 2006-03-16 07:47:12.0 +0100 +@@ -3,7 +3,7 @@ + libc_rtlddir = /lib64 + + # /lib64 and /usr/lib64 are provided by glibc instead base-files: #259302. +-define libc0.1_extra_pkg_install ++define libc6_extra_pkg_install + ln -sf /lib debian/$(curpass)/lib64 + ln -sf lib debian/$(curpass)/usr/lib64 + endef Have you tested that? If yes, I will commit it in the glibc. Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: r1335, ld.so path
On Thu, Mar 16, 2006 at 11:25:25AM +0100, Petr Salinger wrote: > > > +--- glibc-2.3.6/debian/sysdeps/kfreebsd-amd64.mk~2006-03-15 > > > 21:53:59.0 +0100 > > > glibc-2.3.6/debian/sysdeps/kfreebsd-amd64.mk 2006-03-16 > > > 07:47:12.0 +0100 > > > +@@ -3,7 +3,7 @@ > > > + libc_rtlddir = /lib64 > > > + > > > + # /lib64 and /usr/lib64 are provided by glibc instead base-files: > > > #259302. > > > +-define libc0.1_extra_pkg_install > > > ++define libc6_extra_pkg_install > > > + ln -sf /lib debian/$(curpass)/lib64 > > > + ln -sf lib debian/$(curpass)/usr/lib64 > > > + endef > > Have you tested that? If yes, I will commit it in the glibc. > > We should also finish discussion about ld.so path (libc_rtlddir) > > http://lists.alioth.debian.org/pipermail/glibc-bsd-devel/2006-January/001008.html > http://lists.alioth.debian.org/pipermail/glibc-bsd-devel/2006-February/001053.html > http://lists.alioth.debian.org/pipermail/glibc-bsd-devel/2006-February/001062.html > > Now it looks like almost all uses > $ ldd /usr/bin/man > libgdbm.so.3 => /usr/lib/libgdbm.so.3 (0x000800631000) > libc.so.0.1 => /lib/libc.so.0.1 (0x000800736000) > /lib/ld-kfreebsd-x86-64.so.1 (0x000800518000) > > exception is glibc (r1284) > $ ldd /usr/bin/iconv > libc.so.0.1 => /lib/libc.so.0.1 (0x000800625000) > /lib64/ld-kfreebsd-x86-64.so.1 (0x00080050c000) > Now that all our files but the glibc have the linker in /lib, I have changed rtlddir to /lib on kfreebsd-amd64. A glibc will be uploaded to unstable today or tomorrow, at it also includes the fix for the /lib64 link. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: mass-filing?
Robert Millan a écrit : Hey, I had a look at [1] and noticed a few candidates for mass-filing: - libasound2-dev - libcap-dev - libdevmapper-dev - libgpmg1-dev Well I am not sure you can do mass filling on such candidates. You have first to investigate why they need libasound2-dev or other packages, and (maybe) write a patch. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: mass-filing?
Robert Millan a écrit : On Wed, Mar 22, 2006 at 12:22:15PM +0100, Aurelien Jarno wrote: Robert Millan a ?crit : Hey, I had a look at [1] and noticed a few candidates for mass-filing: - libasound2-dev - libcap-dev - libdevmapper-dev - libgpmg1-dev Well I am not sure you can do mass filling on such candidates. You have first to investigate why they need libasound2-dev or other packages, and (maybe) write a patch. How about a generic explanation on how to fix the bug, including dpkg-architecture calls, etc? Well even with that you will fill bugs on alsa specific packages... -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: request for upload of net-snmp into unreleased
Robert Millan wrote: On Wed, May 03, 2006 at 08:28:59PM +0200, Petr Salinger wrote: Hi. I just wrote dirty hack for net-snmp, committed in SVN. Nice work. This was a though one. Please could someone built it and upload for both kfreebsd-i386 and kfreebsd-amd64. Uploaded to i386 (I don't have amd64 at hand atm). I will do the upload of amd64 after midnight, so that both uploads do not conflict. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: upload gcj-4.1 into unreleased ?
Petr Salinger wrote: Hi, there is posibility to add also java for kfreebsd. Patches are committed in SVN. Only binary-arch packages can be built directly now, because gjdoc (needed for arch all packages) build-depends on gij. Moreover on gij-4.0, but it should change in near future, see http://lists.debian.org/debian-devel-announce/2006/05/msg8.html In testsuite really fails only FAIL: Throw_2 execution - source compiled test FAIL: Throw_2 execution - gij test FAIL: Throw_2 execution - bytecode->native test FAIL: Throw_2 -O3 execution - source compiled test FAIL: Throw_2 execution - gij test FAIL: Throw_2 -O3 execution - bytecode->native test Andrew Haley on [EMAIL PROTECTED] wrote: The Throw_2 failures probably indicate that signal unwinding doesn't work. See gcc/config/i386/linux-unwind.h for a Linux example of how to do it. It's a real failure, but it's not likely to present a problem for most applications. So it looks like it might be usefull as is. At least for later bootstraping of gjdoc ;-) Could someone (with upload rights) built it and upload for kfreebsd-i386. In theory, it might also built for kfreebsd-amd64, but I did't try it. Thanks for your work! I will give a try, but the network really sucks here at debconf6, so I am not sure I will be able to do it. I will tell you more in a few hours. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian GNU/Linux developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: kfreebsd fixes & Maintainer: Debian QA Group
Petr Salinger wrote: Hi, following packages with kfreebsd fixes (either in BTS or in unreleased) have as Maintainer Debian QA Group <[EMAIL PROTECTED]> easy one: eco5000 libtool - #342672 gtk-engines-thinice libtool guppi libtool libjpeg-mmx add kfreebsd-amd64 - #361614 rioutil kbsd [libtool + 2x defined(__GLIBC__)] sourcenav cfg - #342444 libmusicbrainz-2.1 #368970 Could be possible to upload them fixed into official Debian ? Done. kexiis being integrated with koffice For this one, I have filled a request of removal. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: warning for kfreebsd-amd64 users
On Sat, Jul 29, 2006 at 08:49:05AM +0200, Petr Salinger wrote: > Hi, > > I strongly suggest to kfreebsd-amd64 users to not upgrade > any important packages, especially not sysvinit to version 2.86.ds1-15. > > All packages built by binutils_2.17-1 or binutils_2.17-2 are affected. > They simply will not run: > "ELF binary type "0" not known." > "cannot execute binary file" > > Binutils for kfreebsd-amd64 needs patch > http://glibc-bsd.alioth.debian.org/patches/binutils.diff > applied, especially part Thanks for the patch. Could you send it upstream, if not already done? I have uploaded a version of binutils with your patch. I have started to rebuild the last built packages with the bad version of binutils, but I don't know when I will be able to upload them. If there is still some packages wrongly built in 2/3 days, please send me a list, I will trigger binNMU for them. Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: kfreebsd future
Petr Salinger a écrit : Hi all, especially Aurelien and Robert. I would like to slightly discuss future of GNU/kFreeBSD. - FreeBSD 6.x based * kfreebsd-kernel-headers * freebsd-libs * freebsd-utils * ... GNU/kFreeBSD have to switch FreeBSD base to newer version some day. From my POV, it would be usefull to do it now. kfreebsd-amd64 have only 6.x kernels, half of popcon registered users already uses 6.x kernels on kfreebsd-i386. kfreebsd-image-5.4-1-486 1 kfreebsd-image-5.4-1-586 1 kfreebsd-image-5.4-1-686 3 kfreebsd-image-6.0-1-686 1 kfreebsd-image-6.1-1-586 1 kfreebsd-image-6.1-1-686 3 Does anyone have problems with kfreebsd-image-6.1 ? My only problem is that I would like to see kfreebsd-6.1 in the Debian archive before asking for removal of version 5.4. To get kfreebsd-6.1 in the archive, we need freebsd6-utils first, and thus libbsd. There are still problems with libbsd. Quoting Guillem Jover: 11:58 < braindmg> aurel32: about libbsd I'll talk to you later 11:58 < braindmg> (oh, ok I've 10 mins) 11:59 < braindmg> aurel32: I'm mostly concerned about the library name, there was previously such a lib generated from glibc 11:59 < braindmg> and some packages may end up linking against it accidentally (like inetutils) 12:00 < braindmg> then there's the license, I found there's 4 clause licensed files in the lib May I adapt kfreebsd-kernel-headers to use kfreebsd-source-6.1 ? If it is still possible to use 5.4 kernels then, that looks ok to me. - outdated install CD == http://glibc-bsd.alioth.debian.org/install-cd/ Would be possible to regenerated these images ? Ideally, with kfreebsd-image-6.1 as kernel and with recent glibc. I think Robert Millan is the person to contact for that. I have no idea how the images are generated. - kfreebsd-i386 & etch === Release or not to release ? Well I think we are too far for something releasable, moreover kfreebsd-i386 is still not in the archive. - kfreebsd-i386 on ftp.debian.org as official arch === Are there any objection from ftpmasters ? AFAIK, no there aren't any objection. Last time we (with Guillem Jover) spoke about that with them, it was ok in principle. It was roughly a year ago. Or do they just ignore ArchiveQualification pages and #369797 ? It seems they are ignoring that, also for other architectures such as m32r or armeb. - status mail to debian-devel-announce === Some publicity might help to quicker applying of patches, or to integrate them during NMU, or may be even with ArchiveQualification. This is something to do, what to say? I am not sure we have a lot to add since the last mail to debian-devel-announce. For some time I hopped it would be possible to announce a kfreebsd-amd64 developement machine, but we have no news of the offer I received through Martin Michlmayr for an amd64 machine. No response to ping either :( However if you have some ideas, we can prepare a mail. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: kfreebsd future
Petr Salinger a écrit : - kfreebsd-i386 & etch === Release or not to release ? Well I think we are too far for something releasable IMHO, it would be good to produce some snapshot for evaluation or first try. All packages from such snapshot should be installable. Unstable is sometimes unstable ;-) In "etch based" snapshot there would be kfreebsd specific packages (kfreebsd-common, kfreebsd-6, kfreebsd-kernel-headers, ...) and subset of packages from etch, with either same source as in etch (majority of packages) or with etch based source (currently i.e. gnupg, ifupdown, pam, expat, ...). moreover kfreebsd-i386 is still not in the archive. Yes, this makes creation of such snapshot much harder :-( Maybe it is possible to implement that into mini-dak, the software running on gnuab.org. Guillem, what do you think? - status mail to debian-devel-announce === However if you have some ideas, we can prepare a mail. Well, "Bits from the Debian GNU/kFreeBSD porters" might contain something like bellow. Petr Hello, Debian world! This is a status update for the Debian GNU/FreeBSD port[1]. Compared to previous status mail [2], this port now consists of two architectures: kfreebsd-i386 and kfreebsd-amd64. Currently we focus mainly on kfreebsd-i386. Status -- * We have uptodate toolchain and most of the core Debian packages are in good shape. Unfortunately this is not true anymore :( You have been faster than me to report a bug and send a patch to bug #385974. Thanks! For libxul-dev, http://glibc-bsd.alioth.debian.org/patches/nsprpub.diff is a good stard. * An unofficial buildd is up and running for both kfreebsd-i386 and kfreebsd-amd64, buildd logs for kfreebsd-i386 are available at [3]. Unfortunately, after a dozen of mails/pings on irc to Andreas Barth, I failed to get the kfreebsd-amd64 logs on experimental.ftbfs.de. * There is a DD-accessible porting machine running kfreebsd-i386. http://io.debian.net/ * For kfreebsd-i386, 80% packages have been built, 75% are up-to-date. See [4], [5]. * We want to be in official archive, see #369797, http://wiki.debian.org/ArchiveQualification/kfreebsd-i386. We need your help - * as ftpmaster, could you please take a look at #369797. * as package maintainer, could you please look at BTS whether there is a kfreebsd submission and integrate it. In many cases, it is sufficient to (auto)update config.sub/config.guess or use recent libtool. See [6]. * as NMUer, please could you look at kfreebsd submission in BTS (in addition to usual translations). Of cource, you can also verify whether your package builds [3] ^^ course and/or use porting machine. We would also appreciate hosting a shell server and/or donate hardware for kfreebsd-amd64. Currently we are thinking about releasing etch based snapshot as we have reasonable subset of etch packages [5]. The Debian GNU/kFreeBSD porters Looks nice to me. What the other think? Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: kfreebsd future
Petr Salinger a écrit : - status mail to debian-devel-announce === However if you have some ideas, we can prepare a mail. Well, "Bits from the Debian GNU/kFreeBSD porters" might contain something like bellow. Petr Hello, Debian world! This is a status update for the Debian GNU/FreeBSD port[1]. Compared to previous status mail [2], this port now consists of two architectures: kfreebsd-i386 and kfreebsd-amd64. Currently we focus mainly on kfreebsd-i386. Status -- * We have uptodate toolchain and most of the core Debian packages are in good shape. Unfortunately this is not true anymore :( You have been faster than me to report a bug and send a patch to bug #385974. Thanks! It should be fixed in just uploaded gcc-4.1/gcj-4.1 4.1.1ds1-14. For libxul-dev, http://glibc-bsd.alioth.debian.org/patches/nsprpub.diff is a good stard. xulrunner.diff is now available on alioth. Thanks a lot for your work! BTW, how is http://glibc-bsd.alioth.debian.org/ synced vith web subdir in subversion ? It is rather outdated now. Could be possible to upload xulrunner, kdebase and xorg-server into unreleased ? Done for xulrunner In progress for xorg-server For kdebase, I have just discovered that g++ is totally broken on GNU/kFreeBSD, so I am unable to build it until I find the fix. When trying to build a C++ program, I get: /usr/lib/gcc/i486-kfreebsd-gnu/4.1.2/libstdc++.so: undefined reference to [EMAIL PROTECTED]' Looks nice to me. What the other think? No body object against ;-) So, please, send it. I will send it soon, I am waiting to have some news on the new buildd. The machine called hagrid is donated and hosted by the ETH Zürich, and it has built a few packages during the night from Saturday to Sunday. If it works again, I will add that to the mail. It is a good point to have our build daemons at different location. Also because I sometimes need to poweroff the one that I have at home (they are under my bed). Bye, Aurelien -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [EMAIL PROTECTED] | [EMAIL PROTECTED] `-people.debian.org/~aurel32 | www.aurel32.net -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
