Commit Graph

307 Commits

Author SHA1 Message Date
Alan Cox
7e9d944218 If a physical page is mapped by two or more virtual addresses, transmitted
by the zero-copy sockets method, and written to before the transmission
completes, we need to destroy all of the existing mappings to the page,
not just the one that we fault on.  Otherwise, the mappings will no longer
be to the same page and changes made through one of the mappings will not
be visible through the others.

Observed by: tegge
2005-11-08 06:33:21 +00:00
Alan Cox
674b706ea0 Consider the zero-copy transmission of a page that was wired by mlock(2).
If a copy-on-write fault occurs on the page, the new copy should inherit
a part of the original page's wire count.

Submitted by: tegge
MFC after: 1 week
2005-11-01 04:30:21 +00:00
Dag-Erling Smørgrav
3803b26bae As alc pointed out to me, vm_page.c 1.305 was incomplete: uma_startup()
still uses the constant UMA_BOOT_PAGES.  Change it to accept boot_pages
as an additional argument.

MFC after:	2 weeks
2005-10-08 21:03:54 +00:00
Dag-Erling Smørgrav
cfa22bcc4c Introduce the vm.boot_pages tunable and sysctl, which controls the number
of pages reserved to bootstrap the kernel memory allocator.

MFC after:	2 weeks
2005-08-12 12:24:19 +00:00
Jeff Roberson
761dbeb66f - In vm_page_insert() hold the backing vnode when the first page
is inserted.
 - In vm_page_remove() drop the backing vnode when the last page
   is removed.
 - Don't check the vnode to see if it must be reclaimed on every
   call to vm_page_free_toq() as we only check it now when it is
   actually required.  This saves us two lock operations per call.

Sponsored by:	Isilon Systems, Inc.
2005-03-15 14:14:09 +00:00
Alan Cox
46fbc58202 Transfer responsibility for freeing the page taken from the cache
queue and (possibly) unlocking the containing object from
vm_page_alloc() to vm_page_select_cache().  Recent optimizations to
vm_map_pmap_enter() (see vm_map.c revisions 1.362 and 1.363) and
pmap_enter_quick() have resulted in panic()s because vm_page_alloc()
mistakenly unlocked objects that had not been locked by
vm_page_select_cache().

Reported by: Peter Holm and Kris Kennaway
2005-01-07 05:02:19 +00:00
Warner Losh
60727d8b86 /* -> /*- for license, minor formatting changes 2005-01-07 02:29:27 +00:00
Alan Cox
0869d38ba6 Assert that page allocations during an interrupt specify
VM_ALLOC_INTERRUPT.

Assert that pages removed from the cache queue are not busy.
2004-12-31 19:50:45 +00:00
Alan Cox
7aa2190c8e Access to the page's busy field is (now) synchronized by the containing
object's lock.  Therefore, the assertion that the page queues lock is held
can be removed from vm_page_io_start().
2004-12-29 04:18:22 +00:00
Alan Cox
40198b3c04 Assert that the vm object is locked on entry to vm_page_sleep_if_busy();
remove some unneeded code.
2004-12-26 21:46:44 +00:00
Alan Cox
d19ef81437 The synchronization provided by vm object locking has eliminated the
need for most calls to vm_page_busy().  Specifically, most calls to
vm_page_busy() occur immediately prior to a call to vm_page_remove().
In such cases, the containing vm object is locked across both calls.
Consequently, the setting of the vm page's PG_BUSY flag is not even
visible to other threads that are following the synchronization
protocol.

This change (1) eliminates the calls to vm_page_busy() that
immediately precede a call to vm_page_remove() or functions, such as
vm_page_free() and vm_page_rename(), that call it and (2) relaxes the
requirement in vm_page_remove() that the vm page's PG_BUSY flag is
set.  Now, the vm page's PG_BUSY flag is set only when the vm object
lock is released while the vm page is still in transition.  Typically,
this is when it is undergoing I/O.
2004-11-03 20:17:31 +00:00
Alan Cox
f4d49654ae Assert that the containing vm object is locked in vm_page_cache() and
vm_page_try_to_cache().
2004-10-28 05:26:21 +00:00
Alan Cox
63bb7041cc Assert that the containing vm object is locked in vm_page_flash(). 2004-10-25 19:52:44 +00:00
Alan Cox
75d0533847 Assert that the containing vm object is locked in vm_page_busy() and
vm_page_wakeup().
2004-10-24 23:53:47 +00:00
Alan Cox
0f9f9bcb53 Introduce VM_ALLOC_NOBUSY, an option to vm_page_alloc() and vm_page_grab()
that indicates that the caller does not want a page with its busy flag set.
In many places, the global page queues lock is acquired and released just
to clear the busy flag on a just allocated page.  Both the allocation of
the page and the clearing of the busy flag occur while the containing vm
object is locked.  So, the busy flag might as well never be set.
2004-10-24 06:15:36 +00:00
Alan Cox
1e96d2a217 Correct two errors in PG_BUSY management by vm_page_cowfault(). Both
errors are in rarely executed paths.
1. Each time the retry_alloc path is taken, the PG_BUSY must be set again.
   Otherwise vm_page_remove() panics.
2. There is no need to set PG_BUSY on the newly allocated page before
   freeing it.  The page already has PG_BUSY set by vm_page_alloc().
   Setting it again could cause an assertion failure.

MFC after: 2 weeks
2004-10-18 08:11:59 +00:00
Alan Cox
36aeb90e34 Assert that the containing object is locked in vm_page_io_start() and
vm_page_io_finish().  The motivation being to transition synchronization of
the vm_page's busy field from the global page queues lock to the per-object
lock.
2004-10-17 22:33:40 +00:00
Poul-Henning Kamp
7ce1979be6 Add new a function isa_dma_init() which returns an errno when it fails
and which takes a M_WAITOK/M_NOWAIT flag argument.

Add compatibility isa_dmainit() macro which whines loudly if
isa_dma_init() fails.

Problem uncovered by:	tegge
2004-09-15 12:09:50 +00:00
Alan Cox
a087914310 Advance the state of pmap locking on alpha, amd64, and i386.
- Enable recursion on the page queues lock.  This allows calls to
   vm_page_alloc(VM_ALLOC_NORMAL) and UMA's obj_alloc() with the page
   queues lock held.  Such calls are made to allocate page table pages
   and pv entries.
 - The previous change enables a partial reversion of vm/vm_page.c
   revision 1.216, i.e., the call to vm_page_alloc() by vm_page_cowfault()
   now specifies VM_ALLOC_NORMAL rather than VM_ALLOC_INTERRUPT.
 - Add partial locking to pmap_copy().  (As a side-effect, pmap_copy()
   should now be faster on i386 SMP because it no longer generates IPIs
   for TLB shootdown on the other processors.)
 - Complete the locking of pmap_enter() and pmap_enter_quick().  (As of now,
   all changes to a user-level pmap on alpha, amd64, and i386 are performed
   with appropriate locking.)
2004-07-29 18:56:31 +00:00
Brian Feldman
d951b75210 Fix a race in vm_page_sleep_if_busy(). Due to vm_object locking
being incomplete, it currently has to know how to drop and pick back
up the vm_object's mutex if it has to sleep and drop the page queue
mutex.  The problem with this is that if the page is busy, while we
are sleeping, the page can be freed and object disappear.  When trying
to lock m->object, we'd get a stale or NULL pointer and crash.

The object is now cached, but this makes the assumption that
the object is referenced in some manner and will not itself
disappear while it is unlocked.  Since this only happens if
the object is locked, I had to remove an assumption earlier in
contigmalloc() that reversed the order of locking the object and
doing vm_page_sleep_if_busy(), not the normal order.
2004-07-21 23:56:09 +00:00
Alan Cox
e832aafc51 - Eliminate the pte object from the pmap. Instead, page table pages are
allocated as "no object" pages.  Similar changes were made to the amd64
   and i386 pmap last year.  The primary reason being that maintaining
   a pte object leads to lock order violations.  A secondary reason being
   that the pte object is redundant, i.e., the page table itself can be
   used to lookup page table pages.  (Historical note: The pte object
   predates our ability to allocate "no object" pages.  Thus, the pte
   object was a necessary evil.)
 - Unconditionally check the vm object lock's status in vm_page_remove().
   Previously, this assertion could not be made on Alpha due to its use
   of a pte object.
2004-07-19 18:12:04 +00:00
Alan Cox
790bdd0f2e Increase the scope of the page queues lock in vm_page_alloc() to cover
a diagnostic check that accesses the cache queue count.
2004-07-10 22:12:49 +00:00
Alan Cox
0a2df4773c Remove spl() calls. Update comments to reflect the removal of spl() calls.
Remove '\n' from panic() format strings.  Remove some blank lines.
2004-06-19 04:19:47 +00:00
Alan Cox
d45f21f31a Do not preset PG_BUSY on VM_ALLOC_NOOBJ pages. Such pages are not
accessible through an object.  Thus, PG_BUSY serves no purpose.
2004-06-17 06:16:58 +00:00
Alan Cox
4be14af9cf To date, unwiring a fictitious page has produced a panic. The reason
being that PHYS_TO_VM_PAGE() returns the wrong vm_page for fictitious
pages but unwiring uses PHYS_TO_VM_PAGE().  The resulting panic
reported an unexpected wired count.  Rather than attempting to fix
PHYS_TO_VM_PAGE(), this fix takes advantage of the properties of
fictitious pages.  Specifically, fictitious pages will never be
completely unwired.  Therefore, we can keep a fictitious page's wired
count forever set to one and thereby avoid the use of
PHYS_TO_VM_PAGE() when we know that we're working with a fictitious
page, just not which one.

In collaboration with: green@, tegge@
PR: kern/29915
2004-05-22 04:53:51 +00:00
Alan Cox
1bb816d3d1 Restructure vm_page_select_cache() so that adding assertions is easy.
Some of the conditions that caused vm_page_select_cache() to deactivate a
page were wrong.  For example, deactivating an unmanaged or wired page is a
nop.  Thus, if vm_page_select_cache() had ever encountered an unmanaged or
wired page, it would have looped forever.  Now, we assert that the page is
neither unmanaged nor wired.
2004-05-12 04:27:18 +00:00
Alan Cox
3f39cca96b Cache queue pages are not mapped. Thus, the pmap_remove_all() by
vm_page_alloc() is unnecessary.
2004-05-09 01:00:15 +00:00
Alan Cox
2ec91846fd Update the comment describing vm_page_grab() to reflect the previous
revision and correct some of its style errors.
2004-04-24 21:36:23 +00:00
Alan Cox
7ef6ba5d27 Push down the responsibility for zeroing a physical page from the
caller to vm_page_grab().  Although this gives VM_ALLOC_ZERO a
different meaning for vm_page_grab() than for vm_page_alloc(), I feel
such change is necessary to accomplish other goals.  Specifically, I
want to make the PG_ZERO flag immutable between the time it is
allocated by vm_page_alloc() and freed by vm_page_free() or
vm_page_free_zero() to avoid locking overheads.  Once we gave up on
the ability to automatically recognize a zeroed page upon entry to
vm_page_free(), the ability to mutate the PG_ZERO flag became useless.
Instead, I would like to say that "Once a page becomes valid, its
PG_ZERO flag must be ignored."
2004-04-24 20:53:55 +00:00
Warner Losh
05eb3785e7 Remove advertising clause from University of California Regent's license,
per letter dated July 22, 1999.

Approved by: core
2004-04-06 20:15:37 +00:00
Alan Cox
889eb0fc62 Eliminate unused arguments from vm_page_startup(). 2004-04-04 23:33:36 +00:00
Alan Cox
ca3b447732 Modify contigmalloc1() so that the free page queues lock is not held when
vm_page_free() is called.  The problem with holding this lock is that it is
a spin lock and vm_page_free() may attempt the acquisition of a different
default-type lock.
2004-03-02 08:25:58 +00:00
Alan Cox
0f75a97722 - Correct a long-standing race condition in vm_page_try_to_free() that
could result in a dirty page being unintentionally freed.
 - Simplify the dirty page check in vm_page_dontneed().

Reviewed by:	tegge
MFC after:	7 days
2004-02-19 07:43:55 +00:00
Alan Cox
84d98bf699 - Correct a long-standing race condition in vm_page_try_to_cache() that
could result in a panic "vm_page_cache: caching a dirty page, ...":
   Access to the page must be restricted or removed before calling
   vm_page_cache().  This race condition is identical in nature to that
   which was addressed by vm_pageout.c's revision 1.251.
 - Simplify the code surrounding the fix to this same race condition
   in vm_pageout.c's revision 1.251.  There should be no behavioral
   change.  Reviewed by: tegge

MFC after:	7 days
2004-02-14 08:54:37 +00:00
Alan Cox
65bae14d77 - Enable recursive acquisition of the mutex synchronizing access to the
free pages queue.  This is presently needed by contigmalloc1().
 - Move a sanity check against attempted double allocation of two pages
   to the same vm object offset from vm_page_alloc() to vm_page_insert().
   This provides better protection because double allocation could occur
   through a direct call to vm_page_insert(), such as that by
   vm_page_rename().
 - Modify contigmalloc1() to hold the mutex synchronizing access to the
   free pages queue while it scans vm_page_array in search of free pages.
 - Correct a potential leak of pages by contigmalloc1() that I introduced
   in revision 1.20: We must convert all cache queue pages to free pages
   before we begin removing free pages from the free queue.  Otherwise,
   if we have to restart the scan because we are unable to acquire the
   vm object lock that is necessary to convert a cache queue page to a
   free page, we leak those free pages already removed from the free queue.
2004-01-08 20:48:26 +00:00
Alan Cox
4804edb44f In vm_page_lookup() check the root of the vm object's splay tree for the
desired page before calling vm_page_splay().
2003-12-31 19:02:01 +00:00
Alan Cox
bcdaad7fe7 Simplify vm_page_grab(): Don't bother with the generation check. If the
vm object hasn't changed, the desired page will be at or near the root
of the vm object's splay tree, making vm_page_lookup() cheap.  (The only
lock required for vm_page_lookup() is already held.)  If, however, the
vm object has changed and retry was requested, eliminating the generation
check also eliminates a pointless acquisition and release of the page
queues lock.
2003-12-31 01:44:45 +00:00
Alan Cox
9582cd94cb - Create an unmapped guard page to trap access to vm_page_array[-1].
This guard page would have trapped the problems with the MFC of the PAE
   support to RELENG_4 at an earlier point in the sequence of events.

Submitted by:	tegge
2003-12-22 02:04:08 +00:00
Alan Cox
de33beddd5 - Additional vm object locking in vm_object_split()
- New vm object locking assertions in vm_page_insert() and
   vm_object_set_writeable_dirty()
2003-11-01 04:54:23 +00:00
Alan Cox
ab42316c2f - Retire vm_pageout_page_free(). Instead, use vm_page_select_cache() from
vm_pageout_scan().  Rationale: I don't like leaving a busy page in the
   cache queue with neither the vm object nor the vm page queues lock held.
 - Assert that the page is active in vm_pageout_page_stats().
2003-10-22 18:41:32 +00:00
Alan Cox
0d42c05ff4 - Assert that the containing vm object is locked in
vm_page_set_validclean().  (This function reads and modifies the
   vm page's valid field, which is synchronized by the lock on the
   containing vm object.)
2003-10-21 19:36:51 +00:00
Alan Cox
fee181a696 - Remove some long unused code. 2003-10-20 18:57:01 +00:00
Alan Cox
669890eaeb Retire vm_page_copy(). Its reason for being ended when peter@ modified
pmap_copy_page() et al. to accept a vm_page_t rather than a physical
address.  Also, this change will facilitate locking access to the vm page's
valid field.
2003-10-08 05:35:12 +00:00
Alan Cox
5a3970febf Assert that the containing vm object's lock is held in
vm_page_set_invalid().
2003-10-05 06:58:07 +00:00
Alan Cox
874f526de6 Assert that the containing vm object's lock is held in
vm_page_zero_invalid().
2003-10-04 21:56:27 +00:00
Alan Cox
bf0da100d6 - Extend the scope the vm object lock to cover calls to
vm_page_is_valid().
 - Assert that the lock on the containing vm object is held in
   vm_page_is_valid().
2003-10-04 19:23:29 +00:00
Alan Cox
50028aa7d2 In vm_page_remove(), assert that the vm object is locked, unless an Alpha.
(The Alpha still requires updates to its pmap.)
2003-09-28 04:50:48 +00:00
Alan Cox
95aad59a53 Initialize the page's pindex field even for VM_ALLOC_NOOBJ allocations.
(This field is useful for implementing sanity checks even if the page does
not belong to an object.)
2003-09-22 00:56:13 +00:00
Alan Cox
2370c6d40c Recent pmap changes permit the use of a more precise locking assertion
in vm_page_lookup().
2003-08-28 23:23:04 +00:00
Alan Cox
529e15ed69 Held pages, just like wired pages, should not be added to the cache queues.
Submitted by:	tegge
2003-08-23 20:29:29 +00:00