| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Page migration's __unmap_and_move(), and rmap's try_to_unmap(), were
created for use on pages almost certainly mapped into userspace. But
nowadays compaction often applies them to unmapped page cache pages: which
may exacerbate contention on i_mmap_rwsem quite unnecessarily, since
try_to_unmap_file() makes no preliminary page_mapped() check.
Now check page_mapped() in __unmap_and_move(); and avoid repeating the
same overhead in rmap_walk_file() - don't remove_migration_ptes() when we
never inserted any.
(The PageAnon(page) comment blocks now look even sillier than before, but
clean that up on some other occasion. And note in passing that
try_to_unmap_one() does not use a migration entry when PageSwapCache, so
remove_migration_ptes() will then not update that swap entry to newpage
pte: not a big deal, but something else to clean up later.)
Davidlohr remarked in "mm,fs: introduce helpers around the i_mmap_mutex"
conversion to i_mmap_rwsem, that "The biggest winner of these changes is
migration": a part of the reason might be all of that unnecessary taking
of i_mmap_mutex in page migration; and it's rather a shame that I didn't
get around to sending this patch in before his - this one is much less
useful after Davidlohr's conversion to rwsem, but still good.
Signed-off-by: Hugh Dickins <[email protected]>
Cc: Davidlohr Bueso <[email protected]>
Cc: Rik van Riel <[email protected]>
Cc: Mel Gorman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Since commit 058504edd026 ("fs/seq_file: fallback to vmalloc allocation"),
seq_buf_alloc() falls back to vmalloc() when the kmalloc() for contiguous
memory fails. This was done to address order-4 slab allocations for
reading /proc/stat on large machines and noticed because
PAGE_ALLOC_COSTLY_ORDER < 4, so there is no infinite loop in the page
allocator when allocating new slab for such high-order allocations.
Contiguous memory isn't necessary for caller of seq_buf_alloc(), however.
Other GFP_KERNEL high-order allocations that are <=
PAGE_ALLOC_COSTLY_ORDER will simply loop forever in the page allocator and
oom kill processes as a result.
We don't want to kill processes so that we can allocate contiguous memory
in situations when contiguous memory isn't necessary.
This patch does the kmalloc() allocation with __GFP_NORETRY for high-order
allocations. This still utilizes memory compaction and direct reclaim in
the allocation path, the only difference is that it will fail immediately
instead of oom kill processes when out of memory.
[[email protected]: add comment]
Signed-off-by: David Rientjes <[email protected]>
Cc: Heiko Carstens <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Al Viro <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
The slab shrinkers are currently invoked from the zonelist walkers in
kswapd, direct reclaim, and zone reclaim, all of which roughly gauge the
eligible LRU pages and assemble a nodemask to pass to NUMA-aware
shrinkers, which then again have to walk over the nodemask. This is
redundant code, extra runtime work, and fairly inaccurate when it comes to
the estimation of actually scannable LRU pages. The code duplication will
only get worse when making the shrinkers cgroup-aware and requiring them
to have out-of-band cgroup hierarchy walks as well.
Instead, invoke the shrinkers from shrink_zone(), which is where all
reclaimers end up, to avoid this duplication.
Take the count for eligible LRU pages out of get_scan_count(), which
considers many more factors than just the availability of swap space, like
zone_reclaimable_pages() currently does. Accumulate the number over all
visited lruvecs to get the per-zone value.
Some nodes have multiple zones due to memory addressing restrictions. To
avoid putting too much pressure on the shrinkers, only invoke them once
for each such node, using the class zone of the allocation as the pivot
zone.
For now, this integrates the slab shrinking better into the reclaim logic
and gets rid of duplicative invocations from kswapd, direct reclaim, and
zone reclaim. It also prepares for cgroup-awareness, allowing
memcg-capable shrinkers to be added at the lruvec level without much
duplication of both code and runtime work.
This changes kswapd behavior, which used to invoke the shrinkers for each
zone, but with scan ratios gathered from the entire node, resulting in
meaningless pressure quantities on multi-zone nodes.
Zone reclaim behavior also changes. It used to shrink slabs until the
same amount of pages were shrunk as were reclaimed from the LRUs. Now it
merely invokes the shrinkers once with the zone's scan ratio, which makes
the shrinkers go easier on caches that implement aging and would prefer
feeding back pressure from recently used slab objects to unused LRU pages.
[[email protected]: assure class zone is populated]
Signed-off-by: Johannes Weiner <[email protected]>
Cc: Dave Chinner <[email protected]>
Signed-off-by: Vladimir Davydov <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
These flushes deal with sequence number overflows, such as for long lived
threads. These are rare, but interesting from a debugging PoV. As such,
display the number of flushes when vmacache debugging is enabled.
Signed-off-by: Davidlohr Bueso <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
page owner is for the tracking about who allocated each page. This
document explains what is the page owner feature and what is the merit of
it. And, simple HOW-TO is also explained. See the document for detailed
information.
Signed-off-by: Joonsoo Kim <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: Michal Nazarewicz <[email protected]>
Cc: Jungsoo Son <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Extended memory to store page owner information is initialized some time
later than that page allocator starts. Until initialization, many pages
can be allocated and they have no owner information. This make debugging
using page owner harder, so some fixup will be helpful.
This patch fixes up this situation by setting fake owner information
immediately after page extension is initialized. Information doesn't tell
the right owner, but, at least, it can tell whether page is allocated or
not, more correctly.
On my testing, this patch catches 13343 early allocated pages, although
they are mostly allocated from page extension feature. Anyway, after
then, there is no page left that it is allocated and has no page owner
flag.
Signed-off-by: Joonsoo Kim <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: Michal Nazarewicz <[email protected]>
Cc: Jungsoo Son <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
This is the page owner tracking code which is introduced so far ago. It
is resident on Andrew's tree, though, nobody tried to upstream so it
remain as is. Our company uses this feature actively to debug memory leak
or to find a memory hogger so I decide to upstream this feature.
This functionality help us to know who allocates the page. When
allocating a page, we store some information about allocation in extra
memory. Later, if we need to know status of all pages, we can get and
analyze it from this stored information.
In previous version of this feature, extra memory is statically defined in
struct page, but, in this version, extra memory is allocated outside of
struct page. It enables us to turn on/off this feature at boottime
without considerable memory waste.
Although we already have tracepoint for tracing page allocation/free,
using it to analyze page owner is rather complex. We need to enlarge the
trace buffer for preventing overlapping until userspace program launched.
And, launched program continually dump out the trace buffer for later
analysis and it would change system behaviour with more possibility rather
than just keeping it in memory, so bad for debug.
Moreover, we can use page_owner feature further for various purposes. For
example, we can use it for fragmentation statistics implemented in this
patch. And, I also plan to implement some CMA failure debugging feature
using this interface.
I'd like to give the credit for all developers contributed this feature,
but, it's not easy because I don't know exact history. Sorry about that.
Below is people who has "Signed-off-by" in the patches in Andrew's tree.
Contributor:
Alexander Nyberg <[email protected]>
Mel Gorman <[email protected]>
Dave Hansen <[email protected]>
Minchan Kim <[email protected]>
Michal Nazarewicz <[email protected]>
Andrew Morton <[email protected]>
Jungsoo Son <[email protected]>
Signed-off-by: Joonsoo Kim <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: Michal Nazarewicz <[email protected]>
Cc: Jungsoo Son <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Current stacktrace only have the function for console output. page_owner
that will be introduced in following patch needs to print the output of
stacktrace into the buffer for our own output format so so new function,
snprint_stack_trace(), is needed.
Signed-off-by: Joonsoo Kim <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: Michal Nazarewicz <[email protected]>
Cc: Jungsoo Son <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
do_mmap_private() in nommu.c try to allocate physically contiguous pages
with arbitrary size in some cases and we now have good abstract function
to do exactly same thing, alloc_pages_exact(). So, change to use it.
There is no functional change. This is the preparation step for support
page owner feature accurately.
Signed-off-by: Joonsoo Kim <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: Michal Nazarewicz <[email protected]>
Cc: Jungsoo Son <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Now, we have prepared to avoid using debug-pagealloc in boottime. So
introduce new kernel-parameter to disable debug-pagealloc in boottime, and
makes related functions to be disabled in this case.
Only non-intuitive part is change of guard page functions. Because guard
page is effective only if debug-pagealloc is enabled, turning off
according to debug-pagealloc is reasonable thing to do.
Signed-off-by: Joonsoo Kim <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: Michal Nazarewicz <[email protected]>
Cc: Jungsoo Son <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Until now, debug-pagealloc needs extra flags in struct page, so we need to
recompile whole source code when we decide to use it. This is really
painful, because it takes some time to recompile and sometimes rebuild is
not possible due to third party module depending on struct page. So, we
can't use this good feature in many cases.
Now, we have the page extension feature that allows us to insert extra
flags to outside of struct page. This gets rid of third party module
issue mentioned above. And, this allows us to determine if we need extra
memory for this page extension in boottime. With these property, we can
avoid using debug-pagealloc in boottime with low computational overhead in
the kernel built with CONFIG_DEBUG_PAGEALLOC. This will help our
development process greatly.
This patch is the preparation step to achive above goal. debug-pagealloc
originally uses extra field of struct page, but, after this patch, it will
use field of struct page_ext. Because memory for page_ext is allocated
later than initialization of page allocator in CONFIG_SPARSEMEM, we should
disable debug-pagealloc feature temporarily until initialization of
page_ext. This patch implements this.
Signed-off-by: Joonsoo Kim <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: Michal Nazarewicz <[email protected]>
Cc: Jungsoo Son <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
When we debug something, we'd like to insert some information to every
page. For this purpose, we sometimes modify struct page itself. But,
this has drawbacks. First, it requires re-compile. This makes us
hesitate to use the powerful debug feature so development process is
slowed down. And, second, sometimes it is impossible to rebuild the
kernel due to third party module dependency. At third, system behaviour
would be largely different after re-compile, because it changes size of
struct page greatly and this structure is accessed by every part of
kernel. Keeping this as it is would be better to reproduce errornous
situation.
This feature is intended to overcome above mentioned problems. This
feature allocates memory for extended data per page in certain place
rather than the struct page itself. This memory can be accessed by the
accessor functions provided by this code. During the boot process, it
checks whether allocation of huge chunk of memory is needed or not. If
not, it avoids allocating memory at all. With this advantage, we can
include this feature into the kernel in default and can avoid rebuild and
solve related problems.
Until now, memcg uses this technique. But, now, memcg decides to embed
their variable to struct page itself and it's code to extend struct page
has been removed. I'd like to use this code to develop debug feature, so
this patch resurrect it.
To help these things to work well, this patch introduces two callbacks for
clients. One is the need callback which is mandatory if user wants to
avoid useless memory allocation at boot-time. The other is optional, init
callback, which is used to do proper initialization after memory is
allocated. Detailed explanation about purpose of these functions is in
code comment. Please refer it.
Others are completely same with previous extension code in memcg.
Signed-off-by: Joonsoo Kim <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: Michal Nazarewicz <[email protected]>
Cc: Jungsoo Son <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
GFP_USER, GFP_HIGHUSER and GFP_HIGHUSER_MOVABLE are escalatedly confined
defined, also implied by their names:
GFP_USER = GFP_USER
GFP_USER + __GFP_HIGHMEM = GFP_HIGHUSER
GFP_USER + __GFP_HIGHMEM + __GFP_MOVABLE = GFP_HIGHUSER_MOVABLE
So just make GFP_HIGHUSER and GFP_HIGHUSER_MOVABLE escalatedly defined to
reflect this fact. It also makes the definition clear and texturally warn
on any furture break-up of this escalated relastionship.
Signed-off-by: Jianyu Zhan <[email protected]>
Acked-by: Johannes Weiner <[email protected]>
Acked-by: Kirill A. Shutemov <[email protected]>
Acked-by: David Rientjes <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
include/linux/kmemleak.h: In function 'kmemleak_alloc_recursive':
include/linux/kmemleak.h:43: error: 'SLAB_NOLEAKTRACE' undeclared (first use in this function)
Cc: Joonsoo Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
The gfp was passed in but never used in this function.
Signed-off-by: Zhang Zhen <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
swp_entry_t being defined in include/linux/swap.h instead of
include/linux/mm_types.h causes cyclic include dependency later when
include/linux/page_cgroup.h is included from writeback path. Move the
definition to include/linux/mm_types.h.
While at it, reformat the comment above it.
Signed-off-by: Tejun Heo <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Johannes Weiner <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
This is just a small optimization. The start_pfn can be obtained directly
by phys_index << PFN_SECTION_SHIFT. So the call of page_to_pfn() is
redundant and remove it.
Signed-off-by: Zhang Zhen <[email protected]>
Acked-by: Yasuaki Ishimatsu <[email protected]>
Acked-by: David Rientjes <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: Wang Nan <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
This could be useful for debug in the future if we want to track
major/minor faults more closely, and also avoids the put_page trick we
used with gup.
In order to do this, we also track the task struct in the PASID state
structure. This lets us update the appropriate task stats after the fault
has been handled, and may aid with debug in the future as well.
Signed-off-by: Jesse Barnes <[email protected]>
Tested-by: Oded Gabbay <[email protected]>
Cc: Joerg Roedel <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
This lets drivers like the AMD IOMMUv2 driver handle faults a bit more
simply, rather than doing tricks with page refs and get_user_pages().
Signed-off-by: Jesse Barnes <[email protected]>
Cc: Oded Gabbay <[email protected]>
Cc: Joerg Roedel <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
This function is only called during initialization.
Signed-off-by: Luiz Capitulino <[email protected]>
Cc: Andi Kleen <[email protected]>
Acked-by: David Rientjes <[email protected]>
Cc: Rik van Riel <[email protected]>
Cc: Yasuaki Ishimatsu <[email protected]>
Cc: Yinghai Lu <[email protected]>
Cc: Davidlohr Bueso <[email protected]>
Acked-by: Naoya Horiguchi <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
No reason to duplicate the code of an existing macro.
Signed-off-by: Luiz Capitulino <[email protected]>
Cc: Andi Kleen <[email protected]>
Acked-by: David Rientjes <[email protected]>
Cc: Rik van Riel <[email protected]>
Cc: Yasuaki Ishimatsu <[email protected]>
Cc: Yinghai Lu <[email protected]>
Cc: Davidlohr Bueso <[email protected]>
Acked-by: Naoya Horiguchi <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
The hugepages= entry in kernel-parameters.txt states that 1GB pages can
only be allocated at boot time and not freed afterwards. This is not
true since commit 944d9fec8d7a ("hugetlb: add support for gigantic page
allocation at runtime"), at least for x86_64.
Instead of adding arch-specifc observations to the hugepages= entry,
this commit just drops the out of date information. Further information
about arch-specific support and available features can be obtained in
the hugetlb documentation.
Signed-off-by: Luiz Capitulino <[email protected]>
Cc: Andi Kleen <[email protected]>
Acked-by: David Rientjes <[email protected]>
Cc: Rik van Riel <[email protected]>
Cc: Yasuaki Ishimatsu <[email protected]>
Cc: Yinghai Lu <[email protected]>
Cc: Davidlohr Bueso <[email protected]>
Acked-by: Naoya Horiguchi <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
It isn't supposed to stack, so turn it into a bit-field to save 4 bytes on
the task_struct.
Also, remove the memcg_stop/resume_kmem_account helpers - it is clearer to
set/clear the flag inline. Regarding the overwhelming comment to the
helpers, which is removed by this patch too, we already have a compact yet
accurate explanation in memcg_schedule_cache_create, no need in yet
another one.
Signed-off-by: Vladimir Davydov <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Michal Hocko <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
__memcg_kmem_get_cache can recurse if it calls kmalloc (which it does if
the cgroup's kmem cache doesn't exist), because kmalloc may call
__memcg_kmem_get_cache internally again. To avoid the recursion, we use
the task_struct->memcg_kmem_skip_account flag.
However, there's no need checking the flag in memcg_kmem_newpage_charge,
because there's no way how this function could result in recursion, if
called from memcg_kmem_get_cache. So let's remove the redundant code.
Signed-off-by: Vladimir Davydov <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Michal Hocko <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
The only such flag is KMEM_ACCOUNTED_ACTIVE, but it's set iff
mem_cgroup->kmemcg_id is initialized, so we can check kmemcg_id instead of
having a separate flags field.
Signed-off-by: Vladimir Davydov <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Michal Hocko <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
When the encountered pte is a swap entry, the current code handles two
cases: migration and normal swapentry, but we have a third case: hwpoison
page.
This patch adds hwpoison page handle, consider hwpoison page incore as
same as migration.
[[email protected]: coding-style fixes]
Signed-off-by: Weijie Yang <[email protected]>
Acked-by: Johannes Weiner <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Hugh Dickins <[email protected]>
Cc: Rik van Riel <[email protected]>
Acked-by: Naoya Horiguchi <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Call page_to_pgoff() to get the page offset once we are sure we actually
need it, and any very obvious initial function checks have passed.
Trivial micro-optimization, and potentially save some cycles.
Signed-off-by: Davidlohr Bueso <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Page guard is used by debug-pagealloc feature. Currently, it is
open-coded, but, I think that more abstraction of it makes core page
allocator code more readable.
There is no functional difference.
Signed-off-by: Joonsoo Kim <[email protected]>
Acked-by: Vlastimil Babka <[email protected]>
Cc: Gioh Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
There is a lot of duplication in the rubric around actually setting or
clearing a mem region flag. Create a new helper function to do this and
reduce each of memblock_mark_hotplug() and memblock_clear_hotplug() to a
single line.
This will be useful if someone were to add a new mem region flag - which
I hope to be doing some day soon. But it looks like a plausible cleanup
even without that - so I'd like to get it out of the way now.
Signed-off-by: Tony Luck <[email protected]>
Cc: Santosh Shilimkar <[email protected]>
Cc: Tang Chen <[email protected]>
Cc: Grygorii Strashko <[email protected]>
Cc: Zhang Yanfei <[email protected]>
Cc: Philipp Hachtmann <[email protected]>
Cc: Yinghai Lu <[email protected]>
Cc: Emil Medve <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
task_struct->memcg_kmem_skip_account was initially introduced to avoid
recursion during kmem cache creation: memcg_kmem_get_cache, which is
called by kmem_cache_alloc to determine the per-memcg cache to account
allocation to, may issue lazy cache creation if the needed cache doesn't
exist, which means issuing yet another kmem_cache_alloc. We can't just
pass a flag to the nested kmem_cache_alloc disabling kmem accounting,
because there are hidden allocations, e.g. in INIT_WORK. So we
introduced a flag on the task_struct, memcg_kmem_skip_account, making
memcg_kmem_get_cache return immediately.
By its nature, the flag may also be used to disable accounting for
allocations shared among different cgroups, and currently it is used this
way in memcg_activate_kmem. Using it like this looks like abusing it to
me. If we want to disable accounting for some allocations (which we will
definitely want one day), we should either add GFP_NO_MEMCG or GFP_MEMCG
flag in order to blacklist/whitelist some allocations.
For now, let's simply remove memcg_stop/resume_kmem_account from
memcg_activate_kmem.
Signed-off-by: Vladimir Davydov <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Michal Hocko <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
We already assured the current task has mm in memcg_kmem_should_charge,
no need to double check.
Signed-off-by: Vladimir Davydov <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Michal Hocko <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
cpuset code stopped using cgroup_lock in favor of cpuset_mutex long ago.
Signed-off-by: Vladimir Davydov <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Michal Hocko <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
The alignment in cma_alloc() was done w.r.t. the bitmap. This is a
problem when, for example:
- a device requires 16M (order 12) alignment
- the CMA region is not 16 M aligned
In such a case, can result with the CMA region starting at, say,
0x2f800000 but any allocation you make from there will be aligned from
there. Requesting an allocation of 32 M with 16 M alignment will result
in an allocation from 0x2f800000 to 0x31800000, which doesn't work very
well if your strange device requires 16M alignment.
Change to use bitmap_find_next_zero_area_off() to account for the
difference in alignment at reserve-time and alloc-time.
Signed-off-by: Gregory Fong <[email protected]>
Acked-by: Michal Nazarewicz <[email protected]>
Cc: Marek Szyprowski <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Cc: Kukjin Kim <[email protected]>
Cc: Laurent Pinchart <[email protected]>
Cc: Laura Abbott <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Add a bitmap_find_next_zero_area_off() function which works like
bitmap_find_next_zero_area() function except it allows an offset to be
specified when alignment is checked. This lets caller request a bit such
that its number plus the offset is aligned according to the mask.
[[email protected]: Retrieved from https://patchwork.linuxtv.org/patch/6254/ and updated documentation]
Signed-off-by: Michal Nazarewicz <[email protected]>
Signed-off-by: Kyungmin Park <[email protected]>
Signed-off-by: Marek Szyprowski <[email protected]>
Signed-off-by: Gregory Fong <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Cc: Kukjin Kim <[email protected]>
Cc: Laurent Pinchart <[email protected]>
Cc: Laura Abbott <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
The unmap_mapping_range family of functions do the unmapping of user pages
(ultimately via zap_page_range_single) without touching the actual
interval tree, thus share the lock.
Signed-off-by: Davidlohr Bueso <[email protected]>
Cc: "Kirill A. Shutemov" <[email protected]>
Acked-by: Hugh Dickins <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Cc: Peter Zijlstra (Intel) <[email protected]>
Cc: Rik van Riel <[email protected]>
Cc: Srikar Dronamraju <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Shrinking/truncate logic can call nommu_shrink_inode_mappings() to verify
that any shared mappings of the inode in question aren't broken (dead
zone). afaict the only user being ramfs to handle the size change
attribute.
Pretty much a no-brainer to share the lock.
Signed-off-by: Davidlohr Bueso <[email protected]>
Acked-by: "Kirill A. Shutemov" <[email protected]>
Acked-by: Hugh Dickins <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Cc: Rik van Riel <[email protected]>
Cc: Srikar Dronamraju <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
No brainer conversion: collect_procs_file() only schedules a process for
later kill, share the lock, similarly to the anon vma variant.
Signed-off-by: Davidlohr Bueso <[email protected]>
Acked-by: "Kirill A. Shutemov" <[email protected]>
Acked-by: Hugh Dickins <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Cc: Rik van Riel <[email protected]>
Cc: Srikar Dronamraju <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
__xip_unmap() will remove the xip sparse page from the cache and take down
pte mapping, without altering the interval tree, thus share the
i_mmap_rwsem when searching for the ptes to unmap.
Additionally, tidy up the function a bit and make variables only local to
the interval tree walk loop.
Signed-off-by: Davidlohr Bueso <[email protected]>
Acked-by: "Kirill A. Shutemov" <[email protected]>
Acked-by: Hugh Dickins <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Cc: Rik van Riel <[email protected]>
Cc: Srikar Dronamraju <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Both register and unregister call build_map_info() in order to create the
list of mappings before installing or removing breakpoints for every mm
which maps file backed memory. As such, there is no reason to hold the
i_mmap_rwsem exclusively, so share it and allow concurrent readers to
build the mapping data.
Signed-off-by: Davidlohr Bueso <[email protected]>
Acked-by: Srikar Dronamraju <[email protected]>
Acked-by: "Kirill A. Shutemov" <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Acked-by: Hugh Dickins <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Cc: Rik van Riel <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Similarly to the anon memory counterpart, we can share the mapping's lock
ownership as the interval tree is not modified when doing doing the walk,
only the file page.
Signed-off-by: Davidlohr Bueso <[email protected]>
Acked-by: Rik van Riel <[email protected]>
Acked-by: "Kirill A. Shutemov" <[email protected]>
Acked-by: Hugh Dickins <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Cc: Srikar Dronamraju <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
The i_mmap_mutex is a close cousin of the anon vma lock, both protecting
similar data, one for file backed pages and the other for anon memory. To
this end, this lock can also be a rwsem. In addition, there are some
important opportunities to share the lock when there are no tree
modifications.
This conversion is straightforward. For now, all users take the write
lock.
[[email protected]: update fremap.c]
Signed-off-by: Davidlohr Bueso <[email protected]>
Reviewed-by: Rik van Riel <[email protected]>
Acked-by: "Kirill A. Shutemov" <[email protected]>
Acked-by: Hugh Dickins <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Cc: Srikar Dronamraju <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Signed-off-by: Stephen Rothwell <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Convert all open coded mutex_lock/unlock calls to the
i_mmap_[lock/unlock]_write() helpers.
Signed-off-by: Davidlohr Bueso <[email protected]>
Acked-by: Rik van Riel <[email protected]>
Acked-by: "Kirill A. Shutemov" <[email protected]>
Acked-by: Hugh Dickins <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Cc: Srikar Dronamraju <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
This series is a continuation of the conversion of the i_mmap_mutex to
rwsem, following what we have for the anon memory counterpart. With
Hugh's feedback from the first iteration.
Ultimately, the most obvious paths that require exclusive ownership of the
lock is when we modify the VMA interval tree, via
vma_interval_tree_insert() and vma_interval_tree_remove() families. Cases
such as unmapping, where the ptes content is changed but the tree remains
untouched should make it safe to share the i_mmap_rwsem.
As such, the code of course is straightforward, however the devil is very
much in the details. While its been tested on a number of workloads
without anything exploding, I would not be surprised if there are some
less documented/known assumptions about the lock that could suffer from
these changes. Or maybe I'm just missing something, but either way I
believe its at the point where it could use more eyes and hopefully some
time in linux-next.
Because the lock type conversion is the heart of this patchset,
its worth noting a few comparisons between mutex vs rwsem (xadd):
(i) Same size, no extra footprint.
(ii) Both have CONFIG_XXX_SPIN_ON_OWNER capabilities for
exclusive lock ownership.
(iii) Both can be slightly unfair wrt exclusive ownership, with
writer lock stealing properties, not necessarily respecting
FIFO order for granting the lock when contended.
(iv) Mutexes can be slightly faster than rwsems when
the lock is non-contended.
(v) Both suck at performance for debug (slowpaths), which
shouldn't matter anyway.
Sharing the lock is obviously beneficial, and sem writer ownership is
close enough to mutexes. The biggest winner of these changes is
migration.
As for concrete numbers, the following performance results are for a
4-socket 60-core IvyBridge-EX with 130Gb of RAM.
Both alltests and disk (xfs+ramdisk) workloads of aim7 suite do quite well
with this set, with a steady ~60% throughput (jpm) increase for alltests
and up to ~30% for disk for high amounts of concurrency. Lower counts of
workload users (< 100) does not show much difference at all, so at least
no regressions.
3.18-rc1 3.18-rc1-i_mmap_rwsem
alltests-100 17918.72 ( 0.00%) 28417.97 ( 58.59%)
alltests-200 16529.39 ( 0.00%) 26807.92 ( 62.18%)
alltests-300 16591.17 ( 0.00%) 26878.08 ( 62.00%)
alltests-400 16490.37 ( 0.00%) 26664.63 ( 61.70%)
alltests-500 16593.17 ( 0.00%) 26433.72 ( 59.30%)
alltests-600 16508.56 ( 0.00%) 26409.20 ( 59.97%)
alltests-700 16508.19 ( 0.00%) 26298.58 ( 59.31%)
alltests-800 16437.58 ( 0.00%) 26433.02 ( 60.81%)
alltests-900 16418.35 ( 0.00%) 26241.61 ( 59.83%)
alltests-1000 16369.00 ( 0.00%) 26195.76 ( 60.03%)
alltests-1100 16330.11 ( 0.00%) 26133.46 ( 60.03%)
alltests-1200 16341.30 ( 0.00%) 26084.03 ( 59.62%)
alltests-1300 16304.75 ( 0.00%) 26024.74 ( 59.61%)
alltests-1400 16231.08 ( 0.00%) 25952.35 ( 59.89%)
alltests-1500 16168.06 ( 0.00%) 25850.58 ( 59.89%)
alltests-1600 16142.56 ( 0.00%) 25767.42 ( 59.62%)
alltests-1700 16118.91 ( 0.00%) 25689.58 ( 59.38%)
alltests-1800 16068.06 ( 0.00%) 25599.71 ( 59.32%)
alltests-1900 16046.94 ( 0.00%) 25525.92 ( 59.07%)
alltests-2000 16007.26 ( 0.00%) 25513.07 ( 59.38%)
disk-100 7582.14 ( 0.00%) 7257.48 ( -4.28%)
disk-200 6962.44 ( 0.00%) 7109.15 ( 2.11%)
disk-300 6435.93 ( 0.00%) 6904.75 ( 7.28%)
disk-400 6370.84 ( 0.00%) 6861.26 ( 7.70%)
disk-500 6353.42 ( 0.00%) 6846.71 ( 7.76%)
disk-600 6368.82 ( 0.00%) 6806.75 ( 6.88%)
disk-700 6331.37 ( 0.00%) 6796.01 ( 7.34%)
disk-800 6324.22 ( 0.00%) 6788.00 ( 7.33%)
disk-900 6253.52 ( 0.00%) 6750.43 ( 7.95%)
disk-1000 6242.53 ( 0.00%) 6855.11 ( 9.81%)
disk-1100 6234.75 ( 0.00%) 6858.47 ( 10.00%)
disk-1200 6312.76 ( 0.00%) 6845.13 ( 8.43%)
disk-1300 6309.95 ( 0.00%) 6834.51 ( 8.31%)
disk-1400 6171.76 ( 0.00%) 6787.09 ( 9.97%)
disk-1500 6139.81 ( 0.00%) 6761.09 ( 10.12%)
disk-1600 4807.12 ( 0.00%) 6725.33 ( 39.90%)
disk-1700 4669.50 ( 0.00%) 5985.38 ( 28.18%)
disk-1800 4663.51 ( 0.00%) 5972.99 ( 28.08%)
disk-1900 4674.31 ( 0.00%) 5949.94 ( 27.29%)
disk-2000 4668.36 ( 0.00%) 5834.93 ( 24.99%)
In addition, a 67.5% increase in successfully migrated NUMA pages, thus
improving node locality.
The patch layout is simple but designed for bisection (in case reversion
is needed if the changes break upstream) and easier review:
o Patches 1-4 convert the i_mmap lock from mutex to rwsem.
o Patches 5-10 share the lock in specific paths, each patch
details the rationale behind why it should be safe.
This patchset has been tested with: postgres 9.4 (with brand new hugetlb
support), hugetlbfs test suite (all tests pass, in fact more tests pass
with these changes than with an upstream kernel), ltp, aim7 benchmarks,
memcached and iozone with the -B option for mmap'ing. *Untested* paths
are nommu, memory-failure, uprobes and xip.
This patch (of 8):
Various parts of the kernel acquire and release this mutex, so add
i_mmap_lock_write() and immap_unlock_write() helper functions that will
encapsulate this logic. The next patch will make use of these.
Signed-off-by: Davidlohr Bueso <[email protected]>
Reviewed-by: Rik van Riel <[email protected]>
Acked-by: "Kirill A. Shutemov" <[email protected]>
Acked-by: Hugh Dickins <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Cc: Srikar Dronamraju <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
My current email address will be gone shortly, update my email to be a
gmail one.
Signed-off-by: Xiubo Li <[email protected]>
Cc: Timur Tabi <[email protected]>
Cc: Takashi Iwai <[email protected]>
Acked-by: Nicolin Chen <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
Commit 7654e9d4fd8f ("drivers/rtc/rtc-snvs: fix suspend/resume")
replaces SIMPLE_DEV_PM_OPS with direct declaration of snvs_rtc_pm_ops,
but does so outside #ifdef CONFIG_PM_SLEEP. This causes the driver
build to fail if CONFIG_PM_SLEEP is not configured.
Fixes: 7654e9d4fd8f ("drivers/rtc/rtc-snvs: fix suspend/resume")
Signed-off-by: Guenter Roeck <[email protected]>
Cc: Sanchayan Maity <[email protected]>
Cc: Alessandro Zummo <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
When the vgic initializes its internal state it does so based on the
number of VCPUs available at the time. If we allow KVM to create more
VCPUs after the VGIC has been initialized, we are likely to error out in
unfortunate ways later, perform buffer overflows etc.
Acked-by: Marc Zyngier <[email protected]>
Reviewed-by: Eric Auger <[email protected]>
Signed-off-by: Christoffer Dall <[email protected]>
|
|
Some code paths will need to check to see if the internal state of the
vgic has been initialized (such as when creating new VCPUs), so
introduce such a macro that checks the nr_cpus field which is set when
the vgic has been initialized.
Also set nr_cpus = 0 in kvm_vgic_destroy, because the error path in
vgic_init() will call this function, and code should never errornously
assume the vgic to be properly initialized after an error.
Acked-by: Marc Zyngier <[email protected]>
Reviewed-by: Eric Auger <[email protected]>
Signed-off-by: Christoffer Dall <[email protected]>
|
|
The vgic_initialized() macro currently returns the state of the
vgic->ready flag, which indicates if the vgic is ready to be used when
running a VM, not specifically if its internal state has been
initialized.
Rename the macro accordingly in preparation for a more nuanced
initialization flow.
Acked-by: Marc Zyngier <[email protected]>
Reviewed-by: Eric Auger <[email protected]>
Signed-off-by: Christoffer Dall <[email protected]>
|
|
VGIC initialization currently happens in three phases:
(1) kvm_vgic_create() (triggered by userspace GIC creation)
(2) vgic_init_maps() (triggered by userspace GIC register read/write
requests, or from kvm_vgic_init() if not already run)
(3) kvm_vgic_init() (triggered by first VM run)
We were doing initialization of some state to correspond with the
state of a freshly-reset GIC in kvm_vgic_init(); this is too late,
since it will overwrite changes made by userspace using the
register access APIs before the VM is run. Move this initialization
earlier, into the vgic_init_maps() phase.
This fixes a bug where QEMU could successfully restore a saved
VM state snapshot into a VM that had already been run, but could
not restore it "from cold" using the -loadvm command line option
(the symptoms being that the restored VM would run but interrupts
were ignored).
Finally rename vgic_init_maps to vgic_init and renamed kvm_vgic_init to
kvm_vgic_map_resources.
[ This patch is originally written by Peter Maydell, but I have
modified it somewhat heavily, renaming various bits and moving code
around. If something is broken, I am to be blamed. - Christoffer ]
Acked-by: Marc Zyngier <[email protected]>
Reviewed-by: Eric Auger <[email protected]>
Signed-off-by: Peter Maydell <[email protected]>
Signed-off-by: Christoffer Dall <[email protected]>
|
|
Introduce a new function to unmap user RAM regions in the stage2 page
tables. This is needed on reboot (or when the guest turns off the MMU)
to ensure we fault in pages again and make the dcache, RAM, and icache
coherent.
Using unmap_stage2_range for the whole guest physical range does not
work, because that unmaps IO regions (such as the GIC) which will not be
recreated or in the best case faulted in on a page-by-page basis.
Call this function on secondary and subsequent calls to the
KVM_ARM_VCPU_INIT ioctl so that a reset VCPU will detect the guest
Stage-1 MMU is off when faulting in pages and make the caches coherent.
Acked-by: Marc Zyngier <[email protected]>
Signed-off-by: Christoffer Dall <[email protected]>
|