Commit Graph

579 Commits

Author SHA1 Message Date
Kirk McKusick
57a91f6fb0 During fastpath processing for removal of a short-lived inode, the
set of restrictions for cancelling an inode dependency (inodedep)
is somewhat stronger than originally coded. Since this check appears
in two places, we codify it into the function check_inode_unwritten
which we then call from the two sites, one freeing blocks and the
other freeing directory entries.

Submitted by:	Steinar Haug via Matthew Dillon
2000-01-18 01:33:05 +00:00
Kirk McKusick
4c6adb0622 Need to reorganize the flushing of directory entry (pagedep) dependencies
so that they never try to lock an inode corresponding to ".." as this
can lead to deadlock. We observe that any inode with an updated link count
is always pushed into its buffer at the time of the link count change, so
we do not need to do a VOP_UPDATE, but merely find its buffer and write it.
The only time we need to get the inode itself is from the result of a
mkdir whose name will never be ".." and hence locking such an inode will
never request a lock above us in the filesystem tree. Thanks to Brian
Fundakowski Feldman for providing the test program that tickled soft updates
into hanging in "inode" sleep.

Submitted by:	Brian Fundakowski Feldman <green@FreeBSD.org>
2000-01-18 01:30:03 +00:00
Kirk McKusick
105ef72c55 Better bounding on softdep_flushfiles; other minor tweeks to checks. 2000-01-17 06:35:11 +00:00
Kirk McKusick
107d5039ef Must track multiple uncommitted renames until one ultimately gets
committed to disk or is removed.
2000-01-17 06:28:18 +00:00
Matthew Dillon
173cce7c8e Non-operational change, fix compiler warning.
Reviewed by:  mckusick
2000-01-14 04:39:28 +00:00
Kirk McKusick
d7127837a2 Confirming Peter's fix (locking 101: release the lock before you go
to sleep). Locking 101, part 2: do not look at buffer contents after
you have been asleep. There is no telling what wonderous changes may
have occurred.
2000-01-13 20:03:22 +00:00
Peter Wemm
7f473504e6 Free the global softupdates lock prior to tsleep() in getdirtybuf().
This seems to be responsible for a bunch of panics where the process
sleeps and something else finds softupdates "locked" when it shouldn't
be.  This commit is unreviewed, but has been a big help here.
Previously my boxes would panic pretty much on the first fsync() that
wrote something to disk.
2000-01-13 18:48:12 +00:00
Kirk McKusick
1c2ceb2880 Because cylinder group blocks are now written in background,
it is no longer sufficient to get a lock on a buffer to know
that its write has been completed. We have to first get the
lock on the buffer, then check to see if it is doing a
background write. If it is doing background write, we have
to wait for the background write to finish, then check to see
if that fullfilled our dependency, and if not to start another
write. Luckily the explanation is longer than the fix.
2000-01-13 07:20:01 +00:00
Kirk McKusick
94313add1f A panic occurs during an fsync when a dirty block associated with
a vnode has not been written (which would clear certain of its
dependencies). The problems arises because fsync with MNT_NOWAIT
no longer pushes all the dirty blocks associated with a vnode. It
skips those that require rollbacks, since they will just get instantly
dirty again. Such skipped blocks are marked so that they will not be
skipped a second time (otherwise circular dependencies would never
clear). So, we fsync twice to ensure that everything will be written
at least once.
2000-01-13 07:17:39 +00:00
Kirk McKusick
4ed62fbd7f The only known cause of this panic is running out of disk space.
The problem occurs when an indirect block and a data block are
being allocated at the same time. For example when the 13th block
of the file is written, the filesystem needs to allocate the first
indirect block and a data block. If the indirect block allocation
succeeds, but the data block allocation fails, the error code
dellocates the indirect block as it has nothing at which to point.
Unfortunately, it does not deallocate the indirect block's associated
dependencies which then fail when they find the block unexpectedly
gone (ptr == 0 instead of its expected value). The fix is to fsync
the file before doing the block rollback, as the fsync will flush
out all of the dependencies. Once the rollback is done the file
must be fsync'ed again so that the soft updates code does not find
unexpected changes. This approach is much slower than writing the
code to back out the extraneous dependencies, but running out of
disk space is not expected to be a common occurence, so just getting
it right is the main criterion.

PR:		kern/15063
Submitted by:	Assar Westerlund <assar@stacken.kth.se>
2000-01-11 08:27:00 +00:00
Kirk McKusick
10767f840b We cannot proceed to free the blocks of the file until the dependencies
have been cleaned up by deallocte_dependencies(). Once that is done, it
is safe to post the request to free the blocks. A similar change is also
needed for the freefile case.
2000-01-11 06:52:35 +00:00
Poul-Henning Kamp
ba4ad1fcea Give vn_isdisk() a second argument where it can return a suitable errno.
Suggested by:	bde
2000-01-10 12:04:27 +00:00
Kirk McKusick
26e5527c86 Missing FREE_LOCK call before handle_workitem_freeblocks.
Submitted by:	"Kenneth D. Merry" <ken@kdm.org>
2000-01-10 08:39:03 +00:00
Kirk McKusick
cf60e8e4bf Several performance improvements for soft updates have been added:
1) Fastpath deletions. When a file is being deleted, check to see if it
   was so recently created that its inode has not yet been written to
   disk. If so, the delete can proceed to immediately free the inode.
2) Background writes: No file or block allocations can be done while the
   bitmap is being written to disk. To avoid these stalls, the bitmap is
   copied to another buffer which is written thus leaving the original
   available for futher allocations.
3) Link count tracking. Constantly track the difference in i_effnlink and
   i_nlink so that inodes that have had no change other than i_effnlink
   need not be written.
4) Identify buffers with rollback dependencies so that the buffer flushing
   daemon can choose to skip over them.
2000-01-10 00:24:24 +00:00
Kirk McKusick
f0f7d38386 Keep tighter control of removal dependencies by limiting the number
of dirrem structure rather than the collaterally created freeblks
and freefile structures. Limit the rate of buffer dirtying by the
syncer process during periods of intense file removal.
2000-01-09 23:35:38 +00:00
Kirk McKusick
3f5b28bc07 Reorganize softdep_fsync so that it only does the inode-is-flushed
check before the inode is unlocked while grabbing its parent directory.
Once it is unlocked, other operations may slip in that could make
the inode-is-flushed check fail. Allowing other writes to the inode
before returning from fsync does not break the semantics of fsync
since we have flushed everything that was dirty at the time of the
fsync call.
2000-01-09 23:14:57 +00:00
Kirk McKusick
e2dc60835d Get rid of unreferenced function. 2000-01-09 22:42:42 +00:00
Kirk McKusick
83aaf63ab2 Make static non-exported functions from soft updates. 2000-01-09 22:40:09 +00:00
Peter Wemm
c447342094 Change #ifdef KERNEL to #ifdef _KERNEL in the public headers. "KERNEL"
is an application space macro and the applications are supposed to be free
to use it as they please (but cannot).  This is consistant with the other
BSD's who made this change quite some time ago.  More commits to come.
1999-12-29 05:07:58 +00:00
Bruce Evans
7e58bfacbe Update the unclean flag for mount -u. I forgot to handle this case
when I made the absence of the clean flag sticky in rev.1.88.  This
was a problem main for "mount /".  There is no way to mount "/" for
writing without using mount -u (normally implicitly), so after
"mount -f /" of an unclean filesystem, the absence of the clean flag
was sticky forever.
1999-12-23 15:42:14 +00:00
Eivind Eklund
369dc8ceb8 Change incorrect NULLs to 0s 1999-12-21 11:14:12 +00:00
Robert Watson
91f37dcba1 Second pass commit to introduce new ACL and Extended Attribute system
calls, vnops, vfsops, both in /kern, and to individual file systems that
require a vfsop_ array entry.

Reviewed by:	eivind
1999-12-19 06:08:07 +00:00
Kirk McKusick
6a4152243f The function request_cleanup() had a tsleep() with PCATCH. It is
quite dangerous, since the process may hold locks at the point,
and if it is stopped in that tsleep the machine may hang. Because
the sleep is so short, the PCATCH is not required here, so it has
been removed. For the future, the FreeBSD team needs to decide
whether it is still reasonable to stop a process in tsleep, as that
may affect any other code that uses PCATCH while holding kernel locks.

Submitted by:	Dmitrij Tejblum <tejblum@arc.hq.cti.ru>
Reviewed by:	Kirk McKusick <mckusick@mckusick.com>
1999-12-16 22:02:09 +00:00
Eivind Eklund
762e6b856c Introduce NDFREE (and remove VOP_ABORTOP) 1999-12-15 23:02:35 +00:00
Eivind Eklund
6bdfe06ad9 Lock reporting and assertion changes.
* lockstatus() and VOP_ISLOCKED() gets a new process argument and a new
  return value: LK_EXCLOTHER, when the lock is held exclusively by another
  process.
* The ASSERT_VOP_(UN)LOCKED family is extended to use what this gives them
* Extend the vnode_if.src format to allow more exact specification than
  locked/unlocked.

This commit should not do any semantic changes unless you are using
DEBUG_VFS_LOCKS.

Discussed with:	grog, mch, peter, phk
Reviewed by:	peter
1999-12-11 16:13:02 +00:00
Bill Fumerola
43cd4e8815 Remove the 'alpha, use at your own risk' death-statement.
Reviewed by:	mckusick (verbally at FreeBSDcon)
1999-12-03 00:40:31 +00:00
Bill Fumerola
cfa5001489 Fix typo, add $FreeBSD$ 1999-12-03 00:34:26 +00:00
Kirk McKusick
9f54c05286 Preferentially allocate the first indirect block in the same
cylinder group as the inode. This makes a 15% difference in
read speed for files in the 96K to 500K size range.
1999-12-01 19:33:12 +00:00
Poul-Henning Kamp
71e4fff823 Retire MFS_ROOT and MFS_ROOT_SIZE options from the MFS implementation.
Add MD_ROOT and MD_ROOT_SIZE options to the md driver.

Make the md driver handle MFS_ROOT and MFS_ROOT_SIZE options for compatibility.

Add md driver to GENERIC, PCCARD and LINT.

This is a cleanup which removes the need for some of the worse hacks in
MFS:  We really want to have a rootvnode but MFS on a preloaded image
doesn't really have one.  md is a true device, so it is less trouble.

This has been tested with make release, and if people remember to add
the "md" pseudo-device to their kernels, PicoBSD should be just fine
as well.  If people have no other use for MFS, it can be removed from
the kernel.
1999-11-26 20:08:44 +00:00
Poul-Henning Kamp
38224dcd59 Convert various pieces of code to use vn_isdisk() rather than checking
for vp->v_type == VBLK.

In ccd: we don't need to call VOP_GETATTR to find the type of a vnode.

Reviewed by:    sos
1999-11-22 10:33:55 +00:00
Eivind Eklund
b2f2b704d0 We do not have ffs_checkexp, so remove the prototype 1999-11-20 16:44:44 +00:00
Poul-Henning Kamp
0429e37ade struct mountlist and struct mount.mnt_list have no business being
a CIRCLEQ.  Change them to TAILQ_HEAD and TAILQ_ENTRY respectively.

This removes ugly  mp != (void*)&mountlist  comparisons.

Requested by:   phk
Submitted by:   Jake Burkholder jake@checker.org
PR:             14967
1999-11-20 10:00:46 +00:00
Peter Wemm
63034ded71 Fix a warning (unused static declaration without MFS_ROOT) 1999-11-18 08:49:40 +00:00
Eivind Eklund
dd8c04f4c7 Remove WILLRELE from VOP_SYMLINK
Note: Previous commit to these files (except coda_vnops and devfs_vnops)
that claimed to remove WILLRELE from VOP_RENAME actually removed it from
VOP_MKNOD.
1999-11-13 20:58:17 +00:00
Eivind Eklund
edfe736df9 Remove WILLRELE from VOP_RENAME 1999-11-12 03:34:28 +00:00
Poul-Henning Kamp
698f9cf828 Next step in the device cleanup process.
Correctly lock vnodes when calling VOP_OPEN() from filesystem mount code.

Unify spec_open() for bdev and cdev cases.

Remove the disabled bdev specific read/write code.
1999-11-09 14:15:33 +00:00
Bruce Evans
5bd5c8b9e5 Quick fix for breakage of ext2fs link counts as reported by stat(2) by
the soft updates changes: only report the link count to be i_effnlink
in ufs_getattr() for file systems that maintain i_effnlink.

Tested by:	Mike Dracopoulos <mdraco@math.uoa.gr>
1999-11-03 12:05:39 +00:00
Mike Smith
88d4183b84 Make MFS work with the new root filesystem search process.
In order to achieve this, root filesystem mount is moved from
SI_ORDER_FIRST to SI_ORDER_SECOND in the SI_SUB_MOUNT_ROOT sysinit
group.  Now, modules which wish to usurp the default root mount
can use SI_ORDER_FIRST.

A compiled-in or preloaded MFS filesystem will become the root
filesystem unless the vfs.root.mountfrom environment variable refers
to a valid bootable device.  This will normally only be the case when
the kernel and MFS image have been loaded from a disk which has a
valid /etc/fstab file.  In this case, the variable should be manually
overridden in the loader, or the kernel booted with -a.  In either
case "mfs:" should be supplied as the new value.

Also fix a typo in one DFLTROOT case that would not have compiled.
1999-11-03 11:02:47 +00:00
Mike Smith
6d14782861 Newline-terminate the complaint message about not being able to find
the root vnode pointer.
1999-11-01 23:57:28 +00:00
Matthew Dillon
f9eb66d73a Add sysctl debug.dircheck to allow directory sanity checking to be turned
on with a sysctl.

    Fix two bugs in ufs_lookup that can cause deadlocks due to out-of-order
    locking.  This fix was tested for a few days prior to commit.
1999-10-30 00:51:14 +00:00
Poul-Henning Kamp
923502ff91 useracc() the prequel:
Merge the contents (less some trivial bordering the silly comments)
of <vm/vm_prot.h> and <vm/vm_inherit.h> into <vm/vm.h>.  This puts
the #defines for the vm_inherit_t and vm_prot_t types next to their
typedefs.

This paves the road for the commit to follow shortly: change
useracc() to use VM_PROT_{READ|WRITE} rather than B_{READ|WRITE}
as argument.
1999-10-29 18:09:36 +00:00
Poul-Henning Kamp
b89392e703 Remove the D_NOCLUSTER[RW] options which were added because vn had
problems.  Now that Matt has fixed vn, this can go.  The vn driver
should have used d_maxio (now si_iosize_max) anyway.
1999-09-30 07:11:30 +00:00
Poul-Henning Kamp
1b5464ef9d Remove v_maxio from struct vnode.
Replace it with mnt_iosize_max in struct mount.

Nits from:	bde
1999-09-29 20:05:33 +00:00
Marcel Moolenaar
2c42a14602 sigset_t change (part 2 of 5)
-----------------------------

The core of the signalling code has been rewritten to operate
on the new sigset_t. No methodological changes have been made.
Most references to a sigset_t object are through macros (see
signalvar.h) to create a level of abstraction and to provide
a basis for further improvements.

The NSIG constant has not been changed to reflect the maximum
number of signals possible. The reason is that it breaks
programs (especially shells) which assume that all signals
have a non-null name in sys_signame. See src/bin/sh/trap.c
for an example. Instead _SIG_MAXSIG has been introduced to
hold the maximum signal possible with the new sigset_t.

struct sigprop has been moved from signalvar.h to kern_sig.c
because a) it is only used there, and b) access must be done
though function sigprop(). The latter because the table doesn't
holds properties for all signals, but only for the first NSIG
signals.

signal.h has been reorganized to make reading easier and to
add the new and/or modified structures. The "old" structures
are moved to signalvar.h to prevent namespace polution.

Especially the coda filesystem suffers from the change, because
it contained lines like (p->p_sigmask == SIGIO), which is easy
to do for integral types, but not for compound types.

NOTE: kdump (and port linux_kdump) must be recompiled.

Thanks to Garrett Wollman and Daniel Eischen for pressing the
importance of changing sigreturn as well.
1999-09-29 15:03:48 +00:00
Poul-Henning Kamp
d6a0e38a1b Remove five now unused fields from struct cdevsw. They should never
have been there in the first place.  A GENERIC kernel shrinks almost 1k.

Add a slightly different safetybelt under nostop for tty drivers.

Add some missing FreeBSD tags
1999-09-25 18:24:47 +00:00
Matthew Dillon
67ddfcaf69 More removals of vnode->v_lastr, replaced by preexisting seqcount
heuristic to detect sequential operation.

    VM-related forced clustering code removed from ufs in preparation for a
    commit to vm/vm_fault.c that does it more generally.

Reviewed by: David Greenman <dg@root.com>, Alan Cox <alc@cs.rice.edu>
1999-09-20 23:27:58 +00:00
Poul-Henning Kamp
faad302913 Fix a harmless bug I introduced, simplify a bit more while here. 1999-09-20 21:14:43 +00:00
Poul-Henning Kamp
fae03f66d1 Step one of replacing devsw->d_maxio with si_bsize_max.
Rename dev->si_bsize_max to si_iosize_max and set it in spec_open
if the device didn't.

Set vp->v_maxio from dev->si_bsize_max in spec_open rather than
in ufs_bmap.c
1999-09-20 19:57:28 +00:00
Bruce Evans
887ba12fc5 Removed diskerr()'s unused d_name arg and updated callers. This fixes
warnings caused by the arg having the wrong type (not const enough).
The arg was also wrong (a full name instead of a short one) for calls
from from subr_diskmbr.c and pc98/diskslice_machdep.c.
1999-09-13 12:59:41 +00:00
Alfred Perlstein
c24fda81c9 Seperate the export check in VFS_FHTOVP, exports are now checked via
VFS_CHECKEXP.

Add fh(open|stat|stafs) syscalls to allow userland to query filesystems
based on (network) filehandle.

Obtained from:	NetBSD
1999-09-11 00:46:08 +00:00