Commit Graph

90 Commits

Author SHA1 Message Date
Julian Elischer
3f6f17ee1e Submitted by: Archie and me.
We encountered an interesting situation where the superblock for
a file system got written to disk with the "fs_fmod" flag set to
one. It appears that this flag is normally supposed to be cleared
during ffs_sync(), but we experienced a crash, or some other weird
occurrence that left it on the disk set to 1.

Later this partition was mounted read-only... and the fs_fmod
field was never cleared, causing ffs_sync() to panic "rofs mod"
when trying to unmount that filesystem (ffs_vfsops.c: line 790).

fix:
set this bit to 0 when you load the superblock from disk.
(see more complete mail on this to hackers)
1996-11-13 01:45:56 +00:00
David Greenman
e75838f701 Eliminate an unnecessary synchronous write (and an 8K bcopy+bzero) when
truncating/deleting large files.

Reviewed by:	mckusick, dyson
Submitted by:	Kirk McKusick <mckusick@mckusick.com>, modified for
		FreeBSD by me.
1996-11-05 08:19:40 +00:00
Bruce Evans
ccb174975a Fixed lblktosize(). It overflowed at 2G. This bug only affected
ufs_read() and ufs_write().

Found by: looking at warnings for comparing the result of lblktosize()
(which is usually daddr_t = long) with file sizes (which are u_quad_t
for ufs).  File sizes should probably be off_t's to avoid warnings
when the are compared with file offsets, so the fixed lblktosize()
casts to off_t instead of u_quad_t.

Added definition of smalllblksize().  It is the same as the old
lblksize() and is more efficient for small block numbers on 32-bit
machines.

Use smalllblktosize() instead of its expansion in blksize() and
dblksize().  This keeps the line length short and makes it more
obvious that the shift can't overflow.
1996-10-12 22:12:51 +00:00
Nate Williams
030e2e9ebb In sys/time.h, struct timespec is defined as:
/*
         * Structure defined by POSIX.4 to be like a timeval.
         */
        struct timespec {
                time_t  ts_sec;         /* seconds */
                long    ts_nsec;        /* and nanoseconds */
        };

        The correct names of the fields are tv_sec and tv_nsec.

Reminded by:	James Drobina <jdrobina@infinet.com>
1996-09-19 18:21:32 +00:00
Peter Wemm
7b8830a51e Argh, I have had one "uid 0 on /: file system full" too many. The problem
is that it doesn't say _what_ did it! (the core dumped console message
is very useful for listing the process name and pid).  This adds similar
information.
1996-09-17 14:38:16 +00:00
John Dyson
c645dc1239 Fix a VOP_UNLOCK panic when using options DIAGNOSTIC during dismount. 1996-09-07 17:34:57 +00:00
John Dyson
6476c0d204 Even though this looks like it, this is not a complex code change.
The interface into the "VMIO" system has changed to be more consistant
and robust.  Essentially, it is now no longer necessary to call vn_open
to get merged VM/Buffer cache operation, and exceptional conditions
such as merged operation of VBLK devices is simpler and more correct.

This code corrects a potentially large set of problems including the
problems with ktrace output and loaded systems, file create/deletes,
etc.

Most of the changes to NFS are cosmetic and name changes, eliminating
a layer of subroutine calls.  The direct calls to vput/vrele have
been re-instituted for better cross platform compatibility.

Reviewed by: davidg
1996-08-21 21:56:23 +00:00
Bruce Evans
6ab46d52a5 Don't use NULL in non-pointer contexts. 1996-07-12 04:12:25 +00:00
David Greenman
2f9bae59d6 Moved the fsnode MALLOC to before the call to getnewvnode() so that the
process won't possibly block before filling in the fsnode pointer (v_data)
which might be dereferenced during a sync since the vnode is put on the
mnt_vnodelist by getnewvnode.

Pointed out by Matt Day <mday@artisoft.com>
1996-06-12 03:37:57 +00:00
Gary Palmer
6ddbf1e299 Clean up various compiler warnings. Most (if not all) were benign
Reviewed by:	bde
1996-05-08 04:29:08 +00:00
Bruce Evans
2043dc9a22 Removed bogus _BEGIN_DECLS/_END_DECLS.
Removed unused struct tag declarations in cloned code.

Added or cleaned up idempotency ifdefs.
1996-05-01 02:16:17 +00:00
John Dyson
847a3ba792 Handle the bogus device that MFS uses as its VBLK device. We now don't
try to VMIO open it on MFS mounts.  This will fix the mfs_badops
panic.
1996-03-02 22:18:34 +00:00
John Dyson
91477adc6e Enable VMIO for non-VDIR metadata and block device. 1996-03-02 03:45:12 +00:00
Bruce Evans
e6302eab11 Removed vestigial support for the obsolete FIFO option. In ext2fs
it caused null pointer panics for all fifo operations unless FIFO
was defined.
1996-02-25 20:12:36 +00:00
Mike Pritchard
6c5e9bbdf5 Fix a bunch of spelling errors in the comment fields of
a bunch of system include files.
1996-01-30 23:02:38 +00:00
John Dyson
bd7e5f992e Eliminated many redundant vm_map_lookup operations for vm_mmap.
Speed up for vfs_bio -- addition of a routine bqrelse to greatly diminish
	overhead for merged cache.
Efficiency improvement for vfs_cluster.  It used to do alot of redundant
	calls to cluster_rbuild.
Correct the ordering for vrele of .text and release of credentials.
Use the selective tlb update for 486/586/P6.
Numerous fixes to the size of objects allocated for files.  Additionally,
	fixes in the various pagers.
Fixes for proper positioning of vnode_pager_setsize in msdosfs and ext2fs.
Fixes in the swap pager for exhausted resources.  The pageout code
	will not as readily thrash.
Change the page queue flags (PG_ACTIVE, PG_INACTIVE, PG_FREE, PG_CACHE) into
	page queue indices (PQ_ACTIVE, PQ_INACTIVE, PQ_FREE, PQ_CACHE),
	thereby improving efficiency of several routines.
Eliminate even more unnecessary vm_page_protect operations.
Significantly speed up process forks.
Make vm_object_page_clean more efficient, thereby eliminating the pause
	that happens every 30seconds.
Make sequential clustered writes B_ASYNC instead of B_DELWRI even in the
	case of filesystems mounted async.
Fix a panic with busy pages when write clustering is done for non-VMIO
	buffers.
1996-01-19 04:00:31 +00:00
Bruce Evans
51ea8b57b8 Partially fixed negative and truncated "Avail" counts in df output.
This fixes PR943.

ffs/ffs_vfsops.c:
ffs_statfs() multiplied by (100 - minfree) as part of calculating the
minfree percentage (complemented in 100%), so with the standard minfree
of 8, it was broken for file systems of size >= 1TB/92 = 11GB.  Use the
standard freespace() macro instead.  This also fixes a rounding bug (the
"Avail" count was sometimes 1 too small).

ffs/* (not fixed):
The freespace() macro multiplies by minfree, so with the standard
minfree of 8, it is broken for file systems of size >= 1TB/8 = 128GB.
This bug is more serious since it affects block allocation.

ffs/ffs_alloc.c (not fixed):
Ordinary users are sometimes allowed to allocate 1 (partial) block
too many so that the "Avail" count goes negative.  E.g., if there is
1 fragment available and the file is fairly large, one more full
block is allocated.

df/df.c:
ufs_df() used/uses essentially the same code as ffs_statfs(), so it
had/has the same bugs.

ufs_df() gratuitously replaced "Avail" counts of < 0 by 0, so it
gave different results for non-mounted file systems in this case.
1996-01-14 18:55:09 +00:00
Garrett Wollman
01733a9b6d Convert QUOTA to new-style option. 1996-01-05 18:31:58 +00:00
Garrett Wollman
0e41ee3037 Convert DDB to new-style option. 1996-01-04 21:13:23 +00:00
Poul-Henning Kamp
b8dce649f1 Staticize. 1995-12-17 21:14:36 +00:00
Peter Wemm
dbe525ffb0 Silence a harmless warning... 1995-12-15 03:36:25 +00:00
John Dyson
a316d390bd Changes to support 1Tb filesizes. Pages are now named by an
(object,index) pair instead of (object,offset) pair.
1995-12-11 04:58:34 +00:00
David Greenman
efeaf95a41 Untangled the vm.h include file spaghetti. 1995-12-07 12:48:31 +00:00
Bruce Evans
57a4e3fa8a Completed function declarations and/or added prototypes and/or #includes
to get the prototypes.
1995-12-03 11:17:15 +00:00
Poul-Henning Kamp
c03020b22f Fix compiler warnings. 1995-11-20 12:25:37 +00:00
John Dyson
aa3960192e General fixes to the vfs clustring code:
1) Make cluster buffer list be a non-malloced chain.  This eliminates
yet another 'evil' M_WAITOK and generally cleans up the code.
2) Fix write clustering for ext2fs.  It was just broken.  Also, ffs
clustering had an efficiency problem that more bawrites were happening
than should have been.
3) Make changes to buf.h to support the above, plus remove b_pfcent
at the request of David Greenman.

Note that the reallocblocks code is disabled pending rewrite for
the cluster buffer list changes.
1995-11-19 19:55:26 +00:00
Poul-Henning Kamp
af8364b0ad Get rid of the last debug sysctl variables of the old style. 1995-11-14 09:40:06 +00:00
Bruce Evans
f57e65478d Introduced a type `vop_t' for vnode operation functions and used
it 1138 times (:-() in casts and a few more times in declarations.
This change is null for the i386.

The type has to be `typedef int vop_t(void *)' and not `typedef
int vop_t()' because `gcc -Wstrict-prototypes' warns about the
latter.  Since vnode op functions are called with args of different
(struct pointer) types, neither of these function types is any use
for type checking of the arg, so it would be preferable not to use
the complete function type, especially since using the complete
type requires adding 1138 casts to avoid compiler warnings and
another 40+ casts to reverse the function pointer conversions before
calling the functions.
1995-11-09 08:17:23 +00:00
John Dyson
4dc4e0d228 Make MNT_ASYNC more effective for UFS. It should not be too much more
dangerous than the original MNT_ASYNC.  There might be some minor
security considerations due to data writes not being posted as promptly
as before.  Meta-data operations are still not quite as fast as Linux,
but streaming I/O is still higher.
1995-11-05 21:01:15 +00:00
John Dyson
2c4488fce3 Finalize GETPAGES layering scheme. Move the device GETPAGES
interface into specfs code.  No need at this point to modify the
PUTPAGES stuff except in the layered-type (NULL/UNION) filesystems.
1995-10-23 02:23:29 +00:00
John Dyson
10d86e5e1f Re-enable read clustering. 1995-09-25 06:00:59 +00:00
David Greenman
ecf0c3ce4f Shit! I changed the wrong doclusterread! ...Thanks to Steven Wallace and
Poul-Henning for convincing me that I should look at my mistake! :-)
1995-09-22 06:02:40 +00:00
David Greenman
aee3c74900 Disable file read clustering until the bug(s) in vfs_cluster.c are fixed.
This should temporarily fix the sig 10/11 problems that people have been
having for the past 3 weeks.
1995-09-22 00:05:46 +00:00
David Greenman
9eda220c0b Slight optimization for the standard case of rotdelay=0. 1995-09-08 17:16:32 +00:00
John Dyson
3441c37e0b Added indirect pointer for ffs_getpages, and added external declaration. 1995-09-06 05:41:17 +00:00
John Dyson
c83ebe7781 Added VOP_GETPAGES/VOP_PUTPAGES and also the "backwards" block count
for VOP_BMAP.  Updated affected filesystems...
1995-09-04 00:21:16 +00:00
Julian Elischer
2b14f991e6 Reviewed by: julian with quick glances by bruce and others
Submitted by:	terry (terry lambert)
This is  a composite of 3 patch sets submitted by terry.
they are:
New low-level init code that supports loadbal modules better
some cleanups in the namei code to help terry in 16-bit character support
some changes to the mount-root code to make it a little more
modular..

NOTE: mounting root off cdrom or NFS MIGHT be broken as I haven't been able
to test those cases..

certainly mounting root of disk still works just fine..
mfs should work but is untested. (tomorrows task)

The low level init stuff includes a total rewrite of init_main.c
to make it possible for new modules to have an init phase by simply
adding an entry to a TEXT_SET (or is it DATA_SET) list. thus a new module can
be added to the kernel without editing any other files other than the
'files' file.
1995-08-28 09:19:25 +00:00
Bruce Evans
4238af7cd2 Don't call VOP_UPDATE() with volatile timestamps. 1995-08-25 19:40:32 +00:00
David Greenman
a9e479e2fa Honor -async mount option when doing the inode update.
Obtained from:	4.4BSD-Lite2
1995-08-16 13:16:58 +00:00
David Greenman
628641f8a6 Converted mountlist to a CIRCLEQ.
Partially obtained from: 4.4BSD-Lite2
1995-08-11 11:31:18 +00:00
David Greenman
39e68756fb Use bdwrite() rather than brelse(). The cylinder group bitmap modification
is not preserved otherwise.
Note that this is a no-op in FreeBSD, however, as we have doreallocblks
disabled.

Submitted by:	Kirk McKusick
1995-08-07 08:16:32 +00:00
David Greenman
b6dedae69b Removed redundant call to vm_object_page_clean - this is already done
in vfs_msync().
1995-08-06 11:56:42 +00:00
David Greenman
7221ecc0ba Use the correct flags (IO_SYNC -> B_SYNC) when deciding to do a sync or
async write in the section that changes the filesize. The bug resulted
in the updates always being async.

Obtained from:	4.4BSD-Lite2
1995-08-04 05:49:17 +00:00
David Greenman
8997d94f79 Since ufs_ihashget can block, the lock must be checked for each time
the function returns. Also, moved lock into .bss and made minor cosmetic
changes.

Submitted by:	Bruce Evans
1995-07-21 16:20:20 +00:00
David Greenman
2094ddb6f0 Implement a lock in ffs_vget to prevent a race condition where two processes
try allocate the same inode/vnode, causing a duplicate.

Submitted by:	Matt Dillon, slightly reworked by me.
1995-07-21 03:52:40 +00:00
David Greenman
24a1cce34f NOTE: libkvm, w, ps, 'top', and any other utility which depends on struct
proc or any VM system structure will have to be rebuilt!!!

Much needed overhaul of the VM system. Included in this first round of
changes:

1) Improved pager interfaces: init, alloc, dealloc, getpages, putpages,
   haspage, and sync operations are supported. The haspage interface now
   provides information about clusterability. All pager routines now take
   struct vm_object's instead of "pagers".

2) Improved data structures. In the previous paradigm, there is constant
   confusion caused by pagers being both a data structure ("allocate a
   pager") and a collection of routines. The idea of a pager structure has
   escentially been eliminated. Objects now have types, and this type is
   used to index the appropriate pager. In most cases, items in the pager
   structure were duplicated in the object data structure and thus were
   unnecessary. In the few cases that remained, a un_pager structure union
   was created in the object to contain these items.

3) Because of the cleanup of #1 & #2, a lot of unnecessary layering can now
   be removed. For instance, vm_object_enter(), vm_object_lookup(),
   vm_object_remove(), and the associated object hash list were some of the
   things that were removed.

4) simple_lock's removed. Discussion with several people reveals that the
   SMP locking primitives used in the VM system aren't likely the mechanism
   that we'll be adopting. Even if it were, the locking that was in the code
   was very inadequate and would have to be mostly re-done anyway. The
   locking in a uni-processor kernel was a no-op but went a long way toward
   making the code difficult to read and debug.

5) Places that attempted to kludge-up the fact that we don't have kernel
   thread support have been fixed to reflect the reality that we are really
   dealing with processes, not threads. The VM system didn't have complete
   thread support, so the comments and mis-named routines were just wrong.
   We now use tsleep and wakeup directly in the lock routines, for instance.

6) Where appropriate, the pagers have been improved, especially in the
   pager_alloc routines. Most of the pager_allocs have been rewritten and
   are now faster and easier to maintain.

7) The pagedaemon pageout clustering algorithm has been rewritten and
   now tries harder to output an even number of pages before and after
   the requested page. This is sort of the reverse of the ideal pagein
   algorithm and should provide better overall performance.

8) Unnecessary (incorrect) casts to caddr_t in calls to tsleep & wakeup
   have been removed. Some other unnecessary casts have also been removed.

9) Some almost useless debugging code removed.

10) Terminology of shadow objects vs. backing objects straightened out.
    The fact that the vm_object data structure escentially had this
    backwards really confused things. The use of "shadow" and "backing
    object" throughout the code is now internally consistent and correct
    in the Mach terminology.

11) Several minor bug fixes, including one in the vm daemon that caused
    0 RSS objects to not get purged as intended.

12) A "default pager" has now been created which cleans up the transition
    of objects to the "swap" type. The previous checks throughout the code
    for swp->pg_data != NULL were really ugly. This change also provides
    the rudiments for future backing of "anonymous" memory by something
    other than the swap pager (via the vnode pager, for example), and it
    allows the decision about which of these pagers to use to be made
    dynamically (although will need some additional decision code to do
    this, of course).

13) (dyson) MAP_COPY has been deprecated and the corresponding "copy
    object" code has been removed. MAP_COPY was undocumented and non-
    standard. It was furthermore broken in several ways which caused its
    behavior to degrade to MAP_PRIVATE. Binaries that use MAP_COPY will
    continue to work correctly, but via the slightly different semantics
    of MAP_PRIVATE.

14) (dyson) Sharing maps have been removed. It's marginal usefulness in a
    threads design can be worked around in other ways. Both #12 and #13
    were done to simplify the code and improve readability and maintain-
    ability. (As were most all of these changes)

TODO:

1) Rewrite most of the vnode pager to use VOP_GETPAGES/PUTPAGES. Doing
   this will reduce the vnode pager to a mere fraction of its current size.

2) Rewrite vm_fault and the swap/vnode pagers to use the clustering
   information provided by the new haspage pager interface. This will
   substantially reduce the overhead by eliminating a large number of
   VOP_BMAP() calls. The VOP_BMAP() filesystem interface should be
   improved to provide both a "behind" and "ahead" indication of
   contiguousness.

3) Implement the extended features of pager_haspage in swap_pager_haspage().
   It currently just says 0 pages ahead/behind.

4) Re-implement the swap device (swstrategy) in a more elegant way, perhaps
   via a much more general mechanism that could also be used for disk
   striping of regular filesystems.

5) Do something to improve the architecture of vm_object_collapse(). The
   fact that it makes calls into the swap pager and knows too much about
   how the swap pager operates really bothers me. It also doesn't allow
   for collapsing of non-swap pager objects ("unnamed" objects backed by
   other pagers).
1995-07-13 08:48:48 +00:00
David Greenman
aa2cabb958 1) Converted v_vmdata to v_object.
2) Removed unnecessary vm_object_lookup()/pager_cache(object, TRUE) pairs
   after vnode_pager_alloc() calls - the object is already guaranteed to be
   persistent.
3) Removed some gratuitous casts.
1995-06-28 12:01:13 +00:00
Rodney W. Grimes
9b2e535452 Remove trailing whitespace. 1995-05-30 08:16:23 +00:00
David Greenman
347a022fa7 Kill bogus vnode_pager_setsize(). It was being called at the wrong time
and resulted in the object size being too small. This caused bad things
to happen later when the file was mapped.

Reviewed by:	John Dyson
1995-05-28 04:32:23 +00:00
David Greenman
61f5d51062 Changes to fix the following bugs:
1) Files weren't properly synced on filesystems other than UFS. In some
   cases, this lead to lost data. Most likely would be noticed on NFS.
   The fix is to make the VM page sync/object_clean general rather than
   in each filesystem.
2) Mixing regular and mmaped file I/O on NFS was very broken. It caused
   chunks of files to end up as zeroes rather than the intended contents.
   The fix was to fix several race conditions and to kludge up the
   "b_dirtyoff" and "b_dirtyend" that NFS relies upon - paying attention
   to page modifications that occurred via the mmapping.

Reviewed by:	David Greenman
Submitted by:	John Dyson
1995-05-21 21:39:31 +00:00