22 Mar, 2006
40 commits
-
Remove __put_page from outside the core mm/. It is dangerous because it does
not handle compound pages nicely, and misses 1->0 transitions. If a user
later appears that really needs the extra speed we can reevaluate.Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Remove page_count and __put_page from x86-64 pageattr
Signed-off-by: Nick Piggin
Acked-by: Andi Kleen
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Use page->lru.next to implement the singly linked list of pages rather than
the struct deferred_page which needs to be allocated and freed for each
page.Signed-off-by: Nick Piggin
Acked-by: Andi Kleen
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Stop using __put_page and page_count in i386 pageattr.c
Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
sg increments the refcount of constituent pages in its higher order memory
allocations when they are about to be mapped by userspace. This is done so
the subsequent get_page/put_page when doing the mapping and unmapping does not
free the page.Move over to the preferred way, that is, using compound pages instead. This
fixes a whole class of possible obscure bugs where a get_user_pages on a
constituent page may outlast the user mappings or even the driver.Signed-off-by: Nick Piggin
Cc: Hugh Dickins
Cc: Douglas Gilbert
Cc: James Bottomley
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Now that it's madvisable, remove two pieces of VM_DONTCOPY bogosity:
1. There was and is no logical reason why VM_DONTCOPY should be in the
list of flags which forbid vma merging (and those drivers which set
it are also setting VM_IO, which itself forbids the merge).2. It's hard to understand the purpose of the VM_HUGETLB, VM_DONTCOPY
block in vm_stat_account: but never mind, it's under CONFIG_HUGETLB,
which (unlike CONFIG_HUGETLB_PAGE or CONFIG_HUGETLBFS) has never been
defined.Signed-off-by: Hugh Dickins
Cc: William Lee Irwin III
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
In shrink_inactive_list(), nr_scan is not accounted when nr_taken is 0.
But 0 pages taken does not mean 0 pages scanned.Move the goto statement below the accounting code to fix it.
Signed-off-by: Wu Fengguang
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
In isolate_lru_pages(), *scanned reports one more scan because the scan
counter is increased one more time on exit of the while-loop.Change the while-loop to for-loop to fix it.
Signed-off-by: Nick Piggin
Signed-off-by: Wu Fengguang
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Add some comments to explain how zone reclaim works. And it fixes the
following issues:- PF_SWAPWRITE needs to be set for RECLAIM_SWAP to be able to write
out pages to swap. Currently RECLAIM_SWAP may not do that.- remove setting nr_reclaimed pages after slab reclaim since the slab shrinking
code does not use that and the nr_reclaimed pages is just right for the
intended follow up action.Signed-off-by: Christoph Lameter
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
We have:
try_to_free_pages
->shrink_caches(struct zone **zones, ..)
->shrink_zone(struct zone *, ...)
->shrink_cache(struct zone *, ...)
->shrink_list(struct list_head *, ...)
->refill_inactive_list((struct zone *, ...)which is fairly irrational.
Rename things so that we have
try_to_free_pages
->shrink_zones(struct zone **zones, ..)
->shrink_zone(struct zone *, ...)
->shrink_inactive_list(struct zone *, ...)
->shrink_page_list(struct list_head *, ...)
->shrink_active_list(struct zone *, ...)Cc: Nick Piggin
Cc: Christoph Lameter
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Change all the vmscan functions to retunr the number-of-reclaimed pages and
remove scan_conrtol.nr_reclaimed.Saves ten-odd bytes of text and makes things clearer and more consistent.
The patch also changes the behaviour of zone_reclaim() when it falls back to slab shrinking. Christoph says
"Setting this to one means that we will rescan and shrink the slab for
each allocation if we are out of zone memory and RECLAIM_SLAB is set. Plus
if we do an order 0 allocation we do not go off node as intended."We better set this to zero. This means the allocation will go offnode
despite us having potentially freed lots of memory on the zone. Future
allocations can then again be done from this zone."Cc: Nick Piggin
Cc: Christoph Lameter
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Turn basically everything in vmscan.c into `unsigned long'. This is to avoid
the possibility that some piece of code in there might decide to operate upon
more than 4G (or even 2G) of pages in one hit.This might be silly, but we'll need it one day.
Cc: Christoph Lameter
Cc: Nick Piggin
Signed-off-by: Rafael J. Wysocki
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Initialise as much of scan_control as possible at the declaration site. This
tidies things up a bit and assures us that all unmentioned fields are zeroed
out.Signed-off-by: Christoph Lameter
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Make nr_to_scan and priority a parameter instead of putting it into scan
control. This allows various small optimizations and IMHO makes the code
easier to read.Signed-off-by: Christoph Lameter
Cc: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Slab duplicates on_each_cpu().
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
When on_each_cpu() runs the callback on other CPUs, it runs with local
interrupts disabled. So we should run the function with local interrupts
disabled on this CPU, too.And do the same for UP, so the callback is run in the same environment on both
UP and SMP. (strictly it should do preempt_disable() too, but I think
local_irq_disable is sufficiently equivalent).Also uninlines on_each_cpu(). softirq.c was the most appropriate file I could
find, but it doesn't seem to justify creating a new file.Oh, and fix up that comment over (under?) x86's smp_call_function(). It
drives me nuts.Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
SLAB_NO_REAP is documented as an option that will cause this slab not to be
reaped under memory pressure. However, that is not what happens. The only
thing that SLAB_NO_REAP controls at the moment is the reclaim of the unused
slab elements that were allocated in batch in cache_reap(). Cache_reap()
is run every few seconds independently of memory pressure.Could we remove the whole thing? Its only used by three slabs anyways and
I cannot find a reason for having this option.There is an additional problem with SLAB_NO_REAP. If set then the recovery
of objects from alien caches is switched off. Objects not freed on the
same node where they were initially allocated will only be reused if a
certain amount of objects accumulates from one alien node (not very likely)
or if the cache is explicitly shrunk. (Strangely __cache_shrink does not
check for SLAB_NO_REAP)Getting rid of SLAB_NO_REAP fixes the problems with alien cache freeing.
Signed-off-by: Christoph Lameter
Cc: Pekka Enberg
Cc: Manfred Spraul
Cc: Mark Fasheh
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Fix kernel-doc warnings in mm/slab.c.
Signed-off-by: Randy Dunlap
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
We have struct kmem_cache now so use it instead of the old typedef.
Signed-off-by: Pekka Enberg
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Remove cachep->spinlock. Locking has moved to the kmem_list3 and most of
the structures protected earlier by cachep->spinlock is now protected by
the l3->list_lock. slab cache tunables like batchcount are accessed always
with the cache_chain_mutex held.Patch tested on SMP and NUMA kernels with dbench processes running,
constant onlining/offlining, and constant cache tuning, all at the same
time.Signed-off-by: Ravikiran Thirumalai
Cc: Christoph Lameter
Cc: Pekka Enberg
Cc: Manfred Spraul
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
slab.c has become a bit revolting again. Try to repair it.
- Coding style fixes
- Don't do assignments-in-if-statements.
- Don't typecast assignments to/from void*
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Extract setup_cpu_cache() function from kmem_cache_create() to make the
latter a little less complex.Signed-off-by: Pekka Enberg
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Clean up the object to index mapping that has been spread around mm/slab.c.
Signed-off-by: Pekka Enberg
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Since size_t has the same size as a long on all architectures, it's enough
for overflow checks to check against ULONG_MAX.This change could allow a compiler better optimization (especially in the
n=1 case).The practical effect seems to be positive, but quite small:
text data bss dec hex filename
21762380 5859870 1848928 29471178 1c1b1ca vmlinux-old
21762211 5859870 1848928 29471009 1c1b121 vmlinux-patchedSigned-off-by: Adrian Bunk
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Insert "fresh" huge pages into the hugepage allocator by the same means as
they are freed back into it. This reduces code size and allows
enqueue_huge_page to be inlined into the hugepage free fastpath.Eliminate occurances of hugepages on the free list with non-zero refcount.
This can allow stricter refcount checks in future. Also required for
lockless pagecache.Signed-off-by: Nick Piggin
"This patch also eliminates a leak "cleaned up" by re-clobbering the
refcount on every allocation from the hugepage freelists. With respect to
the lockless pagecache, the crucial aspect is to eliminate unconditional
set_page_count() to 0 on pages with potentially nonzero refcounts, though
closer inspection suggests the assignments removed are entirely spurious."Acked-by: William Irwin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The bootmem code added to page_alloc.c duplicated some page freeing code
that it really doesn't need to because it is not so performance critical.While we're here, make prefetching work properly by actually prefetching
the page we're about to use before prefetching ahead to the next one (ie.
get the most important transaction started first). Also prefetch just a
single page ahead rather than leaving a gap of 16.Jack Steiner reported no problems with SGI's ia64 simulator.
Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Clarify that preemption needs to be guarded against with the
__xxx_page_state functions.Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Have an explicit mm call to split higher order pages into individual pages.
Should help to avoid bugs and be more explicit about the code's intention.Signed-off-by: Nick Piggin
Cc: Russell King
Cc: David Howells
Cc: Ralf Baechle
Cc: Benjamin Herrenschmidt
Cc: Paul Mundt
Cc: "David S. Miller"
Cc: Chris Zankel
Signed-off-by: Yoichi Yuasa
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
- Don't return uninitialised stack values in case of allocation failure
- Don't bother clearing PageCompound because __GFP_COMP wasn't specified
Increment over the pte page rather than one pte entry in
pte_alloc_one_kernel- Actually increment the page pointer in pte_alloc_one
- Compile fixes, typos.
Signed-off-by: Nick Piggin
Acked-by: Chris Zankel
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
atomic_add_unless (atomic_inc_not_zero) no longer requires an offset refcount
to function correctly.Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The VM has an interesting race where a page refcount can drop to zero, but it
is still on the LRU lists for a short time. This was solved by testing a 0->1
refcount transition when picking up pages from the LRU, and dropping the
refcount in that case.Instead, use atomic_add_unless to ensure we never pick up a 0 refcount page
from the LRU, thus a 0 refcount page will never have its refcount elevated
until it is allocated again.Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Atomic operation removal from slab
Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
More atomic operation removal from page allocator
Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
In the page release paths, we can be sure that nobody will mess with our
page->flags because the refcount has dropped to 0. So no need for atomic
operations here.Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
PG_active is protected by zone->lru_lock, it does not need TestSet/TestClear
operations.Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
PG_lru is protected by zone->lru_lock. It does not need TestSet/TestClear
operations.Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
If vmscan finds a zero refcount page on the lru list, never ClearPageLRU
it. This means the release code need not hold ->lru_lock to stabilise
PageLRU, so that lock may be skipped entirely when releasing !PageLRU pages
(because we know PageLRU won't have been temporarily cleared by vmscan,
which was previously guaranteed by holding the lock to synchronise against
vmscan).Signed-off-by: Nick Piggin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
set_pgdir isn't needed anymore for a very long time. Remove the leftover
implementation on sh64 and the stub on s390.Signed-off-by: Christoph Hellwig
Cc: Heiko Carstens
Cc: Martin Schwidefsky
Cc: Paul Mundt
Cc: Richard Curnow
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Do not use platform_device_register_simple() as it is going away, define
dcdbas_driver and implement ->probe() and ->remove() functions so manual
binding and unbinding will work with this driver.Also switch to using attribute_group when creating sysfs attributes and
make sure to check and handle errors; explicitely remove attributes when
detaching driver.Signed-off-by: Dmitry Torokhov
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Do not use platform_device_register_simple() as it is going away.
Signed-off-by: Dmitry Torokhov
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
