When /dev/loop-control exists and *device is empty,
the mount may fail if a concurrent mount is running.
function old new delta
set_loop 809 807 -2
Signed-off-by: Xiaoming Ni <nixiaoming@huawei.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Starting with linux kernel v5.4 squashfs has a more strict parameter
checking implemented. Unlike util-linux mount, busybox never supported
the sizelimit option but simply forwards it to the kernel.
Since v5.4 mounting will fail with
squashfs: Unknown parameter 'sizelimit'
Support the sizelimit parameter by setting it in the LOOP_SET_STATUS64
structure before handing it to the kernel.
While at it also add support for the offset option, which currently will
always be set to 0.
function old new delta
cut_out_ull_opt - 167 +167
singlemount 1230 1266 +36
set_loop 834 862 +28
losetup_main 479 483 +4
------------------------------------------------------------------------------
(add/remove: 1/0 grow/shrink: 3/0 up/down: 235/0) Total: 235 bytes
Signed-off-by: Steffen Trumtrar <s.trumtrar@pengutronix.de>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
The "autolooped" mount (mount [-oloop] IMAGE /DIR/DIR)
always creates AUTOCLEARed loopdevs, so that umounting
drops them (and this does not require any code in the
umount userspace).
This happens since circa linux-2.6.25:
commit 96c5865559cee0f9cbc5173f3c949f6ce3525581
Date: Wed Feb 6 01:36:27 2008 -0800
Subject: Allow auto-destruction of loop devices
IOW: in this case, umount does not have to use -d
to drop the loopdev.
The explicit loop mount (mount /dev/loopN /DIR/DIR)
does not do this. In this case, umount without -d
should not drop loopdev.
Unfortunately, bbox umount currently always implies -d,
this probably needs fixing.
function old new delta
set_loop 537 597 +60
singlemount 1101 1138 +37
losetup_main 419 432 +13
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 3/0 up/down: 110/0) Total: 110 bytes
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
loopinfo.lo_file_name is not enough to uniquely identify a file on a system with
multiple mount namespaces. We could conceivably change this to dedup on
(lo_rdevice, lo_inode), but, as the comment above the deleted code notes, this
whole approach of reusing devices is racy anyway, so it seems better to stop
doing it entirely.
Signed-off-by: Kevin Wallace <k@igneous.io>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
This change retains "or later" state! No licensing _changes_ here,
only form is adjusted (article, space between "GPL" and "v2" and so on).
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
platform.h: fix wrong check for endianness, fix lchown
aliasing to chown on uclibc.
Code seems to not be affected in my testing.
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
things like xasprintf() into xfuncs.c, remove xprint_file_by_name() (it only
had one user), clean up lots of #includes... General cleanup pass. What I've
been doing for the last couple days.
And it conflicts! I've removed httpd.c from this checkin due to somebody else
touching that file. It builds for me. I have to catch a bus. (Now you know
why I'm looking forward to Mercurial.)
messages, C) can show the current association (if any) when called
with only one argument. Update the documentation a lot too.
Remind me to add a test suite for this thing. I think I've figured out
how to handle root-only testsuites...
appended to an executable that's being run (yes, I'm doing this) you get
EPERM, but mounting readonly fixes it. Doing the fallback all the time
shouldn't hurt, and is one less test.
things down a bit, fixed a number of funky corner cases, added support for
several new features (things like mount --move, mount --bind, lazy unounts,
automatic detection of loop mounts, and so on). Probably broke several
other things, but it's fixable. (Bang on it, tell me what doesn't work for
you...)
Note: you no longer need to say "-o loop". It does that for you when
necessary.
Still need to add "user mount" support, which involves making mount suid. Not
too hard to do under the new infrastructure, just haven't done it yet...
The previous code had the following notes, that belong in the version
control comments:
- * 3/21/1999 Charles P. Wright <cpwright@cpwright.com>
- * searches through fstab when -a is passed
- * will try mounting stuff with all fses when passed -t auto
- *
- * 1999-04-17 Dave Cinege...Rewrote -t auto. Fixed ro mtab.
- *
- * 1999-10-07 Erik Andersen <andersen@codepoet.org>.
- * Rewrite of a lot of code. Removed mtab usage (I plan on
- * putting it back as a compile-time option some time),
- * major adjustments to option parsing, and some serious
- * dieting all around.
- *
- * 1999-11-06 mtab support is back - andersee
- *
- * 2000-01-12 Ben Collins <bcollins@debian.org>, Borrowed utils-linux's
- * mount to add loop support.
- *
- * 2000-04-30 Dave Cinege <dcinege@psychosis.com>
- * Rewrote fstab while loop and lower mount section. Can now do
- * single mounts from fstab. Can override fstab options for single
- * mount. Common mount_one call for single mounts and 'all'. Fixed
- * mtab updating and stale entries. Removed 'remount' default.
- *
Is the change on libbb/loop.c which you commited in 2005/1/3 effective
really?
The __GLIBC__ macro and __UCLIBC__ macro are defined in
feature.h in glibc source, so the change may not be effective.
If you want to check this with __GLIBC__, feature.h header is needed.
Some architectures(e.g. PPC series) need to include linux/posix_types.h
in stead of asm/posix_types.h, so the patch which is attached with
this mail include <linux/posix_types.h>.
