Bug#1068239: jq: JSON filters can be fooled by \u0041 or other escapes in object names

Package: jq Version: 1.6-2.1 Severity: important Consider this JSON file: { "\u0041PIModule": "/test2.dll", "APIModule": "/test.dll" } On running jq .APIModule < test.json, the output is "/test.dll". The expected output is "/test2.dll", "/test.dll", or alternately an error

Bug#1068024: Potential solution to your downgrade problem in dpkg

The dpkg -> xz-utils downgrade problem has a suggestion that suggests itself. 1) Downgrade dpkg's dependency to the last known good. It doesn't matter how old so long as it can read the file formats. I understand this is likely to be 5.4.1. 2) Statically link all the decompressor libraries into

Bug#1061404: dpkg read buffer overrun unpacking K (long symbolic) records in data.tar

On Tue, Jan 23, 2024 at 5:40 PM Guillem Jover wrote: > > Hi! > > On Tue, 2024-01-23 at 16:47:34 -0800, Joshua Hudson wrote: > > On Tue, Jan 23, 2024 at 3:16 PM Guillem Jover wrote: > > > On Tue, 2024-01-23 at 13:46:53 -0800, Joshua Hudson wrote: > > > >

Bug#1061404: dpkg read buffer overrun unpacking K (long symbolic) records in data.tar

On Tue, Jan 23, 2024 at 3:16 PM Guillem Jover wrote: > > Hi! > > On Tue, 2024-01-23 at 13:46:53 -0800, Joshua Hudson wrote: > > Package: dpkg > > Version: 1.21.22 > > Severity: important > > > On unpacking a custom .dpkg file with long symbolic links, I fou

Bug#1061404: dpkg read buffer overrun unpacking K (long symbolic) records in data.tar

Package: dpkg Version: 1.21.22 Severity: important Dear Maintainer, On unpacking a custom .dpkg file with long symbolic links, I found a bunch of symbolic links ending in right, and one with copyright. The overrun made all the links exactly the same length; suggesting reuse of some kind of

Bug#1050578:

Bug verified *upstream* on stock kernel version 6.4.12

Bug#1050483:

Trying on sid (2023-08-25) objdump fails with: "./mv: file format not recognized" and readelf still works. objdump version from sid: GNU objdump (GNU Binutils for Debian) 2.41

Bug#1037516: ircd-hybrid hates winbindd users

Package: ircd-hybrid Version: 1:8.2.43+dfsg.1-1 Severity: important Tags: newcomer upstream Dear Maintainer, When connecting to ircd-hybrid on localhost, users from a Windows domain get kicked out with invalid user. I tracked this down to the source code; ircd-hybrid believes that \ is an

Bug#1008507:

diff -ur glibc.orig/glibc/sysdeps/unix/sysv/linux/pathconf.c glibc/glibc/sysdeps/unix/sysv/linux/pathconf.c --- glibc.orig/glibc/sysdeps/unix/sysv/linux/pathconf.c2022-03-29 17:50:12.558027042 -0700 +++ glibc/glibc/sysdeps/unix/sysv/linux/pathconf.c2022-03-29 17:52:33.262157543 -0700 @@

Bug#1008507:

/* reproduce glibc bug */ /* * I don't typically care what the return of pathconf(..., _PC_LINK_MAX) * is; I just care whether or not it's greater than 1 because I'm checking * whether the filesystem supports hard links or not. */ #include #include #include #include #include #include

Bug#969489: systemd: systemctl can hang up waiting for device nodes that are already mounted

I'd really rather not run systems at all but pulseaudio is just too broken now. On Friday, September 4, 2020, Joshua Hudson wrote: > I'm not convinced it's dmraid because udev reports the nodes up. It looks > like systems only thinks the first device name on the list is alive. > Unfo

Bug#969489: systemd: systemctl can hang up waiting for device nodes that are already mounted

I'm not convinced it's dmraid because udev reports the nodes up. It looks like systems only thinks the first device name on the list is alive. Unfortunately that name isn't stable across boots. On Fri, Sep 4, 2020, 12:58 PM Michael Biebl wrote: > On Thu, 3 Sep 2020 16:06:28 -0700 Joshua Hud

Bug#969489: systemd: systemctl can hang up waiting for device nodes that are already mounted

Yeah dmraid's still in use. Is there a way to just get rid of all those device units because they don't do anything right now except cause problems? I've modified the boot order so systemd sees everything already mounted. On 9/3/20, Michael Biebl wrote: > Am 04.09.2020 um 00:58 schrieb Jos

Bug#969489: systemd: systemctl can hang up waiting for device nodes that are already mounted

um 00:10 schrieb Joshua Hudson: >> brw-rw 1 root disk 254, 2 Sep 3 14:51 >> /dev/mapper/isw_cfbejbfeib_Volume12 >> >> udevadm command output attached. >> >> The device isn't dead. It works when accessed directly by name. >> >> $ /sbin/swapon &g

Bug#969489: systemd: systemctl can hang up waiting for device nodes that are already mounted

brw-rw 1 root disk 254, 2 Sep 3 14:51 /dev/mapper/isw_cfbejbfeib_Volume12 udevadm command output attached. The device isn't dead. It works when accessed directly by name. $ /sbin/swapon NAMETYPE SIZE USED PRIO /dev/mapper/isw_cfbejbfeib_Volume12

Bug#946974: systemd: hang waiting for VM ethernet card to come online

01:09:06PM -0700, Joshua Hudson wrote: > >> I suspect it might be confusing since I filed two different eth0-related >> bugs. >> >> In this particular instance, the DHCP server was live, and I could >> have omitted the & from the line in rc.local and it would have

Bug#946974: systemd: hang waiting for VM ethernet card to come online

I suspect it might be confusing since I filed two different eth0-related bugs. In this particular instance, the DHCP server was live, and I could have omitted the & from the line in rc.local and it would have come up only a couple of seconds slower.

Bug#947806: systemd doesn't like my raid, times out waiting for online partitions to come online, and can't continue boot

Hey I got lucky with a google search and found what would be a dupe if it were ever finished. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=805821 Somebody else managed to track the problem down. For some reason systemd is only watching for /dev/dm-* rather than the long dmraid device names.

Bug#947806: systemd doesn't like my raid, times out waiting for online partitions to come online, and can't continue boot

I double-checked with init=/bin/sh. All the devices already exist in /dev/mapper before /linuxrc shuts down the initrd copy of udev. systemd will never see the activation of a udev device because that's already done. They're all active.

Bug#947806: systemd doesn't like my raid, times out waiting for online partitions to come online, and can't continue boot

That turned out to be really easy to track down. The reason you aren't getting any notification of the raid array being activated is it was already activated in initrd. It will not be activated a second time. dmraid has some interesting principles in play here. 1) dmraid has its own idea of

Bug#947811: systemd: don't check *mounted* filesystems by default

> If you want to fsck/mount your partitions manually via a custom script that is run before init, then I would set passno to 0. I've since masked the job that systemd uses to run fsck. It's still a bug because if somebody forgets and leaves a filesystem mounted read-write in the emergency shell

Bug#947806: systemd doesn't like my raid, times out waiting for online partitions to come online, and can't continue boot

>> Dec 31 08:27:23 nova dmraid-activate[554]: ERROR: Cannot retrieve RAID set information for isw_cfbejbfeib_Volume1 > Have you further investigated those error messages? I have now investigated and it's clearly spurious. udev tries to run the following commands: /sbin/dmraid-activate /dev/sda

Bug#947811: systemd: don't check *mounted* filesystems by default

#need_real_mtab /dev/mapper/isw_cfbejbfeib_Volume11 / ext3 defaults 0 1 /dev/mapper/isw_cfbejbfeib_Volume12 none swap sw 0 0 none /tmp tmpfs defaults 0 0 /dev/mapper/isw_cfbejbfeib_Volume13 /home ext3 defaults 0 2 /dev/mapper/isw_cfbejbfeib_Volume14 /home/joshua/imagefiles ext3 defaults 0 3

Bug#947908: pulseaudio: sound doesn't play on first launch of pulseaudio at login; restarting it always results in working sound but will reoccur on next boot

> Is there actually a pulseaudio daemon running? yup. > Where? Not sure what this is supposed to mean. > Is it managed by systemd? I dunno how starting it at login is supposed to work. client.conf: autospawn = no Ok this makes absolutely no sense at all. The problem went away when I took

Bug#947908: pulseaudio: sound doesn't play on first launch of pulseaudio at login; restarting it always results in working sound but will reoccur on next boot

> Does `journalctl --user-unit pulseaudio.service` have anything to say? Says "daemon already running" quite a few times. > You can set the debug flags on daemon.conf to force it to verbose mode. Spectacular: E: [pulseaudio] pid.c: Daemon already running. E: [pulseaudio] main.c:

Bug#947807: systemd: can't install systemd when /etc/machine-id is a symbolic link to /dev/null

# apt-get install systemd-sysv [lots of junk] Initializing machine ID from random generator. Failed to truncate /etc/machine-id: Invalid argument This is trivially verified by # dpkg-reconfigure systemd Initializing machine ID from random generator. Failed to truncate /etc/machine-id: Invalid

Bug#947806: systemd doesn't like my raid, times out waiting for online partitions to come online, and can't continue boot

To let the gravity of the but sink in, this is the pair of scripts that I jammed in front of systemd that got it to somehow behave itself: /sbin/init: #!/bin/sh echo "INIT: rc init is in startup" PATH=/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin /etc/rc.earlyboot || PS1='init#

Bug#861053:

The problem is /boot/grub/grub.cfg contains lines that look like: search --no-floppy --fs-uuid --set=root d847c628-aa10-4bef-92b6-72ebacc07d7b search doesn't work because it starts the protected mode driver, which finds multiple copies of the disk (because it's a full-disk RAID

Bug#861053:

So the real problem is grub.cfg contains lines that look like this: search --no-floppy --fs-uuid --set=root d847c628-aa10-4bef-92b6-72ebacc07d7b The probe tries to activate builtin protected-mode drivers, and the drivers "work", so the probe finds two paths to the disk. Both of them

Bug#946974: systemd: hang waiting for VM ethernet card to come online

Forgive me if I'm not quite so sure, because this is the second time I've had an undiagnosable problem where systemd appeared to be waiting for something to start that was already started. On my home system I had to roll back systemd altogether because it would hang for many minutes on /home and

Bug#943441: kolourpaint: Trying to draw transparent on a zoomed-in view does not work

Package: kolourpaint Version: 4:18.04.0-1 Severity: normal Dear Maintainer, 1) Open PNG image 2) Zoom -> 800% 3) Select foreground color -> transparent 4) Select drawing tool (pen, line, and eraser (background color -> transparent) tried) 5) Try to draw on top of image Drawing fails Note that

Bug#923200: libpolkit-qt5-1-1 depends on libpam-systemd

Should I really have to provide an empty virtual package? I'm using pam_group so sound keeps playing in bkg while current virtual console has another user.

Bug#935853: nouveau.ko: kernel selects nouveau for GE108 GeForce1030 which does not support it

Package: src:linux Version: 4.19.37-5+deb10u2 Severity: important File: nouveau.ko Dear Maintainer, I swapped out a failed graphics card for a NVidia GeForce 1030 GE108; the kernel autodetected the nouveau driver and tried to use it, which it cannot actually do, thus no video at bootup. If I

Bug#931901:

Adding RV635 to the list of affected cards In this case, the card needs firmware, the firmware is on the disk, but KMS remains disabled.

Bug#931901:

Adding RV635 to the list of affected cards

Bug#923207: Turns out this bug is not fixed

Turns out this bug is not fixed. I now have the following sneak dependency lxsession -> lxsession-logout -> lxpolkit -> polkit-1 -> libpam-systemd -> systemd-sysv It looks like it was supposed to be lxsession -> lxsession-logout -> polkit-1-auth-agent (lxpolkit -> polkit-1 -> libpam-systemd ->

Bug#924445: I finally figured this one out

pulseaudio exits by default when there's nothing connected to it. However I have this line in ~/.pulse/default.pa load-module module-native-protocol-tcp auth-ip-acl=127.0.0.1 auth-anonymous=1 So if the next program trying to use it is one of the programs trying to use it over TCP, it doesn't

Bug#930652: systemd removes static ipv4 address from ethernet sporadically

I have attached a copy of /etc/network/interfaces. We're now waiting for another recurrence. interfaces Description: Binary data

Bug#924445: pulseaudio crashes randomly, no diagnostic on terminal, possibly associated with Wesnoth

Sorry for the long delay; looks like my spam filter ate the last reply. Now looks like this on startup: W: [pulseaudio] authkey.c: Failed to open cookie file '/home/joshua/.config/pulse/cookie': No such file or directory W: [pulseaudio] authkey.c: Failed to load authentication key

Bug#924445: pulseaudio crashes randomly, no diagnostic on terminal, possibly associated with Wesnoth

Unfortunately there will never be a coredump because it's not crashing in any usual sense. It just stops. joshua@nova:~ϟ gdb pulseaudio GNU gdb (Debian 8.2.1-2) 8.2.1 Copyright (C) 2018 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later

Bug#780768:

On a desktop wired LAN that's just ridiculous. Since then and another really dumb bug I tore out NetworkManager.

Bug#861783: dpkg: trying to local divert a file diverted by a package does not work

(If the suggestions below end up not being accepted, then you could simply use your local firefox, and create a dummy package to fulfill dependencies with something like equivs.) Ah yes I suppose I was not clear. The firefox-esr package causes all its depends to be installed, which is what allows

Bug#861004: closed by Guido Günther <a...@sigxcpu.org> (Re: Bug#861004: git-buildpackage: gbp buildpackage fails with nonsense errors when binary files have changed since last release)

I expect the tool to treat the provided git repository as the master source. This should result in not needing to fetch a source tarball at all. This tool was recommended to me for how to update a debian package to take a new upstream source. Apparently it can't actually do that.

Bug#653959: citadel-server was auto-selected and installed by dist-upgrade for what appears to be no good reason, and apt-get purge citadel-server failed to back out all system changes

Apt-get dist-upgrade should not be installing new auto-start publicly listening services. On Tuesday, January 24, 2017, Michael Meskes wrote: > severity normal > thanks > > On Sun, Jan 01, 2012 at 02:42:07PM -0800, Joshua wrote: > > Source: citadel-server > > Version: wheezy

Bug#839551:

Patch included; I sent the same patch upstream but they seem checked out. diff -ru stock/rdesktop-1.8.3/doc/rdesktop.1 work/rdesktop-1.8.3/doc/rdesktop.1 --- stock/rdesktop-1.8.3/doc/rdesktop.1 2016-10-01 12:21:55.0 -0700 +++ work/rdesktop-1.8.3/doc/rdesktop.1 2016-10-01

Bug#827388: systemd: systemctl restart winbind starts a broken winbind

In this case, the init script has restart != stop + start. Had systemd called restart it probably would have worked. On 6/15/16, Felipe Sateler <fsate...@debian.org> wrote: > Control: reassign -1 winbind 2:4.1.17+dfsg-2+deb8u2 > > On 15 June 2016 at 11:52, Joshua Hudson <j

Bug#743215: write: you are uid ???, but your login is as uid 0

I am astounded at your inability to understand this. # su -c 'write root' | cat will fail because write is running as bin and checks who owns standard output for a match or not. Now this is a pretty perverse example. There's a breaking change in systemd's cron where the standard output spooler

Bug#788631: /sbin/mkfs.minix: mkfs.minix hang/infinite loop

Bug is FIXED in util-linux 2.27

Bug#791430: udev tries to make /dev/sr? nodes only when cdrom in drive, causing problems

I booted directly to a shell and started the system by hand. After mounting devtmpfs, /dev/sr? were not created. Started udevd and they were. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#791430: udev tries to make /dev/sr? nodes only when cdrom in drive, causing problems

It is important to note that without the device node, attempting to eject /dev/sr0 or /dev/sr1 is going to fail. Some (notably apple) cdroms don't have eject buttons. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact

Bug#791430: udev tries to make /dev/sr? nodes only when cdrom in drive, causing problems

Devices not created by kernel. Verified by stopping udev. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#791430: udev tries to make /dev/sr? nodes only when cdrom in drive, causing problems

Update: problem disappeared after cold boot. Additional information gained from experiment: devtmpfs does not create /dev/sr? nodes on this system. udevd does once started. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact

Bug#788631: /sbin/mkfs.minix: mkfs.minix hang/infinite loop

Update: bug is in the V3 code. Command goes through w/o using -3. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#788631: Info received (Bug#788631: /sbin/mkfs.minix: mkfs.minix hang/infinite loop)

-06-20 20:27:11.732436640 -0700 @@ -49,6 +49,9 @@ * 06.29.11 - Overall cleanups for util-linux and v3 support * Davidlohr Bueso d...@gnu.org * + * 06.20.15 - Do not infinite loop or crash on large devices + * Joshua Hudson joshud...@gmail.com + * * Usage: mkfs

Bug#788631: /sbin/mkfs.minix: mkfs.minix hang/infinite loop

I found the exact cause after reading the source code long enough. It tried to allocate too many inodes, pushing the first zone 65536 blocks. This fails hard due to integer roundoff. Therefore, workaround: mkfs.minix -3 -i specifying a more reasonable number of inodes will allow the

Bug#680929: Got bug, can reproduce 100% of the time

Affected: Debian Jessie i386 Qt: 4.8.6 KDE Development Platform: 4.14.2 Okular: 0.20.2 Reproduces 100% of the time. I am logged in over remote desktop. Desktop snapshot provided in case you need to verify remote desktop.

Bug#743215: I managed to figure out what's going on

I managed to figure out what's going on. write tries to use standard output to determine who the currently logged on user is. Traditional cron would have stdout captured by a spool running as a user process; however systemd cron would have it captured by a systemd process running as root. While

Bug#781869: pulseaudio: sound total hangup when program using audio suddenly dies

On further diagnostic (now that we know for sure what we are looking for), the user logged in to TTY2 is relevant. This eliminates the statistical match for a hard match to the problem. If the TTY is logged in as root or not logged in, happens every time. I'm in the habit of doing all root work

Bug#781869: pulseaudio: sound total hangup when program using audio suddenly dies

Does pulseaudio get unstuck by doing another roundtrip to tty2? Nope. I haven't found an unstick short of X logoff-logon. Killing pulseaudio does not unstick. Will try pull from experimental soon. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of

Bug#781869: pulseaudio: sound total hangup when program using audio suddenly dies

The second is that when you return to X your sound does not return. This is the bug that needs debugging. Output of getfacl suggests permissions are being restored correctly but pulseaudio is getting stuck by the transient. Given the statistical nature of the inital repro with same user on TTY2,

Bug#781869: pulseaudio: sound total hangup when program using audio suddenly dies

Hmmm. systemd-logind is running. Maybe we got a genuine systemd bug that I tried to avoid by not upgrading init. If I just pin the ACLs at 666 (this is safe on a single-user box) and the problem goes away would that confirm where to look? I miss the old console group. A much simpler solution.

Bug#781869: pulseaudio: sound total hangup when program using audio suddenly dies

Indeed. chmod 666 /dev/snd/* makes the problem go away. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#781869: pulseaudio: sound total hangup when program using audio suddenly dies

s/Killing pulseaudio/Killing and restarting pulseaudio/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#779002: closed by Michael Gilbert mgilb...@debian.org (Re: [pkg-wine-party] Bug#779002: wine32 sound does not work on amd64 arch)

I'd advise setting a package recommends as having to file a bug to find it is pretty silly. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#226171: patch

Patch for this bug is in bug 770139. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#770139: Fwd: Bug#770139: lilo chokes on dmraid mirroring

patch provided. You probably don't want to apply it as-is as it breaks other scenarios. I had to completely remove devicemapper support to make it work. It would be completely feasible to make it a compile option and provide two binary packages, but that would be more of a headache. diff -ur

Bug#680649: linux-image-3.0.0-1-amd64: Restoring from hibernate requires rerunning /etc/init.d/kbd start manually to fix broken consoles

Package version changed from 3.0.0-2 to 3.0.0.3. Incidentally I have upgraded to 3.2 and the problem does not occur there either. On 7/7/12, Jonathan Nieder jrnie...@gmail.com wrote: Hi, Joshua Hudson wrote: For some reason a new apt-get dist-upgrade fixed bug. A new kernel package

Bug#647153:

Fix as provided in upstream does not work. I had to re-apply my original patch to work. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#680649: linux-image-3.0.0-1-amd64: Restoring from hibernate requires rerunning /etc/init.d/kbd start manually to fix broken consoles

For some reason a new apt-get dist-upgrade fixed bug. A new kernel package was pulled but the version did not change. On 7/7/12, Jonathan Nieder jrnie...@gmail.com wrote: Hi, Joshua wrote: Version: 3.0.0-3 [...] I have some nonstandard console settings that require running /etc/init.d/kbd

Bug#53829:

With the change to implement /run in squeezy, we can now move mtab off of /etc. Right now, the system is using /etc/mtab - /proc/mounts. I have found a number of tools (not all debian based) that do fopen(/etc/mtab, r+) and disregard /etc/mtab if it cannot be opened as a way of verifying the

Bug#53829:

Herein lies the problem. I managed to move the location to /run/mount/mtab by changing the value of the #define, but that leaves no upgrade path from /proc/mounts due to the bizarre code in /lib/init/mount-functions.sh. The migration should have happened at package install time, not boot time.

Bug#653959: citadel-server was auto-selected and installed by dist-upgrade for what appears to be no good reason, and apt-get purge citadel-server failed to back out all system changes

fiddling around with aptitude why revealed the haul-in is quilt recommending mail-transport-agent another case of the build tools requesting something absolute nonsense. Last time around it was postgres, this time it was citadel On 1/1/12, Joshua joshud...@gmail.com wrote: Source:

Bug#603885:

Appears to be fixed in iceweasel-6.0 -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#625801: Fwd: Bug#625801: apt should allow remove package depended on by metapackage and keep metapackage

-- Forwarded message -- From: Joshua Hudson joshud...@gmail.com Date: Sat, May 7, 2011 at 9:49 AM Subject: Re: Bug#625801: apt should allow remove package depended on by metapackage and keep metapackage To: Julian Andres Klode j...@debian.org Well this is rather brutal

Bug#603885: Fwd: Bug#603885: iceweasel: right click on link sometimes doesn't work

Bug is easily triggered by right-clicking on one of the initial links of the homepage before doing anything else (click the titlebar first if the window doesn't have focus). Bug does not occur in firefox 3.6 compiled fresh from Mozilla. -- To UNSUBSCRIBE, email to