[PATCH 0/1] posix timers: Expand exitsting info in proc file

Hi.

I'm working on the checkpoint-restore project (http://criu.org), on
realisation of posix-timers. To compleatly checkpoint and restore these
timers we need to know which clock they are using. So I d'like to add
this information to existing syscall which shows posix-timers info.

Pavel Tikhomirov (1):
  posix-timers: Show clock ID in proc file

 fs/proc/base.c |    1 +
 1 file changed, 1 insertion(+)

-- 
1.7.9.5

Expand information about posix-timers in /proc/<pid>/timers by adding
info about clock, with which the timer was created. I.e. in the forth
line of timer info after "notify:" line go "ClockID: <clock_id>".

Signed-off-by: Pavel Tikhomirov <redacted>
---
 fs/proc/base.c |    1 +
 1 file changed, 1 insertion(+)

diff --git a/fs/proc/base.c b/fs/proc/base.c
index 2dad4a9..8a38eef 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c

@@ -2079,6 +2079,7 @@ static int show_timer(struct seq_file *m, void *v)
 		nstr[notify & ~SIGEV_THREAD_ID],
 		(notify & SIGEV_THREAD_ID) ? "tid" : "pid",
 		pid_nr_ns(timer->it_pid, tp->ns));
+	seq_printf(m, "ClockID: %d\n", timer->it_clock);
 
 	return 0;
 }

-- 
1.7.9.5

On Fri, May 17, 2013 at 02:12:03AM +0400, Pavel Tikhomirov wrote:

What userspace tool just broke by adding a new field to this file?

thanks,

greg k-h

On 05/17/2013 02:12 AM, Pavel Tikhomirov wrote:

Expand information about posix-timers in /proc/<pid>/timers by adding
info about clock, with which the timer was created. I.e. in the forth
line of timer info after "notify:" line go "ClockID: <clock_id>".

Signed-off-by: Pavel Tikhomirov <redacted>

Acked-by: Pavel Emelyanov <redacted>

On 05/17/2013 07:59 PM, Greg KH wrote:

On Fri, May 17, 2013 at 02:12:03AM +0400, Pavel Tikhomirov wrote:

quoted

Expand information about posix-timers in /proc/<pid>/timers by adding
info about clock, with which the timer was created. I.e. in the forth
line of timer info after "notify:" line go "ClockID: <clock_id>".

Signed-off-by: Pavel Tikhomirov <redacted>
---
 fs/proc/base.c |    1 +
 1 file changed, 1 insertion(+)

diff --git a/fs/proc/base.c b/fs/proc/base.c
index 2dad4a9..8a38eef 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c

@@ -2079,6 +2079,7 @@ static int show_timer(struct seq_file *m, void *v)
 		nstr[notify & ~SIGEV_THREAD_ID],
 		(notify & SIGEV_THREAD_ID) ? "tid" : "pid",
 		pid_nr_ns(timer->it_pid, tp->ns));
+	seq_printf(m, "ClockID: %d\n", timer->it_clock);

What userspace tool just broke by adding a new field to this file?

Zero for two reasons:

1. this proc file appeared this merge window
2. this file format is

  <name>: <value>

and should be parsed line-by-line by applications like e.g. /proc/pid/status
or /proc/pid/smaps. Both these files were extended with fields in the past with
no worries to tools, so should be this new one.

thanks,

greg k-h
.

Thanks,
Pavel

On Fri, May 17, 2013 at 08:12:55PM +0400, Pavel Emelyanov wrote:

On 05/17/2013 07:59 PM, Greg KH wrote:

quoted

On Fri, May 17, 2013 at 02:12:03AM +0400, Pavel Tikhomirov wrote:

quoted

Expand information about posix-timers in /proc/<pid>/timers by adding
info about clock, with which the timer was created. I.e. in the forth
line of timer info after "notify:" line go "ClockID: <clock_id>".

Signed-off-by: Pavel Tikhomirov <redacted>
---
 fs/proc/base.c |    1 +
 1 file changed, 1 insertion(+)

diff --git a/fs/proc/base.c b/fs/proc/base.c
index 2dad4a9..8a38eef 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c

@@ -2079,6 +2079,7 @@ static int show_timer(struct seq_file *m, void *v)
 		nstr[notify & ~SIGEV_THREAD_ID],
 		(notify & SIGEV_THREAD_ID) ? "tid" : "pid",
 		pid_nr_ns(timer->it_pid, tp->ns));
+	seq_printf(m, "ClockID: %d\n", timer->it_clock);

What userspace tool just broke by adding a new field to this file?

Zero for two reasons:

1. this proc file appeared this merge window
2. this file format is

  <name>: <value>

and should be parsed line-by-line by applications like e.g. /proc/pid/status
or /proc/pid/smaps. Both these files were extended with fields in the past with
no worries to tools, so should be this new one.

Ok, just making sure :)