mirror of https://github.com/mkerrisk/man-pages
proc.5: Document inaccurate RSS due to SPLIT_RSS_COUNTING
[mtk: Manually applied patch, because of conflicts with other merged changes; also added an edit suggested by Jann; see the thread at https://lore.kernel.org/linux-man/20201012114940.1317510-1-jannh@google.com/] Since 34e55232e59f7b19050267a05ff1226e5cd122a5 (introduced back in v2.6.34), Linux uses per-thread RSS counters to reduce cache contention on the per-mm counters. With a 4K page size, that means that you can end up with the counters off by up to 252KiB per thread. Example: $ cat rsstest.c #include <stdlib.h> #include <err.h> #include <stdio.h> #include <signal.h> #include <unistd.h> #include <sys/mman.h> #include <sys/eventfd.h> #include <sys/prctl.h> void dump(int pid) { char cmd[1000]; sprintf(cmd, "grep '^VmRSS' /proc/%d/status;" "grep '^Rss:' /proc/%d/smaps_rollup;" "echo", pid, pid ); system(cmd); } int main(void) { eventfd_t dummy; int child_wait = eventfd(0, EFD_SEMAPHORE|EFD_CLOEXEC); int child_resume = eventfd(0, EFD_SEMAPHORE|EFD_CLOEXEC); if (child_wait == -1 || child_resume == -1) err(1, "eventfd"); pid_t child = fork(); if (child == -1) err(1, "fork"); if (child == 0) { if (prctl(PR_SET_PDEATHSIG, SIGKILL)) err(1, "PDEATHSIG"); if (getppid() == 1) exit(0); char *mapping = mmap(NULL, 80 * 0x1000, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0); eventfd_write(child_wait, 1); eventfd_read(child_resume, &dummy); for (int i=0; i<40; i++) mapping[0x1000 * i] = 1; eventfd_write(child_wait, 1); eventfd_read(child_resume, &dummy); for (int i=40; i<80; i++) mapping[0x1000 * i] = 1; eventfd_write(child_wait, 1); eventfd_read(child_resume, &dummy); exit(0); } eventfd_read(child_wait, &dummy); dump(child); eventfd_write(child_resume, 1); eventfd_read(child_wait, &dummy); dump(child); eventfd_write(child_resume, 1); eventfd_read(child_wait, &dummy); dump(child); eventfd_write(child_resume, 1); exit(0); } $ gcc -o rsstest rsstest.c && ./rsstest VmRSS: 68 kB Rss: 616 kB VmRSS: 68 kB Rss: 776 kB VmRSS: 812 kB Rss: 936 kB $ Let's document that those counters aren't entirely accurate. Reported-by: Mark Mossberg <mark.mossberg@gmail.com> Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>
This commit is contained in:
parent
14948ad6ec
commit
20e43cd694
36
man5/proc.5
36
man5/proc.5
|
@ -2265,6 +2265,9 @@ This is just the pages which
|
||||||
count toward text, data, or stack space.
|
count toward text, data, or stack space.
|
||||||
This does not include pages
|
This does not include pages
|
||||||
which have not been demand-loaded in, or which are swapped out.
|
which have not been demand-loaded in, or which are swapped out.
|
||||||
|
This value is inaccurate; see
|
||||||
|
.I /proc/[pid]/statm
|
||||||
|
below.
|
||||||
.TP
|
.TP
|
||||||
(25) \fIrsslim\fP \ %lu
|
(25) \fIrsslim\fP \ %lu
|
||||||
Current soft limit in bytes on the rss of the process;
|
Current soft limit in bytes on the rss of the process;
|
||||||
|
@ -2409,10 +2412,11 @@ The columns are:
|
||||||
size (1) total program size
|
size (1) total program size
|
||||||
(same as VmSize in \fI/proc/[pid]/status\fP)
|
(same as VmSize in \fI/proc/[pid]/status\fP)
|
||||||
resident (2) resident set size
|
resident (2) resident set size
|
||||||
(same as VmRSS in \fI/proc/[pid]/status\fP)
|
(inaccurate; same as VmRSS in \fI/proc/[pid]/status\fP)
|
||||||
shared (3) number of resident shared pages
|
shared (3) number of resident shared pages
|
||||||
(i.e., backed by a file)
|
(i.e., backed by a file)
|
||||||
(same as RssFile+RssShmem in \fI/proc/[pid]/status\fP)
|
(inaccurate; same as RssFile+RssShmem in
|
||||||
|
\fI/proc/[pid]/status\fP)
|
||||||
text (4) text (code)
|
text (4) text (code)
|
||||||
.\" (not including libs; broken, includes data segment)
|
.\" (not including libs; broken, includes data segment)
|
||||||
lib (5) library (unused since Linux 2.6; always 0)
|
lib (5) library (unused since Linux 2.6; always 0)
|
||||||
|
@ -2421,6 +2425,16 @@ data (6) data + stack
|
||||||
dt (7) dirty pages (unused since Linux 2.6; always 0)
|
dt (7) dirty pages (unused since Linux 2.6; always 0)
|
||||||
.EE
|
.EE
|
||||||
.in
|
.in
|
||||||
|
.IP
|
||||||
|
.\" See SPLIT_RSS_COUNTING in the kernel.
|
||||||
|
.\" Inaccuracy is bounded by TASK_RSS_EVENTS_THRESH.
|
||||||
|
Some of these values are inaccurate because
|
||||||
|
of a kernel-internal scalability optimization.
|
||||||
|
If accurate values are required, use
|
||||||
|
.I /proc/[pid]/smaps
|
||||||
|
or
|
||||||
|
.I /proc/[pid]/smaps_rollup
|
||||||
|
instead, which are much slower but provide accurate, detailed information.
|
||||||
.TP
|
.TP
|
||||||
.I /proc/[pid]/status
|
.I /proc/[pid]/status
|
||||||
Provides much of the information in
|
Provides much of the information in
|
||||||
|
@ -2597,6 +2611,9 @@ directly access physical memory.
|
||||||
.TP
|
.TP
|
||||||
.IR VmHWM
|
.IR VmHWM
|
||||||
Peak resident set size ("high water mark").
|
Peak resident set size ("high water mark").
|
||||||
|
This value is inaccurate; see
|
||||||
|
.I /proc/[pid]/statm
|
||||||
|
above.
|
||||||
.TP
|
.TP
|
||||||
.IR VmRSS
|
.IR VmRSS
|
||||||
Resident set size.
|
Resident set size.
|
||||||
|
@ -2605,16 +2622,25 @@ Note that the value here is the sum of
|
||||||
.IR RssFile ,
|
.IR RssFile ,
|
||||||
and
|
and
|
||||||
.IR RssShmem .
|
.IR RssShmem .
|
||||||
|
This value is inaccurate; see
|
||||||
|
.I /proc/[pid]/statm
|
||||||
|
above.
|
||||||
.TP
|
.TP
|
||||||
.IR RssAnon
|
.IR RssAnon
|
||||||
Size of resident anonymous memory.
|
Size of resident anonymous memory.
|
||||||
.\" commit bf9683d6990589390b5178dafe8fd06808869293
|
.\" commit bf9683d6990589390b5178dafe8fd06808869293
|
||||||
(since Linux 4.5).
|
(since Linux 4.5).
|
||||||
|
This value is inaccurate; see
|
||||||
|
.I /proc/[pid]/statm
|
||||||
|
above.
|
||||||
.TP
|
.TP
|
||||||
.IR RssFile
|
.IR RssFile
|
||||||
Size of resident file mappings.
|
Size of resident file mappings.
|
||||||
.\" commit bf9683d6990589390b5178dafe8fd06808869293
|
.\" commit bf9683d6990589390b5178dafe8fd06808869293
|
||||||
(since Linux 4.5).
|
(since Linux 4.5).
|
||||||
|
This value is inaccurate; see
|
||||||
|
.I /proc/[pid]/statm
|
||||||
|
above.
|
||||||
.TP
|
.TP
|
||||||
.IR RssShmem
|
.IR RssShmem
|
||||||
Size of resident shared memory (includes System V shared memory,
|
Size of resident shared memory (includes System V shared memory,
|
||||||
|
@ -2626,6 +2652,9 @@ and shared anonymous mappings).
|
||||||
.TP
|
.TP
|
||||||
.IR VmData ", " VmStk ", " VmExe
|
.IR VmData ", " VmStk ", " VmExe
|
||||||
Size of data, stack, and text segments.
|
Size of data, stack, and text segments.
|
||||||
|
This value is inaccurate; see
|
||||||
|
.I /proc/[pid]/statm
|
||||||
|
above.
|
||||||
.TP
|
.TP
|
||||||
.IR VmLib
|
.IR VmLib
|
||||||
Shared library code size.
|
Shared library code size.
|
||||||
|
@ -2641,6 +2670,9 @@ Size of second-level page tables (added in Linux 4.0; removed in Linux 4.15).
|
||||||
.\" commit b084d4353ff99d824d3bc5a5c2c22c70b1fba722
|
.\" commit b084d4353ff99d824d3bc5a5c2c22c70b1fba722
|
||||||
Swapped-out virtual memory size by anonymous private pages;
|
Swapped-out virtual memory size by anonymous private pages;
|
||||||
shmem swap usage is not included (since Linux 2.6.34).
|
shmem swap usage is not included (since Linux 2.6.34).
|
||||||
|
This value is inaccurate; see
|
||||||
|
.I /proc/[pid]/statm
|
||||||
|
above.
|
||||||
.TP
|
.TP
|
||||||
.IR HugetlbPages
|
.IR HugetlbPages
|
||||||
Size of hugetlb memory portions
|
Size of hugetlb memory portions
|
||||||
|
|
Loading…
Reference in New Issue