linux-stable/Documentation/vm/transhuge.txt
Mel Gorman 444eb2a449 mm: thp: set THP defrag by default to madvise and add a stall-free defrag option
THP defrag is enabled by default to direct reclaim/compact but not wake
kswapd in the event of a THP allocation failure.  The problem is that
THP allocation requests potentially enter reclaim/compaction.  This
potentially incurs a severe stall that is not guaranteed to be offset by
reduced TLB misses.  While there has been considerable effort to reduce
the impact of reclaim/compaction, it is still a high cost and workloads
that should fit in memory fail to do so.  Specifically, a simple
anon/file streaming workload will enter direct reclaim on NUMA at least
even though the working set size is 80% of RAM.  It's been years and
it's time to throw in the towel.

First, this patch defines THP defrag as follows;

 madvise: A failed allocation will direct reclaim/compact if the application requests it
 never:   Neither reclaim/compact nor wake kswapd
 defer:   A failed allocation will wake kswapd/kcompactd
 always:  A failed allocation will direct reclaim/compact (historical behaviour)
          khugepaged defrag will enter direct/reclaim but not wake kswapd.

Next it sets the default defrag option to be "madvise" to only enter
direct reclaim/compaction for applications that specifically requested
it.

Lastly, it removes a check from the page allocator slowpath that is
related to __GFP_THISNODE to allow "defer" to work.  The callers that
really cares are slub/slab and they are updated accordingly.  The slab
one may be surprising because it also corrects a comment as kswapd was
never woken up by that path.

This means that a THP fault will no longer stall for most applications
by default and the ideal for most users that get THP if they are
immediately available.  There are still options for users that prefer a
stall at startup of a new application by either restoring historical
behaviour with "always" or pick a half-way point with "defer" where
kswapd does some of the work in the background and wakes kcompactd if
necessary.  THP defrag for khugepaged remains enabled and will enter
direct/reclaim but no wakeup kswapd or kcompactd.

After this patch a THP allocation failure will quickly fallback and rely
on khugepaged to recover the situation at some time in the future.  In
some cases, this will reduce THP usage but the benefit of THP is hard to
measure and not a universal win where as a stall to reclaim/compaction
is definitely measurable and can be painful.

The first test for this is using "usemem" to read a large file and write
a large anonymous mapping (to avoid the zero page) multiple times.  The
total size of the mappings is 80% of RAM and the benchmark simply
measures how long it takes to complete.  It uses multiple threads to see
if that is a factor.  On UMA, the performance is almost identical so is
not reported but on NUMA, we see this

usemem
                                   4.4.0                 4.4.0
                          kcompactd-v1r1         nodefrag-v1r3
Amean    System-1       102.86 (  0.00%)       46.81 ( 54.50%)
Amean    System-4        37.85 (  0.00%)       34.02 ( 10.12%)
Amean    System-7        48.12 (  0.00%)       46.89 (  2.56%)
Amean    System-12       51.98 (  0.00%)       56.96 ( -9.57%)
Amean    System-21       80.16 (  0.00%)       79.05 (  1.39%)
Amean    System-30      110.71 (  0.00%)      107.17 (  3.20%)
Amean    System-48      127.98 (  0.00%)      124.83 (  2.46%)
Amean    Elapsd-1       185.84 (  0.00%)      105.51 ( 43.23%)
Amean    Elapsd-4        26.19 (  0.00%)       25.58 (  2.33%)
Amean    Elapsd-7        21.65 (  0.00%)       21.62 (  0.16%)
Amean    Elapsd-12       18.58 (  0.00%)       17.94 (  3.43%)
Amean    Elapsd-21       17.53 (  0.00%)       16.60 (  5.33%)
Amean    Elapsd-30       17.45 (  0.00%)       17.13 (  1.84%)
Amean    Elapsd-48       15.40 (  0.00%)       15.27 (  0.82%)

For a single thread, the benchmark completes 43.23% faster with this
patch applied with smaller benefits as the thread increases.  Similar,
notice the large reduction in most cases in system CPU usage.  The
overall CPU time is

               4.4.0       4.4.0
        kcompactd-v1r1 nodefrag-v1r3
User        10357.65    10438.33
System       3988.88     3543.94
Elapsed      2203.01     1634.41

Which is substantial. Now, the reclaim figures

                                 4.4.0       4.4.0
                          kcompactd-v1r1nodefrag-v1r3
Minor Faults                 128458477   278352931
Major Faults                   2174976         225
Swap Ins                      16904701           0
Swap Outs                     17359627           0
Allocation stalls                43611           0
DMA allocs                           0           0
DMA32 allocs                  19832646    19448017
Normal allocs                614488453   580941839
Movable allocs                       0           0
Direct pages scanned          24163800           0
Kswapd pages scanned                 0           0
Kswapd pages reclaimed               0           0
Direct pages reclaimed        20691346           0
Compaction stalls                42263           0
Compaction success                 938           0
Compaction failures              41325           0

This patch eliminates almost all swapping and direct reclaim activity.
There is still overhead but it's from NUMA balancing which does not
identify that it's pointless trying to do anything with this workload.

I also tried the thpscale benchmark which forces a corner case where
compaction can be used heavily and measures the latency of whether base
or huge pages were used

thpscale Fault Latencies
                                       4.4.0                 4.4.0
                              kcompactd-v1r1         nodefrag-v1r3
Amean    fault-base-1      5288.84 (  0.00%)     2817.12 ( 46.73%)
Amean    fault-base-3      6365.53 (  0.00%)     3499.11 ( 45.03%)
Amean    fault-base-5      6526.19 (  0.00%)     4363.06 ( 33.15%)
Amean    fault-base-7      7142.25 (  0.00%)     4858.08 ( 31.98%)
Amean    fault-base-12    13827.64 (  0.00%)    10292.11 ( 25.57%)
Amean    fault-base-18    18235.07 (  0.00%)    13788.84 ( 24.38%)
Amean    fault-base-24    21597.80 (  0.00%)    24388.03 (-12.92%)
Amean    fault-base-30    26754.15 (  0.00%)    19700.55 ( 26.36%)
Amean    fault-base-32    26784.94 (  0.00%)    19513.57 ( 27.15%)
Amean    fault-huge-1      4223.96 (  0.00%)     2178.57 ( 48.42%)
Amean    fault-huge-3      2194.77 (  0.00%)     2149.74 (  2.05%)
Amean    fault-huge-5      2569.60 (  0.00%)     2346.95 (  8.66%)
Amean    fault-huge-7      3612.69 (  0.00%)     2997.70 ( 17.02%)
Amean    fault-huge-12     3301.75 (  0.00%)     6727.02 (-103.74%)
Amean    fault-huge-18     6696.47 (  0.00%)     6685.72 (  0.16%)
Amean    fault-huge-24     8000.72 (  0.00%)     9311.43 (-16.38%)
Amean    fault-huge-30    13305.55 (  0.00%)     9750.45 ( 26.72%)
Amean    fault-huge-32     9981.71 (  0.00%)    10316.06 ( -3.35%)

The average time to fault pages is substantially reduced in the majority
of caseds but with the obvious caveat that fewer THPs are actually used
in this adverse workload

                                   4.4.0                 4.4.0
                          kcompactd-v1r1         nodefrag-v1r3
Percentage huge-1         0.71 (  0.00%)       14.04 (1865.22%)
Percentage huge-3        10.77 (  0.00%)       33.05 (206.85%)
Percentage huge-5        60.39 (  0.00%)       38.51 (-36.23%)
Percentage huge-7        45.97 (  0.00%)       34.57 (-24.79%)
Percentage huge-12       68.12 (  0.00%)       40.07 (-41.17%)
Percentage huge-18       64.93 (  0.00%)       47.82 (-26.35%)
Percentage huge-24       62.69 (  0.00%)       44.23 (-29.44%)
Percentage huge-30       43.49 (  0.00%)       55.38 ( 27.34%)
Percentage huge-32       50.72 (  0.00%)       51.90 (  2.35%)

                                 4.4.0       4.4.0
                          kcompactd-v1r1nodefrag-v1r3
Minor Faults                  37429143    47564000
Major Faults                      1916        1558
Swap Ins                          1466        1079
Swap Outs                      2936863      149626
Allocation stalls                62510           3
DMA allocs                           0           0
DMA32 allocs                   6566458     6401314
Normal allocs                216361697   216538171
Movable allocs                       0           0
Direct pages scanned          25977580       17998
Kswapd pages scanned                 0     3638931
Kswapd pages reclaimed               0      207236
Direct pages reclaimed         8833714          88
Compaction stalls               103349           5
Compaction success                 270           4
Compaction failures             103079           1

Note again that while this does swap as it's an aggressive workload, the
direct relcim activity and allocation stalls is substantially reduced.
There is some kswapd activity but ftrace showed that the kswapd activity
was due to normal wakeups from 4K pages being allocated.
Compaction-related stalls and activity are almost eliminated.

I also tried the stutter benchmark.  For this, I do not have figures for
NUMA but it's something that does impact UMA so I'll report what is
available

stutter
                                 4.4.0                 4.4.0
                        kcompactd-v1r1         nodefrag-v1r3
Min         mmap      7.3571 (  0.00%)      7.3438 (  0.18%)
1st-qrtle   mmap      7.5278 (  0.00%)     17.9200 (-138.05%)
2nd-qrtle   mmap      7.6818 (  0.00%)     21.6055 (-181.25%)
3rd-qrtle   mmap     11.0889 (  0.00%)     21.8881 (-97.39%)
Max-90%     mmap     27.8978 (  0.00%)     22.1632 ( 20.56%)
Max-93%     mmap     28.3202 (  0.00%)     22.3044 ( 21.24%)
Max-95%     mmap     28.5600 (  0.00%)     22.4580 ( 21.37%)
Max-99%     mmap     29.6032 (  0.00%)     25.5216 ( 13.79%)
Max         mmap   4109.7289 (  0.00%)   4813.9832 (-17.14%)
Mean        mmap     12.4474 (  0.00%)     19.3027 (-55.07%)

This benchmark is trying to fault an anonymous mapping while there is a
heavy IO load -- a scenario that desktop users used to complain about
frequently.  This shows a mix because the ideal case of mapping with THP
is not hit as often.  However, note that 99% of the mappings complete
13.79% faster.  The CPU usage here is particularly interesting

               4.4.0       4.4.0
        kcompactd-v1r1nodefrag-v1r3
User           67.50        0.99
System       1327.88       91.30
Elapsed      2079.00     2128.98

And once again we look at the reclaim figures

                                 4.4.0       4.4.0
                          kcompactd-v1r1nodefrag-v1r3
Minor Faults                 335241922  1314582827
Major Faults                       715         819
Swap Ins                             0           0
Swap Outs                            0           0
Allocation stalls               532723           0
DMA allocs                           0           0
DMA32 allocs                1822364341  1177950222
Normal allocs               1815640808  1517844854
Movable allocs                       0           0
Direct pages scanned          21892772           0
Kswapd pages scanned          20015890    41879484
Kswapd pages reclaimed        19961986    41822072
Direct pages reclaimed        21892741           0
Compaction stalls              1065755           0
Compaction success                 514           0
Compaction failures            1065241           0

Allocation stalls and all direct reclaim activity is eliminated as well
as compaction-related stalls.

THP gives impressive gains in some cases but only if they are quickly
available.  We're not going to reach the point where they are completely
free so lets take the costs out of the fast paths finally and defer the
cost to kswapd, kcompactd and khugepaged where it belongs.

Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-03-17 15:09:34 -07:00

460 lines
20 KiB
Text

= Transparent Hugepage Support =
== Objective ==
Performance critical computing applications dealing with large memory
working sets are already running on top of libhugetlbfs and in turn
hugetlbfs. Transparent Hugepage Support is an alternative means of
using huge pages for the backing of virtual memory with huge pages
that supports the automatic promotion and demotion of page sizes and
without the shortcomings of hugetlbfs.
Currently it only works for anonymous memory mappings but in the
future it can expand over the pagecache layer starting with tmpfs.
The reason applications are running faster is because of two
factors. The first factor is almost completely irrelevant and it's not
of significant interest because it'll also have the downside of
requiring larger clear-page copy-page in page faults which is a
potentially negative effect. The first factor consists in taking a
single page fault for each 2M virtual region touched by userland (so
reducing the enter/exit kernel frequency by a 512 times factor). This
only matters the first time the memory is accessed for the lifetime of
a memory mapping. The second long lasting and much more important
factor will affect all subsequent accesses to the memory for the whole
runtime of the application. The second factor consist of two
components: 1) the TLB miss will run faster (especially with
virtualization using nested pagetables but almost always also on bare
metal without virtualization) and 2) a single TLB entry will be
mapping a much larger amount of virtual memory in turn reducing the
number of TLB misses. With virtualization and nested pagetables the
TLB can be mapped of larger size only if both KVM and the Linux guest
are using hugepages but a significant speedup already happens if only
one of the two is using hugepages just because of the fact the TLB
miss is going to run faster.
== Design ==
- "graceful fallback": mm components which don't have transparent hugepage
knowledge fall back to breaking huge pmd mapping into table of ptes and,
if necessary, split a transparent hugepage. Therefore these components
can continue working on the regular pages or regular pte mappings.
- if a hugepage allocation fails because of memory fragmentation,
regular pages should be gracefully allocated instead and mixed in
the same vma without any failure or significant delay and without
userland noticing
- if some task quits and more hugepages become available (either
immediately in the buddy or through the VM), guest physical memory
backed by regular pages should be relocated on hugepages
automatically (with khugepaged)
- it doesn't require memory reservation and in turn it uses hugepages
whenever possible (the only possible reservation here is kernelcore=
to avoid unmovable pages to fragment all the memory but such a tweak
is not specific to transparent hugepage support and it's a generic
feature that applies to all dynamic high order allocations in the
kernel)
- this initial support only offers the feature in the anonymous memory
regions but it'd be ideal to move it to tmpfs and the pagecache
later
Transparent Hugepage Support maximizes the usefulness of free memory
if compared to the reservation approach of hugetlbfs by allowing all
unused memory to be used as cache or other movable (or even unmovable
entities). It doesn't require reservation to prevent hugepage
allocation failures to be noticeable from userland. It allows paging
and all other advanced VM features to be available on the
hugepages. It requires no modifications for applications to take
advantage of it.
Applications however can be further optimized to take advantage of
this feature, like for example they've been optimized before to avoid
a flood of mmap system calls for every malloc(4k). Optimizing userland
is by far not mandatory and khugepaged already can take care of long
lived page allocations even for hugepage unaware applications that
deals with large amounts of memory.
In certain cases when hugepages are enabled system wide, application
may end up allocating more memory resources. An application may mmap a
large region but only touch 1 byte of it, in that case a 2M page might
be allocated instead of a 4k page for no good. This is why it's
possible to disable hugepages system-wide and to only have them inside
MADV_HUGEPAGE madvise regions.
Embedded systems should enable hugepages only inside madvise regions
to eliminate any risk of wasting any precious byte of memory and to
only run faster.
Applications that gets a lot of benefit from hugepages and that don't
risk to lose memory by using hugepages, should use
madvise(MADV_HUGEPAGE) on their critical mmapped regions.
== sysfs ==
Transparent Hugepage Support can be entirely disabled (mostly for
debugging purposes) or only enabled inside MADV_HUGEPAGE regions (to
avoid the risk of consuming more memory resources) or enabled system
wide. This can be achieved with one of:
echo always >/sys/kernel/mm/transparent_hugepage/enabled
echo madvise >/sys/kernel/mm/transparent_hugepage/enabled
echo never >/sys/kernel/mm/transparent_hugepage/enabled
It's also possible to limit defrag efforts in the VM to generate
hugepages in case they're not immediately free to madvise regions or
to never try to defrag memory and simply fallback to regular pages
unless hugepages are immediately available. Clearly if we spend CPU
time to defrag memory, we would expect to gain even more by the fact
we use hugepages later instead of regular pages. This isn't always
guaranteed, but it may be more likely in case the allocation is for a
MADV_HUGEPAGE region.
echo always >/sys/kernel/mm/transparent_hugepage/defrag
echo defer >/sys/kernel/mm/transparent_hugepage/defrag
echo madvise >/sys/kernel/mm/transparent_hugepage/defrag
echo never >/sys/kernel/mm/transparent_hugepage/defrag
"always" means that an application requesting THP will stall on allocation
failure and directly reclaim pages and compact memory in an effort to
allocate a THP immediately. This may be desirable for virtual machines
that benefit heavily from THP use and are willing to delay the VM start
to utilise them.
"defer" means that an application will wake kswapd in the background
to reclaim pages and wake kcompact to compact memory so that THP is
available in the near future. It's the responsibility of khugepaged
to then install the THP pages later.
"madvise" will enter direct reclaim like "always" but only for regions
that are have used madvise(MADV_HUGEPAGE). This is the default behaviour.
"never" should be self-explanatory.
By default kernel tries to use huge zero page on read page fault.
It's possible to disable huge zero page by writing 0 or enable it
back by writing 1:
echo 0 >/sys/kernel/mm/transparent_hugepage/use_zero_page
echo 1 >/sys/kernel/mm/transparent_hugepage/use_zero_page
khugepaged will be automatically started when
transparent_hugepage/enabled is set to "always" or "madvise, and it'll
be automatically shutdown if it's set to "never".
khugepaged runs usually at low frequency so while one may not want to
invoke defrag algorithms synchronously during the page faults, it
should be worth invoking defrag at least in khugepaged. However it's
also possible to disable defrag in khugepaged by writing 0 or enable
defrag in khugepaged by writing 1:
echo 0 >/sys/kernel/mm/transparent_hugepage/khugepaged/defrag
echo 1 >/sys/kernel/mm/transparent_hugepage/khugepaged/defrag
You can also control how many pages khugepaged should scan at each
pass:
/sys/kernel/mm/transparent_hugepage/khugepaged/pages_to_scan
and how many milliseconds to wait in khugepaged between each pass (you
can set this to 0 to run khugepaged at 100% utilization of one core):
/sys/kernel/mm/transparent_hugepage/khugepaged/scan_sleep_millisecs
and how many milliseconds to wait in khugepaged if there's an hugepage
allocation failure to throttle the next allocation attempt.
/sys/kernel/mm/transparent_hugepage/khugepaged/alloc_sleep_millisecs
The khugepaged progress can be seen in the number of pages collapsed:
/sys/kernel/mm/transparent_hugepage/khugepaged/pages_collapsed
for each pass:
/sys/kernel/mm/transparent_hugepage/khugepaged/full_scans
max_ptes_none specifies how many extra small pages (that are
not already mapped) can be allocated when collapsing a group
of small pages into one large page.
/sys/kernel/mm/transparent_hugepage/khugepaged/max_ptes_none
A higher value leads to use additional memory for programs.
A lower value leads to gain less thp performance. Value of
max_ptes_none can waste cpu time very little, you can
ignore it.
max_ptes_swap specifies how many pages can be brought in from
swap when collapsing a group of pages into a transparent huge page.
/sys/kernel/mm/transparent_hugepage/khugepaged/max_ptes_swap
A higher value can cause excessive swap IO and waste
memory. A lower value can prevent THPs from being
collapsed, resulting fewer pages being collapsed into
THPs, and lower memory access performance.
== Boot parameter ==
You can change the sysfs boot time defaults of Transparent Hugepage
Support by passing the parameter "transparent_hugepage=always" or
"transparent_hugepage=madvise" or "transparent_hugepage=never"
(without "") to the kernel command line.
== Need of application restart ==
The transparent_hugepage/enabled values only affect future
behavior. So to make them effective you need to restart any
application that could have been using hugepages. This also applies to
the regions registered in khugepaged.
== Monitoring usage ==
The number of transparent huge pages currently used by the system is
available by reading the AnonHugePages field in /proc/meminfo. To
identify what applications are using transparent huge pages, it is
necessary to read /proc/PID/smaps and count the AnonHugePages fields
for each mapping. Note that reading the smaps file is expensive and
reading it frequently will incur overhead.
There are a number of counters in /proc/vmstat that may be used to
monitor how successfully the system is providing huge pages for use.
thp_fault_alloc is incremented every time a huge page is successfully
allocated to handle a page fault. This applies to both the
first time a page is faulted and for COW faults.
thp_collapse_alloc is incremented by khugepaged when it has found
a range of pages to collapse into one huge page and has
successfully allocated a new huge page to store the data.
thp_fault_fallback is incremented if a page fault fails to allocate
a huge page and instead falls back to using small pages.
thp_collapse_alloc_failed is incremented if khugepaged found a range
of pages that should be collapsed into one huge page but failed
the allocation.
thp_split_page is incremented every time a huge page is split into base
pages. This can happen for a variety of reasons but a common
reason is that a huge page is old and is being reclaimed.
This action implies splitting all PMD the page mapped with.
thp_split_page_failed is is incremented if kernel fails to split huge
page. This can happen if the page was pinned by somebody.
thp_deferred_split_page is incremented when a huge page is put onto split
queue. This happens when a huge page is partially unmapped and
splitting it would free up some memory. Pages on split queue are
going to be split under memory pressure.
thp_split_pmd is incremented every time a PMD split into table of PTEs.
This can happen, for instance, when application calls mprotect() or
munmap() on part of huge page. It doesn't split huge page, only
page table entry.
thp_zero_page_alloc is incremented every time a huge zero page is
successfully allocated. It includes allocations which where
dropped due race with other allocation. Note, it doesn't count
every map of the huge zero page, only its allocation.
thp_zero_page_alloc_failed is incremented if kernel fails to allocate
huge zero page and falls back to using small pages.
As the system ages, allocating huge pages may be expensive as the
system uses memory compaction to copy data around memory to free a
huge page for use. There are some counters in /proc/vmstat to help
monitor this overhead.
compact_stall is incremented every time a process stalls to run
memory compaction so that a huge page is free for use.
compact_success is incremented if the system compacted memory and
freed a huge page for use.
compact_fail is incremented if the system tries to compact memory
but failed.
compact_pages_moved is incremented each time a page is moved. If
this value is increasing rapidly, it implies that the system
is copying a lot of data to satisfy the huge page allocation.
It is possible that the cost of copying exceeds any savings
from reduced TLB misses.
compact_pagemigrate_failed is incremented when the underlying mechanism
for moving a page failed.
compact_blocks_moved is incremented each time memory compaction examines
a huge page aligned range of pages.
It is possible to establish how long the stalls were using the function
tracer to record how long was spent in __alloc_pages_nodemask and
using the mm_page_alloc tracepoint to identify which allocations were
for huge pages.
== get_user_pages and follow_page ==
get_user_pages and follow_page if run on a hugepage, will return the
head or tail pages as usual (exactly as they would do on
hugetlbfs). Most gup users will only care about the actual physical
address of the page and its temporary pinning to release after the I/O
is complete, so they won't ever notice the fact the page is huge. But
if any driver is going to mangle over the page structure of the tail
page (like for checking page->mapping or other bits that are relevant
for the head page and not the tail page), it should be updated to jump
to check head page instead. Taking reference on any head/tail page would
prevent page from being split by anyone.
NOTE: these aren't new constraints to the GUP API, and they match the
same constrains that applies to hugetlbfs too, so any driver capable
of handling GUP on hugetlbfs will also work fine on transparent
hugepage backed mappings.
In case you can't handle compound pages if they're returned by
follow_page, the FOLL_SPLIT bit can be specified as parameter to
follow_page, so that it will split the hugepages before returning
them. Migration for example passes FOLL_SPLIT as parameter to
follow_page because it's not hugepage aware and in fact it can't work
at all on hugetlbfs (but it instead works fine on transparent
hugepages thanks to FOLL_SPLIT). migration simply can't deal with
hugepages being returned (as it's not only checking the pfn of the
page and pinning it during the copy but it pretends to migrate the
memory in regular page sizes and with regular pte/pmd mappings).
== Optimizing the applications ==
To be guaranteed that the kernel will map a 2M page immediately in any
memory region, the mmap region has to be hugepage naturally
aligned. posix_memalign() can provide that guarantee.
== Hugetlbfs ==
You can use hugetlbfs on a kernel that has transparent hugepage
support enabled just fine as always. No difference can be noted in
hugetlbfs other than there will be less overall fragmentation. All
usual features belonging to hugetlbfs are preserved and
unaffected. libhugetlbfs will also work fine as usual.
== Graceful fallback ==
Code walking pagetables but unware about huge pmds can simply call
split_huge_pmd(vma, pmd, addr) where the pmd is the one returned by
pmd_offset. It's trivial to make the code transparent hugepage aware
by just grepping for "pmd_offset" and adding split_huge_pmd where
missing after pmd_offset returns the pmd. Thanks to the graceful
fallback design, with a one liner change, you can avoid to write
hundred if not thousand of lines of complex code to make your code
hugepage aware.
If you're not walking pagetables but you run into a physical hugepage
but you can't handle it natively in your code, you can split it by
calling split_huge_page(page). This is what the Linux VM does before
it tries to swapout the hugepage for example. split_huge_page() can fail
if the page is pinned and you must handle this correctly.
Example to make mremap.c transparent hugepage aware with a one liner
change:
diff --git a/mm/mremap.c b/mm/mremap.c
--- a/mm/mremap.c
+++ b/mm/mremap.c
@@ -41,6 +41,7 @@ static pmd_t *get_old_pmd(struct mm_stru
return NULL;
pmd = pmd_offset(pud, addr);
+ split_huge_pmd(vma, pmd, addr);
if (pmd_none_or_clear_bad(pmd))
return NULL;
== Locking in hugepage aware code ==
We want as much code as possible hugepage aware, as calling
split_huge_page() or split_huge_pmd() has a cost.
To make pagetable walks huge pmd aware, all you need to do is to call
pmd_trans_huge() on the pmd returned by pmd_offset. You must hold the
mmap_sem in read (or write) mode to be sure an huge pmd cannot be
created from under you by khugepaged (khugepaged collapse_huge_page
takes the mmap_sem in write mode in addition to the anon_vma lock). If
pmd_trans_huge returns false, you just fallback in the old code
paths. If instead pmd_trans_huge returns true, you have to take the
page table lock (pmd_lock()) and re-run pmd_trans_huge. Taking the
page table lock will prevent the huge pmd to be converted into a
regular pmd from under you (split_huge_pmd can run in parallel to the
pagetable walk). If the second pmd_trans_huge returns false, you
should just drop the page table lock and fallback to the old code as
before. Otherwise you can proceed to process the huge pmd and the
hugepage natively. Once finished you can drop the page table lock.
== Refcounts and transparent huge pages ==
Refcounting on THP is mostly consistent with refcounting on other compound
pages:
- get_page()/put_page() and GUP operate in head page's ->_count.
- ->_count in tail pages is always zero: get_page_unless_zero() never
succeed on tail pages.
- map/unmap of the pages with PTE entry increment/decrement ->_mapcount
on relevant sub-page of the compound page.
- map/unmap of the whole compound page accounted in compound_mapcount
(stored in first tail page).
PageDoubleMap() indicates that ->_mapcount in all subpages is offset up by one.
This additional reference is required to get race-free detection of unmap of
subpages when we have them mapped with both PMDs and PTEs.
This is optimization required to lower overhead of per-subpage mapcount
tracking. The alternative is alter ->_mapcount in all subpages on each
map/unmap of the whole compound page.
We set PG_double_map when a PMD of the page got split for the first time,
but still have PMD mapping. The addtional references go away with last
compound_mapcount.
split_huge_page internally has to distribute the refcounts in the head
page to the tail pages before clearing all PG_head/tail bits from the page
structures. It can be done easily for refcounts taken by page table
entries. But we don't have enough information on how to distribute any
additional pins (i.e. from get_user_pages). split_huge_page() fails any
requests to split pinned huge page: it expects page count to be equal to
sum of mapcount of all sub-pages plus one (split_huge_page caller must
have reference for head page).
split_huge_page uses migration entries to stabilize page->_count and
page->_mapcount.
We safe against physical memory scanners too: the only legitimate way
scanner can get reference to a page is get_page_unless_zero().
All tail pages has zero ->_count until atomic_add(). It prevent scanner
from geting reference to tail page up to the point. After the atomic_add()
we don't care about ->_count value. We already known how many references
with should uncharge from head page.
For head page get_page_unless_zero() will succeed and we don't mind. It's
clear where reference should go after split: it will stay on head page.
Note that split_huge_pmd() doesn't have any limitation on refcounting:
pmd can be split at any point and never fails.
== Partial unmap and deferred_split_huge_page() ==
Unmapping part of THP (with munmap() or other way) is not going to free
memory immediately. Instead, we detect that a subpage of THP is not in use
in page_remove_rmap() and queue the THP for splitting if memory pressure
comes. Splitting will free up unused subpages.
Splitting the page right away is not an option due to locking context in
the place where we can detect partial unmap. It's also might be
counterproductive since in many cases partial unmap unmap happens during
exit(2) if an THP crosses VMA boundary.
Function deferred_split_huge_page() is used to queue page for splitting.
The splitting itself will happen when we get memory pressure via shrinker
interface.