2019-05-27 08:55:01 +02:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2018-04-10 16:32:14 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
2013-02-27 17:03:15 -08:00
|
|
|
|
2017-02-08 18:51:30 +01:00
|
|
|
|
2017-02-06 10:57:33 +01:00
|
|
|
|
2024-06-07 14:23:56 +02:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
2007-02-14 00:34:12 -08:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2014-12-10 15:45:01 -08:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
fs/proc/internal.h: rearrange struct proc_dir_entry
struct proc_dir_entry became bit messy over years:
* move 16-bit ->mode_t before namelen to get rid of padding
* make ->in_use first field: it seems to be most used resulting in
smaller code on x86_64 (defconfig):
add/remove: 0/0 grow/shrink: 7/13 up/down: 24/-67 (-43)
Function old new delta
proc_readdir_de 451 455 +4
proc_get_inode 282 286 +4
pde_put 65 69 +4
remove_proc_subtree 294 297 +3
remove_proc_entry 297 300 +3
proc_register 295 298 +3
proc_notify_change 94 97 +3
unuse_pde 27 26 -1
proc_reg_write 89 85 -4
proc_reg_unlocked_ioctl 85 81 -4
proc_reg_read 89 85 -4
proc_reg_llseek 87 83 -4
proc_reg_get_unmapped_area 123 119 -4
proc_entry_rundown 139 135 -4
proc_reg_poll 91 85 -6
proc_reg_mmap 79 73 -6
proc_get_link 55 49 -6
proc_reg_release 108 101 -7
proc_reg_open 298 291 -7
close_pdeo 228 218 -10
* move writeable fields together to a first cacheline (on x86_64),
those include
* ->in_use: reference count, taken every open/read/write/close etc
* ->count: reference count, taken at readdir on every entry
* ->pde_openers: tracks (nearly) every open, dirtied
* ->pde_unload_lock: spinlock protecting ->pde_openers
* ->proc_iops, ->proc_fops, ->data: writeonce fields,
used right together with previous group.
* other rarely written fields go into 1st/2nd and 2nd/3rd cacheline on
32-bit and 64-bit respectively.
Additionally on 32-bit, ->subdir, ->subdir_node, ->namelen, ->name go
fully into 2nd cacheline, separated from writeable fields. They are all
used during lookup.
Link: http://lkml.kernel.org/r/20171220215914.GA7877@avx2
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-02-06 15:37:18 -08:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2018-04-10 16:32:14 -07:00
|
|
|
|
fs/proc/internal.h: rearrange struct proc_dir_entry
struct proc_dir_entry became bit messy over years:
* move 16-bit ->mode_t before namelen to get rid of padding
* make ->in_use first field: it seems to be most used resulting in
smaller code on x86_64 (defconfig):
add/remove: 0/0 grow/shrink: 7/13 up/down: 24/-67 (-43)
Function old new delta
proc_readdir_de 451 455 +4
proc_get_inode 282 286 +4
pde_put 65 69 +4
remove_proc_subtree 294 297 +3
remove_proc_entry 297 300 +3
proc_register 295 298 +3
proc_notify_change 94 97 +3
unuse_pde 27 26 -1
proc_reg_write 89 85 -4
proc_reg_unlocked_ioctl 85 81 -4
proc_reg_read 89 85 -4
proc_reg_llseek 87 83 -4
proc_reg_get_unmapped_area 123 119 -4
proc_entry_rundown 139 135 -4
proc_reg_poll 91 85 -6
proc_reg_mmap 79 73 -6
proc_get_link 55 49 -6
proc_reg_release 108 101 -7
proc_reg_open 298 291 -7
close_pdeo 228 218 -10
* move writeable fields together to a first cacheline (on x86_64),
those include
* ->in_use: reference count, taken every open/read/write/close etc
* ->count: reference count, taken at readdir on every entry
* ->pde_openers: tracks (nearly) every open, dirtied
* ->pde_unload_lock: spinlock protecting ->pde_openers
* ->proc_iops, ->proc_fops, ->data: writeonce fields,
used right together with previous group.
* other rarely written fields go into 1st/2nd and 2nd/3rd cacheline on
32-bit and 64-bit respectively.
Additionally on 32-bit, ->subdir, ->subdir_node, ->namelen, ->name go
fully into 2nd cacheline, separated from writeable fields. They are all
used during lookup.
Link: http://lkml.kernel.org/r/20171220215914.GA7877@avx2
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-02-06 15:37:18 -08:00
|
|
|
|
2018-02-06 15:37:21 -08:00
|
|
|
|
|
|
|
|
|
fs/proc/internal.h: rearrange struct proc_dir_entry
struct proc_dir_entry became bit messy over years:
* move 16-bit ->mode_t before namelen to get rid of padding
* make ->in_use first field: it seems to be most used resulting in
smaller code on x86_64 (defconfig):
add/remove: 0/0 grow/shrink: 7/13 up/down: 24/-67 (-43)
Function old new delta
proc_readdir_de 451 455 +4
proc_get_inode 282 286 +4
pde_put 65 69 +4
remove_proc_subtree 294 297 +3
remove_proc_entry 297 300 +3
proc_register 295 298 +3
proc_notify_change 94 97 +3
unuse_pde 27 26 -1
proc_reg_write 89 85 -4
proc_reg_unlocked_ioctl 85 81 -4
proc_reg_read 89 85 -4
proc_reg_llseek 87 83 -4
proc_reg_get_unmapped_area 123 119 -4
proc_entry_rundown 139 135 -4
proc_reg_poll 91 85 -6
proc_reg_mmap 79 73 -6
proc_get_link 55 49 -6
proc_reg_release 108 101 -7
proc_reg_open 298 291 -7
close_pdeo 228 218 -10
* move writeable fields together to a first cacheline (on x86_64),
those include
* ->in_use: reference count, taken every open/read/write/close etc
* ->count: reference count, taken at readdir on every entry
* ->pde_openers: tracks (nearly) every open, dirtied
* ->pde_unload_lock: spinlock protecting ->pde_openers
* ->proc_iops, ->proc_fops, ->data: writeonce fields,
used right together with previous group.
* other rarely written fields go into 1st/2nd and 2nd/3rd cacheline on
32-bit and 64-bit respectively.
Additionally on 32-bit, ->subdir, ->subdir_node, ->namelen, ->name go
fully into 2nd cacheline, separated from writeable fields. They are all
used during lookup.
Link: http://lkml.kernel.org/r/20171220215914.GA7877@avx2
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-02-06 15:37:18 -08:00
|
|
|
|
|
|
|
|
|
proc: decouple proc from VFS with "struct proc_ops"
Currently core /proc code uses "struct file_operations" for custom hooks,
however, VFS doesn't directly call them. Every time VFS expands
file_operations hook set, /proc code bloats for no reason.
Introduce "struct proc_ops" which contains only those hooks which /proc
allows to call into (open, release, read, write, ioctl, mmap, poll). It
doesn't contain module pointer as well.
Save ~184 bytes per usage:
add/remove: 26/26 grow/shrink: 1/4 up/down: 1922/-6674 (-4752)
Function old new delta
sysvipc_proc_ops - 72 +72
...
config_gz_proc_ops - 72 +72
proc_get_inode 289 339 +50
proc_reg_get_unmapped_area 110 107 -3
close_pdeo 227 224 -3
proc_reg_open 289 284 -5
proc_create_data 60 53 -7
rt_cpu_seq_fops 256 - -256
...
default_affinity_proc_fops 256 - -256
Total: Before=5430095, After=5425343, chg -0.09%
Link: http://lkml.kernel.org/r/20191225172228.GA13378@avx2
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-02-03 17:37:14 -08:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2019-02-01 14:20:01 -08:00
|
|
|
|
2018-05-15 15:57:23 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2018-05-18 11:46:15 +01:00
|
|
|
|
fs/proc/internal.h: rearrange struct proc_dir_entry
struct proc_dir_entry became bit messy over years:
* move 16-bit ->mode_t before namelen to get rid of padding
* make ->in_use first field: it seems to be most used resulting in
smaller code on x86_64 (defconfig):
add/remove: 0/0 grow/shrink: 7/13 up/down: 24/-67 (-43)
Function old new delta
proc_readdir_de 451 455 +4
proc_get_inode 282 286 +4
pde_put 65 69 +4
remove_proc_subtree 294 297 +3
remove_proc_entry 297 300 +3
proc_register 295 298 +3
proc_notify_change 94 97 +3
unuse_pde 27 26 -1
proc_reg_write 89 85 -4
proc_reg_unlocked_ioctl 85 81 -4
proc_reg_read 89 85 -4
proc_reg_llseek 87 83 -4
proc_reg_get_unmapped_area 123 119 -4
proc_entry_rundown 139 135 -4
proc_reg_poll 91 85 -6
proc_reg_mmap 79 73 -6
proc_get_link 55 49 -6
proc_reg_release 108 101 -7
proc_reg_open 298 291 -7
close_pdeo 228 218 -10
* move writeable fields together to a first cacheline (on x86_64),
those include
* ->in_use: reference count, taken every open/read/write/close etc
* ->count: reference count, taken at readdir on every entry
* ->pde_openers: tracks (nearly) every open, dirtied
* ->pde_unload_lock: spinlock protecting ->pde_openers
* ->proc_iops, ->proc_fops, ->data: writeonce fields,
used right together with previous group.
* other rarely written fields go into 1st/2nd and 2nd/3rd cacheline on
32-bit and 64-bit respectively.
Additionally on 32-bit, ->subdir, ->subdir_node, ->namelen, ->name go
fully into 2nd cacheline, separated from writeable fields. They are all
used during lookup.
Link: http://lkml.kernel.org/r/20171220215914.GA7877@avx2
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-02-06 15:37:18 -08:00
|
|
|
|
2018-04-24 17:05:17 +02:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2014-12-10 15:45:01 -08:00
|
|
|
|
2018-04-10 16:32:20 -07:00
|
|
|
|
2014-12-10 15:45:01 -08:00
|
|
|
|
2018-04-10 16:31:52 -07:00
|
|
|
|
fs/proc/internal.h: rearrange struct proc_dir_entry
struct proc_dir_entry became bit messy over years:
* move 16-bit ->mode_t before namelen to get rid of padding
* make ->in_use first field: it seems to be most used resulting in
smaller code on x86_64 (defconfig):
add/remove: 0/0 grow/shrink: 7/13 up/down: 24/-67 (-43)
Function old new delta
proc_readdir_de 451 455 +4
proc_get_inode 282 286 +4
pde_put 65 69 +4
remove_proc_subtree 294 297 +3
remove_proc_entry 297 300 +3
proc_register 295 298 +3
proc_notify_change 94 97 +3
unuse_pde 27 26 -1
proc_reg_write 89 85 -4
proc_reg_unlocked_ioctl 85 81 -4
proc_reg_read 89 85 -4
proc_reg_llseek 87 83 -4
proc_reg_get_unmapped_area 123 119 -4
proc_entry_rundown 139 135 -4
proc_reg_poll 91 85 -6
proc_reg_mmap 79 73 -6
proc_get_link 55 49 -6
proc_reg_release 108 101 -7
proc_reg_open 298 291 -7
close_pdeo 228 218 -10
* move writeable fields together to a first cacheline (on x86_64),
those include
* ->in_use: reference count, taken every open/read/write/close etc
* ->count: reference count, taken at readdir on every entry
* ->pde_openers: tracks (nearly) every open, dirtied
* ->pde_unload_lock: spinlock protecting ->pde_openers
* ->proc_iops, ->proc_fops, ->data: writeonce fields,
used right together with previous group.
* other rarely written fields go into 1st/2nd and 2nd/3rd cacheline on
32-bit and 64-bit respectively.
Additionally on 32-bit, ->subdir, ->subdir_node, ->namelen, ->name go
fully into 2nd cacheline, separated from writeable fields. They are all
used during lookup.
Link: http://lkml.kernel.org/r/20171220215914.GA7877@avx2
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-02-06 15:37:18 -08:00
|
|
|
|
proc: faster open/read/close with "permanent" files
Now that "struct proc_ops" exist we can start putting there stuff which
could not fly with VFS "struct file_operations"...
Most of fs/proc/inode.c file is dedicated to make open/read/.../close
reliable in the event of disappearing /proc entries which usually happens
if module is getting removed. Files like /proc/cpuinfo which never
disappear simply do not need such protection.
Save 2 atomic ops, 1 allocation, 1 free per open/read/close sequence for such
"permanent" files.
Enable "permanent" flag for
/proc/cpuinfo
/proc/kmsg
/proc/modules
/proc/slabinfo
/proc/stat
/proc/sysvipc/*
/proc/swaps
More will come once I figure out foolproof way to prevent out module
authors from marking their stuff "permanent" for performance reasons
when it is not.
This should help with scalability: benchmark is "read /proc/cpuinfo R times
by N threads scattered over the system".
N R t, s (before) t, s (after)
-----------------------------------------------------
64 4096 1.582458 1.530502 -3.2%
256 4096 6.371926 6.125168 -3.9%
1024 4096 25.64888 24.47528 -4.6%
Benchmark source:
#include <chrono>
#include <iostream>
#include <thread>
#include <vector>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
const int NR_CPUS = sysconf(_SC_NPROCESSORS_ONLN);
int N;
const char *filename;
int R;
int xxx = 0;
int glue(int n)
{
cpu_set_t m;
CPU_ZERO(&m);
CPU_SET(n, &m);
return sched_setaffinity(0, sizeof(cpu_set_t), &m);
}
void f(int n)
{
glue(n % NR_CPUS);
while (*(volatile int *)&xxx == 0) {
}
for (int i = 0; i < R; i++) {
int fd = open(filename, O_RDONLY);
char buf[4096];
ssize_t rv = read(fd, buf, sizeof(buf));
asm volatile ("" :: "g" (rv));
close(fd);
}
}
int main(int argc, char *argv[])
{
if (argc < 4) {
std::cerr << "usage: " << argv[0] << ' ' << "N /proc/filename R
";
return 1;
}
N = atoi(argv[1]);
filename = argv[2];
R = atoi(argv[3]);
for (int i = 0; i < NR_CPUS; i++) {
if (glue(i) == 0)
break;
}
std::vector<std::thread> T;
T.reserve(N);
for (int i = 0; i < N; i++) {
T.emplace_back(f, i);
}
auto t0 = std::chrono::system_clock::now();
{
*(volatile int *)&xxx = 1;
for (auto& t: T) {
t.join();
}
}
auto t1 = std::chrono::system_clock::now();
std::chrono::duration<double> dt = t1 - t0;
std::cout << dt.count() << '
';
return 0;
}
P.S.:
Explicit randomization marker is added because adding non-function pointer
will silently disable structure layout randomization.
[akpm@linux-foundation.org: coding style fixes]
Reported-by: kbuild test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Joe Perches <joe@perches.com>
Link: http://lkml.kernel.org/r/20200222201539.GA22576@avx2
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-06 20:09:01 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2018-06-13 19:43:19 +01:00
|
|
|
|
2016-10-28 01:22:25 -07:00
|
|
|
|
2011-05-24 17:12:48 -07:00
|
|
|
|
2018-08-21 21:54:09 -07:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2018-06-13 19:43:19 +01:00
|
|
|
|
proc: faster open/read/close with "permanent" files
Now that "struct proc_ops" exist we can start putting there stuff which
could not fly with VFS "struct file_operations"...
Most of fs/proc/inode.c file is dedicated to make open/read/.../close
reliable in the event of disappearing /proc entries which usually happens
if module is getting removed. Files like /proc/cpuinfo which never
disappear simply do not need such protection.
Save 2 atomic ops, 1 allocation, 1 free per open/read/close sequence for such
"permanent" files.
Enable "permanent" flag for
/proc/cpuinfo
/proc/kmsg
/proc/modules
/proc/slabinfo
/proc/stat
/proc/sysvipc/*
/proc/swaps
More will come once I figure out foolproof way to prevent out module
authors from marking their stuff "permanent" for performance reasons
when it is not.
This should help with scalability: benchmark is "read /proc/cpuinfo R times
by N threads scattered over the system".
N R t, s (before) t, s (after)
-----------------------------------------------------
64 4096 1.582458 1.530502 -3.2%
256 4096 6.371926 6.125168 -3.9%
1024 4096 25.64888 24.47528 -4.6%
Benchmark source:
#include <chrono>
#include <iostream>
#include <thread>
#include <vector>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
const int NR_CPUS = sysconf(_SC_NPROCESSORS_ONLN);
int N;
const char *filename;
int R;
int xxx = 0;
int glue(int n)
{
cpu_set_t m;
CPU_ZERO(&m);
CPU_SET(n, &m);
return sched_setaffinity(0, sizeof(cpu_set_t), &m);
}
void f(int n)
{
glue(n % NR_CPUS);
while (*(volatile int *)&xxx == 0) {
}
for (int i = 0; i < R; i++) {
int fd = open(filename, O_RDONLY);
char buf[4096];
ssize_t rv = read(fd, buf, sizeof(buf));
asm volatile ("" :: "g" (rv));
close(fd);
}
}
int main(int argc, char *argv[])
{
if (argc < 4) {
std::cerr << "usage: " << argv[0] << ' ' << "N /proc/filename R
";
return 1;
}
N = atoi(argv[1]);
filename = argv[2];
R = atoi(argv[3]);
for (int i = 0; i < NR_CPUS; i++) {
if (glue(i) == 0)
break;
}
std::vector<std::thread> T;
T.reserve(N);
for (int i = 0; i < N; i++) {
T.emplace_back(f, i);
}
auto t0 = std::chrono::system_clock::now();
{
*(volatile int *)&xxx = 1;
for (auto& t: T) {
t.join();
}
}
auto t1 = std::chrono::system_clock::now();
std::chrono::duration<double> dt = t1 - t0;
std::cout << dt.count() << '
';
return 0;
}
P.S.:
Explicit randomization marker is added because adding non-function pointer
will silently disable structure layout randomization.
[akpm@linux-foundation.org: coding style fixes]
Reported-by: kbuild test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Joe Perches <joe@perches.com>
Link: http://lkml.kernel.org/r/20200222201539.GA22576@avx2
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-06 20:09:01 -07:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2022-09-20 20:35:23 +03:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2018-04-10 16:31:52 -07:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2023-09-12 13:56:48 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2011-05-24 17:12:48 -07:00
|
|
|
|
2016-09-02 00:42:02 +03:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2020-02-19 17:17:34 -06:00
|
|
|
|
2014-11-01 11:10:28 -04:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2016-10-28 01:22:25 -07:00
|
|
|
|
2011-05-24 17:12:48 -07:00
|
|
|
|
2013-04-12 18:03:36 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2018-08-21 21:54:37 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2006-06-26 00:25:56 -07:00
|
|
|
|
2006-06-26 00:25:55 -07:00
|
|
|
|
|
|
|
|
|
2018-08-21 21:54:37 -07:00
|
|
|
|
2006-06-26 00:25:55 -07:00
|
|
|
|
2006-06-26 00:25:56 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
|
|
2017-09-30 14:45:42 -04:00
|
|
|
|
2017-01-03 10:23:11 +13:00
|
|
|
|
2012-08-23 14:43:24 +04:00
|
|
|
|
2017-11-17 15:26:49 -08:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2008-07-25 01:48:29 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
2009-04-07 13:19:18 -04:00
|
|
|
|
2024-06-07 14:23:56 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2024-08-21 18:39:10 +01:00
|
|
|
|
2024-06-07 14:23:56 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2018-05-18 08:47:13 -07:00
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2013-04-12 01:08:50 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2023-01-13 12:49:12 +01:00
|
|
|
|
2021-01-21 14:19:43 +01:00
|
|
|
|
2023-01-13 12:49:11 +01:00
|
|
|
|
2021-01-21 14:19:43 +01:00
|
|
|
|
2020-02-19 18:22:26 -06:00
|
|
|
|
2016-11-10 22:18:28 +01:00
|
|
|
|
2018-05-02 21:26:16 -04:00
|
|
|
|
2013-04-12 01:08:50 +01:00
|
|
|
|
2013-05-16 12:07:31 -04:00
|
|
|
|
2019-03-05 15:50:29 -08:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2013-04-12 01:08:50 +01:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2018-05-03 09:21:05 -04:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2018-06-07 17:10:10 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2008-07-25 01:48:29 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2018-04-24 17:08:36 +02:00
|
|
|
|
|
|
|
|
|
2018-04-24 17:00:52 +02:00
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2018-02-06 15:37:31 -08:00
|
|
|
|
2013-05-16 12:07:31 -04:00
|
|
|
|
2018-02-06 15:37:31 -08:00
|
|
|
|
2009-04-07 13:19:18 -04:00
|
|
|
|
2020-12-15 20:42:42 -08:00
|
|
|
|
2009-12-15 16:45:39 -08:00
|
|
|
|
2018-04-10 16:32:14 -07:00
|
|
|
|
2009-12-15 16:45:39 -08:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2009-04-07 13:19:18 -04:00
|
|
|
|
2015-05-11 16:44:25 -05:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2018-05-18 11:46:15 +01:00
|
|
|
|
2015-05-11 16:44:25 -05:00
|
|
|
|
2009-04-07 13:19:18 -04:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2009-04-07 13:19:18 -04:00
|
|
|
|
2008-07-25 01:48:29 -07:00
|
|
|
|
|
|
|
|
|
2019-12-04 16:50:05 -08:00
|
|
|
|
2016-12-12 16:45:17 -08:00
|
|
|
|
2013-04-04 16:28:47 -04:00
|
|
|
|
2018-04-10 16:31:05 -07:00
|
|
|
|
2015-02-21 22:16:11 -05:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2018-11-01 23:07:25 +00:00
|
|
|
|
2010-03-07 16:41:34 -08:00
|
|
|
|
2018-04-10 16:31:09 -07:00
|
|
|
|
2020-02-21 08:43:23 -06:00
|
|
|
|
2016-12-12 16:45:32 -08:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
2010-03-07 16:41:34 -08:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2010-03-07 16:41:34 -08:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2014-07-31 03:10:50 -07:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2017-02-10 10:35:02 +03:00
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
2017-02-10 10:35:02 +03:00
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2013-04-12 02:29:19 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
mm: add /proc/pid/smaps_rollup
/proc/pid/smaps_rollup is a new proc file that improves the performance
of user programs that determine aggregate memory statistics (e.g., total
PSS) of a process.
Android regularly "samples" the memory usage of various processes in
order to balance its memory pool sizes. This sampling process involves
opening /proc/pid/smaps and summing certain fields. For very large
processes, sampling memory use this way can take several hundred
milliseconds, due mostly to the overhead of the seq_printf calls in
task_mmu.c.
smaps_rollup improves the situation. It contains most of the fields of
/proc/pid/smaps, but instead of a set of fields for each VMA,
smaps_rollup instead contains one synthetic smaps-format entry
representing the whole process. In the single smaps_rollup synthetic
entry, each field is the summation of the corresponding field in all of
the real-smaps VMAs. Using a common format for smaps_rollup and smaps
allows userspace parsers to repurpose parsers meant for use with
non-rollup smaps for smaps_rollup, and it allows userspace to switch
between smaps_rollup and smaps at runtime (say, based on the
availability of smaps_rollup in a given kernel) with minimal fuss.
By using smaps_rollup instead of smaps, a caller can avoid the
significant overhead of formatting, reading, and parsing each of a large
process's potentially very numerous memory mappings. For sampling
system_server's PSS in Android, we measured a 12x speedup, representing
a savings of several hundred milliseconds.
One alternative to a new per-process proc file would have been including
PSS information in /proc/pid/status. We considered this option but
thought that PSS would be too expensive (by a few orders of magnitude)
to collect relative to what's already emitted as part of
/proc/pid/status, and slowing every user of /proc/pid/status for the
sake of readers that happen to want PSS feels wrong.
The code itself works by reusing the existing VMA-walking framework we
use for regular smaps generation and keeping the mem_size_stats
structure around between VMA walks instead of using a fresh one for each
VMA. In this way, summation happens automatically. We let seq_file
walk over the VMAs just as it does for regular smaps and just emit
nothing to the seq_file until we hit the last VMA.
Benchmarks:
using smaps:
iterations:1000 pid:1163 pss:220023808
0m29.46s real 0m08.28s user 0m20.98s system
using smaps_rollup:
iterations:1000 pid:1163 pss:220702720
0m04.39s real 0m00.03s user 0m04.31s system
We're using the PSS samples we collect asynchronously for
system-management tasks like fine-tuning oom_adj_score, memory use
tracking for debugging, application-level memory-use attribution, and
deciding whether we want to kill large processes during system idle
maintenance windows. Android has been using PSS for these purposes for
a long time; as the average process VMA count has increased and and
devices become more efficiency-conscious, PSS-collection inefficiency
has started to matter more. IMHO, it'd be a lot safer to optimize the
existing PSS-collection model, which has been fine-tuned over the years,
instead of changing the memory tracking approach entirely to work around
smaps-generation inefficiency.
Tim said:
: There are two main reasons why Android gathers PSS information:
:
: 1. Android devices can show the user the amount of memory used per
: application via the settings app. This is a less important use case.
:
: 2. We log PSS to help identify leaks in applications. We have found
: an enormous number of bugs (in the Android platform, in Google's own
: apps, and in third-party applications) using this data.
:
: To do this, system_server (the main process in Android userspace) will
: sample the PSS of a process three seconds after it changes state (for
: example, app is launched and becomes the foreground application) and about
: every ten minutes after that. The net result is that PSS collection is
: regularly running on at least one process in the system (usually a few
: times a minute while the screen is on, less when screen is off due to
: suspend). PSS of a process is an incredibly useful stat to track, and we
: aren't going to get rid of it. We've looked at some very hacky approaches
: using RSS ("take the RSS of the target process, subtract the RSS of the
: zygote process that is the parent of all Android apps") to reduce the
: accounting time, but it regularly overestimated the memory used by 20+
: percent. Accordingly, I don't think that there's a good alternative to
: using PSS.
:
: We started looking into PSS collection performance after we noticed random
: frequency spikes while a phone's screen was off; occasionally, one of the
: CPU clusters would ramp to a high frequency because there was 200-300ms of
: constant CPU work from a single thread in the main Android userspace
: process. The work causing the spike (which is reasonable governor
: behavior given the amount of CPU time needed) was always PSS collection.
: As a result, Android is burning more power than we should be on PSS
: collection.
:
: The other issue (and why I'm less sure about improving smaps as a
: long-term solution) is that the number of VMAs per process has increased
: significantly from release to release. After trying to figure out why we
: were seeing these 200-300ms PSS collection times on Android O but had not
: noticed it in previous versions, we found that the number of VMAs in the
: main system process increased by 50% from Android N to Android O (from
: ~1800 to ~2700) and varying increases in every userspace process. Android
: M to N also had an increase in the number of VMAs, although not as much.
: I'm not sure why this is increasing so much over time, but thinking about
: ASLR and ways to make ASLR better, I expect that this will continue to
: increase going forward. I would not be surprised if we hit 5000 VMAs on
: the main Android process (system_server) by 2020.
:
: If we assume that the number of VMAs is going to increase over time, then
: doing anything we can do to reduce the overhead of each VMA during PSS
: collection seems like the right way to go, and that means outputting an
: aggregate statistic (to avoid whatever overhead there is per line in
: writing smaps and in reading each line from userspace).
Link: http://lkml.kernel.org/r/20170812022148.178293-1-dancol@google.com
Signed-off-by: Daniel Colascione <dancol@google.com>
Cc: Tim Murray <timmurray@google.com>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Randy Dunlap <rdunlap@infradead.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Sonny Rao <sonnyrao@chromium.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-09-06 16:25:08 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2014-10-09 15:25:51 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2014-10-09 15:25:26 -07:00
|
|
|
|
2022-09-06 19:48:57 +00:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2016-10-28 01:22:25 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
2014-10-09 15:25:24 -07:00
|
|
|
|
|
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
mm: add /proc/pid/smaps_rollup
/proc/pid/smaps_rollup is a new proc file that improves the performance
of user programs that determine aggregate memory statistics (e.g., total
PSS) of a process.
Android regularly "samples" the memory usage of various processes in
order to balance its memory pool sizes. This sampling process involves
opening /proc/pid/smaps and summing certain fields. For very large
processes, sampling memory use this way can take several hundred
milliseconds, due mostly to the overhead of the seq_printf calls in
task_mmu.c.
smaps_rollup improves the situation. It contains most of the fields of
/proc/pid/smaps, but instead of a set of fields for each VMA,
smaps_rollup instead contains one synthetic smaps-format entry
representing the whole process. In the single smaps_rollup synthetic
entry, each field is the summation of the corresponding field in all of
the real-smaps VMAs. Using a common format for smaps_rollup and smaps
allows userspace parsers to repurpose parsers meant for use with
non-rollup smaps for smaps_rollup, and it allows userspace to switch
between smaps_rollup and smaps at runtime (say, based on the
availability of smaps_rollup in a given kernel) with minimal fuss.
By using smaps_rollup instead of smaps, a caller can avoid the
significant overhead of formatting, reading, and parsing each of a large
process's potentially very numerous memory mappings. For sampling
system_server's PSS in Android, we measured a 12x speedup, representing
a savings of several hundred milliseconds.
One alternative to a new per-process proc file would have been including
PSS information in /proc/pid/status. We considered this option but
thought that PSS would be too expensive (by a few orders of magnitude)
to collect relative to what's already emitted as part of
/proc/pid/status, and slowing every user of /proc/pid/status for the
sake of readers that happen to want PSS feels wrong.
The code itself works by reusing the existing VMA-walking framework we
use for regular smaps generation and keeping the mem_size_stats
structure around between VMA walks instead of using a fresh one for each
VMA. In this way, summation happens automatically. We let seq_file
walk over the VMAs just as it does for regular smaps and just emit
nothing to the seq_file until we hit the last VMA.
Benchmarks:
using smaps:
iterations:1000 pid:1163 pss:220023808
0m29.46s real 0m08.28s user 0m20.98s system
using smaps_rollup:
iterations:1000 pid:1163 pss:220702720
0m04.39s real 0m00.03s user 0m04.31s system
We're using the PSS samples we collect asynchronously for
system-management tasks like fine-tuning oom_adj_score, memory use
tracking for debugging, application-level memory-use attribution, and
deciding whether we want to kill large processes during system idle
maintenance windows. Android has been using PSS for these purposes for
a long time; as the average process VMA count has increased and and
devices become more efficiency-conscious, PSS-collection inefficiency
has started to matter more. IMHO, it'd be a lot safer to optimize the
existing PSS-collection model, which has been fine-tuned over the years,
instead of changing the memory tracking approach entirely to work around
smaps-generation inefficiency.
Tim said:
: There are two main reasons why Android gathers PSS information:
:
: 1. Android devices can show the user the amount of memory used per
: application via the settings app. This is a less important use case.
:
: 2. We log PSS to help identify leaks in applications. We have found
: an enormous number of bugs (in the Android platform, in Google's own
: apps, and in third-party applications) using this data.
:
: To do this, system_server (the main process in Android userspace) will
: sample the PSS of a process three seconds after it changes state (for
: example, app is launched and becomes the foreground application) and about
: every ten minutes after that. The net result is that PSS collection is
: regularly running on at least one process in the system (usually a few
: times a minute while the screen is on, less when screen is off due to
: suspend). PSS of a process is an incredibly useful stat to track, and we
: aren't going to get rid of it. We've looked at some very hacky approaches
: using RSS ("take the RSS of the target process, subtract the RSS of the
: zygote process that is the parent of all Android apps") to reduce the
: accounting time, but it regularly overestimated the memory used by 20+
: percent. Accordingly, I don't think that there's a good alternative to
: using PSS.
:
: We started looking into PSS collection performance after we noticed random
: frequency spikes while a phone's screen was off; occasionally, one of the
: CPU clusters would ramp to a high frequency because there was 200-300ms of
: constant CPU work from a single thread in the main Android userspace
: process. The work causing the spike (which is reasonable governor
: behavior given the amount of CPU time needed) was always PSS collection.
: As a result, Android is burning more power than we should be on PSS
: collection.
:
: The other issue (and why I'm less sure about improving smaps as a
: long-term solution) is that the number of VMAs per process has increased
: significantly from release to release. After trying to figure out why we
: were seeing these 200-300ms PSS collection times on Android O but had not
: noticed it in previous versions, we found that the number of VMAs in the
: main system process increased by 50% from Android N to Android O (from
: ~1800 to ~2700) and varying increases in every userspace process. Android
: M to N also had an increase in the number of VMAs, although not as much.
: I'm not sure why this is increasing so much over time, but thinking about
: ASLR and ways to make ASLR better, I expect that this will continue to
: increase going forward. I would not be surprised if we hit 5000 VMAs on
: the main Android process (system_server) by 2020.
:
: If we assume that the number of VMAs is going to increase over time, then
: doing anything we can do to reduce the overhead of each VMA during PSS
: collection seems like the right way to go, and that means outputting an
: aggregate statistic (to avoid whatever overhead there is per line in
: writing smaps and in reading each line from userspace).
Link: http://lkml.kernel.org/r/20170812022148.178293-1-dancol@google.com
Signed-off-by: Daniel Colascione <dancol@google.com>
Cc: Tim Murray <timmurray@google.com>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Randy Dunlap <rdunlap@infradead.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Sonny Rao <sonnyrao@chromium.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-09-06 16:25:08 -07:00
|
|
|
|
2013-04-11 13:34:43 +01:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2020-12-15 20:42:39 -08:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2024-08-06 18:02:29 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
