Change logs for apport source package in Utopic

  • apport (2.14.7-0ubuntu8.5) utopic-security; urgency=medium
    
      * SECURITY UPDATE: When /proc/sys/fs/suid_dumpable is enabled, crashing a
        program that is suid root or not readable for the user would create
        root-owned core files in the current directory of that program.  Creating
        specially crafted core files in /etc/logrotate.d or similar could then
        lead to arbitrary code execution with root privileges.  Now core files do
        not get written for these kinds of programs, in accordance with the
        intention of core(5).
        Thanks to Sander Bos for discovering this issue!
        (CVE-2015-1324, LP: #1452239)
      * SECURITY UPDATE: When writing a core dump file for a crashed packaged
        program, don't close and reopen the .crash report file but just rewind and
        re-read it. This prevents the user from modifying the .crash report file
        while "apport" is running to inject data and creating crafted core dump
        files. In conjunction with the above vulnerability of writing core dump
        files to arbitrary directories this could be exploited to gain root
        privileges.
        Thanks to Philip Pettersson for discovering this issue!
        (CVE-2015-1325, LP: #1453900)
      * test_signal_crashes(): Drop hardcoded /tmp/ path in do_crash(),
        test_nonwritable_cwd() uses a different dir.
    
     -- Martin Pitt <email address hidden>  Wed, 13 May 2015 11:59:03 +0200
  • apport (2.14.7-0ubuntu8.4) utopic-security; urgency=medium
    
      * SECURITY UPDATE: insecure /proc/net/unix parsing (LP: #1444518)
        - data/apport: temporarily disable container support until it can be
          re-written in a secure manner.
        - CVE number pending
     -- Marc Deslauriers <email address hidden>   Thu, 16 Apr 2015 07:40:49 -0400
  • apport (2.14.7-0ubuntu8.3) utopic-security; urgency=medium
    
      * SECURITY UPDATE: privilege escalation through namespaces and crafted
        chroot (LP: #1438345)
        - data/apport: If crash comes from a container, rather than
          chrooting into it, detect what LXC container it is and then use the
          attach_wait API call to execute apport in the container.
        - data/apport: Don't fail when encountering unicode characters.
          (Thanks to Martin Pitt)
        - test/test_signal_crashes.py: Test for the unicode fix.
          (Thanks to Martin Pitt)
        - CVE-2015-1318
     -- Stephane Graber <email address hidden>   Wed, 08 Apr 2015 13:16:32 -0400
  • apport (2.14.7-0ubuntu8.2) utopic-proposed; urgency=medium
    
      * apport/ui.py: Only provide a UI to hooks if the crash db will accept the
        report. This avoids asking questions if the report is merely sent to
        whoopsie for Ubuntu stable releases. (LP: #1084979)
     -- Brian Murray <email address hidden>   Thu, 05 Feb 2015 09:39:32 -0800
  • apport (2.14.7-0ubuntu8.1) utopic-proposed; urgency=medium
    
      * hookutils, attach_root_command_outputs(): Fix UnicodeDecodeError crash for
        non-textual values. Cherry-picked from trunk r2890. (LP: #1370259)
     -- Martin Pitt <email address hidden>   Thu, 08 Jan 2015 11:23:53 +0100
  • apport (2.14.7-0ubuntu8) utopic; urgency=medium
    
      * etc/init.d/apport: Stop setting $PATH in the init.d script. It breaks
        assumptions from /lib/lsb/init-functions.d/ which might call other tools
        which are not in /bin; also, we generally shouldn't meddle with $PATH in
        individual scripts.  (LP: #1372665)
     -- Martin Pitt <email address hidden>   Mon, 20 Oct 2014 14:51:41 -0400
  • apport (2.14.7-0ubuntu7) utopic; urgency=medium
    
      * Disable Launchpad crash upload for final Ubuntu 14.10.
     -- Martin Pitt <email address hidden>   Wed, 15 Oct 2014 15:37:02 +0200
  • apport (2.14.7-0ubuntu6) utopic; urgency=medium
    
      * data/recoverable_problem: Fix a Traceback with how the ValueError was
        matched. (LP: #1345569)
     -- Brian Murray <email address hidden>   Mon, 13 Oct 2014 10:02:21 -0700
  • apport (2.14.7-0ubuntu5) utopic; urgency=medium
    
      [ Evan Dandrea]
      * data/recoverable_problem: Handle the parent process going away while
        we're attempting to read from proc. (LP: #1345569)
     -- Brian Murray <email address hidden>   Tue, 07 Oct 2014 15:26:30 -0700
  • apport (2.14.7-0ubuntu3) utopic; urgency=medium
    
      [ Steve Langasek ]
      * Refactor apport-noui/whoopsie-upload-all to behave more reliably in
        case of overlapping crash processing (LP: #1354318):
        - debian/apport-noui.upstart: refactor to make this an 'instance' job
          for each incoming .crash file, and drop the racy handling of non-root
          .crash files (as well as the unnecessary 'env MATCH' line).
        - data/whoopsie-upload-all: refactor report processing to ensure that
          whoopsie-upload-all can be called multiple times in parallel without
          causing any .crash file to be processed more than once.
        - data/whoopsie-upload-all: handle setting ownership of files in
          process_report() instead of relying on this script being called by a
          particular user.
        - data/whoopsie-upload-all: don't spin in wait_uploaded() watching for
          .uploaded files if the corresponding .upload file has been removed out
          from under us.
        - data/whoopsie-upload-all: by default, return immediately instead of
          waiting to see if whoopsie processes all of the crashes.
    
      [ Brian Murray ]
      * data/whoopsie-upload-all: indicate that all reports have been uploaded
        even those that were marked for upload earlier.
     -- Brian Murray <email address hidden>   Thu, 02 Oct 2014 08:33:49 -0700
  • apport (2.14.7-0ubuntu2) utopic; urgency=medium
    
      [ Brian Murray ]
      * data/general-hooks/ubuntu.py: check to see if system-image-cli is
        available rather than if /etc/system-image/client.ini exists.
    
      [ Martin Pitt ]
      * ui.py: Robustify check if apport-retrace is installed. This brings back
        the "Examine locally" UI option. (LP: #1358734) (Cherry-picked from
        trunk).
     -- Martin Pitt <email address hidden>   Thu, 04 Sep 2014 12:30:27 +0200
  • apport (2.14.7-0ubuntu1) utopic; urgency=medium
    
      * New upstream release:
        - Fix interpretation of core dump ulimits: they are in bytes, not KiB.
          (LP: #1346497)
        - apport-retrace: Don't require specifying an operation; default to
          updating the given .crash file. (LP: #1361242)
        - Write report even on UnreportableReasons, so that whoopsie can still
          upload them. (LP: #1360417)
        - apt/dpkg install_packages(): Write a "packages.txt" into sandbox root
          directory to keep track of installed packages and their versions.
          Prerequisite for LP #1352591.
        - apt/dpkg install_packages(): Avoid re-downloading/installing packages
          which are already installed into a permanent sandbox. Prerequisite for
          LP #1352591.
        - sandboxutils.py, make_sandbox(): Drop check for already existing files
          when dynamically resolving libraries and ExecutablePaths; with that,
          these packages would never get updated in a permanent sandbox. The new
          version tracking from above now takes care of that. (LP: #1352591)
        - Fix report.test_add_proc_info test to stop assuming that pid 1 is named
          "init", as one can specify a different one on the kernel command line.
        - report.py, add_gdb_info(): Check for truncated core dumps, and set
          UnreportableReason and raise an IOError on them. Handle this in
          apport-retrace and whoopsie-upload-all to fail properly instead of
          silently producing broken Stacktraces. (LP: #1354571)
     -- Martin Pitt <email address hidden>   Fri, 29 Aug 2014 12:38:54 +0200
  • apport (2.14.6-0ubuntu2) utopic; urgency=medium
    
      * Cherry-pick from trunk: Adjust backend_apt_dpkg.test_get_file_package()
        test: newer util-linux versions do not provide /etc/blkid.tab any more.
     -- Martin Pitt <email address hidden>   Tue, 19 Aug 2014 10:38:44 +0200
  • apport (2.14.6-0ubuntu1) utopic; urgency=medium
    
      * New upstream bug fix release:
        - general-hooks/generic.py: Fix hook crash if there are non-distro
          libraries and no user interface.
        - collect_info(): Don't assume that reports have a ProblemType. Fixes
          crashes with apport-collect. (LP: #1325729)
        - apport-retrace: Declare -s/-g/-o as mutually exclusive, to get proper
          error messages instead of silent misbehaviour. (LP: #1352450)
        - apport-gtk: Drop usage of deprecated GTK stock items. (LP: #1348116)
     -- Martin Pitt <email address hidden>   Mon, 18 Aug 2014 07:57:16 +0200
  • apport (2.14.5-0ubuntu4) utopic; urgency=medium
    
      [ Steve Langasek ]
      * Fix invalid shebang lines for apport-noui maintainer scripts which I
        somehow overlooked.
    
      [ Martin Pitt ]
      * Add data/general-hooks/powerpc.py: Collect some PowerPC[64] information.
        Thanks to Thierry FAUCK! (LP: #1336462)
     -- Martin Pitt <email address hidden>   Fri, 08 Aug 2014 15:13:16 +0200
  • apport (2.14.5-0ubuntu3) utopic; urgency=medium
    
      * apport-noui: make the package installation automatically enable
        autosubmission, which doesn't currently work on the phone without it;
        and update the package description accordingly.  LP: #1351137.
     -- Steve Langasek <email address hidden>   Fri, 01 Aug 2014 15:06:49 -0700
  • apport (2.14.5-0ubuntu2) utopic; urgency=medium
    
      * data/general-hooks/ubuntu.py: Check for /etc/system-image/client.ini
        instead of /var/log/system-image/client.log since the latter is only
        readable by root.
     -- Brian Murray <email address hidden>   Tue, 29 Jul 2014 13:20:54 -0700
  • apport (2.14.5-0ubuntu1) utopic; urgency=medium
    
      * New upstream bug fix release. Changes since last snapshot:
        - apt/dpkg get_file_package(): If checking for uninstalled packages, don't
          use package information from the host system, but always look it up in
          the correct indexes. Otherwise this returns wrong results when files
          move to a different package between releases. Thanks Brian Murray!
          (LP: #1336062)
        - apt/dpkg install_packages(): Disable fetching apt translation indexes,
          to save some bandwidth.
        - whoopsie-upload-all: Ignore crash reports with incomplete core dumps
          instead of crashing on them. Thanks Brian Murray. (LP: #1329520)
        - etc/default/apport: Fix comment to use "service" instead of calling the
          init.d script directly.
        - whoopsie-upload-all: Collect hooks information to gather ApportVersion,
          NonfreeKernelModules, and SystemImageInfo. Do this before collecting
          package data to minimize hook run time. (LP: #1349579)
        - Adjust report.test_get_logind_session test to work with systemd >= 205.
        - Fix report.test_check_interpreted_twistd test to skip instead of fail if
          twisted is not installed.
    
      * Drop automatix.py general hook. Automatix is long gone.
      * etc/apport/blacklist.d/apport: Drop long obsolete npviewer/npplayer
        entries.
      * Drop unnecessary python-gi build dependency.
      * Drop python-twisted-core build dependency. It's just used for one test,
        and will still run in autopkgtest. (Avoids current twisted
        uninstallability in main)
     -- Martin Pitt <email address hidden>   Tue, 29 Jul 2014 12:26:23 +0200
  • apport (2.14.4-0ubuntu3) utopic; urgency=medium
    
      * In apport-kde recommend gdb-minimal before gdb
        LP: #1347565 "apport recommends gdb"
     -- Jonathan Riddell <email address hidden>   Wed, 23 Jul 2014 12:35:16 +0200
  • apport (2.14.4-0ubuntu2) utopic; urgency=medium
    
      * Merge from trunk:
        - apport-bug: Stop checking the autoreport flag and calling
          whoopsie-upload-all; these two are different tasks, and that breaks bug
          reporting. (LP: #1339663)
     -- Martin Pitt <email address hidden>   Thu, 10 Jul 2014 17:18:33 +0200
  • apport (2.14.4-0ubuntu1) utopic; urgency=medium
    
      * New upstream bug fix release:
        - Adjust code to match latest pep8 checker.
        - Report.crash_signature_addresses(): Drop kernel architecture from
          StacktraceAddressSignature field. It isn't useful there (at most the ELF
          architecture, but we don't really need that either). This makes it
          easier to regenerate broken signatures from existing reports on
          different architectures. (LP: #1336565)
      * Fix PEP-8 errors in Ubuntu hooks.
     -- Martin Pitt <email address hidden>   Thu, 03 Jul 2014 06:50:06 +0200
  • apport (2.14.3-0ubuntu2) utopic; urgency=medium
    
      * Turn on Launchpad crash submission for the devel series.
     -- Martin Pitt <email address hidden>   Wed, 18 Jun 2014 15:40:26 +0200
  • apport (2.14.3-0ubuntu1) utopic; urgency=medium
    
      * New upstream bug fix release:
        - Add kernel package version to the various kernel-related hooks. Thanks
          Brian Murray. (LP: #1316845)
        - Use package name in duplicate signature for recoverable problems. Thanks
          Brian Murray. (LP: #1316763)
        - Have whoopsie-upload-all upload recoverable problems. Thanks Brian
          Murray.  (LP: #1319099)
     -- Martin Pitt <email address hidden>   Fri, 30 May 2014 15:35:53 +0200
  • apport (2.14.2-0ubuntu4) utopic; urgency=medium
    
      * data/package-hooks/source_linux.py: create a duplicate signature for
        suspend resume failures. (LP: #1316841)
     -- Brian Murray <email address hidden>   Fri, 16 May 2014 09:46:51 -0700
  • apport (2.14.2-0ubuntu3) utopic; urgency=medium
    
      * debian/apport-noui.upstart: remove early exit (LP: #1235436)
      * debian/apport-noui.dirs: create /var/lib/apport (LP: #1235436)
     -- Brian Murray <email address hidden>   Wed, 14 May 2014 12:26:39 -0700
  • apport (2.14.2-0ubuntu2) utopic; urgency=medium
    
      * Merge "backend_apt_dpkg.test_install_packages_versioned: Fix namespacing
        of ver" fix from trunk, to better debug the autopkgtest failure on i386.
     -- Martin Pitt <email address hidden>   Fri, 02 May 2014 17:30:39 +0200
  • apport (2.14.2-0ubuntu1) utopic; urgency=medium
    
      * New upstream release:
        - Move error handling for invalid .crash files into collect_info(), so
          that it also applies when using the "Show Details..." button in the UI.
          Otherwise the UI just hangs eternally at this point when encountering
          broken core dumps. (LP: #1282349)
        - apt/dpkg install_packages(): Try to install the requested package
          version instead of always picking the most recent one. This should
          improve retracing results for older crash reports. Thanks to Brian
          Murray for inspiring this.
        - sandboxutils.py, make_sandbox(): When determining packages to install
          from ProcMaps, look up and use the package versions from the reporter,
          to improve retracing results. Thanks Brian Murray for the initial patch!
        - iwlwifi_error_dump: Make reports private, and subscribe
          canonical-kernel-team.  Thanks Seth Forshee. (LP: #1313818)
        - signal_crashes test: Time out after 5 seconds if the test process does
          not terminate on the specified signal, to avoid eternal hangs.
        - signal_crashes test: Ensure that we don't inherit an ignored SIGQUIT
          from the caller.
      * Switch Vcs-* to utopic branch.
      * Add build-essential test dependency, some tests call gcc.
      * Install iwl_wifi_error_dump and corresponding udev rules. (Second part of
        LP: #1313818)
     -- Martin Pitt <email address hidden>   Wed, 30 Apr 2014 14:07:17 +0200
  • apport (2.14.1-0ubuntu3) trusty; urgency=medium
    
      * Cherry-pick from trunk: Delay the import of the glob and re modules in the
        python apport hook, and only import them when needed. Speeds up
        interpreter startup time by 50%. (LP: #1307684)
     -- Matthias Klose <email address hidden>   Tue, 15 Apr 2014 08:42:00 +0200