* [PATCH 1/3] vmscan: report vm_flags in page_referenced()
2009-06-08 9:10 [PATCH 0/3] make mapped executable pages the first class citizen (with test cases) Wu Fengguang
@ 2009-06-08 9:10 ` Wu Fengguang
2009-06-08 9:10 ` [PATCH 2/3] vmscan: make mapped executable pages the first class citizen Wu Fengguang
` (2 subsequent siblings)
3 siblings, 0 replies; 10+ messages in thread
From: Wu Fengguang @ 2009-06-08 9:10 UTC (permalink / raw)
To: Andrew Morton
Cc: KOSAKI Motohiro, Peter Zijlstra, Wu Fengguang, Andi Kleen,
Christoph Lameter, Elladan, Nick Piggin, Johannes Weiner,
Rik van Riel, tytso@mit.edu, linux-mm@kvack.org,
minchan.kim@gmail.com
[-- Attachment #1: mm-vmscan-report-vm_flags-in-page_referenced.patch --]
[-- Type: text/plain, Size: 6565 bytes --]
Collect vma->vm_flags of the VMAs that actually referenced the page.
This is preparing for more informed reclaim heuristics,
eg. to protect executable file pages more aggressively.
For now only the VM_EXEC bit will be used by the caller.
Thanks to Johannes, Peter and Minchan for all the good tips.
Acked-by: Peter Zijlstra <peterz@infradead.org>
Reviewed-by: Rik van Riel <riel@redhat.com>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Reviewed-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
---
include/linux/rmap.h | 5 +++--
mm/rmap.c | 37 ++++++++++++++++++++++++++-----------
mm/vmscan.c | 7 +++++--
3 files changed, 34 insertions(+), 15 deletions(-)
--- linux.orig/include/linux/rmap.h
+++ linux/include/linux/rmap.h
@@ -83,7 +83,8 @@ static inline void page_dup_rmap(struct
/*
* Called from mm/vmscan.c to handle paging out
*/
-int page_referenced(struct page *, int is_locked, struct mem_cgroup *cnt);
+int page_referenced(struct page *, int is_locked,
+ struct mem_cgroup *cnt, unsigned long *vm_flags);
int try_to_unmap(struct page *, int ignore_refs);
/*
@@ -121,7 +122,7 @@ int page_wrprotect(struct page *page, in
#define anon_vma_prepare(vma) (0)
#define anon_vma_link(vma) do {} while (0)
-#define page_referenced(page,l,cnt) TestClearPageReferenced(page)
+#define page_referenced(page, locked, cnt, flags) TestClearPageReferenced(page)
#define try_to_unmap(page, refs) SWAP_FAIL
static inline int page_mkclean(struct page *page)
--- linux.orig/mm/rmap.c
+++ linux/mm/rmap.c
@@ -333,7 +333,9 @@ static int page_mapped_in_vma(struct pag
* repeatedly from either page_referenced_anon or page_referenced_file.
*/
static int page_referenced_one(struct page *page,
- struct vm_area_struct *vma, unsigned int *mapcount)
+ struct vm_area_struct *vma,
+ unsigned int *mapcount,
+ unsigned long *vm_flags)
{
struct mm_struct *mm = vma->vm_mm;
unsigned long address;
@@ -381,11 +383,14 @@ out_unmap:
(*mapcount)--;
pte_unmap_unlock(pte, ptl);
out:
+ if (referenced)
+ *vm_flags |= vma->vm_flags;
return referenced;
}
static int page_referenced_anon(struct page *page,
- struct mem_cgroup *mem_cont)
+ struct mem_cgroup *mem_cont,
+ unsigned long *vm_flags)
{
unsigned int mapcount;
struct anon_vma *anon_vma;
@@ -405,7 +410,8 @@ static int page_referenced_anon(struct p
*/
if (mem_cont && !mm_match_cgroup(vma->vm_mm, mem_cont))
continue;
- referenced += page_referenced_one(page, vma, &mapcount);
+ referenced += page_referenced_one(page, vma,
+ &mapcount, vm_flags);
if (!mapcount)
break;
}
@@ -418,6 +424,7 @@ static int page_referenced_anon(struct p
* page_referenced_file - referenced check for object-based rmap
* @page: the page we're checking references on.
* @mem_cont: target memory controller
+ * @vm_flags: collect encountered vma->vm_flags who actually referenced the page
*
* For an object-based mapped page, find all the places it is mapped and
* check/clear the referenced flag. This is done by following the page->mapping
@@ -427,7 +434,8 @@ static int page_referenced_anon(struct p
* This function is only called from page_referenced for object-based pages.
*/
static int page_referenced_file(struct page *page,
- struct mem_cgroup *mem_cont)
+ struct mem_cgroup *mem_cont,
+ unsigned long *vm_flags)
{
unsigned int mapcount;
struct address_space *mapping = page->mapping;
@@ -467,7 +475,8 @@ static int page_referenced_file(struct p
*/
if (mem_cont && !mm_match_cgroup(vma->vm_mm, mem_cont))
continue;
- referenced += page_referenced_one(page, vma, &mapcount);
+ referenced += page_referenced_one(page, vma,
+ &mapcount, vm_flags);
if (!mapcount)
break;
}
@@ -481,29 +490,35 @@ static int page_referenced_file(struct p
* @page: the page to test
* @is_locked: caller holds lock on the page
* @mem_cont: target memory controller
+ * @vm_flags: collect encountered vma->vm_flags who actually referenced the page
*
* Quick test_and_clear_referenced for all mappings to a page,
* returns the number of ptes which referenced the page.
*/
-int page_referenced(struct page *page, int is_locked,
- struct mem_cgroup *mem_cont)
+int page_referenced(struct page *page,
+ int is_locked,
+ struct mem_cgroup *mem_cont,
+ unsigned long *vm_flags)
{
int referenced = 0;
if (TestClearPageReferenced(page))
referenced++;
+ *vm_flags = 0;
if (page_mapped(page) && page->mapping) {
if (PageAnon(page))
- referenced += page_referenced_anon(page, mem_cont);
+ referenced += page_referenced_anon(page, mem_cont,
+ vm_flags);
else if (is_locked)
- referenced += page_referenced_file(page, mem_cont);
+ referenced += page_referenced_file(page, mem_cont,
+ vm_flags);
else if (!trylock_page(page))
referenced++;
else {
if (page->mapping)
- referenced +=
- page_referenced_file(page, mem_cont);
+ referenced += page_referenced_file(page,
+ mem_cont, vm_flags);
unlock_page(page);
}
}
--- linux.orig/mm/vmscan.c
+++ linux/mm/vmscan.c
@@ -584,6 +584,7 @@ static unsigned long shrink_page_list(st
struct pagevec freed_pvec;
int pgactivate = 0;
unsigned long nr_reclaimed = 0;
+ unsigned long vm_flags;
cond_resched();
@@ -634,7 +635,8 @@ static unsigned long shrink_page_list(st
goto keep_locked;
}
- referenced = page_referenced(page, 1, sc->mem_cgroup);
+ referenced = page_referenced(page, 1,
+ sc->mem_cgroup, &vm_flags);
/* In active use or really unfreeable? Activate it. */
if (sc->order <= PAGE_ALLOC_COSTLY_ORDER &&
referenced && page_mapping_inuse(page))
@@ -1215,6 +1217,7 @@ static void shrink_active_list(unsigned
{
unsigned long pgmoved;
unsigned long pgscanned;
+ unsigned long vm_flags;
LIST_HEAD(l_hold); /* The pages which were snipped off */
LIST_HEAD(l_inactive);
struct page *page;
@@ -1255,7 +1258,7 @@ static void shrink_active_list(unsigned
/* page_referenced clears PageReferenced */
if (page_mapping_inuse(page) &&
- page_referenced(page, 0, sc->mem_cgroup))
+ page_referenced(page, 0, sc->mem_cgroup, &vm_flags))
pgmoved++;
list_add(&page->lru, &l_inactive);
--
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 10+ messages in thread* [PATCH 2/3] vmscan: make mapped executable pages the first class citizen
2009-06-08 9:10 [PATCH 0/3] make mapped executable pages the first class citizen (with test cases) Wu Fengguang
2009-06-08 9:10 ` [PATCH 1/3] vmscan: report vm_flags in page_referenced() Wu Fengguang
@ 2009-06-08 9:10 ` Wu Fengguang
2009-06-08 15:34 ` Christoph Lameter
2009-06-08 9:10 ` [PATCH 3/3] vmscan: merge duplicate code in shrink_active_list() Wu Fengguang
2009-07-10 7:24 ` [PATCH 0/3] make mapped executable pages the first class citizen (with test cases) Nai Xia
3 siblings, 1 reply; 10+ messages in thread
From: Wu Fengguang @ 2009-06-08 9:10 UTC (permalink / raw)
To: Andrew Morton
Cc: KOSAKI Motohiro, Elladan, Nick Piggin, Andi Kleen,
Christoph Lameter, Rik van Riel, Peter Zijlstra, Wu Fengguang,
Johannes Weiner, tytso@mit.edu, linux-mm@kvack.org,
minchan.kim@gmail.com
[-- Attachment #1: mm-vmscan-protect-exec-referenced.patch --]
[-- Type: text/plain, Size: 19158 bytes --]
Protect referenced PROT_EXEC mapped pages from being deactivated.
PROT_EXEC(or its internal presentation VM_EXEC) pages normally belong to some
currently running executables and their linked libraries, they shall really be
cached aggressively to provide good user experiences.
Thanks to Johannes Weiner for the advice to reuse the VMA walk in
page_referenced() to get the PROT_EXEC bit.
[more details]
( The consequences of this patch will have to be discussed together with
Rik van Riel's recent patch "vmscan: evict use-once pages first". )
( Some of the good points and insights are taken into this changelog.
Thanks to all the involved people for the great LKML discussions. )
the problem
-----------
For a typical desktop, the most precious working set is composed of
*actively accessed*
(1) memory mapped executables
(2) and their anonymous pages
(3) and other files
(4) and the dcache/icache/.. slabs
while the least important data are
(5) infrequently used or use-once files
For a typical desktop, one major problem is busty and large amount of (5)
use-once files flushing out the working set.
Inside the working set, (4) dcache/icache have already been too sticky ;-)
So we only have to care (2) anonymous and (1)(3) file pages.
anonymous pages
---------------
Anonymous pages are effectively immune to the streaming IO attack, because we
now have separate file/anon LRU lists. When the use-once files crowd into the
file LRU, the list's "quality" is significantly lowered. Therefore the scan
balance policy in get_scan_ratio() will choose to scan the (low quality) file
LRU much more frequently than the anon LRU.
file pages
----------
Rik proposed to *not* scan the active file LRU when the inactive list grows
larger than active list. This guarantees that when there are use-once streaming
IO, and the working set is not too large(so that active_size < inactive_size),
the active file LRU will *not* be scanned at all. So the not-too-large working
set can be well protected.
But there are also situations where the file working set is a bit large so that
(active_size >= inactive_size), or the streaming IOs are not purely use-once.
In these cases, the active list will be scanned slowly. Because the current
shrink_active_list() policy is to deactivate active pages regardless of their
referenced bits. The deactivated pages become susceptible to the streaming IO
attack: the inactive list could be scanned fast (500MB / 50MBps = 10s) so that
the deactivated pages don't have enough time to get re-referenced. Because a
user tend to switch between windows in intervals from seconds to minutes.
This patch holds mapped executable pages in the active list as long as they
are referenced during each full scan of the active list. Because the active
list is normally scanned much slower, they get longer grace time (eg. 100s)
for further references, which better matches the pace of user operations.
Therefore this patch greatly prolongs the in-cache time of executable code,
when there are moderate memory pressures.
before patch: guaranteed to be cached if reference intervals < I
after patch: guaranteed to be cached if reference intervals < I+A
(except when randomly reclaimed by the lumpy reclaim)
where
A = time to fully scan the active file LRU
I = time to fully scan the inactive file LRU
Note that normally A >> I.
side effects
------------
This patch is safe in general, it restores the pre-2.6.28 mmap() behavior
but in a much smaller and well targeted scope.
One may worry about some one to abuse the PROT_EXEC heuristic. But as
Andrew Morton stated, there are other tricks to getting that sort of boost.
Another concern is the PROT_EXEC mapped pages growing large in rare cases,
and therefore hurting reclaim efficiency. But a sane application targeted for
large audience will never use PROT_EXEC for data mappings. If some home made
application tries to abuse that bit, it shall be aware of the consequences.
If it is abused to scale of 2/3 total memory, it gains nothing but overheads.
benchmarks
----------
1) memory tight desktop
1.1) brief summary
- clock time and major faults are reduced by 50%;
- pswpin numbers are reduced to ~1/3.
That means X desktop responsiveness is doubled under high memory/swap pressure.
1.2) test scenario
- nfsroot gnome desktop with 512M physical memory
- run some programs, and switch between the existing windows
after starting each new program.
1.3) progress timing (seconds)
before after programs
0.02 0.02 N xeyes
0.75 0.76 N firefox
2.02 1.88 N nautilus
3.36 3.17 N nautilus --browser
5.26 4.89 N gthumb
7.12 6.47 N gedit
9.22 8.16 N xpdf /usr/share/doc/shared-mime-info/shared-mime-info-spec.pdf
13.58 12.55 N xterm
15.87 14.57 N mlterm
18.63 17.06 N gnome-terminal
21.16 18.90 N urxvt
26.24 23.48 N gnome-system-monitor
28.72 26.52 N gnome-help
32.15 29.65 N gnome-dictionary
39.66 36.12 N /usr/games/sol
43.16 39.27 N /usr/games/gnometris
48.65 42.56 N /usr/games/gnect
53.31 47.03 N /usr/games/gtali
58.60 52.05 N /usr/games/iagno
65.77 55.42 N /usr/games/gnotravex
70.76 61.47 N /usr/games/mahjongg
76.15 67.11 N /usr/games/gnome-sudoku
86.32 75.15 N /usr/games/glines
92.21 79.70 N /usr/games/glchess
103.79 88.48 N /usr/games/gnomine
113.84 96.51 N /usr/games/gnotski
124.40 102.19 N /usr/games/gnibbles
137.41 114.93 N /usr/games/gnobots2
155.53 125.02 N /usr/games/blackjack
179.85 135.11 N /usr/games/same-gnome
224.49 154.50 N /usr/bin/gnome-window-properties
248.44 162.09 N /usr/bin/gnome-default-applications-properties
282.62 173.29 N /usr/bin/gnome-at-properties
323.72 188.21 N /usr/bin/gnome-typing-monitor
363.99 199.93 N /usr/bin/gnome-at-visual
394.21 206.95 N /usr/bin/gnome-sound-properties
435.14 224.49 N /usr/bin/gnome-at-mobility
463.05 234.11 N /usr/bin/gnome-keybinding-properties
503.75 248.59 N /usr/bin/gnome-about-me
554.00 276.27 N /usr/bin/gnome-display-properties
615.48 304.39 N /usr/bin/gnome-network-preferences
693.03 342.01 N /usr/bin/gnome-mouse-properties
759.90 388.58 N /usr/bin/gnome-appearance-properties
937.90 508.47 N /usr/bin/gnome-control-center
1109.75 587.57 N /usr/bin/gnome-keyboard-properties
1399.05 758.16 N : oocalc
1524.64 830.03 N : oodraw
1684.31 900.03 N : ooimpress
1874.04 993.91 N : oomath
2115.12 1081.89 N : ooweb
2369.02 1161.99 N : oowriter
Note that the last ": oo*" commands are actually commented out.
1.4) vmstat numbers (some relevant ones are marked with *)
before after
nr_free_pages 1293 3898
nr_inactive_anon 59956 53460
nr_active_anon 26815 30026
nr_inactive_file 2657 3218
nr_active_file 2019 2806
nr_unevictable 4 4
nr_mlock 4 4
nr_anon_pages 26706 27859
*nr_mapped 3542 4469
nr_file_pages 72232 67681
nr_dirty 1 0
nr_writeback 123 19
nr_slab_reclaimable 3375 3534
nr_slab_unreclaimable 11405 10665
nr_page_table_pages 8106 7864
nr_unstable 0 0
nr_bounce 0 0
*nr_vmscan_write 394776 230839
nr_writeback_temp 0 0
numa_hit 6843353 3318676
numa_miss 0 0
numa_foreign 0 0
numa_interleave 1719 1719
numa_local 6843353 3318676
numa_other 0 0
*pgpgin 5954683 2057175
*pgpgout 1578276 922744
*pswpin 1486615 512238
*pswpout 394568 230685
pgalloc_dma 277432 56602
pgalloc_dma32 6769477 3310348
pgalloc_normal 0 0
pgalloc_movable 0 0
pgfree 7048396 3371118
pgactivate 2036343 1471492
pgdeactivate 2189691 1612829
pgfault 3702176 3100702
*pgmajfault 452116 201343
pgrefill_dma 12185 7127
pgrefill_dma32 334384 653703
pgrefill_normal 0 0
pgrefill_movable 0 0
pgsteal_dma 74214 22179
pgsteal_dma32 3334164 1638029
pgsteal_normal 0 0
pgsteal_movable 0 0
pgscan_kswapd_dma 1081421 1216199
pgscan_kswapd_dma32 58979118 46002810
pgscan_kswapd_normal 0 0
pgscan_kswapd_movable 0 0
pgscan_direct_dma 2015438 1086109
pgscan_direct_dma32 55787823 36101597
pgscan_direct_normal 0 0
pgscan_direct_movable 0 0
pginodesteal 3461 7281
slabs_scanned 564864 527616
kswapd_steal 2889797 1448082
kswapd_inodesteal 14827 14835
pageoutrun 43459 21562
allocstall 9653 4032
pgrotated 384216 228631
1.5) free numbers at the end of the tests
before patch:
total used free shared buffers cached
Mem: 474 467 7 0 0 236
-/+ buffers/cache: 230 243
Swap: 1023 418 605
after patch:
total used free shared buffers cached
Mem: 474 457 16 0 0 236
-/+ buffers/cache: 221 253
Swap: 1023 404 619
2) memory flushing in a file server
2.1) brief summary
The number of major faults from 50 to 3 during 10% cache hot reads.
That means this patch successfully stops major faults when the active file list
is slowly scanned when there are partially cache hot streaming IO.
2.2) test scenario
Do 100000 pread(size=110 pages, offset=(i*100) pages), where 10% of the pages
will be activated:
for i in `seq 0 100 10000000`; do echo $i 110; done > pattern-hot-10
iotrace.rb --load pattern-hot-10 --play /b/sparse
vmmon nr_mapped nr_active_file nr_inactive_file pgmajfault pgdeactivate pgfree
and monitor /proc/vmstat during the time. The test box has 2G memory.
I carried out tests on fresh booted console as well as X desktop,
and fetched the vmstat numbers on
(1) begin: shortly after the big read IO starts;
(2) end: just before the big read IO stops;
(3) restore: the big read IO stops and the zsh working set restored
(4) restore X: after IO, switch back and forth between the urxvt and firefox
windows to restore their working set.
2.3) console mode results
nr_mapped nr_active_file nr_inactive_file pgmajfault pgdeactivate pgfree
2.6.29 VM_EXEC protection ON:
begin: 2481 2237 8694 630 0 574299
end: 275 231976 233914 633 776271 20933042
restore: 370 232154 234524 691 777183 20958453
2.6.29 VM_EXEC protection ON (second run):
begin: 2434 2237 8493 629 0 574195
end: 284 231970 233536 632 771918 20896129
restore: 399 232218 234789 690 774526 20957909
2.6.30-rc4-mm VM_EXEC protection OFF:
begin: 2479 2344 9659 210 0 579643
end: 284 232010 234142 260 772776 20917184
restore: 379 232159 234371 301 774888 20967849
The above console numbers show that
- The startup pgmajfault of 2.6.30-rc4-mm is merely 1/3 that of 2.6.29.
I'd attribute that improvement to the mmap readahead improvements :-)
- The pgmajfault increment during the file copy is 633-630=3 vs 260-210=50.
That's a huge improvement - which means with the VM_EXEC protection logic,
active mmap pages is pretty safe even under partially cache hot streaming IO.
- when active:inactive file lru size reaches 1:1, their scan rates is 1:20.8
under 10% cache hot IO. (computed with formula Dpgdeactivate:Dpgfree)
That roughly means the active mmap pages get 20.8 more chances to get
re-referenced to stay in memory.
- The absolute nr_mapped drops considerably to 1/9 during the big IO, and the
dropped pages are mostly inactive ones. The patch has almost no impact in
this aspect, that means it won't unnecessarily increase memory pressure.
(In contrast, your 20% mmap protection ratio will keep them all, and
therefore eliminate the extra 41 major faults to restore working set
of zsh etc.)
The iotrace.rb read throughput is
151.194384MB/s 284.198252s 100001x 450560b --load pattern-hot-10 --play /b/sparse
which means the inactive list is rotated at the speed of 250MB/s,
so a full scan of which takes about 3.5 seconds, while a full scan
of active file list takes about 77 seconds.
2.4) X mode results
We can reach roughly the same conclusions for X desktop:
nr_mapped nr_active_file nr_inactive_file pgmajfault pgdeactivate pgfree
2.6.30-rc4-mm VM_EXEC protection ON:
begin: 9740 8920 64075 561 0 678360
end: 768 218254 220029 565 798953 21057006
restore: 857 218543 220987 606 799462 21075710
restore X: 2414 218560 225344 797 799462 21080795
2.6.30-rc4-mm VM_EXEC protection OFF:
begin: 9368 5035 26389 554 0 633391
end: 770 218449 221230 661 646472 17832500
restore: 1113 218466 220978 710 649881 17905235
restore X: 2687 218650 225484 947 802700 21083584
- the absolute nr_mapped drops considerably (to 1/13 of the original size)
during the streaming IO.
- the delta of pgmajfault is 3 vs 107 during IO, or 236 vs 393
during the whole process.
CC: Elladan <elladan@eskimo.com>
CC: Nick Piggin <npiggin@suse.de>
CC: Andi Kleen <andi@firstfloor.org>
CC: Christoph Lameter <cl@linux-foundation.org>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Reviewed-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
---
mm/vmscan.c | 52 +++++++++++++++++++++++++++++++++++++++++++-------
1 file changed, 45 insertions(+), 7 deletions(-)
--- linux.orig/mm/vmscan.c
+++ linux/mm/vmscan.c
@@ -1219,6 +1219,7 @@ static void shrink_active_list(unsigned
unsigned long pgscanned;
unsigned long vm_flags;
LIST_HEAD(l_hold); /* The pages which were snipped off */
+ LIST_HEAD(l_active);
LIST_HEAD(l_inactive);
struct page *page;
struct pagevec pvec;
@@ -1258,28 +1259,42 @@ static void shrink_active_list(unsigned
/* page_referenced clears PageReferenced */
if (page_mapping_inuse(page) &&
- page_referenced(page, 0, sc->mem_cgroup, &vm_flags))
+ page_referenced(page, 0, sc->mem_cgroup, &vm_flags)) {
pgmoved++;
+ /*
+ * Identify referenced, file-backed active pages and
+ * give them one more trip around the active list. So
+ * that executable code get better chances to stay in
+ * memory under moderate memory pressure. Anon pages
+ * are not likely to be evicted by use-once streaming
+ * IO, plus JVM can create lots of anon VM_EXEC pages,
+ * so we ignore them here.
+ */
+ if ((vm_flags & VM_EXEC) && !PageAnon(page)) {
+ list_add(&page->lru, &l_active);
+ continue;
+ }
+ }
list_add(&page->lru, &l_inactive);
}
/*
- * Move the pages to the [file or anon] inactive list.
+ * Move pages back to the lru list.
*/
pagevec_init(&pvec, 1);
- lru = LRU_BASE + file * LRU_FILE;
spin_lock_irq(&zone->lru_lock);
/*
- * Count referenced pages from currently used mappings as
- * rotated, even though they are moved to the inactive list.
- * This helps balance scan pressure between file and anonymous
- * pages in get_scan_ratio.
+ * Count referenced pages from currently used mappings as rotated,
+ * even though only some of them are actually re-activated. This
+ * helps balance scan pressure between file and anonymous pages in
+ * get_scan_ratio.
*/
reclaim_stat->recent_rotated[!!file] += pgmoved;
pgmoved = 0; /* count pages moved to inactive list */
+ lru = LRU_BASE + file * LRU_FILE;
while (!list_empty(&l_inactive)) {
page = lru_to_page(&l_inactive);
prefetchw_prev_lru_page(page, &l_inactive, flags);
@@ -1302,6 +1317,29 @@ static void shrink_active_list(unsigned
__mod_zone_page_state(zone, NR_LRU_BASE + lru, pgmoved);
__count_zone_vm_events(PGREFILL, zone, pgscanned);
__count_vm_events(PGDEACTIVATE, pgmoved);
+
+ pgmoved = 0; /* count pages moved back to active list */
+ lru = LRU_ACTIVE + file * LRU_FILE;
+ while (!list_empty(&l_active)) {
+ page = lru_to_page(&l_active);
+ prefetchw_prev_lru_page(page, &l_active, flags);
+ VM_BUG_ON(PageLRU(page));
+ SetPageLRU(page);
+ VM_BUG_ON(!PageActive(page));
+
+ list_move(&page->lru, &zone->lru[lru].list);
+ mem_cgroup_add_lru_list(page, lru);
+ pgmoved++;
+ if (!pagevec_add(&pvec, page)) {
+ spin_unlock_irq(&zone->lru_lock);
+ if (buffer_heads_over_limit)
+ pagevec_strip(&pvec);
+ __pagevec_release(&pvec);
+ spin_lock_irq(&zone->lru_lock);
+ }
+ }
+ __mod_zone_page_state(zone, NR_LRU_BASE + lru, pgmoved);
+
spin_unlock_irq(&zone->lru_lock);
if (buffer_heads_over_limit)
pagevec_strip(&pvec);
--
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 10+ messages in thread* Re: [PATCH 2/3] vmscan: make mapped executable pages the first class citizen
2009-06-08 9:10 ` [PATCH 2/3] vmscan: make mapped executable pages the first class citizen Wu Fengguang
@ 2009-06-08 15:34 ` Christoph Lameter
2009-06-08 17:30 ` Nai Xia
2009-06-09 3:28 ` Wu Fengguang
0 siblings, 2 replies; 10+ messages in thread
From: Christoph Lameter @ 2009-06-08 15:34 UTC (permalink / raw)
To: Wu Fengguang
Cc: Andrew Morton, KOSAKI Motohiro, Elladan, Nick Piggin, Andi Kleen,
Rik van Riel, Peter Zijlstra, Johannes Weiner, tytso@mit.edu,
linux-mm@kvack.org, minchan.kim@gmail.com
On Mon, 8 Jun 2009, Wu Fengguang wrote:
> 1.2) test scenario
>
> - nfsroot gnome desktop with 512M physical memory
> - run some programs, and switch between the existing windows
> after starting each new program.
Is there a predefined sequence or does this vary between tests? Scripted?
What percentage of time is saved in the test after due to the
modifications?
Around 20%?
> (1) begin: shortly after the big read IO starts;
> (2) end: just before the big read IO stops;
> (3) restore: the big read IO stops and the zsh working set restored
> (4) restore X: after IO, switch back and forth between the urxvt and firefox
> windows to restore their working set.
Any action done on the firefox sessions? Or just switch to a firefox
session that needs to redraw?
> The above console numbers show that
>
> - The startup pgmajfault of 2.6.30-rc4-mm is merely 1/3 that of 2.6.29.
> I'd attribute that improvement to the mmap readahead improvements :-)
So there are other effects,,, You not measuring the effect only this
patchset?
> - The pgmajfault increment during the file copy is 633-630=3 vs 260-210=50.
> That's a huge improvement - which means with the VM_EXEC protection logic,
> active mmap pages is pretty safe even under partially cache hot streaming IO.
Looks good.
> - The absolute nr_mapped drops considerably to 1/9 during the big IO, and the
> dropped pages are mostly inactive ones. The patch has almost no impact in
> this aspect, that means it won't unnecessarily increase memory pressure.
> (In contrast, your 20% mmap protection ratio will keep them all, and
> therefore eliminate the extra 41 major faults to restore working set
> of zsh etc.)
Good.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 2/3] vmscan: make mapped executable pages the first class citizen
2009-06-08 15:34 ` Christoph Lameter
@ 2009-06-08 17:30 ` Nai Xia
2009-06-09 3:28 ` Wu Fengguang
1 sibling, 0 replies; 10+ messages in thread
From: Nai Xia @ 2009-06-08 17:30 UTC (permalink / raw)
To: Christoph Lameter
Cc: Wu Fengguang, Andrew Morton, KOSAKI Motohiro, Elladan,
Nick Piggin, Andi Kleen, Rik van Riel, Peter Zijlstra,
Johannes Weiner, tytso@mit.edu, linux-mm@kvack.org,
minchan.kim@gmail.com
On Mon, Jun 8, 2009 at 11:34 PM, Christoph
Lameter<cl@linux-foundation.org> wrote:
> On Mon, 8 Jun 2009, Wu Fengguang wrote:
>
>> 1.2) test scenario
>>
>> - nfsroot gnome desktop with 512M physical memory
>> - run some programs, and switch between the existing windows
>> after starting each new program.
>
> Is there a predefined sequence or does this vary between tests? Scripted?
>
> What percentage of time is saved in the test after due to the
> modifications?
> Around 20%?
I think measuring the percentage of saved time may not be a good idea.
The major underlying factor for time of swithing GUI windows may vary
application to application, distribution to distribution and machine to
machine. It's not reproducable.
I am having a ridiculous timing for swithing from any window to window
of slickedit, because of its damn slow redrawing method.
I bet this patch will gain at most 1% on timing for this case. :)
>
>> (1) begin: shortly after the big read IO starts;
>> (2) end: just before the big read IO stops;
>> (3) restore: the big read IO stops and the zsh working set restored
>> (4) restore X: after IO, switch back and forth between the urxvt and firefox
>> windows to restore their working set.
>
> Any action done on the firefox sessions? Or just switch to a firefox
> session that needs to redraw?
>
>> The above console numbers show that
>>
>> - The startup pgmajfault of 2.6.30-rc4-mm is merely 1/3 that of 2.6.29.
>> I'd attribute that improvement to the mmap readahead improvements :-)
>
> So there are other effects,,, You not measuring the effect only this
> patchset?
>
>> - The pgmajfault increment during the file copy is 633-630=3 vs 260-210=50.
>> That's a huge improvement - which means with the VM_EXEC protection logic,
>> active mmap pages is pretty safe even under partially cache hot streaming IO.
>
> Looks good.
>
>> - The absolute nr_mapped drops considerably to 1/9 during the big IO, and the
>> dropped pages are mostly inactive ones. The patch has almost no impact in
>> this aspect, that means it won't unnecessarily increase memory pressure.
>> (In contrast, your 20% mmap protection ratio will keep them all, and
>> therefore eliminate the extra 41 major faults to restore working set
>> of zsh etc.)
>
> Good.
>
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo@kvack.org. For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
>
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 2/3] vmscan: make mapped executable pages the first class citizen
2009-06-08 15:34 ` Christoph Lameter
2009-06-08 17:30 ` Nai Xia
@ 2009-06-09 3:28 ` Wu Fengguang
1 sibling, 0 replies; 10+ messages in thread
From: Wu Fengguang @ 2009-06-09 3:28 UTC (permalink / raw)
To: Christoph Lameter
Cc: Andrew Morton, KOSAKI Motohiro, Elladan, Nick Piggin, Andi Kleen,
Rik van Riel, Peter Zijlstra, Johannes Weiner, tytso@mit.edu,
linux-mm@kvack.org, minchan.kim@gmail.com
[-- Attachment #1: Type: text/plain, Size: 4800 bytes --]
On Mon, Jun 08, 2009 at 11:34:06PM +0800, Christoph Lameter wrote:
> On Mon, 8 Jun 2009, Wu Fengguang wrote:
>
> > 1.2) test scenario
> >
> > - nfsroot gnome desktop with 512M physical memory
> > - run some programs, and switch between the existing windows
> > after starting each new program.
>
> Is there a predefined sequence or does this vary between tests? Scripted?
Yes it's scripted testing and has a predefined sequence.
The scripts are attached for your reference.
> What percentage of time is saved in the test after due to the
> modifications?
> Around 20%?
It's 50%, hehe. I've posted the startup times for each program:
before after programs
0.02 0.02 N xeyes
0.75 0.76 N firefox
2.02 1.88 N nautilus
3.36 3.17 N nautilus --browser
5.26 4.89 N gthumb
7.12 6.47 N gedit
9.22 8.16 N xpdf /usr/share/doc/shared-mime-info/shared-mime-info-spec.pdf
13.58 12.55 N xterm
15.87 14.57 N mlterm
18.63 17.06 N gnome-terminal
21.16 18.90 N urxvt
26.24 23.48 N gnome-system-monitor
28.72 26.52 N gnome-help
32.15 29.65 N gnome-dictionary
39.66 36.12 N /usr/games/sol
43.16 39.27 N /usr/games/gnometris
48.65 42.56 N /usr/games/gnect
53.31 47.03 N /usr/games/gtali
58.60 52.05 N /usr/games/iagno
65.77 55.42 N /usr/games/gnotravex
70.76 61.47 N /usr/games/mahjongg
76.15 67.11 N /usr/games/gnome-sudoku
86.32 75.15 N /usr/games/glines
92.21 79.70 N /usr/games/glchess
103.79 88.48 N /usr/games/gnomine
113.84 96.51 N /usr/games/gnotski
124.40 102.19 N /usr/games/gnibbles
137.41 114.93 N /usr/games/gnobots2
155.53 125.02 N /usr/games/blackjack
179.85 135.11 N /usr/games/same-gnome
224.49 154.50 N /usr/bin/gnome-window-properties
248.44 162.09 N /usr/bin/gnome-default-applications-properties
282.62 173.29 N /usr/bin/gnome-at-properties
323.72 188.21 N /usr/bin/gnome-typing-monitor
363.99 199.93 N /usr/bin/gnome-at-visual
394.21 206.95 N /usr/bin/gnome-sound-properties
435.14 224.49 N /usr/bin/gnome-at-mobility
463.05 234.11 N /usr/bin/gnome-keybinding-properties
503.75 248.59 N /usr/bin/gnome-about-me
554.00 276.27 N /usr/bin/gnome-display-properties
615.48 304.39 N /usr/bin/gnome-network-preferences
693.03 342.01 N /usr/bin/gnome-mouse-properties
759.90 388.58 N /usr/bin/gnome-appearance-properties
937.90 508.47 N /usr/bin/gnome-control-center
1109.75 587.57 N /usr/bin/gnome-keyboard-properties
1399.05 758.16 N : oocalc
1524.64 830.03 N : oodraw
1684.31 900.03 N : ooimpress
1874.04 993.91 N : oomath
2115.12 1081.89 N : ooweb
2369.02 1161.99 N : oowriter
> > (1) begin: shortly after the big read IO starts;
> > (2) end: just before the big read IO stops;
> > (3) restore: the big read IO stops and the zsh working set restored
> > (4) restore X: after IO, switch back and forth between the urxvt and firefox
> > windows to restore their working set.
>
> Any action done on the firefox sessions? Or just switch to a firefox
> session that needs to redraw?
After starting each new program, a new tab is opened in firefox to render a
simple web page. It's the same web page, so firefox may actually cache it.
> > The above console numbers show that
> >
> > - The startup pgmajfault of 2.6.30-rc4-mm is merely 1/3 that of 2.6.29.
> > I'd attribute that improvement to the mmap readahead improvements :-)
>
> So there are other effects,,, You not measuring the effect only this
> patchset?
Yes there are additional effects in the .29 vs .30 comparisons.
But the following .30 vs .30 comparisons in X can lead to the same conclusions
except for this additional effect.
> > - The pgmajfault increment during the file copy is 633-630=3 vs 260-210=50.
> > That's a huge improvement - which means with the VM_EXEC protection logic,
> > active mmap pages is pretty safe even under partially cache hot streaming IO.
>
> Looks good.
>
> > - The absolute nr_mapped drops considerably to 1/9 during the big IO, and the
> > dropped pages are mostly inactive ones. The patch has almost no impact in
> > this aspect, that means it won't unnecessarily increase memory pressure.
> > (In contrast, your 20% mmap protection ratio will keep them all, and
> > therefore eliminate the extra 41 major faults to restore working set
> > of zsh etc.)
>
> Good.
Thanks,
Fengguang
[-- Attachment #2: run-many-x-apps.sh --]
[-- Type: application/x-sh, Size: 1735 bytes --]
[-- Attachment #3: test-mmap-exec-prot.sh --]
[-- Type: application/x-sh, Size: 203 bytes --]
^ permalink raw reply [flat|nested] 10+ messages in thread
* [PATCH 3/3] vmscan: merge duplicate code in shrink_active_list()
2009-06-08 9:10 [PATCH 0/3] make mapped executable pages the first class citizen (with test cases) Wu Fengguang
2009-06-08 9:10 ` [PATCH 1/3] vmscan: report vm_flags in page_referenced() Wu Fengguang
2009-06-08 9:10 ` [PATCH 2/3] vmscan: make mapped executable pages the first class citizen Wu Fengguang
@ 2009-06-08 9:10 ` Wu Fengguang
2009-07-10 7:24 ` [PATCH 0/3] make mapped executable pages the first class citizen (with test cases) Nai Xia
3 siblings, 0 replies; 10+ messages in thread
From: Wu Fengguang @ 2009-06-08 9:10 UTC (permalink / raw)
To: Andrew Morton
Cc: KOSAKI Motohiro, Pekka Enberg, Peter Zijlstra, Wu Fengguang,
Andi Kleen, Christoph Lameter, Elladan, Nick Piggin,
Johannes Weiner, Rik van Riel, tytso@mit.edu, linux-mm@kvack.org,
minchan.kim@gmail.com
[-- Attachment #1: mm-vmscan-reduce-code.patch --]
[-- Type: text/plain, Size: 5649 bytes --]
The "move pages to active list" and "move pages to inactive list"
code blocks are mostly identical and can be served by a function.
Thanks to Andrew Morton for pointing this out.
Note that buffer_heads_over_limit check will also be carried out
for re-activated pages, which is slightly different from pre-2.6.28
kernels. Also, Rik's "vmscan: evict use-once pages first" patch
could totally stop scans of active file list when memory pressure is low.
So the net effect could be, the number of buffer heads is now more
likely to grow large.
However that's fine according to Johannes' comments:
I don't think that this could be harmful. We just preserve the buffer
mappings of what we consider the working set and with low memory
pressure, as you say, this set is not big.
As to stripping of reactivated pages: the only pages we re-activate
for now are those VM_EXEC mapped ones. Since we don't expect IO from
or to these pages, removing the buffer mappings in case they grow too
large should be okay, I guess.
CC: Pekka Enberg <penberg@cs.helsinki.fi>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Reviewed-by: Rik van Riel <riel@redhat.com>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Reviewed-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
---
mm/vmscan.c | 95 ++++++++++++++++++++++----------------------------
1 file changed, 42 insertions(+), 53 deletions(-)
--- linux.orig/mm/vmscan.c
+++ linux/mm/vmscan.c
@@ -1211,6 +1211,43 @@ static inline void note_zone_scanning_pr
* But we had to alter page->flags anyway.
*/
+static void move_active_pages_to_lru(struct zone *zone,
+ struct list_head *list,
+ enum lru_list lru)
+{
+ unsigned long pgmoved = 0;
+ struct pagevec pvec;
+ struct page *page;
+
+ pagevec_init(&pvec, 1);
+
+ while (!list_empty(list)) {
+ page = lru_to_page(list);
+ prefetchw_prev_lru_page(page, list, flags);
+
+ VM_BUG_ON(PageLRU(page));
+ SetPageLRU(page);
+
+ VM_BUG_ON(!PageActive(page));
+ if (!is_active_lru(lru))
+ ClearPageActive(page); /* we are de-activating */
+
+ list_move(&page->lru, &zone->lru[lru].list);
+ mem_cgroup_add_lru_list(page, lru);
+ pgmoved++;
+
+ if (!pagevec_add(&pvec, page) || list_empty(list)) {
+ spin_unlock_irq(&zone->lru_lock);
+ if (buffer_heads_over_limit)
+ pagevec_strip(&pvec);
+ __pagevec_release(&pvec);
+ spin_lock_irq(&zone->lru_lock);
+ }
+ }
+ __mod_zone_page_state(zone, NR_LRU_BASE + lru, pgmoved);
+ if (!is_active_lru(lru))
+ __count_vm_events(PGDEACTIVATE, pgmoved);
+}
static void shrink_active_list(unsigned long nr_pages, struct zone *zone,
struct scan_control *sc, int priority, int file)
@@ -1222,8 +1259,6 @@ static void shrink_active_list(unsigned
LIST_HEAD(l_active);
LIST_HEAD(l_inactive);
struct page *page;
- struct pagevec pvec;
- enum lru_list lru;
struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc);
lru_add_drain();
@@ -1240,6 +1275,7 @@ static void shrink_active_list(unsigned
}
reclaim_stat->recent_scanned[!!file] += pgmoved;
+ __count_zone_vm_events(PGREFILL, zone, pgscanned);
if (file)
__mod_zone_page_state(zone, NR_ACTIVE_FILE, -pgmoved);
else
@@ -1282,8 +1318,6 @@ static void shrink_active_list(unsigned
/*
* Move pages back to the lru list.
*/
- pagevec_init(&pvec, 1);
-
spin_lock_irq(&zone->lru_lock);
/*
* Count referenced pages from currently used mappings as rotated,
@@ -1293,57 +1327,12 @@ static void shrink_active_list(unsigned
*/
reclaim_stat->recent_rotated[!!file] += pgmoved;
- pgmoved = 0; /* count pages moved to inactive list */
- lru = LRU_BASE + file * LRU_FILE;
- while (!list_empty(&l_inactive)) {
- page = lru_to_page(&l_inactive);
- prefetchw_prev_lru_page(page, &l_inactive, flags);
- VM_BUG_ON(PageLRU(page));
- SetPageLRU(page);
- VM_BUG_ON(!PageActive(page));
- ClearPageActive(page);
-
- list_move(&page->lru, &zone->lru[lru].list);
- mem_cgroup_add_lru_list(page, lru);
- pgmoved++;
- if (!pagevec_add(&pvec, page)) {
- spin_unlock_irq(&zone->lru_lock);
- if (buffer_heads_over_limit)
- pagevec_strip(&pvec);
- __pagevec_release(&pvec);
- spin_lock_irq(&zone->lru_lock);
- }
- }
- __mod_zone_page_state(zone, NR_LRU_BASE + lru, pgmoved);
- __count_zone_vm_events(PGREFILL, zone, pgscanned);
- __count_vm_events(PGDEACTIVATE, pgmoved);
-
- pgmoved = 0; /* count pages moved back to active list */
- lru = LRU_ACTIVE + file * LRU_FILE;
- while (!list_empty(&l_active)) {
- page = lru_to_page(&l_active);
- prefetchw_prev_lru_page(page, &l_active, flags);
- VM_BUG_ON(PageLRU(page));
- SetPageLRU(page);
- VM_BUG_ON(!PageActive(page));
-
- list_move(&page->lru, &zone->lru[lru].list);
- mem_cgroup_add_lru_list(page, lru);
- pgmoved++;
- if (!pagevec_add(&pvec, page)) {
- spin_unlock_irq(&zone->lru_lock);
- if (buffer_heads_over_limit)
- pagevec_strip(&pvec);
- __pagevec_release(&pvec);
- spin_lock_irq(&zone->lru_lock);
- }
- }
- __mod_zone_page_state(zone, NR_LRU_BASE + lru, pgmoved);
+ move_active_pages_to_lru(zone, &l_active,
+ LRU_ACTIVE + file * LRU_FILE);
+ move_active_pages_to_lru(zone, &l_inactive,
+ LRU_BASE + file * LRU_FILE);
spin_unlock_irq(&zone->lru_lock);
- if (buffer_heads_over_limit)
- pagevec_strip(&pvec);
- pagevec_release(&pvec);
}
static int inactive_anon_is_low_global(struct zone *zone)
--
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 10+ messages in thread* Re: [PATCH 0/3] make mapped executable pages the first class citizen (with test cases)
2009-06-08 9:10 [PATCH 0/3] make mapped executable pages the first class citizen (with test cases) Wu Fengguang
` (2 preceding siblings ...)
2009-06-08 9:10 ` [PATCH 3/3] vmscan: merge duplicate code in shrink_active_list() Wu Fengguang
@ 2009-07-10 7:24 ` Nai Xia
2009-07-10 8:34 ` Wu Fengguang
3 siblings, 1 reply; 10+ messages in thread
From: Nai Xia @ 2009-07-10 7:24 UTC (permalink / raw)
To: Wu Fengguang
Cc: Andrew Morton, KOSAKI Motohiro, Andi Kleen, Christoph Lameter,
Elladan, Nick Piggin, Johannes Weiner, Peter Zijlstra,
Rik van Riel, tytso@mit.edu, linux-mm@kvack.org,
minchan.kim@gmail.com
Hi,
I was able to launch some tests with SPEC cpu2006.
The benchmark was based on mmotm
commit 0b7292956dbdfb212abf6e3c9cfb41e9471e1081 on a intel Q6600 box with
4G ram. The kernel cmdline mem=500M was used to see how good exec-prot can
be under memory stress.
Following are the results:
Estimated
Base Base Base
Benchmarks Ref. Run Time Ratio
mmotm with 500M
400.perlbench 9770 671 14.6 *
401.bzip2 9650 1011 9.55 *
403.gcc 8050 774 10.4 *
462.libquantum 20720 1213 17.1 *
mmot-prot with 500M
400.perlbench 9770 658 14.8 *
401.bzip2 9650 1007 9.58 *
403.gcc 8050 749 10.8 *
462.libquantum 20720 1116 18.6 *
mmotm with 4G ( allowing the full working sets)
400.perlbench 9770 594 16.5 *
401.bzip2 9650 828 11.7 *
403.gcc 8050 523 15.4 *
462.libquantum 20720 1121 18.5 *
It's worth noting that SPEC documented "The CPU2006 benchmarks
(code + workload) have been designed to fit within about 1GB of
physical memory",
and the exec vm sizes of these programs are as below:
perlbench 956KB
bzip2 56KB
gcc 3008KB
libquantum 36KB
Are we expecting to see more good results for cpu-bound programs (e.g.
scientific ones)
with large number of exec pages ?
Best Regards,
Nai Xia
On Mon, Jun 8, 2009 at 5:10 PM, Wu Fengguang<fengguang.wu@intel.com> wrote:
> Andrew,
>
> I managed to back this patchset with two test cases :)
>
> They demonstrated that
> - X desktop responsiveness can be *doubled* under high memory/swap pressure
> - it can almost stop major faults when the active file list is slowly scanned
> because of undergoing partially cache hot streaming IO
>
> The details are included in the changelog.
>
> Thanks,
> Fengguang
> --
>
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo@kvack.org. For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
>
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 10+ messages in thread* Re: [PATCH 0/3] make mapped executable pages the first class citizen (with test cases)
2009-07-10 7:24 ` [PATCH 0/3] make mapped executable pages the first class citizen (with test cases) Nai Xia
@ 2009-07-10 8:34 ` Wu Fengguang
2009-07-10 16:50 ` Nai Xia
0 siblings, 1 reply; 10+ messages in thread
From: Wu Fengguang @ 2009-07-10 8:34 UTC (permalink / raw)
To: Nai Xia
Cc: Andrew Morton, KOSAKI Motohiro, Andi Kleen, Christoph Lameter,
Elladan, Nick Piggin, Johannes Weiner, Peter Zijlstra,
Rik van Riel, tytso@mit.edu, linux-mm@kvack.org,
minchan.kim@gmail.com
On Fri, Jul 10, 2009 at 03:24:29PM +0800, Nai Xia wrote:
> Hi,
>
> I was able to launch some tests with SPEC cpu2006.
> The benchmark was based on mmotm
> commit 0b7292956dbdfb212abf6e3c9cfb41e9471e1081 on a intel Q6600 box with
> 4G ram. The kernel cmdline mem=500M was used to see how good exec-prot can
> be under memory stress.
Thank you for the testings, Nai!
> Following are the results:
>
> Estimated
> Base Base Base
> Benchmarks Ref. Run Time Ratio
>
> mmotm with 500M
> 400.perlbench 9770 671 14.6 *
> 401.bzip2 9650 1011 9.55 *
> 403.gcc 8050 774 10.4 *
> 462.libquantum 20720 1213 17.1 *
>
>
> mmot-prot with 500M
> 400.perlbench 9770 658 14.8 *
> 401.bzip2 9650 1007 9.58 *
> 403.gcc 8050 749 10.8 *
> 462.libquantum 20720 1116 18.6 *
>
> mmotm with 4G ( allowing the full working sets)
> 400.perlbench 9770 594 16.5 *
> 401.bzip2 9650 828 11.7 *
> 403.gcc 8050 523 15.4 *
> 462.libquantum 20720 1121 18.5 *
mmotm mmotm-prot mmotm-4G mmotm-prot mmotm-4G
14.6 14.8 16.5 +1.4% +13.0%
9.55 9.58 11.7 +0.3% +22.5%
10.4 10.8 15.4 +3.8% +48.1%
17.1 18.6 18.5 +8.8% +8.2%
So it's mostly small improvements.
> It's worth noting that SPEC documented "The CPU2006 benchmarks
> (code + workload) have been designed to fit within about 1GB of
> physical memory",
> and the exec vm sizes of these programs are as below:
> perlbench 956KB
> bzip2 56KB
> gcc 3008KB
> libquantum 36KB
>
>
> Are we expecting to see more good results for cpu-bound programs (e.g.
> scientific ones)
> with large number of exec pages ?
Not likely. Scientific computing is typically equipped with lots of
memory and the footprint of the program itself is relatively small.
The exec-mmap protection mainly helps when some exec pages/programs
have been inactive for some minutes and then go active. That's the
typically desktop use pattern.
Thanks,
Fengguang
> On Mon, Jun 8, 2009 at 5:10 PM, Wu Fengguang<fengguang.wu@intel.com> wrote:
> > Andrew,
> >
> > I managed to back this patchset with two test cases :)
> >
> > They demonstrated that
> > - X desktop responsiveness can be *doubled* under high memory/swap pressure
> > - it can almost stop major faults when the active file list is slowly scanned
> > A because of undergoing partially cache hot streaming IO
> >
> > The details are included in the changelog.
> >
> > Thanks,
> > Fengguang
> > --
> >
> > --
> > To unsubscribe, send a message with 'unsubscribe linux-mm' in
> > the body to majordomo@kvack.org. A For more info on Linux MM,
> > see: http://www.linux-mm.org/ .
> > Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
> >
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH 0/3] make mapped executable pages the first class citizen (with test cases)
2009-07-10 8:34 ` Wu Fengguang
@ 2009-07-10 16:50 ` Nai Xia
0 siblings, 0 replies; 10+ messages in thread
From: Nai Xia @ 2009-07-10 16:50 UTC (permalink / raw)
To: Wu Fengguang
Cc: Andrew Morton, KOSAKI Motohiro, Andi Kleen, Christoph Lameter,
Elladan, Nick Piggin, Johannes Weiner, Peter Zijlstra,
Rik van Riel, tytso@mit.edu, linux-mm@kvack.org,
minchan.kim@gmail.com
On Fri, Jul 10, 2009 at 4:34 PM, Wu Fengguang<fengguang.wu@intel.com> wrote:
> On Fri, Jul 10, 2009 at 03:24:29PM +0800, Nai Xia wrote:
>> Hi,
>>
>> I was able to launch some tests with SPEC cpu2006.
>> The benchmark was based on mmotm
>> commit 0b7292956dbdfb212abf6e3c9cfb41e9471e1081 on a intel Q6600 box with
>> 4G ram. The kernel cmdline mem=500M was used to see how good exec-prot can
>> be under memory stress.
>
> Thank you for the testings, Nai!
You are welcome :)
>
>> Following are the results:
>>
>> Estimated
>> Base Base Base
>> Benchmarks Ref. Run Time Ratio
>>
>> mmotm with 500M
>> 400.perlbench 9770 671 14.6 *
>> 401.bzip2 9650 1011 9.55 *
>> 403.gcc 8050 774 10.4 *
>> 462.libquantum 20720 1213 17.1 *
>>
>>
>> mmot-prot with 500M
>> 400.perlbench 9770 658 14.8 *
>> 401.bzip2 9650 1007 9.58 *
>> 403.gcc 8050 749 10.8 *
>> 462.libquantum 20720 1116 18.6 *
>>
>> mmotm with 4G ( allowing the full working sets)
>> 400.perlbench 9770 594 16.5 *
>> 401.bzip2 9650 828 11.7 *
>> 403.gcc 8050 523 15.4 *
>> 462.libquantum 20720 1121 18.5 *
>
> mmotm mmotm-prot mmotm-4G mmotm-prot mmotm-4G
> 14.6 14.8 16.5 +1.4% +13.0%
> 9.55 9.58 11.7 +0.3% +22.5%
> 10.4 10.8 15.4 +3.8% +48.1%
> 17.1 18.6 18.5 +8.8% +8.2%
>
> So it's mostly small improvements.
>
>> It's worth noting that SPEC documented "The CPU2006 benchmarks
>> (code + workload) have been designed to fit within about 1GB of
>> physical memory",
>> and the exec vm sizes of these programs are as below:
>> perlbench 956KB
>> bzip2 56KB
>> gcc 3008KB
>> libquantum 36KB
>>
>>
>> Are we expecting to see more good results for cpu-bound programs (e.g.
>> scientific ones)
>> with large number of exec pages ?
>
> Not likely. Scientific computing is typically equipped with lots of
> memory and the footprint of the program itself is relatively small.
OK, well, maybe as long as there is still swapping, improvement is
possible. Actually, in the above cases like bzip2, its exec footprint
is already quite small compared to the percentage of the improvement.
Let me see if I am lucky enough to have someone majoring in computing chemistry
in our Univ. give a benchmark. :) You know they have relatively small
machines doing
small personal computing jobs and sometimes swapping still matters.
>
> The exec-mmap protection mainly helps when some exec pages/programs
> have been inactive for some minutes and then go active. That's the
> typically desktop use pattern.
OK. Still it's good to see that this patch can improve more than 20% on average
on non-typical cases, hehe.
Regards,
Nai
>
> Thanks,
> Fengguang
>
>> On Mon, Jun 8, 2009 at 5:10 PM, Wu Fengguang<fengguang.wu@intel.com> wrote:
>> > Andrew,
>> >
>> > I managed to back this patchset with two test cases :)
>> >
>> > They demonstrated that
>> > - X desktop responsiveness can be *doubled* under high memory/swap pressure
>> > - it can almost stop major faults when the active file list is slowly scanned
>> > because of undergoing partially cache hot streaming IO
>> >
>> > The details are included in the changelog.
>> >
>> > Thanks,
>> > Fengguang
>> > --
>> >
>> > --
>> > To unsubscribe, send a message with 'unsubscribe linux-mm' in
>> > the body to majordomo@kvack.org. For more info on Linux MM,
>> > see: http://www.linux-mm.org/ .
>> > Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
>> >
>
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 10+ messages in thread