[9fans] setitimer equivalent

[9fans] setitimer equivalent

[yoann padioleau]

Hi,

I’m trying to port some code to plan9 using APE
but I can’t find the setitimer function in any include/ape/ files.
What is the equivalent to setitimer under plan9?

Re: [9fans] setitimer equivalent

[James A. Robinson]

[-- Attachment #1: Type: text/plain, Size: 465 bytes --]

I'm not familiar with a function named settimer, is that from Windows?
Given the name, would sleep/alarm fit your requirements?

http://plan9.bell-labs.com/magic/man2html/2/sleep

On Tue, Apr 28, 2015 at 9:31 PM yoann padioleau <aryx.padator@gmail.com>
wrote:

> Hi,
>
> I’m trying to port some code to plan9 using APE
> but I can’t find the setitimer function in any include/ape/ files.
> What is the equivalent to setitimer under plan9?
>
>
>

[-- Attachment #2: Type: text/html, Size: 792 bytes --]

Re: [9fans] setitimer equivalent

[lucio]

> I'm not familiar with a function named settimer, is that from Windows?
> Given the name, would sleep/alarm fit your requirements?

According to the Ubuntu man pages, setitimer(2) operates on one of
three interval timers: ral, virtual and profile.  It could take quite
a bit of effort to port this to APE.  I see that NetBSD echoes the
Linux documentation.

Lucio.

Re: [9fans] setitimer equivalent

[James A. Robinson]

[-- Attachment #1: Type: text/plain, Size: 700 bytes --]

Sorry, I missed the extra 'i' in there.  So is the requirement to get all
three of those time domains supported?  Supporting all three looks to me
like something one would need for building real time systems or a profiling
tool?

On Tue, Apr 28, 2015 at 9:57 PM <lucio@proxima.alt.za> wrote:

> > I'm not familiar with a function named settimer, is that from Windows?
> > Given the name, would sleep/alarm fit your requirements?
>
> According to the Ubuntu man pages, setitimer(2) operates on one of
> three interval timers: ral, virtual and profile.  It could take quite
> a bit of effort to port this to APE.  I see that NetBSD echoes the
> Linux documentation.
>
> Lucio.
>
>
>

[-- Attachment #2: Type: text/html, Size: 963 bytes --]

Re: [9fans] setitimer equivalent

[lucio]

> Sorry, I missed the extra 'i' in there.  So is the requirement to get all
> three of those time domains supported?  Supporting all three looks to me
> like something one would need for building real time systems or a profiling
> tool?

For the record, the NetBSD man pages suggest that Posix has declared
setitimer and getitimer deprecated.

Lucio.

Re: [9fans] setitimer equivalent

[erik quanstrom]

On Tue Apr 28 21:30:24 PDT 2015, aryx.padator@gmail.com wrote:
> Hi,
> 
> I’m trying to port some code to plan9 using APE
> but I can’t find the setitimer function in any include/ape/ files.
> What is the equivalent to setitimer under plan9?

if you're ok on using nsec() as your time source, select(3) as posix would
have it, will do.

- erik

Re: [9fans] setitimer equivalent, and sigvtalarm equivalent

[yoann padioleau]

Hi,

I’m not sure I understand your proposal.

But looking more at the code I need to port, an interpreter, I think
the main issue is that the interpreter relies on sigvtalarm as
a way to preempt the interpreted program while allowing
the user program to use only sigalarm. So more generally
the problem is how can I have two alarms in plan9? I see only
one alarm signal in ape, SIGALRM. There is no SIGVTALRM.



> On Apr 29, 2015, at 9:47 AM, erik quanstrom <quanstro@quanstro.net> wrote:
> 
> On Tue Apr 28 21:30:24 PDT 2015, aryx.padator@gmail.com wrote:
>> Hi,
>> 
>> I’m trying to port some code to plan9 using APE
>> but I can’t find the setitimer function in any include/ape/ files.
>> What is the equivalent to setitimer under plan9?
> 
> if you're ok on using nsec() as your time source, select(3) as posix would
> have it, will do.
> 
> - erik
> 

Re: [9fans] setitimer equivalent, and sigvtalarm equivalent

[Skip Tavakkolian]

[-- Attachment #1: Type: text/plain, Size: 1253 bytes --]

one way to solve this would be to fork as many procs as there are
timers; each would sleep the length of its timer value and then
postnote an identifying artificial note (vtalarm, fooalarm, etc.) to
the parent process.  the parent process then interprets the artificial
note as a particular type of alarm.

-Skip

> Hi,
> 
> I’m not sure I understand your proposal.
> 
> But looking more at the code I need to port, an interpreter, I think
> the main issue is that the interpreter relies on sigvtalarm as
> a way to preempt the interpreted program while allowing
> the user program to use only sigalarm. So more generally
> the problem is how can I have two alarms in plan9? I see only
> one alarm signal in ape, SIGALRM. There is no SIGVTALRM.
> 
> 
> 
>> On Apr 29, 2015, at 9:47 AM, erik quanstrom <quanstro@quanstro.net> wrote:
>> 
>> On Tue Apr 28 21:30:24 PDT 2015, aryx.padator@gmail.com wrote:
>>> Hi,
>>> 
>>> I’m trying to port some code to plan9 using APE
>>> but I can’t find the setitimer function in any include/ape/ files.
>>> What is the equivalent to setitimer under plan9?
>> 
>> if you're ok on using nsec() as your time source, select(3) as posix would
>> have it, will do.
>> 
>> - erik
>> 

[-- Attachment #2: notesy.c --]
[-- Type: text/plain, Size: 798 bytes --]

#include <u.h>
#include <libc.h>
#include <stdio.h>

void
alarmhandler(void *a, char *msg)
{
	USED(a);
	if(strstr(msg, "alarm") != 0) {
		printf("received alarm signal %s\n", msg);
		noted(NCONT);
	}
}

int
timer(char *msg, int ms)
{
	int pid, apid;

	pid = getpid();
	switch (apid = fork()) {
	case -1:
		sysfatal("fork failed: %r");
	case 0:
		sleep(ms);
		postnote(PNPROC, pid, msg);
		exits(nil);
	default:
		break;
	}
	return apid;
}

void
main(int argc, char **argv)
{
	int pid, t1pid, t2pid, count = 0;
	notify(alarmhandler);

	t1pid = timer("fooalarm", 750); count++;
	t2pid = timer("baralarm", 250); count++;

	alarm(500);
	while (count > 0) {
		pid = waitpid();
		if (pid == t1pid) count--;
		if (pid == t2pid) count--;
	}

	exits(nil);
}

Re: [9fans] setitimer equivalent, and sigvtalarm equivalent

[erik quanstrom]

> I’m not sure I understand your proposal.
> 
> But looking more at the code I need to port, an interpreter, I think
> the main issue is that the interpreter relies on sigvtalarm as
> a way to preempt the interpreted program while allowing
> the user program to use only sigalarm. So more generally
> the problem is how can I have two alarms in plan9? I see only
> one alarm signal in ape, SIGALRM. There is no SIGVTALRM.

the standard way to do this is as the kernel does.  the kernel typically
keeps one highres timer (per core), and a linked list of events sorted by alarm
time.  when the alarm fires, do the action associated with each alarm that has
passed and reload periodic timers.  this is in port/portclock.c.  i'm sure there
are implementations of this algorithm elsewhere, too.

- erik

Re: [9fans] setitimer equivalent, and sigvtalarm equivalent

[cinap_lenrek]

won't you need some kind of flow control in this? just hammering
notes onto a process constantly might prevent that process from
doing any progress and even stack overflow as signal handlers
are reentrant on unix/ape. i remember that these timer
signals where handled specially in the linux kernel when i
implemented support for it in linuxemu that tries to prevent
this.

whats the program using interval timers for? maybe one can
rewrite the program todo stuff with multiple processes instead
of using signals?

--
cinap

Re: [9fans] setitimer equivalent, and sigvtalarm equivalent

[yoann padioleau]

Hi,

Yes I think you’re right. Indeed the machine has one
timer interrupt and it’s enough for the kernel to demultiplex
it over many processes who each can have different alarm
requirments.

Thx.

> On May 2, 2015, at 8:06 AM, erik quanstrom <quanstro@quanstro.net> wrote:
> 
>> I’m not sure I understand your proposal.
>> 
>> But looking more at the code I need to port, an interpreter, I think
>> the main issue is that the interpreter relies on sigvtalarm as
>> a way to preempt the interpreted program while allowing
>> the user program to use only sigalarm. So more generally
>> the problem is how can I have two alarms in plan9? I see only
>> one alarm signal in ape, SIGALRM. There is no SIGVTALRM.
> 
> the standard way to do this is as the kernel does.  the kernel typically
> keeps one highres timer (per core), and a linked list of events sorted by alarm
> time.  when the alarm fires, do the action associated with each alarm that has
> passed and reload periodic timers.  this is in port/portclock.c.  i'm sure there
> are implementations of this algorithm elsewhere, too.
> 
> - erik
>