Pawel Jakub Dawidek wrote:
On Thu, Mar 05, 2009 at 12:06:21PM -0600, Alan Cox wrote:
Pawel Jakub Dawidek wrote:
This assertion is triggered on my 7-STABLE machine. It was fine when it
was running 6.1, but now after upgrade it panics. If this assert is
bogus could you MFC it to 7-STABLE
On Thu, Mar 05, 2009 at 12:06:21PM -0600, Alan Cox wrote:
> Pawel Jakub Dawidek wrote:
> >This assertion is triggered on my 7-STABLE machine. It was fine when it
> >was running 6.1, but now after upgrade it panics. If this assert is
> >bogus could you MFC it to 7-STABLE as well?
> >
> >
>
> Hmm.
Pawel Jakub Dawidek wrote:
On Wed, Dec 19, 2007 at 05:47:51AM +, Alan Cox wrote:
alc 2007-12-19 05:47:51 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Eliminate redundant code from vm_page_startup().
Revision ChangesPath
On Wed, Dec 19, 2007 at 05:47:51AM +, Alan Cox wrote:
> alc 2007-12-19 05:47:51 UTC
>
> FreeBSD src repository
>
> Modified files:
> sys/vm vm_page.c
> Log:
> Eliminate redundant code from vm_page_startup().
>
> Revision ChangesPath
> 1.363 +
emaste 2008-09-26 18:44:40 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c vm_page.h
Log:
SVN rev 183389 on 2008-09-26 18:44:40Z by emaste
Move CTASSERT from header file to source file, per implementation note now
in the CTASSERT man page.
R
jhb 2008-09-23 15:59:38 UTC
FreeBSD src repository
Modified files:(Branch: RELENG_7)
sys/vm vm_page.c
Log:
SVN rev 183302 on 2008-09-23 15:59:38Z by jhb
MFC: 179623
Essentially, neither madvise(..., MADV_DONTNEED) nor madvise(..., MADV_FREE)
wor
kmacy 2008-08-17 23:40:29 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
SVN rev 181811 on 2008-08-17 23:40:29Z by kmacy
Work around differences in page allocation for initial page tables on xen
MFC after: 1 month
Revision Ch
In message <[EMAIL PROTECTED]>, Alan Cox writes:
>Poul-Henning Kamp wrote:
>>> Essentially, neither madvise(..., MADV_DONTNEED) nor madvise(...,
>>> MADV_FREE)
>>> work. (Moreover, I don't believe that they have ever worked as intended.)
>>
>> I remember they did have a measurable but minor im
Poul-Henning Kamp wrote:
In message <[EMAIL PROTECTED]>, Alan Cox writes:
alc 2008-06-06 18:38:43 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
SVN rev 179623 on 2008-06-06 18:38:43Z by alc
Essentially, neither madvise(..., MADV_DONTNEED)
In message <[EMAIL PROTECTED]>, Alan Cox writes:
>alc 2008-06-06 18:38:43 UTC
>
> FreeBSD src repository
>
> Modified files:
>sys/vm vm_page.c
> Log:
> SVN rev 179623 on 2008-06-06 18:38:43Z by alc
>
> Essentially, neither madvise(..., MADV_DONTNEED) nor madvise(..
alc 2008-06-06 18:38:43 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
SVN rev 179623 on 2008-06-06 18:38:43Z by alc
Essentially, neither madvise(..., MADV_DONTNEED) nor madvise(..., MADV_FREE)
work. (Moreover, I don't believe that they
On Thu, 15 May 2008 18:52:31 + (UTC), Alan Cox <[EMAIL PROTECTED]> wrote:
> alc 2008-05-15 18:52:31 UTC
>
> FreeBSD src repository
>
> Modified files:
> sys/vm vm_page.c
> Log:
> Don't call vm_reserv_alloc_page() on device-backed objects.
> Otherwise, the sys
On Thu, May 15, 2008 at 06:52:31PM +, Alan Cox wrote:
> Modified files:
> sys/vm vm_page.c
> Log:
> Don't call vm_reserv_alloc_page() on device-backed objects. Otherwise, the
> system may panic because there is no reservation structure corresponding to
> the physic
alc 2008-05-15 18:52:31 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Don't call vm_reserv_alloc_page() on device-backed objects. Otherwise, the
system may panic because there is no reservation structure corresponding to
the physical addr
alc 2008-04-06 18:09:28 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c vm_phys.c vm_reserv.c
vm_reserv.h
Log:
Introduce vm_reserv_reclaim_contig(). This function is used by
contigmalloc(9) as a last resort to steal pages
alc 2008-03-19 20:24:36 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c vm_page.h vm_pageout.c
Log:
Rename vm_pageq_requeue() to vm_page_requeue() on account of its recent
migration to vm/vm_page.c.
Revision ChangesPath
1.368 +2 -2
alc 2008-01-02 04:43:47 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Defer setting either PG_CACHED or PG_FREE until after the free page
queues lock is acquired. Otherwise, the state of a reservation's
pages' flags and its population cou
On Thu, Dec 20, 2007 at 10:45:55PM +, Alan Cox wrote:
> alc 2007-12-20 22:45:54 UTC
>
> FreeBSD src repository
>
> Modified files:
> sys/vm vm_page.c vm_phys.c vm_phys.h
> Log:
> Modify vm_phys_unfree_page() so that it no longer requires the given
> page t
alc 2007-12-20 22:45:54 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c vm_phys.c vm_phys.h
Log:
Modify vm_phys_unfree_page() so that it no longer requires the given
page to be in the free lists. Instead, it now returns TRUE if it
removed the pa
alc 2007-12-19 05:47:51 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Eliminate redundant code from vm_page_startup().
Revision ChangesPath
1.363 +0 -17 src/sys/vm/vm_page.c
___
c
alc 2007-12-11 21:20:34 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Simplify vm_page_free_toq().
Revision ChangesPath
1.362 +3 -5 src/sys/vm/vm_page.c
___
cvs-all@freebsd.org m
alc 2007-12-02 07:43:43 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Correct a comment.
Revision ChangesPath
1.361 +1 -1 src/sys/vm/vm_page.c
___
cvs-all@freebsd.org mailing lis
alc 2007-11-28 22:23:35 UTC
FreeBSD src repository
Modified files:(Branch: RELENG_7)
sys/vm vm_page.c
Log:
MFC revision 1.360
When reactivating a cached page, reset the page's pool to the default
pool. (Not doing this before was a performance pe
alc 2007-11-21 23:22:10 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
When reactivating a cached page, reset the page's pool to the default
pool. (Not doing this before was a performance pessimization but not
a cause for panic.)
Revi
kib 2007-11-08 14:03:00 UTC
FreeBSD src repository
Modified files:(Branch: RELENG_7)
sys/vm vm_page.c
Log:
MFC rev. 1.359:
Do not cache the page belonging to the swap object that has no backing
store allocated in the pager.
Approved by:re (k
kib 2007-11-05 10:25:13 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
The intent of the freeing the (zeroed) page in vm_page_cache() for
default object rather than cache it was to have
vm_pager_has_page(object, pindex, ...) == FALSE to imp
alc 2007-10-31 02:27:29 UTC
FreeBSD src repository
Modified files:(Branch: RELENG_7)
sys/vm vm_page.c
Log:
MFC revision 1.358
Change vm_page_cache_transfer() such that it does not transfer pages
that would have an offset beyond the end of the tar
alc 2007-10-27 00:09:30 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Change vm_page_cache_transfer() such that it does not transfer pages
that would have an offset beyond the end of the target object. Such
pages should remain in the sour
alc 2007-10-08 18:01:39 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
In the rare case that vm_page_cache() actually frees the given page,
it must first ensure that the page is no longer mapped. This is
trivially accomplished by calling p
alc 2007-07-14 21:21:18 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c vm_phys.c vm_phys.h
Log:
Eliminate two unused functions: vm_phys_alloc_pages() and
vm_phys_free_pages(). Rename vm_phys_alloc_pages_locked() to
vm_phys_alloc_pages() and vm_
alc 2007-07-10 18:41:34 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Correct a problem in the ZERO_COPY_SOCKETS option, specifically, in
vm_page_cowfault(). Initially, if vm_page_cowfault() sleeps, the given
page is wired, preventing it
mjacob 2007-06-17 04:19:05 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Don't declare inline a function which isn't.
Revision ChangesPath
1.350 +1 -1 src/sys/vm/vm_page.c
___
cvs-a
alc 2007-06-16 21:07:51 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
If attempting to cache a "busy", panic instead of printing a diagnostic
message and returning.
Revision ChangesPath
1.349 +1 -2 src/sys/vm/vm_page.c
_
alc 2007-02-18 05:54:43 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Enable vm_page_free() and vm_page_free_zero() to be called on some pages
without the page queues lock being held, specifically, pages that are not
contained in a vm obje
alc 2007-02-17 19:37:00 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Remove a stale comment. Add punctuation to a nearby comment.
Revision ChangesPath
1.339 +1 -6 src/sys/vm/vm_page.c
alc 2007-02-15 05:43:38 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Relax the page queue lock assertions in vm_page_remove() and
vm_page_free_toq() to account for recent changes that allow
vm_page_free_toq() to be called on some pages wi
alc 2007-02-14 07:05:55 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Avoid the unnecessary acquisition of the free page queues lock when a page
is actually being added to the hold queue, not the free queue. At the same
time, avoid unnece
alc 2007-02-11 05:18:40 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c vm_zeroidle.c
Log:
Use the free page queue mutex instead of the page queue mutex to
synchronize sleeping and waking of the zero idle thread.
Revision ChangesPath
1.
alc 2007-02-07 06:37:30 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c vm_pageout.c
Log:
Change the pagedaemon, vm_wait(), and vm_waitpfault() to sleep on the
vm page queue free mutex instead of the vm page queue mutex.
Revision ChangesP
kmacy 2006-12-08 08:44:47 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Remove the requirement that phys_avail be sorted in ascending order
by explicitly finding the lowest and highest addresses when calculating
the size of the vm_pages arra
alc 2006-11-08 19:11:54 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
I misplaced the assertion that was added to vm_page_startup() in the
previous change. Correct its placement.
Revision ChangesPath
1.332 +6 -6 src/sys
alc 2006-11-08 18:43:47 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Simplify the construction of the free queues in vm_page_startup(). Add
an assertion to test a hypothesis concerning other redundant computation
in vm_page_startup().
alc 2006-09-29 04:51:16 UTC
FreeBSD src repository
Modified files:(Branch: RELENG_6)
sys/vm vm_page.c
Log:
MFC revisions 1.313 and 1.328:
When the VM needs to allocated physical memory pages (for non interrupt use)
and it has not plenty of free pa
kensmith2006-09-29 00:20:56 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Fix two minor style(9) nits in v1.313 which were noticed during an
MFC review. alc@ will be MFCing V1.313 plus style fix to RELENG_6.
Revision ChangesPath
1.3
alc 2006-08-27 19:50:13 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c vm_page.h
Log:
Refactor vm_page_sleep_if_busy() so that the test for a busy page is
inlined and a procedure call is made in the rare case, i.e., when it is
necessary to sleep
alc 2006-08-21 00:34:31 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Page flags are reset on (re)allocation. There is no need to clear any
flags except for PG_ZERO in vm_page_free_toq().
Revision ChangesPath
1.326 +0 -8
alc 2006-08-10 04:48:29 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Change vm_page_cowfault() so that it doesn't allocate a pre-busied page.
Revision ChangesPath
1.323 +1 -2 src/sys/vm/vm_page.c
_
alc 2006-08-06 00:15:40 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Change vm_page_sleep_if_busy() so that it no longer requires the caller to
hold the page queues lock.
Revision ChangesPath
1.321 +2 -1 src/sys/vm/vm_p
jhb 2006-07-05 19:08:49 UTC
FreeBSD src repository
Modified files:(Branch: RELENG_6)
sys/vm vm_page.c vm_pageq.c
Log:
MFC: Move the code to handle the vm.blacklist tunable up a layer.
Revision ChangesPath
1.304.2.3 +39 -1 src/sys/vm/vm_
jhb 2006-06-23 16:44:24 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c vm_pageq.c
Log:
Move the code to handle the vm.blacklist tunable up a layer into
vm_page_startup(). As a result, we now only lookup the tunable once
instead of looking it up
ps 2006-05-31 22:55:23 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Fix minidumps to include pages allocated via pmap_map on amd64.
These pages are allocated from the direct map, and were not previous
tracked. This included the vm_page_
ups 2006-02-15 22:29:53 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
When the VM needs to allocated physical memory pages (for non interrupt use)
and it has not plenty of free pages it tries to free pages in the cache queue.
Unfortunately
alc 2006-01-31 19:06:52 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Change #if defined(DIAGNOSTIC) to KASSERT.
Revision ChangesPath
1.312 +3 -4 src/sys/vm/vm_page.c
___
cvs-all
alc 2006-01-24 07:21:38 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
In vm_page_set_invalid() invalidate all of the page's mappings as soon as
any part of the page's contents is invalidated.
Submitted by: tegge
Revision Changes
alc 2005-12-13 19:59:10 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
Assert that the page that is given to vm_page_free_toq() does not have any
managed mappings.
Revision ChangesPath
1.309 +2 -0 src/sys/vm/vm_page.c
___
alc 2005-11-13 07:38:16 UTC
FreeBSD src repository
Modified files:(Branch: RELENG_6)
sys/vm vm_page.c
Log:
MFC revisions 1.307 and 1.308
Consider the zero-copy transmission of a page that was wired by mlock(2).
If a copy-on-write fault occurs on
On Tue, 8 Nov 2005, Alan Cox wrote:
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.
alc 2005-11-08 06:33:21 UTC
FreeBSD src repository
Modified files:
sys/vm vm_page.c
Log:
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
58 matches
Mail list logo
