Bug#932169: dante-server: systemd fails to start danted service
Hi, I can confirm this behavior on armhf. Removing "lib64" from the danted.service file fixes it. Lukas
Bug#752789: initramfs-tools: mkinitramfs doesn't honor /usr/share/initramfs-tools/modules
On Mon, Sep 29, 2014 at 7:47 PM, Ben Hutchings b...@decadent.org.uk wrote: I think the best way to deal with this would be to add a comment clarifying which file path is actually read. Yes, that would be really helpful! -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#752789: initramfs-tools: mkinitramfs doesn't honor /usr/share/initramfs-tools/modules
On Mon, Sep 29, 2014 at 7:47 PM, Ben Hutchings b...@decadent.org.uk wrote: /usr/share/initramfs-tools/modules is the 'shipped' version of /etc/initramfs-tools/modules, and is copied to the latter file if it does not already exist. The comment is of course correct in the copy. And user-edittable configuration files are always installed in /etc, not /usr. Normally we would include /etc/initramfs-tools/modules in the package as a conffile, and then dpkg would take care of preserving any customised version. However, the installer may in some cases add modules to this file, which could result in dpkg later claiming that it's been edited by the user. I think the best way to deal with this would be to add a comment clarifying which file path is actually read. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#756495: ddpt breaks on input from pipe
Package: ddpt Version: 0.92-1 Severity: normal Hi, ddpt 0.92 breaks on input from pipe: $ tar -Sxf foo.tgz --to-command ddpt if=- of=/dev/sdb obs=512 oflag=sparse,pt 103+0 records in 103+0 records out 0 bypassed records out time to transfer data: 0.024940 secs at 2.11 MB/sec 116kB 0:00:00 [2.72MB/s] tar: foo: Cannot seek to 1049600: Broken pipe This has been resolved in ddpt 0.94 so an upgrade to that version would be very much appreciated. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#669017: Fwd: bootlogd messes up Plymouth's text theme during boot
Hi, I also see that plymouth and bootlogd tend to break each other. This is probably because plymouth (and probably bootlogd, too) try to redirect the console. It would be nice to declare a break on plymouth in the bootlogd package (and vice versa). Since plymouth also logs the output to /var/log/boot.log these two packages don't really make sense to be installed in parallel. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#753000: hello: Use dh-autoreconf
Package: src:hello Severity: normal Tags: patch Version: 2.9-1 Hi, hello uses autotools but the packaging doesn't use dh-autoreconf. I've attached a patch that adds dh-autoreconf and then needed additional build dependencies. The patch that modifies config.guess and config.sub is probably also obsolete if dh-autoreconf is used. Regards, Lukas --- a/hello-2.9/debian/control +++ b/hello-2.9/debian/control @@ -3,7 +3,7 @@ Section: devel Priority: optional Maintainer: Santiago Vila sanv...@debian.org Standards-Version: 3.9.5 -Build-Depends: debhelper (= 9.20120311) +Build-Depends: debhelper (= 9.20120311), dh-autoreconf, help2man, texinfo Homepage: http://www.gnu.org/software/hello/ Package: hello --- a/debian/rules +++ b/debian/rules @@ -1,6 +1,6 @@ #!/usr/bin/make -f %: - dh $@ + dh $@ --with autoreconf override_dh_auto_clean: [ ! -f Makefile ] || $(MAKE) distclean
Bug#752746: plymouth delays splash screen for 5 s
Package: plymouth Version: 0.9.0-3 Severity: normal Hi, plymouth 0.9.0 gained the feature to delay the splash screen for 5 s. If a boot is faster than 5 s no boot splash is shown at all. I'm not really convinced that the feature is a good idea because it seems like regression when it's actually a feature. I propose to do one of the following things: 1. Set ShowDelay=0 in /usr/share/plymouth/plymouth.defaults (and /etc/plymouth/plymouthd.conf for that matter) to disable that feature. 2. Explicitly set ShowDelay=5 in the two files as well (so it's obvious that a new configuration option exists) and document that in a NEWS entry in the plymouth package. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#752789: initramfs-tools: mkinitramfs doesn't honor /usr/share/initramfs-tools/modules
Package: initramfs-tools Version: 0.115 Severity: normal Hi, mkinitramfs (the tool that is called from update-initramfs) doesn't honor /usr/share/initramfs-tools/modules, it only honors /etc/initramfs-tools/modules and /usr/share/initramfs-tools/modules.d. This is unfortunate because /usr/share/initramfs-tools/modules explicitly states that the modules listed in that file are included in the initramfs. The file /usr/share/initramfs-tools/modules should therefore be either added to the list of files that are processed or removed altogether. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#751488: initramfs-tools: Shell spawned despite panic=0
Control: found -1 0.115 This bug is still present in the latest available version. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#752802: src:plymouth: unrecognized configure options
Package: src:plymouth Version: 0.9.0-3 Severity: minor Hi, when messing with the source package of plymouth I noticed that the following two options have been removed: configure: WARNING: unrecognized options: --disable-libkms, --disable-tests You might want to remove them from debian/rules as well. --disable-libkms seems to have been refactored into --enable-drm-renderer. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#751275: Bug#752575: plymouth: excludes drm.ko from initramfs
Hi Simon, On Wed, Jun 25, 2014 at 11:56 AM, Simon McVittie s...@debian.org wrote: To avoid bugs similar to this happening again, I think it's worth pointing out that you might have misunderstood copy_modules_dir's syntax: the description of those changes in debian/changelog is reversed. The second and subsequent arguments to copy_modules_dir are directories or modules to *exclude*, so Lukas' patch *removed* drm from the initramfs (despite being documented as adding it), and my reversion of that patch added it back (which was correct in the commit message I sent, but is described as removing it in the debian/changelog that you uploaded). Sorry if I wasn't clear about this before. Additionally, the arguments to copy_modules_dir are passed to find(1) in the form find ... -name drm -prune -o ..., so copy_modules_dir kernel/drivers/gpu/drm drm means copy everything from kernel/drivers/gpu/drm, but exclude directories named drm, which is why it resulted in nothing being copied into the initramfs at all. If you want to exclude individual modules, without excluding entire directories, the way to do that seems to be copy_modules_dir ... drm.ko (but please don't do that, because i915.ko, radeon.ko etc. depend on drm.ko, so that would reintroduce #752575). thanks for clarifying that. I also noticed that and commented in #751275 but it was too late back then. I suspect the best solution for uvesafb/fglrx users like Lukas might be to blacklist drm.ko so it is never loaded. Perhaps fglrx should ship /etc/modprobe.d/fglrx-blacklist.conf containing blacklist drm, blacklist radeon? I'm not sure whether that would break dual-graphics-card machines, though: I've only heard of machines that combine nVidia discrete graphics and Intel integrated graphics, but similar AMD/Intel combinations might also exist. It seems that i915 has to be blacklisted as well because i915 loads drm and blacklist drm doesn't prevent that. Shipping that in a default configuration would break too many things, I suppose. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#751275: drm.ko in initramfs vs. fglrx
On Wed, Jun 25, 2014 at 12:43 PM, Simon McVittie s...@debian.org wrote: It looks as though Debian's kernels have AGP built-in rather than modular since 2.6.32-10. You mentioned that you're using a non-Debian kernel on an embedded device; if you're going to maintain your own kernel anyway (e.g. if the 3.14 kernel in backports is not suitable for you), perhaps you could do the same? If you can't reproduce this bug with a Debian kernel (or a more Debian-like configuration) perhaps #751275 can be closed as not affecting Debian? I'm using a custom built 3.10.44 that is based on 3.10.11-1 from snapshots.debian.org. All patches have been rediffed and the configuration ought to be the same. Failing that, a local blacklist seems like a reasonable solution for your particular situation. I'm using a blacklist for that particular hardware now which seems to work. However, I'm still struggling to come up with a unified plymouth setup (Wheezy) that works on a few different hardware platforms (i915, radeon/fglrx). Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#751275: plymouth: Kernel error message regarding DRM shown during boot
reopen quit Hi, I've just noticed that my patch has quite drastic consequences: If it's applied no drm modules are copied into the initramfs at all. This breaks drm plymouth when using i915, for example. I'll look into it to find a better solution. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#751488: initramfs-tools: Shell spawned despite panic=0
Package: initramfs-tools
Version: 0.109.1
Severity: critical
Tags: patch
Hi,
I've set panic=0 as a kernel cmdline argument which should trigger a
reboot instead of spawning a shell. However, the reboot seems to be
uneffective and a shell is spawned nevertheless. This is unpleasing
since spawn=0 is marketed as a security feature in
initramfs-tools(8):
panic sets an timeout on panic. panic=sec is a documented
security feature: it disables the debug shell.
Output on screen:
Loading, please wait ...
Spawning shell within the initramfs
Rebooting automatically due to panic= boot argument
BusyBox v1.20.2 (Debian 1:1.20.0-7) built-in shell (ash)
Enter 'help' for a list of built-in commands.
/bin/sh: can't access tty; job control turned off
(initramfs) _
The commands halt, reboot, etc. don't work either.
To fix the security impact of an open shell I propose to at least add a
return after the reboot command so that if the reboot is effectively a
NOP still no shell is spawned.
diff --git a/scripts/functions b/scripts/functions
index 5352f1d..de64494 100644
--- a/scripts/functions
+++ b/scripts/functions
@@ -43,6 +43,7 @@ panic()
echo Rebooting automatically due to panic= boot argument
sleep ${panic}
reboot
+ return
fi
modprobe -v i8042 || true
modprobe -v atkbd || true
Regards,
Lukas
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#751275: plymouth: Kernel error message regarding DRM shown during boot
Package: src:plymouth
Version: 0.8.5.1-5
Severity: minor
Tags: patch
Hi,
I'm using Debian Wheezy and plymouth 0.8.5.1-5 with Kernel 3.10.40 on
an embedded system equipped with a Radeon E6760. For better graphics
performance I've installed fglrx and use uvesafb to get plymouth working.
Quite early in the boot process a kernel error message appears on the
screen and destroys the otherwise slick boot process:
[1.497550] [drm:drm_pci_agp_init] *ERROR* Cannot initialize
the agpgart module.
[1.497629] DRM: Fill_in_dev failed.
I found out that the drm kernel module is not copied into the
initramfs which apparently causes the error message.
I've attached a patch that addresses that problem and works quite well
for me.
Regards,
Lukas
From a856e7b1e08471d941fe5ca41f6ce67f1f223e03 Mon Sep 17 00:00:00 2001
From: Lukas Anzinger lu...@lukasanzinger.at
Date: Wed, 11 Jun 2014 18:12:24 +0200
Subject: [PATCH] Copy module drm into initramfs to fix annoying drm
initializing error message.
This fixes the following error message:
[1.497550] [drm:drm_pci_agp_init] *ERROR* Cannot initialize the agpgart module.
[1.497629] DRM: Fill_in_dev failed.
---
debian/local/plymouth.hook | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/debian/local/plymouth.hook b/debian/local/plymouth.hook
index b499a30..c386704 100644
--- a/debian/local/plymouth.hook
+++ b/debian/local/plymouth.hook
@@ -122,7 +122,7 @@ case ${THEME_NAME} in
copy_exec /usr/lib/@DEB_HOST_MULTIARCH@/plymouth/renderers/drm.so
# add drm modules
- copy_modules_dir kernel/drivers/gpu/drm mga r128 savage sis tdfx via
+ copy_modules_dir kernel/drivers/gpu/drm drm mga r128 savage sis tdfx via
;;
esac
--
2.0.0
Bug#751155: mutagen: new upstream version 1.23
Source: mutagen Version: 1.22-1 Severity: wishlist Dear Maintainer, there's a new upstream version 1.23 available that adds support for AIFF. It's of particular interest to me because mutagen 1.23 is a dependency of the yet to be released beets v1.3.7. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#715534: git is built without support for python
Hi, I'm still interested in having this issue fixed, and obviously I'm not the only one. Is there anything I can do to make it happen? Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#747510: fglrx-legacy-source: Module doesn't compile on 3.10
Package: fglrx-legacy-source
Version: 8.97.100.7-4
Severity: normal
Hi,
the module unfortunately doesn't compile on Linux 3.10, although 8.97.100.7-4
ought to have support for that kernel version.
I've attached a patch that fixes the build errors and now I can finally run
fglrx on Wheezy. I've not tested compiling it on Sid/Jessie.c
Regards,
Lukas
From 6b258e7e7a344a8b3ad4089085bce7cf3e9eedc9 Mon Sep 17 00:00:00 2001
From: Lukas Anzinger lu...@lukasanzinger.at
Date: Fri, 9 May 2014 12:19:54 +0200
Subject: [PATCH] Fix support for Linux 3.10.
---
common/lib/modules/fglrx/build_mod/firegl_public.c | 10 --
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/common/lib/modules/fglrx/build_mod/firegl_public.c b/common/lib/modules/fglrx/build_mod/firegl_public.c
index c33de6d..08ee543 100644
--- a/common/lib/modules/fglrx/build_mod/firegl_public.c
+++ b/common/lib/modules/fglrx/build_mod/firegl_public.c
@@ -191,6 +191,10 @@
#include asm/fpu-internal.h
#endif
+#if LINUX_VERSION_CODE = KERNEL_VERSION(3,4,0)
+#include asm/fpu-internal.h
+#endif
+
#include firegl_public.h
#include kcl_osconfig.h
#include kcl_io.h
@@ -529,10 +533,12 @@ static int firegl_debug_proc_write_wrap(void* file, const char *buffer, unsigned
#else
static int firegl_debug_proc_read_wrap(struct seq_file *m, void* data)
{
-return firegl_debug_proc_read(m-buf, m-from, m-index, m-size, m-size - m-count, data);
+char *start = m-buf + m-from;
+int eof = m-size - m-count;
+return firegl_debug_proc_read(m-buf, start, m-index, m-size, eof, data);
}
-static ssize_t firegl_debug_proc_write_wrap(struct file *file, const char *buffer, size_t count, void *data)
+static kcl_ssize_t firegl_debug_proc_write_wrap(struct file *file, const char *buffer, kcl_size_t count, kcl_off_t *data)
#endif
{
return firegl_debug_proc_write(file, buffer, count, data);
--
1.9.1
Bug#743307: varnish: hash sum of tar balls in mirrors differ
Package: varnish Version: 2.1.3-8+deb6u2 Severity: normal Hi, the hash sums of varnish_2.1.3.orig.tar.gz from security.debian.org and the one from ftp.de.debian.org are different. $ wget -q -O - http://security.debian.org/debian-security/pool/updates/main/v/varnish/varnish_2.1.3.orig.tar.gz | sha1sum - 730eea869b6aabace6e3e638c743d722f5f67fd7 - $ wget -q -O - http://ftp.de.debian.org/debian/pool/main/v/varnish/varnish_2.1.3.orig.tar.gz | sha1sum - 574e9d94405a017102a0b53156d12197a669f053 - This is quite annoying for me because I combine all squeeze distributions (squeeze, squeeze-updates, squeeze/updates in case of debian-security) to one using reprepro. Reprepro complains that it can't import varnish 2.1.3-8+deb6u2 because varnish_2.1.3.orig.tar.gz from security.debian.org has a different hash sum than varnish_2.1.3.orig.tar.gz from ftp.de.debian.org. I'd really appreciate an upload of the package using the orig tar ball uploaded to security.debian.org. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#743245: vbrfix: Complete rewrite with nice Qt GUI available in SVN
Package: vbrfix Version: 0.24-7 Severity: wishlist Tags: patch Hi, although vbrfix hasn't seen a release in quite a while and the homepage of the original author is down he left a complete redesign and rewrite including a fancy Qt GUI in his SVN tree. It hasn't been updated since four years now as well but it's still a major improvement (not only code wise but also from a feature point of view) over the currently packaged 0.24. I've updated your packaging and introduced a second GUI binary package. You might want to take a look at it at https://github.com/Lukas0907/vbrfix. I'd be very glad if you considered merging my changes. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#644610: apt: Erroneous warning on signed snapshots
Hi, a quick workaround is to add /. to the distribution field in the sources.list: deb http://localhost/repo development/snapshots/test1/. main apt transforms the distribution field and removes everything that comes before the last /. So the transformed distribution would then be development/snapshots/test1 because /. is simply stripped for comparison. Your HTTP server should handle the appended /. perfectly fine. I've tried it with nginx and apt 0.8.10.3+squeeze1 and it worked for me. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#644610: apt: Erroneous warning on signed snapshots
Tags: patch
Hi,
I've created a patch for apt against 0.9.16.1 that changes they way apt checks
the codename/suite to a more sane but still backwards compatible way.
For details, please take a look at the attached patch.
Please note that reprepro is still broken on its own in one special case: If
one wants to generate a snapshot of codename wheezy with suite stable you end
up with a snapshot with the codename wheezy and the suite
wheezy/snapshots/123. The suite field should actually point to
stable/snapshots/123 because that's what it is. I've also prepared a patch
for reprepro.
Regards,
Lukas
From 5e79d6113c3a262cd91bc78d53371db90ba9042a Mon Sep 17 00:00:00 2001
From: Lukas Anzinger lu...@lukasanzinger.at
Date: Fri, 21 Mar 2014 12:18:11 +0100
Subject: [PATCH] When comparing Suite and Codename from Release compare untransformed dist first.
A suite or codename entry in the Release file is checked against the
distribution field in the sources.list entry that lead to the download of that
Release file. This distribution entry can contain slashes in the distribution
field:
deb http://security.debian.org/debian wheezy/updates main
However, the Release file may only contain wheezy in the Codename field and
not wheezy/updates. So a transformation needs to take place that removes the
last / and everything that comes after (e.g. /updates). This fails, however,
for valid cases like a reprepro snapshot where the given Codename contains
slashes but is perfectly fine and doesn't need to be transformed. Since that
transformation is essentially just a workaround for special cases like the
security repository, it should be checked if the literal Codename without any
transformations happened is valid and only if isn't the dist should be checked
against the transformated one.
This way special cases like security.debian.org are handled and reprepro
snapshots work too.
---
apt-pkg/acquire-item.cc | 8 ++--
apt-pkg/indexrecords.cc | 6 ++
apt-pkg/indexrecords.h | 1 +
3 files changed, 13 insertions(+), 2 deletions(-)
diff --git a/apt-pkg/acquire-item.cc b/apt-pkg/acquire-item.cc
index 30743ad..bd524b4 100644
--- a/apt-pkg/acquire-item.cc
+++ b/apt-pkg/acquire-item.cc
@@ -1684,7 +1684,10 @@ bool pkgAcqMetaIndex::VerifyVendor(string Message) /*{{{*/
std::cerr Transformed Dist: Transformed std::endl;
}
- if (MetaIndexParser-CheckDist(Transformed) == false)
+ // Only compare with transformed dist if check with untransformed expected
+ // dist fails.
+ if (MetaIndexParser-CheckExpectedDist() == false
+ MetaIndexParser-CheckDist(Transformed) == false)
{
// This might become fatal one day
// Status = StatAuthError;
@@ -1694,8 +1697,9 @@ bool pkgAcqMetaIndex::VerifyVendor(string Message) /*{{{*/
// return false;
if (!Transformed.empty())
{
- _error-Warning(_(Conflicting distribution: %s (expected %s but got %s)),
+ _error-Warning(_(Conflicting distribution: %s (expected %s or %s but got %s)),
Desc.Description.c_str(),
+ MetaIndexParser-GetExpectedDist().c_str(),
Transformed.c_str(),
MetaIndexParser-GetDist().c_str());
}
diff --git a/apt-pkg/indexrecords.cc b/apt-pkg/indexrecords.cc
index 5353d10..e7fb1c8 100644
--- a/apt-pkg/indexrecords.cc
+++ b/apt-pkg/indexrecords.cc
@@ -37,6 +37,12 @@ APT_PURE string indexRecords::GetSuite() const
return this-Suite;
}
+APT_PURE bool indexRecords::CheckExpectedDist() const
+{
+ return (this-Dist == this-ExpectedDist
+ || this-Suite == this-ExpectedDist);
+}
+
APT_PURE bool indexRecords::CheckDist(const string MaybeDist) const
{
return (this-Dist == MaybeDist
diff --git a/apt-pkg/indexrecords.h b/apt-pkg/indexrecords.h
index e31f889..12d4971 100644
--- a/apt-pkg/indexrecords.h
+++ b/apt-pkg/indexrecords.h
@@ -51,6 +51,7 @@ class indexRecords
std::string GetSuite() const;
time_t GetValidUntil() const;
virtual bool CheckDist(const std::string MaybeDist) const;
+ virtual bool CheckExpectedDist() const;
std::string GetExpectedDist() const;
virtual ~indexRecords(){};
};
--
1.8.5.3
Bug#742281: reprepro: Generated Suite field of a snapshot should be based on suite not codename.
Package: reprepro
Version: 4.14.1-1
Severity: normal
Tags: patch
Hi,
The generated suite that is written to the Suite field in the Release file
of a snapshot is constructed using the codename of a distribution. This is
misleading because it mixes codename and suites. The Suite field should
be based on the distribution's suite and the snapshot name. This is also true
for the Codename field that is simply copied from the distribution a
snapshot is based upon.
Despite the comment in code, as of apt 0.9.16.1 the generated fake codename
doesn't suppress the apt warning about not matching fields in Release and
sources.list until #644610 is fixed.
Regards,
Lukas
From: Lukas Anzinger lukas.anzin...@xatronic.at
Date: Fri, 21 Mar 2014 13:09:50 +0100
Subject: Generated Suite field of a snapshot should be based on suite not
codename.
The generated suite that is written to the Suite field in the Release file
of a snapshot was constructed using the codename of a distribution. This is
misleading because it mixes codename and suites. The Suite field should
be based on the distribution's suite and the snapshot name. This is also true
for the Codename field that was simply copied from the distribution a
snapshot is based upon.
Despite the comment in code, as of apt 0.9.16.1 the generated fake codename
doesn't suppress the apt warning about not matching fields in Release and
sources.list until #644610 is fixed.
---
distribution.c | 3 ++-
release.c | 20 ++--
release.h | 2 +-
3 files changed, 17 insertions(+), 8 deletions(-)
diff --git a/distribution.c b/distribution.c
index 6db17da..2d6cdb5 100644
--- a/distribution.c
+++ b/distribution.c
@@ -738,7 +738,8 @@ retvalue distribution_snapshot(struct distribution *distribution, const char *na
assert (distribution != NULL);
- r = release_initsnapshot(distribution-codename, name, release);
+ r = release_initsnapshot(distribution-codename, distribution-suite,
+ name, release);
if (RET_WAS_ERROR(r))
return r;
diff --git a/release.c b/release.c
index 1ebf2bc..4893447 100644
--- a/release.c
+++ b/release.c
@@ -224,7 +224,7 @@ retvalue release_init(struct release **release, const char *codename, const char
return RET_OK;
}
-retvalue release_initsnapshot(const char *codename, const char *name, struct release **release) {
+retvalue release_initsnapshot(const char *codename, const char *suite, const char *name, struct release **release) {
struct release *n;
n = zNEW(struct release);
@@ -235,20 +235,28 @@ retvalue release_initsnapshot(const char *codename, const char *name, struct rel
free(n);
return RET_ERROR_OOM;
}
- /* apt only removes the last /... part but we create two,
- * so stop it generating warnings by faking a suite */
- n-fakesuite = mprintf(%s/snapshots/%s, codename, name);
- if (FAILEDTOALLOC(n-fakesuite)) {
+ n-fakecodename = mprintf(%s/snapshots/%s, codename, name);
+ if (FAILEDTOALLOC(n-fakecodename)) {
free(n-dirofdist);
free(n);
return RET_ERROR_OOM;
}
- n-fakecodename = NULL;
+ n-fakesuite = NULL;
+ if (suite != NULL) {
+ n-fakesuite = mprintf(%s/snapshots/%s, suite, name);
+ if (FAILEDTOALLOC(n-fakesuite)) {
+ free(n-fakecodename);
+ free(n-dirofdist);
+ free(n);
+ return RET_ERROR_OOM;
+ }
+ }
n-fakecomponentprefix = NULL;
n-fakecomponentprefixlen = 0;
n-cachedb = NULL;
n-snapshotname = strdup(name);
if (n-snapshotname == NULL) {
+ free(n-fakecodename);
free(n-fakesuite);
free(n-dirofdist);
free(n);
diff --git a/release.h b/release.h
index a0f10c5..11eb887 100644
--- a/release.h
+++ b/release.h
@@ -27,7 +27,7 @@ typedef unsigned int compressionset; /* 1 indexcompression */
/* Initialize Release generation */
retvalue release_init(struct release **, const char * /*codename*/, /*@null@*/const char * /*suite*/, /*@null@*/const char * /*fakeprefix*/);
/* same but for a snapshot */
-retvalue release_initsnapshot(const char *codename, const char *name, struct release **);
+retvalue release_initsnapshot(const char *codename, const char *suite, const char *name, struct release **);
retvalue release_mkdir(struct release *, const char * /*relativedirectory*/);
Bug#739174: reprepro: Add parameter _addreferences to manually add multiple references at once
Package: reprepro
Version: 4.13.1-1
Severity: wishlist
Tags: patch
Hi,
I've created a patch that makes it possible to manually add multiple
references at once.
I've tested the already implemented _addreference action by referencing all
i386 packages in Squeeze but aborted after waiting for about 20
minutes. Then I wrote _addreferences which takes multiple references
either as command line arguments or on stdin at once. This action
makes it possible to reference all of Squeeze i386 in just about 20 seconds.
Regards,
Lukas
From 8785b08a19cf5b840c94ac2ac15501e3759f5495 Mon Sep 17 00:00:00 2001
From: Lukas Anzinger l.anzin...@gmail.com
Date: Wed, 12 Feb 2014 18:39:18 +0100
Subject: [PATCH] Add parameter _addreferences to manually add multiple
references.
If multiple references should be added using the _addreference command in a
loop or with xargs is extremely slow. _addreferences allows to either pass a
list of references on the command line or via stdin to circumvent ARG_MAX.
---
docs/manual.html | 1 +
docs/reprepro.1 | 5 +
docs/reprepro.bash_completion | 1 +
docs/reprepro.zsh_completion | 8
main.c| 30 ++
5 files changed, 45 insertions(+)
diff --git a/docs/manual.html b/docs/manual.html
index 23faa2c..97e4b51 100644
--- a/docs/manual.html
+++ b/docs/manual.html
@@ -1325,6 +1325,7 @@ Some low level commands to access this are (take a look at the manpage for how t
dt class=commanddumpreferences/dtddprint a list of all references/dd
dt class=command_removereferences/dtddremove everything referenced by a given identifier/dd
dt class=command_addreference/dtddmanually add a reference/dd
+dt class=command_addreferences/dtddmanually add multiple references/dd
/dl
h3files.db / checksums.db/h3
These files contains what reprepro knows about your tt class=dirpool//tt directory,
diff --git a/docs/reprepro.1 b/docs/reprepro.1
index 2ed9298..7c74c31 100644
--- a/docs/reprepro.1
+++ b/docs/reprepro.1
@@ -1092,6 +1092,11 @@ the corresponding Packages or Sources file would get)
.BI _addreference filekey identifier
Manually mark \fIfilekey\fP to be needed by \fIidentifier\fP
.TP
+.BI _addreferences identifier \fR[\fR filekeys \fR]\fR
+Manually mark one or more \fIfilekeys\fP to be needed by \fIidentifier\fP.
+If no command line arguments are given,
+stdin is read and every line treated as one filekey.
+.TP
.BI _removereferences identifier
Remove all references what is needed by
.I identifier.
diff --git a/docs/reprepro.bash_completion b/docs/reprepro.bash_completion
index f339d4e..b548cef 100644
--- a/docs/reprepro.bash_completion
+++ b/docs/reprepro.bash_completion
@@ -315,6 +315,7 @@ _reprepro()
_addchecksums\
_addpackage\
_addreference\
+ _addreferences\
_detect\
_dumpcontents\
_fakeemptyfilelist\
diff --git a/docs/reprepro.zsh_completion b/docs/reprepro.zsh_completion
index 00f47cd..8025420 100644
--- a/docs/reprepro.zsh_completion
+++ b/docs/reprepro.zsh_completion
@@ -266,6 +266,7 @@ hiddencommands=(
_addchecksums:add checksums to database
_addmd5sums:add checksums to database
_addreference:mark a filekey needed by an identifier
+ _addreferences:mark multiple filekeys needed by an identifier
_detect:look if the file belonging to a filekey exists and add to the database.
_dumpcontents:output contents of a part of the repository
_fakeemptyfilelist:create an empty fake filelist cache item for a filekey
@@ -501,6 +502,13 @@ case $state in
_reprepro_identifiers
fi
;;
+ (_addreferences)
+ if [[ $state = first argument ]] ; then
+ _reprepro_identifiers
+ elif [[ $state = second argument ]] ; then
+ _reprepro_filekeys
+ fi
+ ;;
(__uncompress)
if [[ $state = first argument ]] ; then
uncompressions=(.gz .bz2 .lzma .xz .lz)
diff --git a/main.c b/main.c
index 2965341..7009721 100644
--- a/main.c
+++ b/main.c
@@ -617,6 +617,34 @@ ACTION_R(n, n, n, y, addreference) {
return references_increment(argv[1], argv[2]);
}
+ACTION_R(n, n, n, y, addreferences) {
+ char buffer[5000], *nl;
+ int i;
+ retvalue r, ret;
+
+ ret = RET_NOTHING;
+
+ if (argc 2) {
+ for (i = 2 ; i argc ; i++) {
+ const char *filename = argv[i];
+ r = references_increment(filename, argv[1]);
+ RET_UPDATE(ret, r);
+ }
+ } else {
+ while (fgets(buffer, 4999, stdin) != NULL) {
+ nl = strchr(buffer, '\n');
+ if (nl == NULL) {
+return RET_ERROR;
+ }
+ *nl = '\0';
+ r = references_increment(buffer, argv[1]);
+ RET_UPDATE(ret, r);
+ }
+ }
+
+ return ret;
+}
+
static retvalue remove_from_target(struct distribution *distribution, struct trackingdata *trackingdata, struct target *target, int count, const char * const *names, int *todo, bool *gotremoved) {
retvalue result, r;
int i;
@@ -3856,6 +3884,8 @@ static const struct action {
1, 1, _removereferences identifier},
{_addreference, A__R(addreference
Bug#739175: Fwd: reprepro: Add parameter _addreferences to manually add multiple references at once
Package: reprepro
Version: 4.13.1-1
Severity: wishlist
Tags: patch
Hi,
I've created a patch that makes it possible to manually add multiple
references at once.
I've tested the already implemented _addreference action by referencing all
i386 packages in Squeeze but aborted after waiting for about 20
minutes. Then I wrote _addreferences which takes multiple references
either as command line arguments or on stdin at once. This action
makes it possible to reference all of Squeeze i386 in just about 20 seconds.
Regards,
Lukas
From 8785b08a19cf5b840c94ac2ac15501e3759f5495 Mon Sep 17 00:00:00 2001
From: Lukas Anzinger l.anzin...@gmail.com
Date: Wed, 12 Feb 2014 18:39:18 +0100
Subject: [PATCH] Add parameter _addreferences to manually add multiple
references.
If multiple references should be added using the _addreference command in a
loop or with xargs is extremely slow. _addreferences allows to either pass a
list of references on the command line or via stdin to circumvent ARG_MAX.
---
docs/manual.html | 1 +
docs/reprepro.1 | 5 +
docs/reprepro.bash_completion | 1 +
docs/reprepro.zsh_completion | 8
main.c| 30 ++
5 files changed, 45 insertions(+)
diff --git a/docs/manual.html b/docs/manual.html
index 23faa2c..97e4b51 100644
--- a/docs/manual.html
+++ b/docs/manual.html
@@ -1325,6 +1325,7 @@ Some low level commands to access this are (take a look at the manpage for how t
dt class=commanddumpreferences/dtddprint a list of all references/dd
dt class=command_removereferences/dtddremove everything referenced by a given identifier/dd
dt class=command_addreference/dtddmanually add a reference/dd
+dt class=command_addreferences/dtddmanually add multiple references/dd
/dl
h3files.db / checksums.db/h3
These files contains what reprepro knows about your tt class=dirpool//tt directory,
diff --git a/docs/reprepro.1 b/docs/reprepro.1
index 2ed9298..7c74c31 100644
--- a/docs/reprepro.1
+++ b/docs/reprepro.1
@@ -1092,6 +1092,11 @@ the corresponding Packages or Sources file would get)
.BI _addreference filekey identifier
Manually mark \fIfilekey\fP to be needed by \fIidentifier\fP
.TP
+.BI _addreferences identifier \fR[\fR filekeys \fR]\fR
+Manually mark one or more \fIfilekeys\fP to be needed by \fIidentifier\fP.
+If no command line arguments are given,
+stdin is read and every line treated as one filekey.
+.TP
.BI _removereferences identifier
Remove all references what is needed by
.I identifier.
diff --git a/docs/reprepro.bash_completion b/docs/reprepro.bash_completion
index f339d4e..b548cef 100644
--- a/docs/reprepro.bash_completion
+++ b/docs/reprepro.bash_completion
@@ -315,6 +315,7 @@ _reprepro()
_addchecksums\
_addpackage\
_addreference\
+ _addreferences\
_detect\
_dumpcontents\
_fakeemptyfilelist\
diff --git a/docs/reprepro.zsh_completion b/docs/reprepro.zsh_completion
index 00f47cd..8025420 100644
--- a/docs/reprepro.zsh_completion
+++ b/docs/reprepro.zsh_completion
@@ -266,6 +266,7 @@ hiddencommands=(
_addchecksums:add checksums to database
_addmd5sums:add checksums to database
_addreference:mark a filekey needed by an identifier
+ _addreferences:mark multiple filekeys needed by an identifier
_detect:look if the file belonging to a filekey exists and add to the database.
_dumpcontents:output contents of a part of the repository
_fakeemptyfilelist:create an empty fake filelist cache item for a filekey
@@ -501,6 +502,13 @@ case $state in
_reprepro_identifiers
fi
;;
+ (_addreferences)
+ if [[ $state = first argument ]] ; then
+ _reprepro_identifiers
+ elif [[ $state = second argument ]] ; then
+ _reprepro_filekeys
+ fi
+ ;;
(__uncompress)
if [[ $state = first argument ]] ; then
uncompressions=(.gz .bz2 .lzma .xz .lz)
diff --git a/main.c b/main.c
index 2965341..7009721 100644
--- a/main.c
+++ b/main.c
@@ -617,6 +617,34 @@ ACTION_R(n, n, n, y, addreference) {
return references_increment(argv[1], argv[2]);
}
+ACTION_R(n, n, n, y, addreferences) {
+ char buffer[5000], *nl;
+ int i;
+ retvalue r, ret;
+
+ ret = RET_NOTHING;
+
+ if (argc 2) {
+ for (i = 2 ; i argc ; i++) {
+ const char *filename = argv[i];
+ r = references_increment(filename, argv[1]);
+ RET_UPDATE(ret, r);
+ }
+ } else {
+ while (fgets(buffer, 4999, stdin) != NULL) {
+ nl = strchr(buffer, '\n');
+ if (nl == NULL) {
+return RET_ERROR;
+ }
+ *nl = '\0';
+ r = references_increment(buffer, argv[1]);
+ RET_UPDATE(ret, r);
+ }
+ }
+
+ return ret;
+}
+
static retvalue remove_from_target(struct distribution *distribution, struct trackingdata *trackingdata, struct target *target, int count, const char * const *names, int *todo, bool *gotremoved) {
retvalue result, r;
int i;
@@ -3856,6 +3884,8 @@ static const struct action {
1, 1, _removereferences identifier},
{_addreference, A__R(addreference
Bug#720208: cowbuilder: Problem executing command with --execute or --login
Hi, Or at least I wish there were some way to execute a command in a cowbuilder chroot. Maybe there is and I just haven't yet figured out how to use it correctly? The --execute parameter is a bit tedious. It expects a script or statically linked executable that is then cat'ed into the cowbuilder chroot. I actually don't know why it's not simply copied into it. Anyway, it's important to know that it basically copies an executable (script) from the host into the chroot and execute it there. It's not a command that is executed in the chroot! E.g. --execute /bin/bash will fail if your host's bash is linked against a set of libraries that don't exist in your chroot. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#715534: git is built without support for python
Hi Jonathan, although p4 is non-free it could probably be distributed in the non-free component of Debian, couldn't it? When I look at the TOS (http://www.perforce.com/downloads/terms-use) I guess that any other proprietary software that is distributed in non-free has similar license agreements. Anyway I believe that having git compiled with Python support enabled could be useful for different purposes as well. I found the following changelog entry: git-core (1:1.7.0~rc1-1) experimental; urgency=low [...] * debian/rules: NO_PYTHON=1. Packaging the git_remote_helpers Python module will require more work. [...] -- Jonathan Nieder jrnie...@gmail.com Sat, 30 Jan 2010 17:46:04 -0600 What kind of work would it require? Regards, Lukas On Wed, Jan 15, 2014 at 1:45 AM, Jonathan Nieder j...@google.com wrote: severity 715534 wishlist quit Hi Lukas, Lukas Anzinger wrote: when I try to use git-p4 it exists and says that git is built without support for Python. Probably I should remove the git-p4 placeholder. p4 is non-free software, so alas I won't be able to package git-p4 which depends on it in Debian. Hoping that clarifies, Jonathan -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#722712: Converted apt-cacher-ng to debhelper 9
Package: apt-cacher-ng Version: 0.7.18-1 Severity: normal Tags: patch Hi, I've converted your package to debhelper 9; a patched is attached. You might want to run debdiff on the resulting package to check the differences to the current one. The only difference I could find is related to systemd in two of the maintainer scripts. I'd suggest to chmod expire-caller.pl 755 upstream, so we don't have to that in debian/rules. Regards, Lukas converted-apt-cacher-ng-to-debhelper-9.patch Description: Binary data
Bug#721532: aac-enc doesn't include PDF documentation from source package
Package: aac-enc Version: 0.1.1+20130514-2 Hi, it'd be nice if the the PDF documentation would be installed to /usr/share/doc. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#720987: initramfs-tools: network isn't deconfigured which breaks networking initscript
Package: dropbear Version: 2012.55-1.3 Severity: normal Hi, I've been using dropbear to unlock my encrypted root partition remotely via SSH. This works quite fine in general, however, I've stumbled upon the following shortcoming: Dropbear configures the network using configure_networking() from initramfs-tools but never deconfigures it. The initscript networking which is part of the package ifupdown runs ifup during boot which then complains that the interface is already configured. This is quite annoying because I want to use a different network configuration in the initramfs compared to the normal system. configure_networking() uses ipconfig from package klibc-utils to configure the network. I took a look at the source code of ipconfig and found code that brings the interface up (netdev_up()) as well as down (netdev_down()). However, the latter isn't called from anywhere, so it's essentially dead code. In my opinion a fix would require work being done in the following three packages: * klibc-utils - A command line option to bring the interfaces down is needed. * initramfs-tools - A function to deconfigure networking is needed. * dropbear - The introduced function in initramfs-tools must be called to deconfigure networking. Regards, Lukas
Bug#720988: Fwd: initramfs-tools: network isn't deconfigured which breaks networking initscript
Package: dropbear Version: 2012.55-1.3 Severity: normal Hi, I've been using dropbear to unlock my encrypted root partition remotely via SSH. This works quite fine in general, however, I've stumbled upon the following shortcoming: Dropbear configures the network using configure_networking() from initramfs-tools but never deconfigures it. The initscript networking which is part of the package ifupdown runs ifup during boot which then complains that the interface is already configured. This is quite annoying because I want to use a different network configuration in the initramfs compared to the normal system. configure_networking() uses ipconfig from package klibc-utils to configure the network. I took a look at the source code of ipconfig and found code that brings the interface up (netdev_up()) as well as down (netdev_down()). However, the latter isn't called from anywhere, so it's essentially dead code. In my opinion a fix would require work being done in the following three packages: * klibc-utils - A command line option to bring the interfaces down is needed. * initramfs-tools - A function to deconfigure networking is needed. * dropbear - The introduced function in initramfs-tools must be called to deconfigure networking. Regards, Lukas
Bug#718220: fglrx-driver: get-orig-source should download tarball to current directory
Package: fglrx-driver Version: 1:13.4-3 Severity: normal Hi, I've noticed that the get-orig-source target in debian/rules of source package fglrx-driver creates a tarball in a directory above th current one. While this is admittely quite handy, I think it's a violation of §4.9 Debian Policy: get-orig-source: [...] This target fetches the most recent version of the original source package [...] and leaves it in the current directory. Regards, Lukas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#715534: git is built without support for python
Package: git Version: 1:1.8.3.2-1 Severity: normal Hi, when I try to use git-p4 it exists and says that git is built without support for Python. $ git p4 fatal: git was built without support for git-p4 (NO_PYTHON=1). I tried to recompile git with support for Python by removing NO_PYTHON=1 in debian/rules and the build apparently succeeded (all tests were positive and git-p4 works now). I actually don't see a reason why git should be compiled without support for Python but maybe the maintainer can shed some light on the matter. Regards, Lukas -- System Information: Debian Release: wheezy/sid APT prefers raring-updates APT policy: (500, 'raring-updates'), (500, 'raring-security'), (500, 'raring'), (100, 'raring-backports') Architecture: i386 (i686) Kernel: Linux 3.8.0-26-generic (SMP w/4 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages git depends on: ii git-man 1:1.8.3.2-1 ii libc62.17-0ubuntu5 ii libcurl3-gnutls 7.29.0-1ubuntu3.1 ii liberror-perl0.17-1 ii libexpat12.1.0-2 ii libpcre3 1:8.31-2 ii perl-modules 5.14.2-21 ii python 2.7.4-0ubuntu1 ii zlib1g 1:1.2.7.dfsg-13ubuntu2 Versions of packages git recommends: ii less 456-1ubuntu1 ii openssh-client [ssh-client] 1:6.1p1-4 ii patch2.6.1-3ubuntu2 ii rsync3.0.9-4 Versions of packages git suggests: ii gettext-base 0.18.1.1-10ubuntu3 pn git-arch none pn git-bzr none pn git-cvs none pn git-daemon-run | git-daemon-sysvinit none pn git-doc none pn git-elnone ii git-email 1:1.8.3.2-1 pn git-gui none ii git-svn 1:1.8.3.2-1 ii gitk 1:1.8.3.2-1 pn gitwebnone -- no debconf information -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#692611: dash: Using subshell in PS4 variable causes fork bomb
Package: dash Version: 0.5.7-3 Severity: normal Dear Maintainer, to have a timestamp in front of every command that is executed, I tried to set the variable PS4 to '$(date)'. #!/bin/sh set -x export PS4='$(date)' /bin/true This works with bash and results in the following output: % bash ./test.sh + export 'PS4=$(date)' + PS4='$(date)' Wed Nov 7 21:55:10 CET 2012/bin/true If I run the same script in dash it results in a fork bomb. I suppose it's because when PS4 is evaluated and a sub shell is spawned, it inherits PS4 and reevaluates it and so on and so forth. Regards, Lukas -- System Information: Debian Release: wheezy/sid APT prefers unstable APT policy: (500, 'unstable') Architecture: amd64 (x86_64) Kernel: Linux 2.6.32-5-vserver-amd64 (SMP w/24 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages dash depends on: ii debianutils 4.3.4 ii dpkg 1.16.9 ii libc62.13-36 dash recommends no packages. dash suggests no packages. -- debconf information: * dash/sh: true -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#661696: [2.6.33.y] drm/i915: wrong fifo size due to uncareful refactoring which results in an xserver crash at 800x600
Hi Jonathan, yes, it works for me! Regards, Lukas On Sun, Jun 24, 2012 at 6:39 PM, Jonathan Nieder jrnie...@gmail.com wrote: forwarded 661696 https://bugzilla.kernel.org/show_bug.cgi?id=42839 quit (cc-ing Stefan Bader since he maintains the linux-2.6.32.y-drm33.z tree) Hi again, Lukas Anzinger wrote[1]: During a refactoring of the i915 driver a regression has been introduced [e70236a8d3d0] For some chipsets the wrong fifo size is determined which results in lot's of pixel errors when starting the xserver and choosing 800x600 as a resolution. If another resolution is used (eg. 1024x768 or 1280x1024), I don't encounter this problem. [...] Jonathan Nieder wrote: Lukas Anzinger wrote: looks good to me. Do you mean that the patch looks good or that you've tried applying it and the behavior was bad before and good afterward? Ping. Have you tested this patch? From: Adam Jackson a...@redhat.com Date: Fri, 16 Apr 2010 18:20:57 -0400 Subject: drm/i915: Attempt to fix watermark setup on 85x (v2) commit 8f4695ed1c9e068772bcce4cd4ff03f88d57a008 upstream. IS_MOBILE() catches 85x, so we'd always try to use the 9xx FIFO sizing; since there's an explicit 85x version, this seems wrong. v2: Handle 830m correctly too. Signed-off-by: Adam Jackson a...@redhat.com Reviewed-by: Eric Anholt e...@anholt.net Signed-off-by: Eric Anholt e...@anholt.net Signed-off-by: Jonathan Nieder jrnie...@gmail.com In suspense, Jonathan [1] http://bugs.debian.org/661696 -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#661696: drm/i915: wrong fifo size due to uncareful refactoring which results in an xserver crash at 800x600
Hi Jonathan, looks good to me. I've also reported this bug upstream under #42839. Regards, Lukas On Wed, Feb 29, 2012 at 19:36, Jonathan Nieder jrnie...@gmail.com wrote: tags 661696 = upstream patch moreinfo quit Hi Lukas, Lukas Anzinger wrote: During a refactoring of the i915 driver a regression has been introduced ([...]h=e70236a8d3d0a4c100a0b9f7d394d9bda9c56aca): For some chipsets the wrong fifo size is determined which results in lot's of pixel errors when starting the xserver and choosing 800x600 as a resolution. If another resolution is used (eg. 1024x768 or 1280x1024), I don't encounter this problem. I've attached a patch that fixes the problem (no pixel errors anymore) and determines the correct fifo size. Thanks for the report and analysis. Please test the attached patch against 2.6.32.y or a squeeze kernel. If it works, we can send this to Greg for inclusion in the 2.6.32.y series so everyone benefits. Hope that helps, Jonathan -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#661696: drm/i915: wrong fifo size due to uncareful refactoring which results in an xserver crash at 800x600
Package: linux-2.6
Version: 2.6.32-41
Severity: important
Tags: upstream patch
Hi,
I've encountered a nasty bug in the drm/i915 part of Debian's stable kernel.
The code is also present in 2.6.33.20, which is the latest stable upstream
release of 2.6.33.y at the time of writing.
During a refactoring of the i915 driver a regression has been introduced
(http://git.kernel.org/?p=linux/kernel/git/stable/linux-stable.git;a=commit;h=e70236a8d3d0a4c100a0b9f7d394d9bda9c56aca):
For some chipsets the wrong fifo size is determined which results in lot's of
pixel errors when starting the xserver and choosing 800x600 as a resolution. If
another resolution is used (eg. 1024x768 or 1280x1024), I don't encounter this
problem.
I've attached a patch that fixes the problem (no pixel errors anymore) and
determines the correct fifo size.
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -4828,15 +4828,15 @@ static void intel_init_display(struct drm_device *dev)
dev_priv-display.update_wm = i965_update_wm;
else if (IS_I9XX(dev) || IS_MOBILE(dev)) {
dev_priv-display.update_wm = i9xx_update_wm;
- dev_priv-display.get_fifo_size = i9xx_get_fifo_size;
- } else {
if (IS_I85X(dev))
dev_priv-display.get_fifo_size = i85x_get_fifo_size;
else if (IS_845G(dev))
dev_priv-display.get_fifo_size = i845_get_fifo_size;
else
- dev_priv-display.get_fifo_size = i830_get_fifo_size;
+ dev_priv-display.get_fifo_size = i9xx_get_fifo_size;
+ } else {
dev_priv-display.update_wm = i830_update_wm;
+ dev_priv-display.get_fifo_size = i830_get_fifo_size;
}
}
Regards,
Lukas
-- Package-specific info:
** Version:
Linux version 2.6.32-5-686 (Debian 2.6.32-41) (b...@decadent.org.uk) (gcc
version 4.3.5 (Debian 4.3.5-4) ) #1 SMP Mon Jan 16 16:04:25 UTC 2012
** Command line:
BOOT_IMAGE=/boot/vmlinuz-2.6.32-5-686
root=UUID=08a419ad-d4d1-47ff-94fd-6d7c8c5321de ro quiet
** Not tainted
** Kernel log:
[1.637921] sd 0:0:0:0: [sda] 8027712 512-byte logical blocks: (4.11 GB/3.82
GiB)
[1.638005] sd 0:0:0:0: [sda] Write Protect is off
[1.638010] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
[1.638045] sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled,
doesn't support DPO or FUA
[1.638258] sda: sda1 sda2
[1.639892] sd 0:0:0:0: [sda] Attached SCSI disk
[1.740461] usb 1-1: new full speed USB device using uhci_hcd and address 2
[1.744283] kjournald starting. Commit interval 5 seconds
[1.744299] EXT3-fs: mounted filesystem with ordered data mode.
[1.897098] usb 1-1: New USB device found, idVendor=1a40, idProduct=0201
[1.897104] usb 1-1: New USB device strings: Mfr=0, Product=1, SerialNumber=0
[1.897109] usb 1-1: Product: USB 2.0 Hub [MTT]
[1.897256] usb 1-1: configuration #1 chosen from 1 choice
[1.900120] hub 1-1:1.0: USB hub found
[1.902075] hub 1-1:1.0: 7 ports detected
[2.006518] udev[254]: starting version 164
[2.020066] usb 1-2: new full speed USB device using uhci_hcd and address 3
[2.194232] usb 1-2: New USB device found, idVendor=9710, idProduct=7830
[2.194239] usb 1-2: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[2.194244] usb 1-2: Product: USB-MAC Controller
[2.194247] usb 1-2: Manufacturer: Moschip Semiconductor
[2.194251] usb 1-2: SerialNumber: 3b00043d
[2.194414] usb 1-2: configuration #1 chosen from 1 choice
[2.274033] usb 1-1.2: new full speed USB device using uhci_hcd and address 4
[2.350298] pci_hotplug: PCI Hot Plug PCI Core version: 0.5
[2.388923] input: PC Speaker as /devices/platform/pcspkr/input/input0
[2.403022] Marking TSC unstable due to TSC halts in idle
[2.405028] processor LNXCPU:00: registered as cooling_device1
[2.419214] Switching to clocksource acpi_pm
[2.456113] intel_rng: FWH not detected
[2.470309] ACPI: AC Adapter [ADP1] (off-line)
[2.470506] input: Power Button as
/devices/LNXSYSTM:00/LNXSYBUS:00/PNP0C0C:00/input/input1
[2.470522] ACPI: Power Button [PWRB]
[2.470618] input: Power Button as
/devices/LNXSYSTM:00/LNXPWRBN:00/input/input2
[2.470624] ACPI: Power Button [PWRF]
[2.488242] i801_smbus :00:1f.3: PCI INT B - GSI 17 (level, low) - IRQ
17
[2.499895] shpchp: Standard Hot Plug PCI Controller Driver version: 0.4
[2.529165] usb 1-1.3: new low speed USB device using uhci_hcd and address 5
[2.542981] usb 1-2: applying rev.C fixup
[2.549198] usb 1-2: applying rev.C fixup
[2.561229] eth1: register 'MOSCHIP usb-ethernet driver' at
usb-:00:1d.0-2, MOSCHIP 7830/7730 usb-NET adapter, 00:13:3b:00:04:3d
[2.561269] usbcore: registered new interface driver MOSCHIP usb-ethernet
driver
[2.618157] [drm] Initialized drm 1.1.0 20060810
[2.714449]
Bug#570436: curl: Should work now
Package: curl Followup-For: Bug #570436 Hi, I've just tried to rebuild curl with --enable-c-ares and it looks like the IPv6 problems are gone: ### lukas@vm0:~$ curl --version curl 7.23.1 (i486-pc-linux-gnu) libcurl/7.23.1 OpenSSL/1.0.0e zlib/1.2.3.4 c-ares/1.7.5 libidn/1.23 libssh2/1.2.8 librtmp/2.3 Protocols: dict file ftp ftps gopher http https imap imaps ldap pop3 pop3s rtmp rtsp scp sftp smtp smtps telnet tftp Features: AsynchDNS Debug TrackMemory GSS-Negotiate IDN IPv6 Largefile NTLM NTLM_WB SSL libz lukas@vm0:~$ curl http://security.debian.org !DOCTYPE HTML PUBLIC -//IETF//DTD HTML 2.0//EN htmlhead title302 Found/title /headbody h1Found/h1 pThe document has moved a href=http://www.debian.org/security/;here/a./p hr addressApache Server at security.debian.org Port 80/address /body/html lukas@vm0:~$ curl -6 http://security.debian.org curl: (7) Failed to connect to 2001:8d8:580:400:6564:a62:0:2: Network is unreachable ### I don't know on which side the problems have been fixed but it looks good to me at first sight. I think you could try to enable it once again and maybe upload it to experimental first. Regards, Lukas -- System Information: Debian Release: wheezy/sid APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 3.1.0-1-686-pae (SMP w/1 CPU core) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages curl depends on: ii libc6 2.13-21 ii libcurl3 7.23.1-2 ii zlib1g1:1.2.3.4.dfsg-3 curl recommends no packages. curl suggests no packages. -- no debconf information -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#636359: lua5.1 version 5.1.4-10 will sove the problem
Installing latest lua5.1 from unstable did the trick. Thanks!
Bug#636359: prosody: does not start after latest upgrade (complains about missing libs)
Package: prosody Version: 0.8.2-1 Severity: grave Justification: renders package unusable Tonight cron-apt installed a new version of prosody (0.8.2-1) from testing. Unfortunately after the upgrade of the package prosody does not start anymore. It keeps complaining about missing libraries although all dependencies are installed. The output is the following: - # invoke-rc.d prosody restart Restarting Prosody XMPP Server: prosody ** Prosody was unable to find luaexpat This package can be obtained in the following ways: Source: http://www.keplerproject.org/luaexpat/ Debian/Ubuntu:sudo apt-get install liblua5.1-expat0 luarocks: luarocks install luaexpat luaexpat is required for Prosody to run, so we will now exit. More help can be found on our website, at http://prosody.im/doc/depends ** ** Prosody was unable to find luasocket This package can be obtained in the following ways: Source: http://www.tecgraf.puc-rio.br/~diego/professional/luasocket/ Debian/Ubuntu:sudo apt-get install liblua5.1-socket2 luarocks: luarocks install luasocket luasocket is required for Prosody to run, so we will now exit. More help can be found on our website, at http://prosody.im/doc/depends ** ** Prosody was unable to find luafilesystem This package can be obtained in the following ways: Source: http://www.keplerproject.org/luafilesystem/ Debian/Ubuntu:sudo apt-get install liblua5.1-filesystem0 luarocks: luarocks install luafilesystem luafilesystem is required for Prosody to run, so we will now exit. More help can be found on our website, at http://prosody.im/doc/depends ** ** Prosody was unable to find LuaSec This package can be obtained in the following ways: Source: http://www.inf.puc-rio.br/~brunoos/luasec/ Debian/Ubuntu:http://prosody.im/download/start#debian_and_ubuntu luarocks: luarocks install luasec SSL/TLS support will not be available More help can be found on our website, at http://prosody.im/doc/depends ** failed! invoke-rc.d: initscript prosody, action restart failed. - I also tried to purge and reinstall the package, which didn't changed anything, however. I then installed an older version of prosody (0.8.1) from http://snapshot.debian.org which also didn't change anything. However, I successfully migrated to 0.8 a while ago and haven't changed much on the system since then. -- System Information: Debian Release: wheezy/sid APT prefers testing APT policy: (500, 'testing') Architecture: amd64 (x86_64) Kernel: Linux 2.6.39-2-amd64 (SMP w/1 CPU core) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages prosody depends on: ii adduser 3.113 add and remove users and groups ii libc6 2.13-10Embedded GNU C Library: Shared lib ii libidn11 1.22-2 GNU Libidn library, implementation ii liblua5.1-0 5.1.4-5Simple, extensible, embeddable pro ii liblua5.1-expat0 1.2.0-3libexpat bindings for the Lua lang ii liblua5.1-filesystem0 1.5.0-2luafilesystem library for the Lua ii liblua5.1-socket2 2.0.2-6TCP/UDP socket library for Lua 5.1 ii libssl1.0.0 1.0.0d-3 SSL shared libraries ii lua5.15.1.4-5Simple, extensible, embeddable pro ii openssl 1.0.0d-3 Secure Socket Layer (SSL) binary a Versions of packages prosody recommends: ii liblua5.1-event0 0.3.1-3asynchronous event notification li ii liblua5.1-sec10.4-5 SSL socket library for the Lua lan prosody suggests no packages. -- Configuration Files: /etc/prosody/conf.avail/example.com.cfg.lua [Errno 13] Permission denied: u'/etc/prosody/conf.avail/example.com.cfg.lua' /etc/prosody/conf.avail/localhost.cfg.lua [Errno 13] Permission denied: u'/etc/prosody/conf.avail/localhost.cfg.lua' /etc/prosody/prosody.cfg.lua [Errno 13] Permission denied: u'/etc/prosody/prosody.cfg.lua' -- no debconf information
