Commit Graph

297 Commits

Author SHA1 Message Date
kib
6a459eb27c Make the seek a method of the struct fileops.
Tested by:	pho
Sponsored by:	The FreeBSD Foundation
2013-08-21 17:36:01 +00:00
kib
408a640438 Restore the previous sendfile(2) behaviour on the block devices.
Provide valid .fo_sendfile method for several missed struct fileops.

Reviewed by:	glebius
Sponsored by:	The FreeBSD Foundation
2013-08-16 14:22:20 +00:00
avg
4b4c561bbf make path matching in devfs rules consistent and sane (and safer)
Before this change path matching had the following features:
- for device nodes the patterns were matched against full path
- in the above case '/' in a path could be matched by a wildcard
- for directories and links only the last component was matched

So, for example, a pattern like 're*' could match the following entries:
- re0 device
- responder/u0 device
- zvol/recpool directory

Although it was possible to work around this behavior (once it was spotted
and understood), it was very confusing and contrary to documentation.

Now we always match a full path for all types of devfs entries (devices,
directories, links) and a '/' has to be matched explicitly.
This behavior follows the shell globbing rules.

This change is originally developed by Jaakko Heinonen.
Many thanks!

PR:		kern/122838
Submitted by:	jh
MFC after:	4 weeks
2013-07-26 14:25:58 +00:00
gabor
b86fa940aa - Correct mispellings of the word necessary
Submitted by:	Christoph Mallon <christoph.mallon@gmx.de> (via private mail)
2013-04-17 11:42:40 +00:00
kib
92d95b8406 Stop translating the ERESTART error from the open(2) into EINTR.
Posix requires that open(2) is restartable for SA_RESTART.

For non-posix objects, in particular, devfs nodes, still disable
automatic restart of the opens. The open call to a driver could have
significant side effects for the hardware.

Noted and reviewed by:	jilles
Discussed with:	bde
MFC after:	2 weeks
2013-02-07 14:53:33 +00:00
kib
c6bad3bef7 Do not force a writer to the devfs file to drain the buffer writes.
Requested and tested by:	Ian Lepore <freebsd@damnhippie.dyndns.org>
MFC after:	2 weeks
2012-12-23 22:43:27 +00:00
kib
32941467f0 Remove M_USE_RESERVE from the devfs cdp allocator, which is one of two
uses of M_USE_RESERVE in the kernel. This allocation is not special.

Reviewed by:	alc
Tested by:	pho
MFC after:	2 weeks
2012-11-14 19:50:21 +00:00
attilio
d5d551ec46 Complete MPSAFE VFS interface and remove MNTK_MPSAFE flag.
Porters should refer to __FreeBSD_version 1000021 for this change as
it may have happened at the same timeframe.
2012-11-09 18:02:25 +00:00
ed
123cfec6ca Prefer __containerof() above member2struct().
The first does proper checking of the argument types, while the latter
does not.
2012-09-15 19:28:54 +00:00
hselasky
cd2aff7346 Streamline use of cdevpriv and correct some corner cases.
1) It is not useful to call "devfs_clear_cdevpriv()" from
"d_close" callbacks, hence for example read, write, ioctl and
so on might be sleeping at the time of "d_close" being called
and then then freed private data can still be accessed.
Examples: dtrace, linux_compat, ksyms (all fixed by this patch)

2) In sys/dev/drm* there are some cases in which memory will
be freed twice, if open fails, first by code in the open
routine, secondly by the cdevpriv destructor. Move registration
of the cdevpriv to the end of the drm open routines.

3) devfs_clear_cdevpriv() is not called if the "d_open" callback
registered cdevpriv data and the "d_open" callback function
returned an error. Fix this.

Discussed with:	phk
MFC after:	2 weeks
2012-08-15 16:19:39 +00:00
kib
53224f018a Extend the KPI to lock and unlock f_offset member of struct file. It
now fully encapsulates all accesses to f_offset, and extends f_offset
locking to other consumers that need it, in particular, to lseek() and
variants of getdirentries().

Ensure that on 32bit architectures f_offset, which is 64bit quantity,
always read and written under the mtxpool protection. This fixes
apparently easy to trigger race when parallel lseek()s or lseek() and
read/write could destroy file offset.

The already broken ABI emulations, including iBCS and SysV, are not
converted (yet).

Tested by:	pho
No objections from:	jhb
MFC after:    3 weeks
2012-07-02 21:01:03 +00:00
mav
08333340b6 Revert devfs part of r235911. I was unaware about old but unfinished
discussion between kib@ and gibbs@ about it.
2012-05-24 18:19:23 +00:00
mav
96f3e42ce2 MFprojects/zfsd:
Revamp the CAM enclosure services driver.
This updated driver uses an in-kernel daemon to track state changes and
publishes physical path location information\for disk elements into the
CAM device database.

Sponsored by:   Spectra Logic Corporation
Sponsored by:   iXsystems, Inc.
Submitted by:   gibbs, will, mav
2012-05-24 14:07:44 +00:00
mm
ba775c1d2c Add "export" to devfs_opts[] and return EOPNOTSUPP if called with it.
Fixes mountd warnings.

Reported by:	kib
MFC after:	1 week
2012-02-29 16:16:36 +00:00
mm
4825085ea4 To improve control over the use of mount(8) inside a jail(8), introduce
a new jail parameter node with the following parameters:

allow.mount.devfs:
	allow mounting the devfs filesystem inside a jail

allow.mount.nullfs:
	allow mounting the nullfs filesystem inside a jail

Both parameters are disabled by default (equals the behavior before
devfs and nullfs in jails). Administrators have to explicitly allow
mounting devfs and nullfs for each jail. The value "-1" of the
devfs_ruleset parameter is removed in favor of the new allow setting.

Reviewed by:	jamie
Suggested by:	pjd
MFC after:	2 weeks
2012-02-23 18:51:24 +00:00
kib
80ae8fe82c Fix found places where uio_resid is truncated to int.
Add the sysctl debug.iosize_max_clamp, enabled by default. Setting the
sysctl to zero allows to perform the SSIZE_MAX-sized i/o requests from
the usermode.

Discussed with:	bde, das (previous versions)
MFC after:	1 month
2012-02-21 01:05:12 +00:00
ed
2b4f7a9e8a Merge si_name and __si_namebuf.
The si_name pointer always points to the __si_namebuf member inside the
same object. Remove it and rename __si_namebuf to si_name.
2012-02-10 12:40:50 +00:00
mm
1626913ed1 Add support for mounting devfs inside jails.
A new jail(8) option "devfs_ruleset" defines the ruleset enforcement for
mounting devfs inside jails. A value of -1 disables mounting devfs in
jails, a value of zero means no restrictions. Nested jails can only
have mounting devfs disabled or inherit parent's enforcement as jails are
not allowed to view or manipulate devfs(8) rules.

Utilizes new functions introduced in r231265.

Reviewed by:	jamie
MFC after:	1 month
2012-02-09 10:22:08 +00:00
mm
27a58a0dde Introduce the "ruleset=number" option for devfs(5) mounts.
Add support for updating the devfs mount (currently only changing the
ruleset number is supported).
Check mnt_optnew with vfs_filteropt(9).

This new option sets the specified ruleset number as the active ruleset
of the new devfs mount and applies all its rules at mount time. If the
specified ruleset doesn't exist, a new empty ruleset is created.

MFC after:	1 month
2012-02-09 10:09:12 +00:00
jhb
b00203c514 Explicitly use curthread while manipulating td_fpop during last close
of a devfs file descriptor in devfs_close_f().  The passed in td argument
may be NULL if the close was invoked by garbage collection of open
file descriptors in pending control messages in the socket buffer of a
UNIX domain socket after it was closed.

PR:		kern/151758
Submitted by:	Andrey Shidakov  andrey shidakov ru
Submitted by:	Ruben van Staveren  ruben verweg com
Reviewed by:	kib
MFC after:	2 weeks
2011-12-09 17:49:34 +00:00
kib
36fd8d0106 Existing VOP_VPTOCNP() interface has a fatal flow that is critical for
nullfs.  The problem is that resulting vnode is only required to be
held on return from the successfull call to vop, instead of being
referenced.

Nullfs VOP_INACTIVE() method reclaims the vnode, which in combination
with the VOP_VPTOCNP() interface means that the directory vnode
returned from VOP_VPTOCNP() is reclaimed in advance, causing
vn_fullpath() to error with EBADF or like.

Change the interface for VOP_VPTOCNP(), now the dvp must be
referenced. Convert all in-tree implementations of VOP_VPTOCNP(),
which is trivial, because vhold(9) and vref(9) are similar in the
locking prerequisites. Out-of-tree fs implementation of VOP_VPTOCNP(),
if any, should have no trouble with the fix.

Tested by:	pho
Reviewed by:	mckusick
MFC after:	3 weeks (subject of re approval)
2011-11-19 07:50:49 +00:00
eadler
6efa0e7f22 - fix duplicate "a a" in some comments
Submitted by:	eadler
Approved by:	simon
MFC after:	3 days
2011-11-13 17:06:33 +00:00
jhb
1e2d8c9d67 Move the cleanup of f_cdevpriv when the reference count of a devfs
file descriptor drops to zero out of _fdrop() and into devfs_close_f()
as it is only relevant for devfs file descriptors.

Reviewed by:	kib
MFC after:	1 week
2011-11-04 03:39:31 +00:00
kib
5831d7597c Fix kernel panic when d_fdopen csw method is called for NULL fp.
This may happen when kernel consumer calls VOP_OPEN().

Reported by:	Tavis Ormandy <taviso  cmpxchg8b com> through delphij
MFC after:	3 days
2011-11-03 18:55:18 +00:00
kib
12b63c734f Export devfs inode number allocator for the kernel consumers.
Reviewed by:	jhb
MFC after:	2 weeks
2011-10-05 16:50:15 +00:00
kib
011f42054d Add the fo_chown and fo_chmod methods to struct fileops and use them
to implement fchown(2) and fchmod(2) support for several file types
that previously lacked it. Add MAC entries for chown/chmod done on
posix shared memory and (old) in-kernel posix semaphores.

Based on the submission by:	glebius
Reviewed by:	rwatson
Approved by:	re (bz)
2011-08-16 20:07:47 +00:00
kib
3f15a4a91b Do not update mountpoint generation counter to the value which was not
yet acted upon by devfs_populate().

Submitted by:	Kohji Okuno <okuno.kohji jp panasonic com>
Approved by:	re (bz)
MFC after:	1 week
2011-08-09 20:53:33 +00:00
kib
e3d33b853d While fixing the looping of a thread while devfs vnode is reclaimed,
r179247 introduced a possibility of devfs_allocv() returning spurious
ENOENT. If the vnode is selected by vnlru daemon for reclamation, then
devfs_allocv() can get ENOENT from vget() due to devfs_close() dropping
vnode lock around the call to cdevsw d_close method.

Use LK_RETRY in the vget() call, and do some part of the devfs_reclaim()
work in devfs_allocv(), clearing vp->v_data and de->de_vnode. Retry the
allocation of the vnode, now with de->de_vnode == NULL.

The check vp->v_data == NULL at the start of devfs_close() cannot be
affected by the change, since vnode lock must be held while VI_DOOMED
is set, and only dropped after the check.

Reported and tested by:	Kohji Okuno <okuno.kohji jp panasonic com>
Reviewed by:	attilio
MFC after:	3 weeks
2011-07-13 21:07:41 +00:00
jh
dd13f4b76e Don't allow user created symbolic links to cover another entries marked
with DE_USER. If a devfs rule hid such entry, it was possible to create
infinite number of symbolic links with the same name.

Reviewed by:	kib
2010-12-15 16:49:47 +00:00
jh
309885e54e - Assert that dm_lock is exclusively held in devfs_rules_apply() and
in devfs_vmkdir() while adding the entry to de_list of the parent.
- Apply devfs rules to newly created directories and symbolic links.

PR:		kern/125034
Submitted by:	Mateusz Guzik (original version)
2010-12-15 16:42:44 +00:00
jh
bb9ca91194 Handle the special ruleset 0 in devfs_ruleset_use(). An attempt set the
current ruleset to 0 with command "devfs ruleset 0" triggered a KASSERT
in devfs_ruleset_create().

PR:		kern/125030
Submitted by:	Mateusz Guzik
2010-12-12 08:52:13 +00:00
jh
7817d1d259 Format prototypes to follow style(9) more closely.
Discussed with:	kib, phk
2010-10-12 15:58:52 +00:00
jh
400a76c2ba Add a new function devfs_dev_exists() to be able to find out if a
specific devfs path already exists.

The function will be used from kern_conf.c to detect duplicate device
registrations. Callers must hold the devmtx mutex.

Reviewed by:	kib
2010-09-27 18:20:56 +00:00
jh
b6e78e30f2 Add reference counting for devfs paths containing user created symbolic
links. The reference counting is needed to be able to determine if a
specific devfs path exists. For true device file paths we can traverse
the cdevp_list but a separate directory list is needed for user created
symbolic links.

Add a new directory entry flag DE_USER to mark entries which should
unreference their parent directory on deletion.

A new function to traverse cdevp_list and the directory list will be
introduced in a separate commit.

Idea from:	kib
Reviewed by:	kib
2010-09-27 17:47:09 +00:00
jh
4c422ff8ec Modify devfs_fqpn() for future use in devfs path reference counting
code:

- Accept devfs_mount and devfs_dirent as the arguments instead of a
  vnode. This generalizes the function so that it can be used from
  contexts where vnode references are not available.
- Accept NULL cnp argument. No '/' will be appended, if a NULL cnp is
  provided.
- Make the function global and add its prototype to devfs.h.

Reviewed by:	kib
2010-09-21 16:49:02 +00:00
jh
8c48e7bab1 - For consistency, remove "." and ".." entries from de_dlist before
calling devfs_delete() (and thus possibly dropping dm_lock) in
  devfs_rmdir_empty().
- Assert that we don't return doomed entries from devfs_find(). [1]

Suggested by:	kib [1]
Reviewed by:	kib
2010-09-18 18:37:41 +00:00
jh
02ebb2e3c3 Remove empty devfs directories automatically.
devfs_delete() now recursively removes empty parent directories unless
the DEVFS_DEL_NORECURSE flag is specified. devfs_delete() can't be
called anymore with a parent directory vnode lock held because the
possible parent directory deletion needs to lock the vnode. Thus we
unlock the parent directory vnode in devfs_remove() before calling
devfs_delete().

Call devfs_populate_vp() from devfs_symlink() and devfs_vptocnp() as now
directories can get removed.

Add a check for DE_DOOMED flag to devfs_populate_vp() because
devfs_delete() drops dm_lock before the VI_DOOMED vnode flag gets set.
This ensures that devfs_populate_vp() returns an error for directories
which are in progress of deletion.

Reviewed by:	kib
Discussed on:	freebsd-current (mostly silence)
2010-09-15 14:23:55 +00:00
jh
52028f2ca8 Set de_dir for user created symbolic links. This will be needed to be
able to resolve their parent directories.
2010-08-26 16:01:29 +00:00
jh
4a4bf3c511 Call devfs_populate_vp() from devfs_getattr(). It was possible that
fstat(2) returned stale information through an open file descriptor.
2010-08-25 15:29:12 +00:00
jh
6471734221 Introduce and use devfs_populate_vp() to unlock a vnode before calling
devfs_populate(). This is a prerequisite for the automatic removal of
empty directories which will be committed in the future.

Reviewed by:	kib (previous version)
2010-08-22 16:08:12 +00:00
jhb
d4890c88b0 Add dedicated routines to toggle lockmgr flags such as LK_NOSHARE and
LK_CANRECURSE after a lock is created.  Use them to implement macros that
otherwise manipulated the flags directly.  Assert that the associated
lockmgr lock is exclusively locked by the current thread when manipulating
these flags to ensure the flag updates are safe.  This last change required
some minor shuffling in a few filesystems to exclusively lock a brand new
vnode slightly earlier.

Reviewed by:	kib
MFC after:	3 days
2010-08-20 19:46:50 +00:00
jh
61e0ffd171 Call dev_rel() in error paths.
Reported by:	kib
Reviewed by:	kib
MFC after:	2 weeks
2010-08-19 16:39:00 +00:00
jh
2e34ad384b Allow user created symbolic links to cover device files and directories
if the device file appears during or after the link creation.

User created symbolic links are now inserted at the head of the
directory entry list after the "." and ".." entries. A new directory
entry flag DE_COVERED indicates that an entry is covered by a symbolic
link.

PR:		kern/114057
Reviewed by:	kib
Idea from:	kib
Discussed on:	freebsd-current (mostly silence)
2010-08-12 15:29:07 +00:00
kib
d9e9a26839 Enable shared lookups and externed shared ops for devfs.
In collaboration with:	pho
MFC after:	1 month
2010-08-06 09:46:53 +00:00
kib
ba7ee96f4a Add new make_dev_p(9) flag MAKEDEV_ETERNAL to inform devfs that created
cdev will never be destroyed. Propagate the flag to devfs vnodes as
VV_ETERNVALDEV. Use the flags to avoid acquiring devmtx and taking a
thread reference on such nodes.

In collaboration with:	pho
MFC after:	1 month
2010-08-06 09:42:15 +00:00
kib
913fa4ba38 Enable shared locks for the devfs vnodes. Honor the locking mode
requested by lookup(). This should be a nop at the moment.

In collaboration with:	pho
MFC after:	1 month
2010-08-06 09:23:47 +00:00
kib
837d8df157 Initialize VV_ISTTY vnode flag on the devfs vnode creation instead of
doing it on each open.

In collaboration with:	pho
MFC after:	1 month
2010-08-06 09:06:55 +00:00
jh
bc3fa141fa Add a new function devfs_parent_dirent() for resolving devfs parent
directory entry. Use the new function in devfs_fqpn(), devfs_lookupx()
and devfs_vptocnp() instead of manually resolving the parent entry.

Reviewed by:	kib
2010-06-09 15:29:12 +00:00
jh
73a91f4cc7 Don't try to call cdevsw d_close() method when devfs_close() is called
because of insmntque1() failure.

Found with:	stress2
Suggested and reviewed by:	kib
2010-06-01 18:57:21 +00:00
kib
77dcee6926 Add MAKEDEV_NOWAIT flag to make_dev_credf(9), to create a device node
in a no-sleep context. If resource allocation cannot be done without
sleep, make_dev_credf() fails and returns NULL.

Reviewed by:	jh
MFC after:	2 weeks
2010-05-06 19:22:50 +00:00