[Bug 1763086] Re: xfs_logprint can't handle multiply-logged inode fields

Eric Desrochers eric.desrochers at canonical.com
Wed Apr 11 18:04:07 UTC 2018


** Description changed:

  ### DRAFT ####
  [Impact]
  
-  * Under certains condition (such as when selinux is enabled) xfsprogs
- on Trusty may report a error "illegal inode type" and SIGABRT,
- generating a coredump as follow :
+  * Under certains condition (such as when selinux is enabled and
+ probably other ways) xfsprogs on Trusty may report a error "illegal
+ inode type" and SIGABRT, generating a coredump as follow :
  
  sing host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
  Core was generated by `xfs_logprint -c /dev/mapper/image-glance'.
  Program terminated with signal SIGABRT, Aborted.
  
  [Test Case]
  
  1) Create a XFS device (dev/vdb1)
  2) apt-get dist-upgrade
  3) Remove apparmor
  4) Reboot
  5) Installed selinux
  6) Reboot
  7) Run xfs_logprint against /dev/vdb1
     - sudo xfs_logprint -c /dev/vdb1
  
  There is probably other ways but that is the one I used based on the git
  commit log:
  
  "I've tested this by a simple test such as creating one
  file on an selinux box, so that data+attr is set, and
  logprinting"
  
  [Regression Potential]
  
   * This is a rework of "xlog_print_trans_inode()" to handle more than
  one flag on f->ilf_fields if set in order to stop going in error and
  abort when facing the situation. The change has been there for a while
  now (Jan 2013). It never land in Trusty because the package it has was a
  copy from it's predecessor release, saucy.
  
+ As we speak, the switch() statement is doing a Binary AND between
+ ifl_fields & XFS_ILOG_NONCORE
+ 
+ switch (f->ilf_fields & XFS_ILOG_NONCORE) {
+ 
+ XFS_ILOG_NONCORE being the sum of a Binary OR for all the Inode changes
+ to log.
+ 
+ #define XFS_ILOG_NONCORE        (XFS_ILOG_DDATA | XFS_ILOG_DEXT | \
+                                  XFS_ILOG_DBROOT | XFS_ILOG_DEV | \
+                                  XFS_ILOG_UUID | XFS_ILOG_ADATA | \
+                                  XFS_ILOG_AEXT | XFS_ILOG_ABROOT)
+ 
+ When more than 1 flag is set in "ifl_fields". 
+ (Example took from the coredump in gdb "ilf_fields = 133") 
+ 
+ It cannot match any case statement flags, and has no choice but to use
+ default: (when none of the cases are true) and call xlog_panic().
+ 
+ The rework offer a better detection instead of a one size fits all, and
+ provide various switch() statement context and appropriate actions.
+ 
  [Other Info]
  
   * Upstream commit:
  https://git.kernel.org/pub/scm/fs/xfs/xfsprogs-dev.git/commit/?id=dda4129
  
   * Trusty AFFECTED ONLY.
-  
+ 
  $ git describe --contains dda4129
  v3.1.11~29
  
  $ rmadison xfsprogs
  ==> xfsprogs | 3.1.9ubuntu2        | trusty <==
   xfsprogs | 4.3.0+nmu1ubuntu1   | xenial
   xfsprogs | 4.3.0+nmu1ubuntu1.1 | xenial-updates
   xfsprogs | 4.9.0+nmu1ubuntu1   | artful
   xfsprogs | 4.9.0+nmu1ubuntu1   | bionic
  
  [Orig Description]
  It has been brought to my attention that the following :
  
  "
  The command 'xfs_logprint -c <DEVICE>' coredump on Trusty and display the error :
  xlog_print_trans_inode: illegal inode type
  "

-- 
You received this bug notification because you are a member of Ubuntu
Foundations Bugs, which is subscribed to xfsprogs in Ubuntu.
https://bugs.launchpad.net/bugs/1763086

Title:
  xfs_logprint  can't handle multiply-logged inode fields

Status in xfsprogs package in Ubuntu:
  Fix Released
Status in xfsprogs source package in Trusty:
  New

Bug description:
  ### DRAFT ####
  [Impact]

   * Under certains condition (such as when selinux is enabled and
  probably other ways) xfsprogs on Trusty may report a error "illegal
  inode type" and SIGABRT, generating a coredump as follow :

  sing host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
  Core was generated by `xfs_logprint -c /dev/mapper/image-glance'.
  Program terminated with signal SIGABRT, Aborted.

  [Test Case]

  1) Create a XFS device (dev/vdb1)
  2) apt-get dist-upgrade
  3) Remove apparmor
  4) Reboot
  5) Installed selinux
  6) Reboot
  7) Run xfs_logprint against /dev/vdb1
     - sudo xfs_logprint -c /dev/vdb1

  There is probably other ways but that is the one I used based on the
  git commit log:

  "I've tested this by a simple test such as creating one
  file on an selinux box, so that data+attr is set, and
  logprinting"

  [Regression Potential]

   * This is a rework of "xlog_print_trans_inode()" to handle more than
  one flag on f->ilf_fields if set in order to stop going in error and
  abort when facing the situation. The change has been there for a while
  now (Jan 2013). It never land in Trusty because the package it has was
  a copy from it's predecessor release, saucy.

  As we speak, the switch() statement is doing a Binary AND between
  ifl_fields & XFS_ILOG_NONCORE.

  switch (f->ilf_fields & XFS_ILOG_NONCORE) {

  XFS_ILOG_NONCORE being the sum of a Binary OR for all the Inode
  changes to log:

  #define XFS_ILOG_NONCORE        (XFS_ILOG_DDATA | XFS_ILOG_DEXT | \
                                   XFS_ILOG_DBROOT | XFS_ILOG_DEV | \
                                   XFS_ILOG_UUID | XFS_ILOG_ADATA | \
                                   XFS_ILOG_AEXT | XFS_ILOG_ABROOT)

  When more than 1 flag is set in "ifl_fields", it's value become out of
  range. (Example took from the coredump in gdb "ilf_fields = 133")

  ----------------------------------
  #define XFS_ILOG_DDATA  0x002   /* log i_df.if_data */
  #define XFS_ILOG_DEXT   0x004   /* log i_df.if_extents */
  #define XFS_ILOG_DBROOT 0x008   /* log i_df.i_broot */
  #define XFS_ILOG_DEV    0x010   /* log the dev field */
  #define XFS_ILOG_UUID   0x020   /* log the uuid field */
  #define XFS_ILOG_ADATA  0x040   /* log i_af.if_data */
  #define XFS_ILOG_AEXT   0x080   /* log i_af.if_extents */
  #define XFS_ILOG_ABROOT 0x100   /* log i_af.i_broot */

  ifl_field = 133 & XFS_ILOG_NONCORE= 510 = 132
  ----------------------------------

  and cannot match any case statement based on the inode flags value
  above and has no choice but to use default: statement because none of
  them are true and call xlog_panic().

  The rework offer a better detection instead of a one size fits all,
  and provide various switch() statement context and appropriate
  actions.

  It is also capable to handle multiply-logged inode fields.

  [Other Info]

   * Upstream commit:
  https://git.kernel.org/pub/scm/fs/xfs/xfsprogs-dev.git/commit/?id=dda4129

   * Trusty AFFECTED ONLY.

  $ git describe --contains dda4129
  v3.1.11~29

  $ rmadison xfsprogs
  ==> xfsprogs | 3.1.9ubuntu2        | trusty <==
   xfsprogs | 4.3.0+nmu1ubuntu1   | xenial
   xfsprogs | 4.3.0+nmu1ubuntu1.1 | xenial-updates
   xfsprogs | 4.9.0+nmu1ubuntu1   | artful
   xfsprogs | 4.9.0+nmu1ubuntu1   | bionic

  [Orig Description]
  It has been brought to my attention that the following :

  "
  The command 'xfs_logprint -c <DEVICE>' coredump on Trusty and display the error :
  xlog_print_trans_inode: illegal inode type
  "

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/xfsprogs/+bug/1763086/+subscriptions



More information about the foundations-bugs mailing list