Commit Graph

177 Commits

Author SHA1 Message Date
Matthew Dillon
8b03c8ed5e This is a cleanup patch to Peter's new OBJT_PHYS VM object type
and sysv shared memory support for it.  It implements a new
    PG_UNMANAGED flag that has slightly different characteristics
    from PG_FICTICIOUS.

    A new sysctl, kern.ipc.shm_use_phys has been added to enable the
    use of physically-backed sysv shared memory rather then swap-backed.
    Physically backed shm segments are not tracked with PV entries,
    allowing programs which use a large shm segment as a rendezvous
    point to operate without eating an insane amount of KVM in the
    PV entry management.  Read: Oracle.

    Peter's OBJT_PHYS object will also allow us to eventually implement
    page-table sharing and/or 4MB physical page support for such segments.
    We're half way there.
2000-05-29 22:40:54 +00:00
Peter Wemm
0385347c1a Implement an optimization of the VM<->pmap API. Pass vm_page_t's directly
to various pmap_*() functions instead of looking up the physical address
and passing that.  In many cases, the first thing the pmap code was doing
was going to a lot of trouble to get back the original vm_page_t, or
it's shadow pv_table entry.

Inspired by: John Dyson's 1998 patches.

Also:
Eliminate pv_table as a seperate thing and build it into a machine
dependent part of vm_page_t.  This eliminates having a seperate set of
structions that shadow each other in a 1:1 fashion that we often went to
a lot of trouble to translate from one to the other. (see above)
This happens to save 4 bytes of physical memory for each page in the
system.  (8 bytes on the Alpha).

Eliminate the use of the phys_avail[] array to determine if a page is
managed (ie: it has pv_entries etc).  Store this information in a flag.
Things like device_pager set it because they create vm_page_t's on the
fly that do not have pv_entries.  This makes it easier to "unmanage" a
page of physical memory (this will be taken advantage of in subsequent
commits).

Add a function to add a new page to the freelist.  This could be used
for reclaiming the previously wasted pages left over from preloaded
loader(8) files.

Reviewed by:	dillon
2000-05-21 12:50:18 +00:00
Alan Cox
d7414c4446 vm_object_shadow: Remove an incorrect assertion. In obscure circumstances
vm_object_shadow can be called on an object with ref_count > 1 and
OBJ_ONEMAPPING set.  This isn't really a problem for vm_object_shadow.
2000-04-19 16:32:04 +00:00
Philippe Charnier
5929bcfaba Revert spelling mistake I made in the previous commit
Requested by: Alan and Bruce
2000-03-27 20:41:17 +00:00
Philippe Charnier
956f31353c Spelling 2000-03-26 15:20:23 +00:00
Poul-Henning Kamp
db5f635acc Eliminate the undocumented, experimental, non-delivering and highly
dangerous MAX_PERF option.
2000-03-16 08:51:55 +00:00
Matthew Dillon
4f79d873c1 Add MAP_NOSYNC feature to mmap(), and MADV_NOSYNC and MADV_AUTOSYNC to
madvise().

    This feature prevents the update daemon from gratuitously flushing
    dirty pages associated with a mapped file-backed region of memory.  The
    system pager will still page the memory as necessary and the VM system
    will still be fully coherent with the filesystem.  Modifications made
    by other means to the same area of memory, for example by write(), are
    unaffected.  The feature works on a page-granularity basis.

    MAP_NOSYNC allows one to use mmap() to share memory between processes
    without incuring any significant filesystem overhead, putting it in
    the same performance category as SysV Shared memory and anonymous memory.

Reviewed by: julian, alc, dg
1999-12-12 03:19:33 +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
Alan Cox
02577fa23e Remove the last vestiges of "vm_map_t phys_map". It's been unused
since i386/i386/machdep.c rev 1.45 (or 1994 :-) ).
1999-10-29 05:17:20 +00:00
Matthew Dillon
479112dfd9 Remove inappropriate VOP_FSYNC from vm_object_page_clean(). The fsync
syncs the entire underlying file rather then just the requested range,
    resulting in huge inefficiencies when the VM system is articulated in
    a certain way.  The VOP_FSYNC was also found to massively reduce NFS
    performance in certain cases.

    Change MADV_DONTNEED and MADV_FREE to call vm_page_dontneed() instead
    of vm_page_deactivate().  Using vm_page_deactivate() causes all
    inactive and cache pages to be recycled before the dontneed/free page
    is recycled, effectively flushing our entire VM inactive & cache
    queues continuously even if only a few pages are being actively MADV
    free'd and reused (such as occurs with a sequential scan of a
    memory-mapped file).

Reviewed by:	Alan Cox <alc@cs.rice.edu>, David Greenman <dg@root.com>
1999-09-17 05:48:36 +00:00
Peter Wemm
c3aac50f28 $Id$ -> $FreeBSD$ 1999-08-28 01:08:13 +00:00
Alan Cox
76782487f3 Remove the declarations for "vm_map_t io_map". It's been unused
since i386/i386/machdep rev 1.310, i.e., the demise of BOUNCE_BUFFERS.
1999-08-15 23:55:46 +00:00
Alan Cox
aecb0ebbac Remove the declarations for "vm_map_t u_map". It's been unused
since i386/i386/pmap rev 1.190.  (The alpha never used it.)
1999-08-15 21:55:20 +00:00
Alan Cox
193b935887 vm_object_madvise:
Update the comments to match the implementation.

Submitted by:	dillon
1999-08-12 08:22:57 +00:00
Alan Cox
58b4e6cc84 vm_object_madvise:
Support MADV_DONTNEED and MADV_WILLNEED on object types
	besides OBJT_DEFAULT and OBJT_SWAP.

Submitted by:	dillon
1999-08-12 06:33:56 +00:00
Poul-Henning Kamp
ce9edcf5b5 Merge the cons.c and cons.h to the best of my ability. alpha may or
may not compile, I can't test it.
1999-08-09 10:35:05 +00:00
Alan Cox
7f866e4b29 Move the memory access behavior information provided by madvise
from the vm_object to the vm_map.

Submitted by:	dillon
1999-08-01 06:05:09 +00:00
Alan Cox
9b21395a95 Remove vm_object::last_read. It is used by the old swap pager, but
not by the new one, i.e., vm/swap_pager.c rev 1.108.

Reviewed by:	dillon@backplane.com
1999-07-16 05:11:37 +00:00
Alan Cox
32b76dfa8a Cleanup OBJ_ONEMAPPING management.
vm_map.c:
	Don't set OBJ_ONEMAPPING on arbitrary vm objects.  Only default
	and swap type vm objects should have it set.  vm_object_deallocate
	already handles these cases.

vm_object.c:
	If OBJ_ONEMAPPING isn't already clear in vm_object_shadow,
	we are in trouble.  Instead of clearing it, make it
	an assertion that it is already clear.
1999-07-11 18:30:32 +00:00
Peter Wemm
3efc015bae Fix some int/long printf problems for the Alpha 1999-07-01 19:53:43 +00:00
Alan Cox
60ff97b002 Remove vm_object::cache_count and vm_object::wired_count. They are
not used.  (Nor is there any planned use by John who introduced them.)

Reviewed by:	"John S. Dyson" <toor@dyson.iquest.net>
1999-06-20 21:47:02 +00:00
Alan Cox
c7997d57f1 Addendum to 1.155. Verify the existence of the object before checking
its reference count.
1999-05-30 01:12:19 +00:00
Alan Cox
9a2f6362a7 Avoid the creation of unnecessary shadow objects. 1999-05-28 03:39:44 +00:00
Alan Cox
ea41812fe5 Remove prototypes for functions that don't exist anymore (vm_map.h).
Remove a useless argument from vm_map_madvise's interface (vm_map.c,
	vm_map.h, and vm_mmap.c).

Remove a redundant test in vm_uiomove (vm_map.c).

Make two changes to vm_object_coalesce:

1. Determine whether the new range of pages actually overlaps
the existing object's range of pages before calling vm_object_page_remove.
(Prior to this change almost 90% of the calls to vm_object_page_remove
were to remove pages that were beyond the end of the object.)

2. Free any swap space allocated to removed pages.
1999-05-16 05:07:34 +00:00
Alan Cox
a1a54e9fc1 Correct two optimization errors in vm_object_page_remove:
1. The size of vm_object::memq is vm_object::resident_page_count,
not vm_object::size.

2. The "size > 4" test sometimes results in the traversal of a ~1000 page
memq in order to locate ~10 pages.
1999-03-14 06:36:00 +00:00
Matthew Dillon
d1bf5d56b6 Remove unnecessary page protects on map_split and collapse operations.
Fix bug where an object's OBJ_WRITEABLE/OBJ_MIGHTBEDIRTY flags do
    not get set under certain circumstances ( page rename case ).

Reviewed by:	Alan Cox <alc@cs.rice.edu>, John Dyson
1999-02-24 21:26:26 +00:00
Matthew Dillon
1ce137be82 Fix a bug in the new madvise() code that would possibly (improperly)
free swap space out from under a busy page.  This is not legal because
    the swap may be reallocated and I/O issued while I/O is still in
    progress on the same swap page from the madvise()'d object.  This bug
    could only occur under extreme paging conditions but might not cause
    an error until much later.  As a side-benefit, madvise() is now even
    smaller.
1999-02-15 02:03:40 +00:00
Matthew Dillon
41c67e12bd Minor optimization to madvise() MADV_FREE to make page as freeable as
possible without actually unmapping it from the process.

    As of now, I declare madvise() on OBJT_DEFAULT/OBJT_SWAP objects to be
    'working and complete'.
1999-02-12 20:42:19 +00:00
Matthew Dillon
2aaeadf8d9 Fix non-fatal bug in vm_map_insert() which improperly cleared
OBJ_ONEMAPPING in the case where an object is extended by an
    additional vm_map_entry must be allocated.

    In vm_object_madvise(), remove calll to vm_page_cache() in MADV_FREE
    case in order to avoid a page fault on page reuse.  However, we still
    mark the page as clean and destroy any swap backing store.

Submitted by:	Alan Cox <alc@cs.rice.edu>
1999-02-12 09:51:43 +00:00
Matthew Dillon
2ad1a3f729 Revamp vm_object_[q]collapse(). Despite the complexity of this patch,
no major operational changes were made.  The three core object->memq loops
    were moved into a single inline procedure and various operational
    characteristics of the collapse function were documented.
1999-02-08 19:00:15 +00:00
Matthew Dillon
d031cff181 General cleanup. Remove #if 0's and remove useless register qualifiers. 1999-02-08 05:15:54 +00:00
Matthew Dillon
9fdfe602fc Remove MAP_ENTRY_IS_A_MAP 'share' maps. These maps were once used to
attempt to optimize forks but were essentially given-up on due to
    problems and replaced with an explicit dup of the vm_map_entry structure.
    Prior to the removal, they were entirely unused.
1999-02-07 21:48:23 +00:00
Matthew Dillon
9b09fe24a4 When shadowing objects, adjust the page coloring of the shadowing object
such that pages in the combined/shadowed object are consistantly
    colored.

Submitted by:	"John S. Dyson" <dyson@iquest.net>
1999-02-07 08:44:53 +00:00
Matthew Dillon
588059bea0 Fix bug in a KASSERT I introduced in vm_page_qcollapse() rev 1.139.
Since paging is in progress, page scan in vm_page_qcollapse() must be
    protected at atleast splbio() to prevent pages from being ripped out from
    under the scan.
1999-02-04 17:47:52 +00:00
Matthew Dillon
4112823fc7 Submitted by: Alan Cox
The vm_map_insert()/vm_object_coalesce() optimization has been extended
    to include OBJT_SWAP objects as well as OBJT_DEFAULT objects.  This is
    possible because it costs nothing to extend an OBJT_SWAP object with
    the new swapper.  We can't do this with the old swapper.  The old swapper
    used a linear array that would have had to have been reallocated, costing
    time as well as a potential low-memory deadlock.
1999-02-03 01:57:17 +00:00
Matthew Dillon
8aef171243 Fix warnings in preparation for adding -Wall -Wcast-qual to the
kernel compile
1999-01-28 00:57:57 +00:00
Matthew Dillon
8e3ad7c918 Depreciate vm_object_pmap_copy() - nobody uses it. Everyone uses
vm_object_pmap_copt_1() now, apparently.
1999-01-24 01:01:38 +00:00
Matthew Dillon
7bc9e80ecc object->id was badly implemented. It has simply been removed.
object->paging_offset has been removed - it was used to optimize a
    single OBJT_SWAP collapse case yet introduced massive confusion throughout
    vm_object.c.  The optimization was inconsequential except for the
    claim that it didn't have to allocate any memory.  The optimization
    has been removed.

    madvise() has been fixed.  The old madvise() could be made to operate
    on shared objects which is a big no-no.  The new one is much more careful
    in what it modifies.  MADV_FREE was totally broken and has now been fixed.

    vm_page_rename() now automatically dirties a page, so explicit dirtying
    of the page prior to calling vm_page_rename() has been removed.
1999-01-21 09:46:55 +00:00
Matthew Dillon
1c7c3c6a86 This is a rather large commit that encompasses the new swapper,
changes to the VM system to support the new swapper, VM bug
    fixes, several VM optimizations, and some additional revamping of the
    VM code.  The specific bug fixes will be documented with additional
    forced commits.  This commit is somewhat rough in regards to code
    cleanup issues.

Reviewed by:	"John S. Dyson" <root@dyson.iquest.net>, "David Greenman" <dg@root.com>
1999-01-21 08:29:12 +00:00
Eivind Eklund
219cbf59f2 KNFize, by bde. 1999-01-10 01:58:29 +00:00
Eivind Eklund
5526d2d920 Split DIAGNOSTIC -> DIAGNOSTIC, INVARIANTS, and INVARIANT_SUPPORT as
discussed on -hackers.

Introduce 'KASSERT(assertion, ("panic message", args))' for simple
check + panic.

Reviewed by:	msmith
1999-01-08 17:31:30 +00:00
Bruce Evans
289bdf33d3 Ifdefed conditionally used simplock variables. 1999-01-02 11:34:57 +00:00
David Greenman
dd0b2081f4 Implemented zero-copy TCP/IP extensions via sendfile(2) - send a
file to a stream socket. sendfile(2) is similar to implementations in
HP-UX, Linux, and other systems, but the API is more extensive and
addresses many of the complaints that the Apache Group and others have
had with those other implementations. Thanks to Marc Slemko of the
Apache Group for helping me work out the best API for this.
Anyway, this has the "net" result of speeding up sends of files over
TCP/IP sockets by about 10X (that is to say, uses 1/10th of the CPU
cycles) when compared to a traditional read/write loop.
1998-11-05 14:28:26 +00:00
David Greenman
e4b7635de2 Added needed splvm() protection around object page traversal in
vm_object_terminate().
1998-10-27 13:22:51 +00:00
Poul-Henning Kamp
f5ef029e92 Nitpicking and dusting performed on a train. Removes trivial warnings
about unused variables, labels and other lint.
1998-10-25 17:44:59 +00:00
David Greenman
9fcfb650d1 Oops, revert part of last fix. vm_pager_dealloc() can't be called until
after the pages are removed from the object...so fix the problem by
not printing the diagnostic for wired fictitious pages (which is normal).
1998-10-23 05:43:13 +00:00
David Greenman
356863eb01 Fixed two bugs in recent commit: in vm_object_terminate, vm_pager_dealloc
needs to be called prior to freeing remaining pages in the object so that
the device pager has an opportunity to grab its "fake" pages. Also, in
the case of wired pages, the page must be made busy prior to calling
vm_page_remove. This is a difference from 2.2.x that I overlooked when
I brought these changes forward.
1998-10-23 05:25:49 +00:00
David Greenman
0b10ba9822 Make the VM system handle the case where a terminating object contains
legitimately wired pages. Currently we print a diagnostic when this
happens, but this will be removed soon when it will be common for this
to occur with zero-copy TCP/IP buffers.
1998-10-22 02:16:53 +00:00
David Greenman
ce65e68c03 Be more selctive about when we clear p->valid.
Submitted by:	John Dyson <toor@dyson.iquest.net>
1998-09-28 02:40:11 +00:00
Doug Rabson
e69763a315 Cosmetic changes to the PAGE_XXX macros to make them consistent with
the other objects in vm.
1998-09-04 08:06:57 +00:00