linux-stable.git/fs/fat, branch v4.20

fat: truncate inode timestamp updates in setattr

2018-10-31T15:54:14+00:00

setattr_copy can't truncate timestamps correctly for
msdos/vfat, so truncate and copy them ourselves.

Link: http://lkml.kernel.org/r/a2b4701b1125573fafaeaae6802050ca86d6f8cc.1538363961.git.sorenson@redhat.com
Signed-off-by: Frank Sorenson 
Acked-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: change timestamp updates to use fat_truncate_time

2018-10-31T15:54:14+00:00

Convert the inode timestamp updates to use fat_truncate_time.

Link: http://lkml.kernel.org/r/2663d3083c4dd62f00b64612c8eaf5542bb05a4c.1538363961.git.sorenson@redhat.com
Signed-off-by: Frank Sorenson 
Acked-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: add functions to update and truncate timestamps appropriately

2018-10-31T15:54:14+00:00

Add the fat-specific inode_operation ->update_time() and
fat_truncate_time() function to truncate the inode timestamps from 1
nanosecond to the appropriate granularity.

Link: http://lkml.kernel.org/r/38af1ba3c3cf0d7381ce7b63077ef8af75901532.1538363961.git.sorenson@redhat.com
Signed-off-by: Frank Sorenson 
Acked-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: create a function to calculate the timezone offest

2018-10-31T15:54:14+00:00

Patch series "fat: timestamp updates", v5.

fat/msdos timestamps are stored on-disk with several different
granularities, some of them lower resolution than timespec64_trunc() can
provide.  In addition, they are only truncated as they are written to
disk, so the timestamps in-memory for new or modified files/directories
may be different from the same timestamps after a remount, as the
now-truncated times are re-read from the on-disk format.

These patches allow finer granularity for the timestamps where possible
and add fat-specific ->update_time inode operation and fat_truncate_time
functions to truncate each timestamp correctly, giving consistent times
across remounts.

This patch (of 4):

Move the calculation of the number of seconds in the timezone offset to a
common function.

Link: http://lkml.kernel.org/r/3671ff8cff5eeedbb85ebda5e4de0728920db4f6.1538363961.git.sorenson@redhat.com
Signed-off-by: Frank Sorenson 
Acked-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: expand a slightly out-of-date comment

2018-10-31T15:54:14+00:00

The file namei.c seems to have been renamed to namei_msdos.c, so I decided
to update the comment with the correct name, and expand it a bit to tell
the reader what to look for.

Link: http://lkml.kernel.org/r/20180928194947.23932-1-mihir@cs.utexas.edu
Signed-off-by: Mihir Mehta 
Acked-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fs/fat/fatent.c: add cond_resched() to fat_count_free_clusters()

2018-10-13T07:31:03+00:00

On non-preempt kernels this loop can take a long time (more than 50 ticks)
processing through entries.

Link: http://lkml.kernel.org/r/20181010172623.57033-1-khazhy@google.com
Signed-off-by: Khazhismel Kumykov 
Acked-by: OGAWA Hirofumi 
Reviewed-by: Andrew Morton 
Signed-off-by: Andrew Morton 
Signed-off-by: Greg Kroah-Hartman

fat: propagate 64-bit inode timestamps

2018-08-22T17:52:50+00:00

Now that we pass down 64-bit timestamps from VFS, we just need to convert
that correctly into on-disk timestamps.  To make that work correctly, this
changes the last use of time_to_tm() in the kernel to time64_to_tm(),
which also lets use remove that deprecated interfaces.

Similarly, the time_t use in fat_time_fat2unix() truncates the timestamp
on the way in, which can be avoided by using types that are wide enough to
hold the intermediate values during the conversion.

[hirofumi@mail.parknet.co.jp: remove useless temporary variable, needless long long]
Link: http://lkml.kernel.org/r/20180619153646.3637529-1-arnd@arndb.de
Signed-off-by: Arnd Bergmann 
Acked-by: OGAWA Hirofumi 
Cc: Jeff Layton 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: validate ->i_start before using

2018-08-22T17:52:50+00:00

On corrupted FATfs may have invalid ->i_start.  To handle it, this checks
->i_start before using, and return proper error code.

Link: http://lkml.kernel.org/r/87o9f8y1t5.fsf_-_@mail.parknet.co.jp
Signed-off-by: OGAWA Hirofumi 
Reported-by: Anatoly Trosinenko 
Tested-by: Anatoly Trosinenko 
Cc: Alan Cox 
Cc: Al Viro 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: add FITRIM ioctl for FAT file system

2018-08-22T17:52:50+00:00

Add FITRIM ioctl for FAT file system

[witallwang@gmail.com: use u64s]
  Link: http://lkml.kernel.org/r/87h8l37hub.fsf@mail.parknet.co.jp
[hirofumi@mail.parknet.co.jp: bug fixes, coding style fixes, add signal check]
Link: http://lkml.kernel.org/r/87fu10anhj.fsf@mail.parknet.co.jp
Signed-off-by: Wentao Wang 
Signed-off-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: fix memory allocation failure handling of match_strdup()

2018-07-21T19:50:46+00:00

In parse_options(), if match_strdup() failed, parse_options() leaves
opts->iocharset in unexpected state (i.e.  still pointing the freed
string).  And this can be the cause of double free.

To fix, this initialize opts->iocharset always when freeing.

Link: http://lkml.kernel.org/r/8736wp9dzc.fsf@mail.parknet.co.jp
Signed-off-by: OGAWA Hirofumi 
Reported-by: syzbot+90b8e10515ae88228a92@syzkaller.appspotmail.com
Cc: 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds