[prev in list] [next in list] [prev in thread] [next in thread]
List: kde-core-devel
Subject: Re: Review Request: UDisks mount vfat filesystems with flush option.
From: Gary Greene <greeneg () tolharadys ! net>
Date: 2012-06-16 5:25:07
Message-ID: DD5D105E-91ED-4174-964D-CF9255776D83 () tolharadys ! net
[Download RAW message or body]
On Jun 13, 2012, at 2:32 PM, Konstantinos Smanis wrote:
> On Tue, Jun 12, 2012 at 11:33 AM, Maarten De Meyer
> <de.meyer.maarten@gmail.com> wrote:
> >
> > This is an automatically generated e-mail. To reply, visit:
> > http://git.reviewboard.kde.org/r/105224/
> >
> > Review request for kdelibs, Solid and Lukáš Tinkl.
> > By Maarten De Meyer.
> >
> > Description
> >
> > This patch mounts vfat filesystems (mostly usb devices) with the flush
> > option when using the UDisks backend. The hal backend already did this.
> > The flush option makes sure changes are written to the device immediately.
> >
> > Please let me know it this should be implemented in an other (optional)
> > way or if this should be in the UDev rules directly.
> >
> > ps. I do not have commit rights.
> >
> > Testing
> >
> > Compiled without error, mounted multiple devices without problems.
> >
> > Bugs: 273792
> >
> > Diffs
> >
> > solid/solid/backends/udisks/udisksstorageaccess.cpp (7d72a88)
> >
> > View Diff
>
> I tested the patch against 4.8.3, works as expected.
>
> However, I too think that this option should be implemented somewhere
> else in the stack (udisks/udev: whatever is most fit). Although a sane
> default, you should be able to configure it more easily.
I'm erring on the side that this smells of policy, which should NOT be hard coded, \
but rather site or host specified. If I'm a user, I might rather have my \
application's writer thread not block on write, but cache then write via the kernel, \
then so be it (yes, I understand the risk related if I remove the disk before the \
flush has occurred, but in this case, that's what umount is for, which calls flush \
before unmounting the filesystem.....
--
Gary L. Greene, Jr.
===============================================================
Developer and Project Lead for the AltimatOS open source project
See http://www.altimatos.com/ for more information
===============================================================
[Attachment #3 (unknown)]
<html><head></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; \
-webkit-line-break: after-white-space; "><div><div>On Jun 13, 2012, at 2:32 PM, \
Konstantinos Smanis wrote:</div><br class="Apple-interchange-newline"><blockquote \
type="cite"><div>On Tue, Jun 12, 2012 at 11:33 AM, Maarten De Meyer<br><<a \
href="mailto:de.meyer.maarten@gmail.com">de.meyer.maarten@gmail.com</a>> \
wrote:<br><blockquote type="cite"><br></blockquote><blockquote type="cite">This is an \
automatically generated e-mail. To reply, visit:<br></blockquote><blockquote \
type="cite"><a href="http://git.reviewboard.kde.org/r/105224/">http://git.reviewboard.kde.org/r/105224/</a><br></blockquote><blockquote \
type="cite"><br></blockquote><blockquote type="cite">Review request for kdelibs, \
Solid and Lukáš Tinkl.<br></blockquote><blockquote type="cite">By Maarten De \
Meyer.<br></blockquote><blockquote type="cite"><br></blockquote><blockquote \
type="cite">Description<br></blockquote><blockquote \
type="cite"><br></blockquote><blockquote type="cite">This patch mounts vfat \
filesystems (mostly usb devices) with the flush<br></blockquote><blockquote \
type="cite">option when using the UDisks backend. The hal backend already did \
this.<br></blockquote><blockquote type="cite">The flush option makes sure changes are \
written to the device immediately.<br></blockquote><blockquote \
type="cite"><br></blockquote><blockquote type="cite">Please let me know it this \
should be implemented in an other (optional)<br></blockquote><blockquote \
type="cite">way or if this should be in the UDev rules \
directly.<br></blockquote><blockquote type="cite"><br></blockquote><blockquote \
type="cite">ps. I do not have commit rights.<br></blockquote><blockquote \
type="cite"><br></blockquote><blockquote \
type="cite">Testing<br></blockquote><blockquote \
type="cite"><br></blockquote><blockquote type="cite">Compiled without error, mounted \
multiple devices without problems.<br></blockquote><blockquote \
type="cite"><br></blockquote><blockquote type="cite">Bugs: \
273792<br></blockquote><blockquote type="cite"><br></blockquote><blockquote \
type="cite">Diffs<br></blockquote><blockquote \
type="cite"><br></blockquote><blockquote \
type="cite">solid/solid/backends/udisks/udisksstorageaccess.cpp \
(7d72a88)<br></blockquote><blockquote type="cite"><br></blockquote><blockquote \
type="cite">View Diff<br></blockquote><br>I tested the patch against 4.8.3, works as \
expected.<br><br>However, I too think that this option should be implemented \
somewhere<br>else in the stack (udisks/udev: whatever is most fit). Although a \
sane<br>default, you should be able to configure it more \
easily.<br></div></blockquote></div><br><div>I'm erring on the side that this smells \
of policy, which should NOT be hard coded, but rather site or host specified. If I'm \
a user, I might rather have my application's writer thread not block on write, but \
cache then write via the kernel, then so be it (yes, I understand the risk related if \
I remove the disk before the flush has occurred, but in this case, that's what umount \
is for, which calls flush before unmounting the \
filesystem.....</div><div><br></div><div>--</div><div>Gary L. Greene, \
Jr.<br>===============================================================<br>Developer \
and Project Lead for the AltimatOS open source project<br>See <a \
href="http://www.altimatos.com/">http://www.altimatos.com/</a> for more \
information<br>===============================================================<br><br></div></body></html>
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic