Commit Graph

705 Commits

Author SHA1 Message Date
xtraeme 0cbb812de5 Add sysctl options for the syncer:
vfs.sync.delay: max time to delay syncing data
vfs.sync.filedelay: time to delay syncing files
vfs.sync.dirdelay: time to delay syncing directories
vfs.sync.metadelay: time to delay syncing metadata

Note that using a value of 0 is allowed, but it's not
recommended.
2005-09-11 17:55:56 +00:00
chs 0840b7949f in spec_ioctl(), don't dereference v_specinfo if it's NULL.
this is needed due to rev. 1.231 of kern/vfs_subr.c, which now sets
v_specinfo to NULL before changing the vnode's ops vector.
2005-09-11 14:18:54 +00:00
christos 7791a8f18b Also protect the ipsec ioctls from negative offsets to prevent panics
in m_copydata(). Pointed out by Karl Janmar. Move the negative offset
check from kernfs_xread() to kernfs_read().
2005-09-01 06:25:26 +00:00
christos 218f69d99f Don't allow negative offsets when reading the message buffer, because it
can allow reading arbitrary kernel memory.
2005-08-31 09:54:54 +00:00
xtraeme af97f2e875 Remove __P() 2005-08-30 20:08:01 +00:00
christos 50f8955b6e 64 bit inode changes. 2005-08-19 02:04:03 +00:00
yamt 79ff185ac4 don't write-protect wired pages. pointed by Chuck Silvers.
for now, leave a vnode on the syncer's queue, as suggested by him.
2005-07-26 08:06:29 +00:00
erh fbd6fe6c7f Provide a sysctl (vfs.layerfs.debug) to control verbose output when
LAYERFS_DIAGNOSTIC is turned on.
2005-07-24 17:33:24 +00:00
yamt b7bfe82866 update file timestamps for nfsd loaned-read and mmap.
PR/25279.  discussed on tech-kern@.
2005-07-23 12:18:41 +00:00
yamt 01f4919e33 genfs_putpages: don't bother to clean the vnode unless VONWORKLST. 2005-07-17 16:07:19 +00:00
yamt 8af42d8d3c ensure that vnodes with dirty pages are always on syncer's queue.
- genfs_putpages: wait for i/o completion of PG_RELEASED/PG_PAGEOUT pages by
  setting "wasclean" false when encountering them.
  suggested by Stephan Uphoff in PR/24596 (1).

- genfs_putpages: write protect pages when cleaning out, if
  we're going to take the vnode off the syncer's queue.
  uvm_fault: don't write-map pages unless its vnode is already on
  the syncer's queue.

  fix PR/24596 (3) but in the different way from the suggested fix.
  (to keep our current behaviour, ie. not to require explicit msync.
  discussed on tech-kern@.)

- genfs_putpages: don't mistakenly take a vnode off the queue
  by introducing a generation number in genfs_node.
  genfs_getpages: increment the generation number.
  suggested by Stephan Uphoff in PR/24596 (2).

- add some assertions.
2005-07-17 12:27:47 +00:00
yamt 2a6dc9d02d - introduce PGO_NOBLOCKALLOC and use it for ubc mapping
to prevent unnecessary block allocations in the case that
  page size > block size.

- ufs_balloc_range: use VM_PROT_WRITE+PGO_NOBLOCKALLOC rather than
  VM_PROT_READ.
2005-07-17 09:13:35 +00:00
yamt e9e22b28eb genfs_getpages: don't forget to put the vnode onto the syncer's work queue
even in the case of PGO_LOCKED.
2005-07-16 03:54:08 +00:00
yamt 44d128fa8e - constify genfs_ops.
- use member designators.
2005-06-28 09:30:37 +00:00
ws 9d78e0cf36 PR-30566: Poll must not return <sys/errno.h> values.
Start with those places I can easily test.
2005-06-21 14:01:11 +00:00
christos 1979e6e175 rename delay. 2005-05-30 22:13:50 +00:00
christos c107ef9edc - sprinkle const
- avoid shadowed variables.
2005-05-29 21:55:33 +00:00
chs 448875a34c kernfs does not support mmap(), remove code that pretends that it does. 2005-05-20 13:16:54 +00:00
christos 8f3566ce61 PR/29782: Martin Husemann: procfs can not unmount when some process has its
current directory in curproc. Fix from Pedro Martelletto:
We cannot call vgone() from procfs_inactive() if we are coming from
vclean(). that's what's probably causing the deadlock.
2005-04-02 06:15:09 +00:00
thorpej e633e8b61b - Define a VFS_ATTACH() macro that places a reference to a vfsops structure
into the "vfsops" link set.
- Use VFS_ATTACH() where vfsops are declared for individual file systems.
- In vfsinit(), traverse the "vfsops" link set, rather than vfs_list_initial[].
2005-03-29 02:41:05 +00:00
christos bb48399e9b Remove bogus len setting noted by J. Chapman Flack. 2005-03-01 04:39:59 +00:00
christos 1a63592a9b Give more space for cpu info and allocate it dynamically. 2005-02-27 22:29:50 +00:00
perry 477853c351 nuke trailing whitespace 2005-02-26 22:58:54 +00:00
chs d67b9b2ff2 undo the part of rev. 1.93 that turned the past-EOF check into an assertion.
read() can't request pages past EOF, but mmap() can.  apparently I had
disengaged the brain when I said that was ok.
2005-02-16 15:25:33 +00:00
wrstuden e384a44e9d Extend fsync_range(2) to support the FDISKSYNC flag, which requests
that the sync be propogated out through the disk drive caches.
2005-01-25 23:55:20 +00:00
drochner 7d0567768c -in the read-ahead code, avoid to issue read requests at/past EOF
-because noone should request reads past EOF, or writes past EOF which
 are not explicitely marked as file-extending (PGO_PASTEOF), turn
 a boundary check into a KASSERT
approved by Chuck Silvers
2005-01-25 09:50:31 +00:00
thorpej 1c95472d01 Add the system call and VFS infrastructure for file system extended
attributes.

From FreeBSD.
2005-01-02 16:08:28 +00:00
dbj 8962229d27 check for _KERNEL_OPT around opt include 2004-12-22 23:29:51 +00:00
christos 31c81b28f5 Cloning cleanup:
1. make fileops const
2. add 2 new negative errno's to `officially' support the cloning hack:
    - EDUPFD (used to overload ENODEV)
    - EMOVEFD (used to overload ENXIO)
3. Created an fdclone() function to encapsulate the operations needed for
   EMOVEFD, and made all cloners use it.
4. Centralize the local noop/badop fileops functions to:
   fnullop_fcntl, fnullop_poll, fnullop_kqfilter, fbadop_stat
2004-11-30 04:25:43 +00:00
atatat e23f0e2a34 Pass the caller's proc* to soreceive() via auio.uio_procp so that
unp_externalize() is called properly.

Addresses PR kern/28194.
2004-11-12 04:15:29 +00:00
jdolecek d5e68a44ca ptyfs moved from sys/miscfs/ to sys/fs/ 2004-11-11 18:57:21 +00:00
christos 0ecbe2ed07 Cosmetic; cleanup unused junk and fix comments. 2004-11-11 05:46:10 +00:00
christos 7fc8278a34 Add ptyfs. This is experimental. 2004-11-11 01:40:32 +00:00
skrll 89ce37fcdb Backout previous. 2004-10-27 06:55:53 +00:00
skrll 13d6d39771 Don't pass &proc0 in the UIO_SYSSPACE case it is not needed. 2004-10-27 06:51:06 +00:00
yamt 364ea008bb procfs_readdir:
- return correct cookie when buffer size is small.
- simplify logic.
2004-10-04 08:40:18 +00:00
yamt ef763f0a42 procfs_readdir: remove a redundant assignment. 2004-10-04 08:37:06 +00:00
enami d3482ced60 Backout previous; seeing many busy page on the pageq is normal. 2004-10-04 00:28:30 +00:00
enami b6d06fab27 So that not to leave pages busy unnecessarily, bound to specified region
when building cluster if we aren't pagedaemon and clean entire cluster
if we are pagedaemon.
2004-10-03 08:14:25 +00:00
enami a55995c148 Count obj pages freed by pagedaemon. 2004-10-03 07:59:02 +00:00
yamt d79c3679ea procfs_getattr: correct size of /proc/self. 2004-10-02 04:28:57 +00:00
yamt 269a1761b2 procfs_readdir:
- fix a locking problem, using proclist_foreach_call.  PR/27098.
- correct snprintf size argument.
2004-10-01 16:32:16 +00:00
yamt 3cca1d9e20 procfs_readdir: fix an offset handling bug after addition of /proc/self. 2004-10-01 14:09:55 +00:00
yamt f8f70a5eb4 procfs_readdir: use a list macro. 2004-10-01 14:09:14 +00:00
jdolecek 845beacce3 add 'mounts' file for -o linux, which lists all currently mounted
filesystems; Linux glibc statvfs() uses this to get some of mount flags,
and this file is also useful as /emul/linux/etc/mtab (via symlink)
2004-09-20 17:53:08 +00:00
skrll f7155e40f6 There's no need to pass a proc value when using UIO_SYSSPACE with
vn_rdwr(9) and uiomove(9).

OK'd by Jason Thorpe
2004-09-17 14:11:20 +00:00
jdolecek 2993272a0c set mp->mnt_stat.f_namemax on filesystem mount, for use by statvfs 2004-09-13 19:19:44 +00:00
skrll 685703c354 Do previous slightly differently - just pass a struct lwp * and derive the
struct proc *.

OK'd by Jaromir.
2004-08-27 07:02:45 +00:00
jdolecek b1126ead62 fix process used for /proc/<pid>/stat contents - it should be process
<pid>, not the current process looking at the information
2004-08-21 15:59:32 +00:00
mycroft 98a431ae8a Clean up reader/writer counts for the revoke case in fifo_close(). 2004-07-17 20:53:01 +00:00