Author: alc
Date: Tue Feb 26 08:17:34 2013
New Revision: 247314
URL: http://svnweb.freebsd.org/changeset/base/247314
Log:
Be more conservative in auto-sizing and capping the kmem submap. In
fact, use the same values here that we use on 32-bit x86 and MIPS. Some
machines were reported to
On Feb 25, 2013, at 4:36 AM, Andrew Turner wrote:
On Mon, 25 Feb 2013 10:50:19 +0200
Konstantin Belousov kostik...@gmail.com wrote:
On Mon, Feb 25, 2013 at 08:13:13PM +1300, Andrew Turner wrote:
On Thu, 21 Feb 2013 19:02:50 + (UTC)
John Baldwin j...@freebsd.org wrote:
Author: jhb
Author: alc
Date: Tue Feb 26 07:00:24 2013
New Revision: 247312
URL: http://svnweb.freebsd.org/changeset/base/247312
Log:
Eliminate a duplicate #include.
Sponsored by: EMC / Isilon Storage Division
Modified:
head/sys/fs/fuse/fuse_io.c
Modified: head/sys/fs/fuse/fuse_io.c
Author: alc
Date: Tue Feb 26 07:41:34 2013
New Revision: 247313
URL: http://svnweb.freebsd.org/changeset/base/247313
Log:
Eliminate a redundant #include: machine/pmap.h is already included
through vm/pmap.h.
Modified:
head/sys/arm/at91/at91_machdep.c
head/sys/arm/sa11x0/assabet_machdep.c
On 02/26/2013 01:27, Andrew Turner wrote:
On Mon, 25 Feb 2013 15:00:41 -0600
Alan Cox a...@rice.edu wrote:
On Feb 25, 2013, at 4:36 AM, Andrew Turner wrote:
On Mon, 25 Feb 2013 10:50:19 +0200
Konstantin Belousov kostik...@gmail.com wrote:
On Mon, Feb 25, 2013 at 08:13:13PM +1300, Andrew
Author: alc
Date: Fri Feb 22 16:59:52 2013
New Revision: 247153
URL: http://svnweb.freebsd.org/changeset/base/247153
Log:
Eliminate an unused #define.
Modified:
head/sys/powerpc/include/vmparam.h
Modified: head/sys/powerpc/include/vmparam.h
Author: alc
Date: Wed Feb 20 16:48:52 2013
New Revision: 247046
URL: http://svnweb.freebsd.org/changeset/base/247046
Log:
Initialize vm_max_kernel_address on non-FDT platforms. (This should have
been included in r246926.)
The second parameter to pmap_bootstrap() is redundant. Eliminate
Author: alc
Date: Mon Feb 18 01:02:48 2013
New Revision: 246926
URL: http://svnweb.freebsd.org/changeset/base/246926
Log:
On arm, like sparc64, the end of the kernel map varies from one type of
machine to another. Therefore, VM_MAX_KERNEL_ADDRESS can't be a constant.
Instead, #define it to
Author: alc
Date: Mon Feb 18 01:22:20 2013
New Revision: 246929
URL: http://svnweb.freebsd.org/changeset/base/246929
Log:
Place a cap on the size of the kernel's heap, also known as the kmem
submap. Otherwise, after r246204, the auto-scaling logic in kern_malloc.c
tries to create a kmem
On 02/16/2013 00:36, Oleksandr Tymoshenko wrote:
On 2013-02-01, at 2:26 AM, Andre Oppermann an...@freebsd.org wrote:
Author: andre
Date: Fri Feb 1 10:26:31 2013
New Revision: 246204
URL: http://svnweb.freebsd.org/changeset/base/246204
Log:
Add VM_KMEM_SIZE_SCALE parameter set to 2 (50%)
On 02/01/2013 15:57, Andre Oppermann wrote:
On 01.02.2013 22:16, Juli Mallett wrote:
On Fri, Feb 1, 2013 at 1:01 PM, Andre Oppermann an...@freebsd.org
wrote:
On 01.02.2013 21:23, Adrian Chadd wrote:
.. before you make that assumption, please re-visit some the ..
lower-end integrated
On 02/01/2013 04:26, Andre Oppermann wrote:
Author: andre
Date: Fri Feb 1 10:26:31 2013
New Revision: 246204
URL: http://svnweb.freebsd.org/changeset/base/246204
Log:
Add VM_KMEM_SIZE_SCALE parameter set to 2 (50%) for all ARM platforms.
VM_KMEM_SIZE_SCALE specifies which fraction
On 01/07/2013 12:47, Oleksandr Tymoshenko wrote:
On 12/27/2012 6:46 PM, Oleksandr Tymoshenko wrote:
On 12/18/2012 1:59 AM, Alan Cox wrote:
On 12/17/2012 23:40, Oleksandr Tymoshenko wrote:
On 2012-12-08, at 1:21 PM, Alan Cox a...@rice.edu wrote:
On 12/08/2012 14:32, Andre Oppermann wrote
Author: alc
Date: Sat Jan 12 18:06:21 2013
New Revision: 245337
URL: http://svnweb.freebsd.org/changeset/base/245337
Log:
Define VM_KMEM_SIZE_MAX as a fraction of the kernel address space size
rather than a constant so that VM_KMEM_SIZE_MAX will scale automatically
with the kernel address
On 01/12/2013 09:46, Jayachandran C. wrote:
On Fri, Jan 11, 2013 at 11:16 PM, Alan Cox a...@rice.edu wrote:
On 01/11/2013 05:38, Jayachandran C. wrote:
[...]
I see an issue with commit on MIPS XLP platform as well.
With 16 GB physical memory, the ncallout is calculated to be 538881
(since
On 01/12/2013 10:56, Adrian Chadd wrote:
Hi,
I think this outlines a larger scale problem here, which is that way,
way too many things are relying on maxfiles here and it wasn't
properly reviewed or thought out before it made it into the tree.
So, can we either:
* review _all_ the places
On 01/11/2013 05:38, Jayachandran C. wrote:
On Tue, Jan 8, 2013 at 3:12 AM, Andre Oppermann an...@freebsd.org wrote:
On 07.01.2013 20:32, Alan Cox wrote:
On 01/07/2013 12:47, Oleksandr Tymoshenko wrote:
On 12/27/2012 6:46 PM, Oleksandr Tymoshenko wrote:
On 12/18/2012 1:59 AM, Alan Cox wrote
On 01/08/2013 11:21, Konstantin Belousov wrote:
On Tue, Jan 08, 2013 at 10:01:11AM -0700, Ian Lepore wrote:
I'm just learning the armv6/v7 stuff myself right now, but I can answer
your question for our current armv4/v5 implementation...
When there is more than one mapping of a page in v4/v5
On 01/07/2013 12:47, Oleksandr Tymoshenko wrote:
On 12/27/2012 6:46 PM, Oleksandr Tymoshenko wrote:
On 12/18/2012 1:59 AM, Alan Cox wrote:
On 12/17/2012 23:40, Oleksandr Tymoshenko wrote:
On 2012-12-08, at 1:21 PM, Alan Cox a...@rice.edu wrote:
On 12/08/2012 14:32, Andre Oppermann wrote
On 12/17/2012 23:40, Oleksandr Tymoshenko wrote:
On 2012-12-08, at 1:21 PM, Alan Cox a...@rice.edu wrote:
On 12/08/2012 14:32, Andre Oppermann wrote:
.. skipped ..
The trouble seems to come from NSFBUFS which is (512 + maxusers * 16)
resulting in a kernel map of (512 + 400 * 16) * PAGE_SIZE
Author: alc
Date: Wed Dec 19 05:07:27 2012
New Revision: 244417
URL: http://svnweb.freebsd.org/changeset/base/244417
Log:
Eliminate some definitions that haven't been used in a decade or more.
Modified:
head/sys/mips/include/param.h
Modified: head/sys/mips/include/param.h
On 12/09/2012 16:42, Andre Oppermann wrote:
On 09.12.2012 21:35, Alan Cox wrote:
Andre,
I believe that this change did not actually correct the overflow
problem. See below for an explanation.
On 11/29/2012 01:30, Andre Oppermann wrote:
Author: andre
Date: Thu Nov 29 07:30:42 2012
New
Andre,
I believe that this change did not actually correct the overflow
problem. See below for an explanation.
On 11/29/2012 01:30, Andre Oppermann wrote:
Author: andre
Date: Thu Nov 29 07:30:42 2012
New Revision: 243668
URL: http://svnweb.freebsd.org/changeset/base/243668
Log:
Using a
On 12/08/2012 14:32, Andre Oppermann wrote:
On 07.12.2012 23:17, Oleksandr Tymoshenko wrote:
On 12/7/2012 1:44 PM, Andre Oppermann wrote:
On 07.12.2012 22:05, Oleksandr Tymoshenko wrote:
On 12/7/2012 1:53 AM, Andre Oppermann wrote:
On 07.12.2012 10:36, Oleksandr Tymoshenko wrote:
On
Author: alc
Date: Sun Dec 9 00:32:38 2012
New Revision: 244043
URL: http://svnweb.freebsd.org/changeset/base/244043
Log:
In the past four years, we've added two new vm object types. Each time,
similar changes had to be made in various places throughout the machine-
independent virtual
I'm pretty sure that the realmem calculation is going to overflow on
i386/PAE, where the number of bytes of physical memory is greater than
the type long can represent.
On 11/27/2012 15:19, Andre Oppermann wrote:
Author: andre
Date: Tue Nov 27 21:19:58 2012
New Revision: 243631
URL:
Author: alc
Date: Wed Nov 28 18:29:34 2012
New Revision: 243659
URL: http://svnweb.freebsd.org/changeset/base/243659
Log:
Add support for the (relatively) new object type OBJT_MGTDEVICE to
vm_object_set_memattr(). Also, add a safety belt so that
vm_object_set_memattr() doesn't silently
Author: alc
Date: Sun Nov 25 19:42:36 2012
New Revision: 243529
URL: http://svnweb.freebsd.org/changeset/base/243529
Log:
Make a few small changes to vm_map_pmap_enter():
Add detail to the comment describing this function. In particular,
describe what MAP_PREFAULT_PARTIAL does.
Author: alc
Date: Wed Nov 21 06:26:18 2012
New Revision: 243366
URL: http://svnweb.freebsd.org/changeset/base/243366
Log:
Correct an error in r230623. When both VM_ALLOC_NODUMP and VM_ALLOC_ZERO
were specified to vm_page_alloc(), PG_NODUMP wasn't being set on the
allocated page when it
Author: alc
Date: Sat Nov 17 04:00:19 2012
New Revision: 243176
URL: http://svnweb.freebsd.org/changeset/base/243176
Log:
Update a comment to reflect the elimination of the hold queue in r242300.
Modified:
head/sys/vm/vm_page.h
Modified: head/sys/vm/vm_page.h
Author: alc
Date: Fri Nov 16 06:40:40 2012
New Revision: 243137
URL: http://svnweb.freebsd.org/changeset/base/243137
Log:
M_USE_RESERVE is no longer deprecated. It is once again supported.
Reviewed by: kib
Modified:
head/share/man/man9/malloc.9
Modified: head/share/man/man9/malloc.9
Author: alc
Date: Wed Nov 14 17:33:00 2012
New Revision: 243030
URL: http://svnweb.freebsd.org/changeset/base/243030
Log:
The function pmap_alloc_direct_page() unconditionally zeroes the returned
page. Therefore, it is really inappropriate for use by the function
uma_small_alloc(). The
On 11/13/2012 01:49, Andre Oppermann wrote:
On 13.11.2012 03:50, Alan Cox wrote:
Author: alc
Date: Tue Nov 13 02:50:39 2012
New Revision: 242941
URL: http://svnweb.freebsd.org/changeset/base/242941
Log:
Replace the single, global page queues lock with per-queue locks
on the
active
Author: alc
Date: Tue Nov 13 02:50:39 2012
New Revision: 242941
URL: http://svnweb.freebsd.org/changeset/base/242941
Log:
Replace the single, global page queues lock with per-queue locks on the
active and inactive paging queues.
Reviewed by: kib
Modified:
head/sys/sys/vmmeter.h
Author: alc
Date: Sat Nov 3 23:22:49 2012
New Revision: 242535
URL: http://svn.freebsd.org/changeset/base/242535
Log:
Replace all uses of the page queues lock by a R/W lock that is private
to this pmap.
Eliminate two redundant #include's.
Tested by:marcel
Modified:
Author: alc
Date: Thu Nov 1 16:20:02 2012
New Revision: 242434
URL: http://svn.freebsd.org/changeset/base/242434
Log:
In general, we call pmap_remove_all() before calling vm_page_cache(). So,
the call to pmap_remove_all() within vm_page_cache() is usually redundant.
This change eliminates
Author: alc
Date: Mon Oct 29 06:15:04 2012
New Revision: 242300
URL: http://svn.freebsd.org/changeset/base/242300
Log:
Replace the page hold queue, PQ_HOLD, by a new page flag, PG_UNHOLDFREE,
because the queue itself serves no purpose. When a held page is freed,
inserting the page into the
Author: alc
Date: Sat Oct 27 22:43:30 2012
New Revision: 242193
URL: http://svn.freebsd.org/changeset/base/242193
Log:
Eliminate a redundant TLB invalidation from pmap_pv_reclaim().
Modified:
head/sys/mips/mips/pmap.c
Modified: head/sys/mips/mips/pmap.c
On 10/25/2012 11:23, Andre Oppermann wrote:
On 25.10.2012 05:49, Bruce Evans wrote:
On Wed, 24 Oct 2012, Attilio Rao wrote:
On Wed, Oct 24, 2012 at 8:16 PM, Andre Oppermann an...@freebsd.org
wrote:
...
Let's go back and see how we can do this the sanest way. These are
the options I see at
Author: alc
Date: Fri Oct 26 03:02:39 2012
New Revision: 242121
URL: http://svn.freebsd.org/changeset/base/242121
Log:
Port the new PV entry allocator from amd64/i386. This allocator has two
advantages. First, PV entries are roughly half the size. Second, this
allocator doesn't access
On 10/23/2012 11:05, Gleb Smirnoff wrote:
On Tue, Oct 23, 2012 at 05:05:48PM +0200, Andre Oppermann wrote:
A There shouldn't be any users. Zero copy send is broken and
A responsible for random kernel crashes. Zero copy receive isn't
A supported by any modern driver. Both are useless to
There are couple minor grammar issues in the text. See below.
Alan
On 10/20/2012 07:53, Andre Oppermann wrote:
Author: andre
Date: Sat Oct 20 12:53:14 2012
New Revision: 241781
URL: http://svn.freebsd.org/changeset/base/241781
Log:
Hide the unfortunate named sysctl kern.ipc.somaxconn from
On 10/13/2012 08:55, Pawel Jakub Dawidek wrote:
On Fri, Oct 12, 2012 at 11:26:00PM +, Alan Cox wrote:
Author: alc
Date: Fri Oct 12 23:26:00 2012
New Revision: 241498
URL: http://svn.freebsd.org/changeset/base/241498
Log:
Replace all uses of the vm page queues lock by a new R/W lock
Author: alc
Date: Sat Oct 13 18:46:46 2012
New Revision: 241512
URL: http://svn.freebsd.org/changeset/base/241512
Log:
Eliminate the conditional for releasing the page queues lock in
vm_page_sleep(). vm_page_sleep() is no longer called with this lock
held.
Eliminate assertions that
Author: alc
Date: Sat Oct 13 20:19:43 2012
New Revision: 241517
URL: http://svn.freebsd.org/changeset/base/241517
Log:
Move vm_page_requeue() to the only file that uses it.
MFC after:3 weeks
Modified:
head/sys/vm/vm_page.c
head/sys/vm/vm_page.h
head/sys/vm/vm_pageout.c
Author: alc
Date: Sat Oct 13 22:22:53 2012
New Revision: 241518
URL: http://svn.freebsd.org/changeset/base/241518
Log:
MFC r241353, r241356, r241400
To avoid page table page corruption, change pmap_pv_reclaim()'s method
of mapping page table pages.
Add some assertions that were
Author: alc
Date: Sun Oct 14 03:40:08 2012
New Revision: 241520
URL: http://svn.freebsd.org/changeset/base/241520
Log:
Correct an error in pmap_pv_reclaim(). It can legitimately encounter
wired mappings. If it does, it should just skip them.
Modified:
head/sys/mips/mips/pmap.c
Modified:
Author: alc
Date: Fri Oct 12 23:26:00 2012
New Revision: 241498
URL: http://svn.freebsd.org/changeset/base/241498
Log:
Replace all uses of the vm page queues lock by a new R/W lock.
Unfortunately, this lock cannot be defined as static under Xen because it
is (ab)used to serialize queued
Author: alc
Date: Fri Oct 12 03:09:59 2012
New Revision: 241478
URL: http://svn.freebsd.org/changeset/base/241478
Log:
MFC r241155
Update some of the comments. In particular, use sleep in preference
to block where appropriate.
Eliminate some unnecessary casts.
Make a few
Author: alc
Date: Wed Oct 10 17:15:34 2012
New Revision: 241400
URL: http://svn.freebsd.org/changeset/base/241400
Log:
MFi386 r241356
Add several asserts.
MFC after:3 days
Modified:
head/sys/i386/xen/pmap.c
Modified: head/sys/i386/xen/pmap.c
Author: alc
Date: Mon Oct 8 16:57:05 2012
New Revision: 241353
URL: http://svn.freebsd.org/changeset/base/241353
Log:
In a few places, like the implementation of ptrace(), a thread may call
upon pmap_enter() to create a mapping within a different address space,
i.e., not the thread's own
Author: alc
Date: Sun Oct 7 18:07:45 2012
New Revision: 241313
URL: http://svn.freebsd.org/changeset/base/241313
Log:
Optimize the TLB invalidations performed by pmap_protect(). Use the new
range-based invalidation function instead of performing per-page
invalidations. Don't bother
Author: alc
Date: Sat Oct 6 19:05:50 2012
New Revision: 241276
URL: http://svn.freebsd.org/changeset/base/241276
Log:
Correct two pessimizations in pmap_extract_and_hold(). Test the PTE for
having PTE_RO set instead of PTE_D. This avoids some unnecessary failures
by
Author: alc
Date: Sat Oct 6 19:33:52 2012
New Revision: 241287
URL: http://svn.freebsd.org/changeset/base/241287
Log:
In general pmap implementations do not set the wired attribute on
the temporary mappings that are used to implement operations like
pmap_zero_page(). There is no reason
Author: alc
Date: Fri Oct 5 04:35:20 2012
New Revision: 241217
URL: http://svn.freebsd.org/changeset/base/241217
Log:
Eliminate a stale and a duplicated comment.
Modified:
head/sys/mips/mips/pmap.c
Modified: head/sys/mips/mips/pmap.c
Author: alc
Date: Tue Oct 2 07:14:22 2012
New Revision: 241123
URL: http://svn.freebsd.org/changeset/base/241123
Log:
Introduce a new TLB invalidation function for efficiently invalidating
address ranges, and use this function in pmap_remove().
Tested by:jchandra
Modified:
Author: alc
Date: Wed Oct 3 05:06:45 2012
New Revision: 241155
URL: http://svn.freebsd.org/changeset/base/241155
Log:
Tidy up a bit:
Update some of the comments. In particular, use sleep in preference to
block where appropriate.
Eliminate some unnecessary casts.
Make a few
Author: alc
Date: Wed Oct 3 05:42:15 2012
New Revision: 241156
URL: http://svn.freebsd.org/changeset/base/241156
Log:
Reimplement pmap_qremove() using the new TLB invalidation function for
efficiently invalidating address ranges.
Modified:
head/sys/mips/mips/pmap.c
Modified:
Author: alc
Date: Sat Sep 29 08:11:12 2012
New Revision: 241044
URL: http://svn.freebsd.org/changeset/base/241044
Log:
Update a comment to reflect recent locking changes.
Modified:
head/sys/arm/arm/pmap.c
Modified: head/sys/arm/arm/pmap.c
Author: alc
Date: Sat Sep 29 17:20:16 2012
New Revision: 241054
URL: http://svn.freebsd.org/changeset/base/241054
Log:
Add support for mincore(). Specifically, this is an adaptation of the
pmap_mincore() implementation that was added to the original arm pmap
in r235717.
Modified:
Author: alc
Date: Sat Sep 29 19:09:11 2012
New Revision: 241055
URL: http://svn.freebsd.org/changeset/base/241055
Log:
Eliminate unused variables.
Modified:
head/sys/arm/arm/pmap-v6.c
head/sys/arm/arm/pmap.c
Modified: head/sys/arm/arm/pmap-v6.c
Author: alc
Date: Sat Sep 29 22:28:00 2012
New Revision: 241058
URL: http://svn.freebsd.org/changeset/base/241058
Log:
Eliminate an unused declaration.
Modified:
head/sys/arm/include/pmap.h
Modified: head/sys/arm/include/pmap.h
Author: alc
Date: Sun Sep 30 03:54:57 2012
New Revision: 241063
URL: http://svn.freebsd.org/changeset/base/241063
Log:
Stop calling pmap_remove_write() from pmap_remove_all(). Doing so is not
only inefficient but also leads to recursive lock acquisition.
Tested by:ray
Modified:
Author: alc
Date: Fri Sep 28 05:30:59 2012
New Revision: 241020
URL: http://svn.freebsd.org/changeset/base/241020
Log:
Eliminate a stale comment. It describes another use case for the pmap in
Mach that doesn't exist in FreeBSD.
Modified:
head/sys/amd64/amd64/pmap.c
Author: alc
Date: Thu Sep 27 05:39:42 2012
New Revision: 240983
URL: http://svn.freebsd.org/changeset/base/240983
Log:
Implementing pmap_kextract(va) as pmap_extract(kernel_pmap, va) is
problematic because some callers to pmap_kextract() expect its
implementation to be lock-less. In
Author: alc
Date: Wed Sep 26 03:54:55 2012
New Revision: 240947
URL: http://svn.freebsd.org/changeset/base/240947
Log:
MFC r240862
Address a race condition that was introduced in r238212. Unless the page
queues lock is acquired before the page lock is released, there is no
Author: alc
Date: Tue Sep 25 03:59:10 2012
New Revision: 240913
URL: http://svn.freebsd.org/changeset/base/240913
Log:
Eliminate an unused declaration.
Modified:
head/sys/arm/arm/pmap-v6.c
head/sys/arm/arm/pmap.c
Modified: head/sys/arm/arm/pmap-v6.c
Author: alc
Date: Sun Sep 23 17:42:39 2012
New Revision: 240862
URL: http://svn.freebsd.org/changeset/base/240862
Log:
Address a race condition that was introduced in r238212. Unless the page
queues lock is acquired before the page lock is released, there is no
guarantee that the page will
Author: alc
Date: Sat Sep 22 06:54:03 2012
New Revision: 240803
URL: http://svn.freebsd.org/changeset/base/240803
Log:
Since UMA_ZONE_NOFREE is specified when l2zone and l2table_zone are created,
there is no need to release and reacquire the pmap and pvh global locks
around calls to
Author: alc
Date: Thu Sep 20 16:03:04 2012
New Revision: 240754
URL: http://svn.freebsd.org/changeset/base/240754
Log:
MFC r240317
Simplify pmap_unmapdev().
Don't set PTE_W on the page table entry in pmap_kenter{,_attr}() on MIPS.
Modified:
stable/9/sys/amd64/amd64/pmap.c
Author: alc
Date: Thu Sep 20 17:02:04 2012
New Revision: 240757
URL: http://svn.freebsd.org/changeset/base/240757
Log:
MFC r238456 and r238536
Various improvements to vm_contig_grow_cache().
Modified:
stable/9/sys/vm/vm_contig.c
Directory Properties:
stable/9/sys/ (props changed)
Author: alc
Date: Thu Sep 20 18:21:29 2012
New Revision: 240760
URL: http://svn.freebsd.org/changeset/base/240760
Log:
MFC r237168
The page flag PGA_WRITEABLE is set and cleared exclusively by the pmap
layer, but it is read directly by the MI VM layer. This change introduces
Author: alc
Date: Sat Sep 15 17:32:19 2012
New Revision: 240532
URL: http://svn.freebsd.org/changeset/base/240532
Log:
Eliminate an unused malloc type.
Modified:
head/sys/arm/arm/vm_machdep.c
Modified: head/sys/arm/arm/vm_machdep.c
Author: alc
Date: Thu Sep 13 06:16:49 2012
New Revision: 240442
URL: http://svn.freebsd.org/changeset/base/240442
Log:
Simplify the kernel pmap locking in pmap_enter_pv(). While I'm here, tidy
up the comments and whitespace.
Tested by:cognet
Modified:
head/sys/arm/arm/pmap.c
Author: alc
Date: Mon Sep 10 16:11:29 2012
New Revision: 240317
URL: http://svn.freebsd.org/changeset/base/240317
Log:
Simplify pmap_unmapdev(). Since kmem_free() eventually calls pmap_remove(),
pmap_unmapdev()'s own direct efforts to destroy the page table entries are
redundant, so
Author: alc
Date: Mon Sep 10 16:27:19 2012
New Revision: 240321
URL: http://svn.freebsd.org/changeset/base/240321
Log:
Replace all uses of the vm page queues lock by a r/w lock that is private
to this pmap.
Revise some comments.
The file vm/vm_param.h includes the file
Author: alc
Date: Sat Sep 8 17:43:20 2012
New Revision: 240241
URL: http://svn.freebsd.org/changeset/base/240241
Log:
pmap_remove:
Avoid re-walking the page table from the root for every PTE examined.
Tidy up some of pmap_remove()'s helper functions.
pmap_enter:
Set
Author: alc
Date: Fri Aug 31 02:59:44 2012
New Revision: 239934
URL: http://svn.freebsd.org/changeset/base/239934
Log:
Replace all uses of the vm page queues lock by a r/w lock that is private
to this pmap.
Tested by:Ian Lepore
Modified:
head/sys/arm/arm/pmap.c
Modified:
Author: alc
Date: Sat Aug 25 16:55:38 2012
New Revision: 239681
URL: http://svn.freebsd.org/changeset/base/239681
Log:
Retire PV_TABLE_MOD. When we destroy or write protect a dirty mapping,
we call vm_page_dirty(). Maintaining the PV_TABLE_MOD flag, in addition,
serves no useful purpose.
On 08/23/2012 06:46, John Baldwin wrote:
On Wednesday, August 22, 2012 6:48:50 pm Oleksandr Tymoshenko wrote:
Author: gonzo
Date: Wed Aug 22 22:48:50 2012
New Revision: 239597
URL: http://svn.freebsd.org/changeset/base/239597
Log:
Do not change cachable attribute for DMA memory allocated
Author: alc
Date: Fri Aug 17 20:15:01 2012
New Revision: 239361
URL: http://svn.freebsd.org/changeset/base/239361
Log:
Eliminate another vestige of page coloring.
Modified:
head/sys/mips/mips/uma_machdep.c
Modified: head/sys/mips/mips/uma_machdep.c
Author: alc
Date: Fri Aug 17 05:02:29 2012
New Revision: 239352
URL: http://svn.freebsd.org/changeset/base/239352
Log:
Fix two problems with pmap_clear_modify().
First, pmap_clear_modify() is write protecting all mappings to the specified
page, not just clearing the modified bit.
Author: alc
Date: Wed Aug 15 22:12:01 2012
New Revision: 239315
URL: http://svn.freebsd.org/changeset/base/239315
Log:
Correct a KASSERT message.
Submitted by: bde
Modified:
head/sys/kern/vfs_cluster.c
Modified: head/sys/kern/vfs_cluster.c
Author: alc
Date: Wed Aug 15 22:25:57 2012
New Revision: 239316
URL: http://svn.freebsd.org/changeset/base/239316
Log:
Eliminate some unused declarations.
Modified:
head/sys/sys/buf.h
Modified: head/sys/sys/buf.h
==
Author: alc
Date: Wed Aug 15 22:51:01 2012
New Revision: 239317
URL: http://svn.freebsd.org/changeset/base/239317
Log:
Replace all uses of the vm page queues lock by a r/w lock that is private
to this pmap.
Tidy up the #include's.
Remove the (now) unused #define PMAP_SHPGPERPROC.
Author: alc
Date: Thu Aug 16 04:41:15 2012
New Revision: 239321
URL: http://svn.freebsd.org/changeset/base/239321
Log:
Eliminate an unused parameter from init_pte_prot().
Eliminate stray whitespace within init_pte_prot().
Eliminate a gratuitous variable initialization from
Author: alc
Date: Mon Aug 13 17:38:38 2012
New Revision: 239236
URL: http://svn.freebsd.org/changeset/base/239236
Log:
Port the new PV entry allocator from amd64/i386. This allocator has two
advantages. First, PV entries are roughly half the size. Second, this
allocator doesn't access
Author: alc
Date: Thu Aug 9 16:38:17 2012
New Revision: 239152
URL: http://svn.freebsd.org/changeset/base/239152
Log:
Merge r132141 and r111272 from amd64/i386:
Reduce the size of a PV entry by eliminating pv_ptem. There is no need
to store a pointer to the page table page in the PV
On 08/09/2012 05:45, Peter Jeremy wrote:
On 2012-Jul-16 18:13:43 +, Alan Coxa...@freebsd.org wrote:
Various improvements to vm_contig_grow_cache(). Most notably, even when
it can't sleep, it can still move clean pages from the inactive queue to
the cache. Also, when a page is
Author: alc
Date: Fri Aug 10 05:00:50 2012
New Revision: 239170
URL: http://svn.freebsd.org/changeset/base/239170
Log:
Merge r134393 from amd64/i386:
The machine-independent parts of the virtual memory system always pass a
valid pmap to the pmap functions that require one. Remove the
Author: alc
Date: Fri Aug 10 05:47:04 2012
New Revision: 239171
URL: http://svn.freebsd.org/changeset/base/239171
Log:
Eliminate an unnecessary acquisition and release of the page queues lock
from pmap_pte(). PT_SET_MA() is not a queued mapping update, but instead
an immediate mapping
Author: alc
Date: Wed Aug 8 05:28:30 2012
New Revision: 239137
URL: http://svn.freebsd.org/changeset/base/239137
Log:
The assertion that I added in r238889 could legitimately fail when a
debugger creates a breakpoint. Replace that assertion with a narrower
one that still achieves my
Author: alc
Date: Tue Aug 7 04:48:14 2012
New Revision: 239121
URL: http://svn.freebsd.org/changeset/base/239121
Log:
Never sleep on busy pages in vm_pageout_launder(), always skip them. Long
ago, sleeping on busy pages in vm_pageout_launder() made sense. The call
to vm_pageout_flush()
Author: alc
Date: Sun Aug 5 16:59:02 2012
New Revision: 239072
URL: http://svn.freebsd.org/changeset/base/239072
Log:
Shave off a few more cycles from the average execution time of pmap_enter()
by simplifying the control flow and reducing the live range of om.
Modified:
Author: alc
Date: Fri Aug 3 01:48:15 2012
New Revision: 238998
URL: http://svn.freebsd.org/changeset/base/238998
Log:
Inline vm_page_aflags_clear() and vm_page_aflags_set().
Add comments stating that neither these functions nor the flags that they
are used to manipulate are part of the
On 08/02/2012 20:48, Alan Cox wrote:
Author: alc
Date: Fri Aug 3 01:48:15 2012
New Revision: 238998
URL: http://svn.freebsd.org/changeset/base/238998
Log:
Inline vm_page_aflags_clear() and vm_page_aflags_set().
Add comments stating that neither these functions nor the flags
Author: alc
Date: Wed Aug 1 16:04:13 2012
New Revision: 238970
URL: http://svn.freebsd.org/changeset/base/238970
Log:
Revise pmap_enter()'s handling of mapping updates that change the
PTE's PG_M and PG_RW bits but not the physical page frame. First,
only perform vm_page_dirty() on a
Author: alc
Date: Mon Jul 30 20:38:37 2012
New Revision: 238915
URL: http://svn.freebsd.org/changeset/base/238915
Log:
Eliminate an unneeded declaration. (I should have removed this as part
of r227568.)
Modified:
head/sys/vm/vm_page.h
Modified: head/sys/vm/vm_page.h
Author: alc
Date: Sun Jul 29 18:20:49 2012
New Revision: 238889
URL: http://svn.freebsd.org/changeset/base/238889
Log:
Shave off a few more cycles from pmap_enter()'s critical section. In
particular, do a little less work with the PV list lock held.
Modified:
head/sys/amd64/amd64/pmap.c
Author: alc
Date: Tue Jul 24 02:35:30 2012
New Revision: 238732
URL: http://svn.freebsd.org/changeset/base/238732
Log:
Addendum to r238604. If the inactive queue scan isn't restarted, then
the variable addl_page_shortage_init isn't needed.
X-MFC after: r238604
Modified:
601 - 700 of 1285 matches
Mail list logo