http://bugzilla.opensuse.org/show_bug.cgi?id=1106214 http://bugzilla.opensuse.org/show_bug.cgi?id=1106214#c3 Iakov Karpov changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |srid@rkmail.ru --- Comment #3 from Iakov Karpov --- Happens with Leap 15.1 Beta. fstrim fails to trim my read-only windows partition (Why would it try to do it in first place?) -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1106214 http://bugzilla.opensuse.org/show_bug.cgi?id=1106214#c5 Stanislav Brabec changed: What |Removed |Added ---------------------------------------------------------------------------- Status|CONFIRMED |IN_PROGRESS --- Comment #5 from Stanislav Brabec --- Behavior of fstrim on read-only partitions is filesystem specific. For example, btrfs has no problem with trimming on read-only mount, as long as the device itself is R/W. But skipping R/O volumes is safe and logical behavior. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1106214 http://bugzilla.opensuse.org/show_bug.cgi?id=1106214#c7 --- Comment #7 from Stanislav Brabec --- According to the upstream, the bug is primarily problem of the NTFS driver. There are several related fixes in util-linux: fstrim: check for read-only devices on -a/-A https://git.kernel.org/pub/scm/utils/util-linux/util-linux.git/commit/?id=fe... and introduces new option --quiet My fix proposes complete skipping of read-only devices in fstrim -a/-A. Sent to the upstream: bug 1127701: https://www.spinics.net/lists/util-linux-ng/msg15752.html bug 1127701: https://www.spinics.net/lists/util-linux-ng/msg15751.html https://www.spinics.net/lists/util-linux-ng/msg15753.html And accepted: bug 1127701: https://git.kernel.org/pub/scm/utils/util-linux/util-linux.git/commit/?id=40... bug 1127701: https://git.kernel.org/pub/scm/utils/util-linux/util-linux.git/commit/?id=40... https://git.kernel.org/pub/scm/utils/util-linux/util-linux.git/commit/?id=2d... Note that my fix is different than the fe7d63105d923d43c9f478acded0ac888463e456 above: The first one checks for ro mount options, my one checks for writability of the target. My one can skip temporary read-only remounts. Iakov Karpov, comment 3: This behavior is intentional. Regular TRIM should increase performance of SSD discs by pre-erasing of unused blocks. It is safe for rotating devices, as the command is ignored there. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1106214 http://bugzilla.opensuse.org/show_bug.cgi?id=1106214#c9 Stanislav Brabec changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jack@suse.com Flags| |needinfo?(jack@suse.com) --- Comment #9 from Stanislav Brabec --- Iakov Karpov, comment 8: I agree. Trimming the read-only partition sounds wrong, and makes no sense. I assumed that your "Why would it try" refers to the trimming of the partition itself. Surprisingly, many kernel drivers have no problem with trimming of read-only mounts. To Jan Kára: Do you think that kernel should implement EROFS error when trying to FITRIM of read-only file system? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1106214 http://bugzilla.opensuse.org/show_bug.cgi?id=1106214#c16 Stanislav Brabec changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(srid@rkmail.ru) | --- Comment #16 from Stanislav Brabec --- Fixed for: SLE15SP1 SLE15 Leap15.1 Leap15.0 SLE12SP5 SLE12SP4 SLE12SP3 SLE12SP2 Fixed out of the box for: Factory Ignored for: SLE12SP1 SLE12 -- You are receiving this mail because: You are on the CC list for the bug.