[PATCH 1/2] Improve the figure of QSBR

[PATCH 1/2] Improve the figure of QSBR

[Hitoshi Mitake]

This commit improves the description of QSBR figure. It would make
understanding of how context switch and run_on() in the pseudo code
relate a little bit easier.

Signed-off-by: Hitoshi Mitake <mitake.hitoshi@gmail.com>
---
 defer/QSBRGracePeriod.eps | 56 ++++++++++++++++++++++++++++++++++++++---------
 defer/QSBRGracePeriod.fig | 36 ++++++++++++++++++++----------
 2 files changed, 70 insertions(+), 22 deletions(-)

diff --git a/defer/QSBRGracePeriod.eps b/defer/QSBRGracePeriod.eps
index 1d7a97e..6895f40 100644
--- a/defer/QSBRGracePeriod.eps
+++ b/defer/QSBRGracePeriod.eps
@@ -1,8 +1,8 @@
-%!PS-Adobe-2.0 EPSF-2.0
+%!PS-Adobe-3.0 EPSF-3.0
 %%Title: QSBRGracePeriod.fig
-%%Creator: fig2dev Version 3.2 Patchlevel 5c
-%%CreationDate: Thu May 19 06:58:39 2011
-%%BoundingBox: 0 0 221 250
+%%Creator: fig2dev Version 3.2 Patchlevel 5e
+%%CreationDate: Sat Feb 20 17:39:00 2016
+%%BoundingBox: 0 0 232 250
 %Magnification: 1.0000
 %%EndComments
 %%BeginProlog
@@ -94,8 +94,8 @@ end

 /pageheader {
 save
-newpath 0 250 moveto 0 0 lineto 221 0 lineto 221 250 lineto closepath clip newpath
--1.7 259.1 translate
+newpath 0 250 moveto 0 0 lineto 232 0 lineto 232 250 lineto closepath clip newpath
+-0.0 259.1 translate
 1 -1 scale
 $F2psBegin
 10 setmiterlimit
@@ -200,6 +200,33 @@ n 1200 3450 m

 % arrowhead
 n 847 3649 m 764 3741 l 880 3699 l  col0 s
+% Polyline
+gs  clippath
+1370 1848 m 1245 1933 l 1279 1983 l 1404 1897 l 1404 1897 l 1289 1941 l 1370 1848 l cp
+eoclip
+n 2250 1275 m
+ 1275 1950 l gs col0 s gr gr
+
+% arrowhead
+n 1370 1848 m 1289 1941 l 1404 1897 l  col0 s
+% Polyline
+gs  clippath
+2213 3140 m 2311 3255 l 2357 3216 l 2258 3101 l 2258 3101 l 2314 3212 l 2213 3140 l cp
+eoclip
+n 1875 2700 m
+ 2325 3225 l gs col0 s gr gr
+
+% arrowhead
+n 2213 3140 m 2314 3212 l 2258 3101 l  col0 s
+% Polyline
+gs  clippath
+1585 2406 m 1734 2432 l 1744 2372 l 1595 2347 l 1595 2347 l 1709 2397 l 1585 2406 l cp
+eoclip
+n 1275 2325 m
+ 1725 2400 l gs col0 s gr gr
+
+% arrowhead
+n 1585 2406 m 1709 2397 l 1595 2347 l  col0 s
 /NimbusSanL-Regu ff 150.00 scf sf
 2550 1200 m
 gs 1 -1 sc (Context Switch) col0 sh gr
@@ -227,6 +254,15 @@ gs 1 -1 sc  90.0 rot (list_del\(\)) col0 sh gr
 /NimbusMonL-Regu-Bold ff 150.00 scf sf
 675 3750 m
 gs 1 -1 sc  90.0 rot (free\(\)) dup sw pop neg 0 rm  col0 sh gr
+/NimbusSanL-Regu ff 100.00 scf sf
+1275 1500 m
+gs 1 -1 sc (run_on\(1\)) col0 sh gr
+/NimbusSanL-Regu ff 100.00 scf sf
+1350 2325 m
+gs 1 -1 sc (run_on\(2\)) col0 sh gr
+/NimbusSanL-Regu ff 100.00 scf sf
+1950 2700 m
+gs 1 -1 sc (run_on\(3\)) col0 sh gr
 % Ellipse
 n 2400 3375 75 75 0 360 DrawEllipse gs col28 1.00 shd ef gr gs col0 s gr

@@ -242,12 +278,12 @@ n 1800 2550 75 75 0 360 DrawEllipse gs col28 1.00 shd ef gr gs col0 s gr
 % Polyline
 15.000 slw
  [60] 0 sd
-n 600 2100 m
- 2625 2100 l gs col28 s gr  [] 0 sd
-% Polyline
- [60] 0 sd
 n 600 3450 m
  2625 3450 l gs col2 s gr  [] 0 sd
+% Polyline
+ [60] 0 sd
+n 600 2100 m
+ 2625 2100 l gs col28 s gr  [] 0 sd
 % here ends figure;
 pagefooter
 showpage
diff --git a/defer/QSBRGracePeriod.fig b/defer/QSBRGracePeriod.fig
index 3bee0de..77b232a 100644
--- a/defer/QSBRGracePeriod.fig
+++ b/defer/QSBRGracePeriod.fig
@@ -1,4 +1,4 @@
-#FIG 3.2  Produced by xfig version 3.2.5b
+#FIG 3.2  Produced by xfig version 3.2.5c
 Landscape
 Center
 Inches
@@ -28,8 +28,6 @@ Single
 	 1200 900 1200 3900
 2 2 0 1 0 3 50 -1 20 0.000 0 0 -1 0 0 5
 	 1200 2550 1275 2550 1275 2850 1200 2850 1200 2550
-2 1 1 2 28 7 49 -1 -1 4.000 0 0 -1 0 0 2
-	 600 2100 2625 2100
 2 1 1 2 2 7 49 -1 -1 4.000 0 0 -1 0 0 2
 	 600 3450 2625 3450
 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 1 1 2
@@ -49,12 +47,26 @@ Single
 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 1 0 2
 	0 0 1.00 60.00 120.00
 	 1200 3450 750 3750
-4 0 0 50 -1 16 9 0.0000 4 120 1110 2550 1200 Context Switch\001
-4 0 0 50 -1 16 9 0.0000 4 120 480 2550 1725 Reader\001
-4 1 0 50 -1 16 9 1.5708 4 120 900 750 2700 Grace Period\001
-4 1 0 50 -1 16 9 0.0000 4 120 480 1200 825 CPU 1\001
-4 1 0 50 -1 16 9 0.0000 4 120 480 1800 825 CPU 2\001
-4 1 0 50 -1 16 9 0.0000 4 120 480 2400 825 CPU 3\001
-4 0 0 50 -1 14 9 1.5708 4 105 1440 150 1650 wait for readers\001
-4 0 0 50 -1 14 9 1.5708 4 135 900 525 1350 list_del()\001
-4 2 0 50 -1 14 9 1.5708 4 120 540 675 3750 free()\001
+2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 1 0 2
+	0 0 1.00 60.00 120.00
+	 2250 1275 1275 1950
+2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 1 0 2
+	0 0 1.00 60.00 120.00
+	 1875 2700 2325 3225
+2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 1 0 2
+	0 0 1.00 60.00 120.00
+	 1275 2325 1725 2400
+2 1 1 2 28 7 49 -1 -1 4.000 0 0 -1 0 0 2
+	 600 2100 2625 2100
+4 0 0 50 -1 16 9 0.0000 4 135 1260 2550 1200 Context Switch\001
+4 0 0 50 -1 16 9 0.0000 4 135 540 2550 1725 Reader\001
+4 1 0 50 -1 16 9 1.5708 4 135 1080 750 2700 Grace Period\001
+4 1 0 50 -1 16 9 0.0000 4 135 450 1200 825 CPU 1\001
+4 1 0 50 -1 16 9 0.0000 4 135 450 1800 825 CPU 2\001
+4 1 0 50 -1 16 9 0.0000 4 135 450 2400 825 CPU 3\001
+4 0 0 50 -1 14 9 1.5708 4 135 1440 150 1650 wait for readers\001
+4 0 0 50 -1 14 9 1.5708 4 165 900 525 1350 list_del()\001
+4 2 0 50 -1 14 9 1.5708 4 165 540 675 3750 free()\001
+4 0 0 50 -1 0 6 0.0000 4 165 810 1275 1500 run_on(1)\001
+4 0 0 50 -1 0 6 0.0000 4 165 810 1350 2325 run_on(2)\001
+4 0 0 50 -1 0 6 0.0000 4 165 810 1950 2700 run_on(3)\001
-- 
1.9.1

[PATCH 2/2] Let a writer thread go back to the original CPU in the toy RCU pseudocode

[Hitoshi Mitake]

This commit a little bit improves the logic of a writer thread. The
new logic lets the writer go back to the original CPU after run_on()
the final CPU.

Of course it is a trivial improvement and the original two lines
pseudocode is beautiful. If it isn't worth to apply, please ignore it.

Signed-off-by: Hitoshi Mitake <mitake.hitoshi@gmail.com>
---
 defer/rcuintro.tex | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/defer/rcuintro.tex b/defer/rcuintro.tex
index 03d6e12..135922c 100644
--- a/defer/rcuintro.tex
+++ b/defer/rcuintro.tex
@@ -166,8 +166,11 @@ quite complex, a toy implementation is exceedingly simple:
 \begin{minipage}[t]{\columnwidth}
 \scriptsize
 \begin{verbatim}
-  1 for_each_online_cpu(cpu)
-  2   run_on(cpu);
+  1 int orig_cpu = smp_processor_id();
+  2 for_each_online_cpu(cpu)
+  3   run_on(cpu);
+  4 if (orig_cpu != smp_processor_id())
+  5   run_on(orig_cpu);
 \end{verbatim}
 \end{minipage}
 \vspace{5pt}
-- 
1.9.1

Re: [PATCH 2/2] Let a writer thread go back to the original CPU in the toy RCU pseudocode

[Paul E. McKenney]

On Sat, Feb 20, 2016 at 05:52:02PM +0900, Hitoshi Mitake wrote:
> This commit a little bit improves the logic of a writer thread. The
> new logic lets the writer go back to the original CPU after run_on()
> the final CPU.
> 
> Of course it is a trivial improvement and the original two lines
> pseudocode is beautiful. If it isn't worth to apply, please ignore it.
> 
> Signed-off-by: Hitoshi Mitake <mitake.hitoshi@gmail.com>

Hello, Hitoshi,

Thank you for the submission, but each addition of this sort invites
another.  For example, you capture the initial CPU, but suppose that
the thread was originally bound to a pair of CPUs?  Wouldn't you then
want to restore the binding to that pair of CPUs?

But if you did that, suppose that both of those CPUs went offline
while going through the for_each_online_cpu() loop?  Wouldn't you
need to instead bind to some CPUs that were actually online?

And suppose that some other CPU changed the binding of the thread
just after you captured the initial binding?  Wouldn't you want to
reflect that change (give or take CPU hotplug operations) at the
end of the loop?

And it just gets worse from there.  So we need to keep this one
simple, which means that I cannot accept this patch.

							Thanx, Paul

> ---
>  defer/rcuintro.tex | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/defer/rcuintro.tex b/defer/rcuintro.tex
> index 03d6e12..135922c 100644
> --- a/defer/rcuintro.tex
> +++ b/defer/rcuintro.tex
> @@ -166,8 +166,11 @@ quite complex, a toy implementation is exceedingly simple:
>  \begin{minipage}[t]{\columnwidth}
>  \scriptsize
>  \begin{verbatim}
> -  1 for_each_online_cpu(cpu)
> -  2   run_on(cpu);
> +  1 int orig_cpu = smp_processor_id();
> +  2 for_each_online_cpu(cpu)
> +  3   run_on(cpu);
> +  4 if (orig_cpu != smp_processor_id())
> +  5   run_on(orig_cpu);
>  \end{verbatim}
>  \end{minipage}
>  \vspace{5pt}
> -- 
> 1.9.1
> 
> --
> To unsubscribe from this list: send the line "unsubscribe perfbook" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

Re: [PATCH 2/2] Let a writer thread go back to the original CPU in the toy RCU pseudocode

[Hitoshi Mitake]

Hi Paul,

At Sat, 20 Feb 2016 21:30:20 -0800,
Paul E. McKenney wrote:
> 
> On Sat, Feb 20, 2016 at 05:52:02PM +0900, Hitoshi Mitake wrote:
> > This commit a little bit improves the logic of a writer thread. The
> > new logic lets the writer go back to the original CPU after run_on()
> > the final CPU.
> > 
> > Of course it is a trivial improvement and the original two lines
> > pseudocode is beautiful. If it isn't worth to apply, please ignore it.
> > 
> > Signed-off-by: Hitoshi Mitake <mitake.hitoshi@gmail.com>
> 
> Hello, Hitoshi,
> 
> Thank you for the submission, but each addition of this sort invites
> another.  For example, you capture the initial CPU, but suppose that
> the thread was originally bound to a pair of CPUs?  Wouldn't you then
> want to restore the binding to that pair of CPUs?
> 
> But if you did that, suppose that both of those CPUs went offline
> while going through the for_each_online_cpu() loop?  Wouldn't you
> need to instead bind to some CPUs that were actually online?
> 
> And suppose that some other CPU changed the binding of the thread
> just after you captured the initial binding?  Wouldn't you want to
> reflect that change (give or take CPU hotplug operations) at the
> end of the loop?
> 
> And it just gets worse from there.  So we need to keep this one
> simple, which means that I cannot accept this patch.

Thanks a lot for your review. I couldn't consider about atomicity of
the CPU state, the affinity of the thread, etc. I assumed the
atomicity is provided and this assumption is wrong.

Again, thanks for your time and review. Your review improved my
understanding of RCU :)

# BTW, how do you think about the change of the figure in the 1st
# patch? Of couse I'm not sure it is worth to apply or not.

Thanks,
Hitoshi

> 
> 							Thanx, Paul
> 
> > ---
> >  defer/rcuintro.tex | 7 +++++--
> >  1 file changed, 5 insertions(+), 2 deletions(-)
> > 
> > diff --git a/defer/rcuintro.tex b/defer/rcuintro.tex
> > index 03d6e12..135922c 100644
> > --- a/defer/rcuintro.tex
> > +++ b/defer/rcuintro.tex
> > @@ -166,8 +166,11 @@ quite complex, a toy implementation is exceedingly simple:
> >  \begin{minipage}[t]{\columnwidth}
> >  \scriptsize
> >  \begin{verbatim}
> > -  1 for_each_online_cpu(cpu)
> > -  2   run_on(cpu);
> > +  1 int orig_cpu = smp_processor_id();
> > +  2 for_each_online_cpu(cpu)
> > +  3   run_on(cpu);
> > +  4 if (orig_cpu != smp_processor_id())
> > +  5   run_on(orig_cpu);
> >  \end{verbatim}
> >  \end{minipage}
> >  \vspace{5pt}
> > -- 
> > 1.9.1
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe perfbook" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > 
> 

Re: [PATCH 1/2] Improve the figure of QSBR

[Hitoshi Mitake]

At Sat, 20 Feb 2016 17:52:01 +0900,
Hitoshi Mitake wrote:
> 
> This commit improves the description of QSBR figure. It would make
> understanding of how context switch and run_on() in the pseudo code
> relate a little bit easier.

Sorry, I found an invalid point in the figure. I'll send v2 later.

Thanks,
Hitoshi

> 
> Signed-off-by: Hitoshi Mitake <mitake.hitoshi@gmail.com>
> ---
>  defer/QSBRGracePeriod.eps | 56 ++++++++++++++++++++++++++++++++++++++---------
>  defer/QSBRGracePeriod.fig | 36 ++++++++++++++++++++----------
>  2 files changed, 70 insertions(+), 22 deletions(-)
> 
> diff --git a/defer/QSBRGracePeriod.eps b/defer/QSBRGracePeriod.eps
> index 1d7a97e..6895f40 100644
> --- a/defer/QSBRGracePeriod.eps
> +++ b/defer/QSBRGracePeriod.eps
> @@ -1,8 +1,8 @@
> -%!PS-Adobe-2.0 EPSF-2.0
> +%!PS-Adobe-3.0 EPSF-3.0
>  %%Title: QSBRGracePeriod.fig
> -%%Creator: fig2dev Version 3.2 Patchlevel 5c
> -%%CreationDate: Thu May 19 06:58:39 2011
> -%%BoundingBox: 0 0 221 250
> +%%Creator: fig2dev Version 3.2 Patchlevel 5e
> +%%CreationDate: Sat Feb 20 17:39:00 2016
> +%%BoundingBox: 0 0 232 250
>  %Magnification: 1.0000
>  %%EndComments
>  %%BeginProlog
> @@ -94,8 +94,8 @@ end
>  
>  /pageheader {
>  save
> -newpath 0 250 moveto 0 0 lineto 221 0 lineto 221 250 lineto closepath clip newpath
> --1.7 259.1 translate
> +newpath 0 250 moveto 0 0 lineto 232 0 lineto 232 250 lineto closepath clip newpath
> +-0.0 259.1 translate
>  1 -1 scale
>  $F2psBegin
>  10 setmiterlimit
> @@ -200,6 +200,33 @@ n 1200 3450 m
>  
>  % arrowhead
>  n 847 3649 m 764 3741 l 880 3699 l  col0 s
> +% Polyline
> +gs  clippath
> +1370 1848 m 1245 1933 l 1279 1983 l 1404 1897 l 1404 1897 l 1289 1941 l 1370 1848 l cp
> +eoclip
> +n 2250 1275 m
> + 1275 1950 l gs col0 s gr gr
> +
> +% arrowhead
> +n 1370 1848 m 1289 1941 l 1404 1897 l  col0 s
> +% Polyline
> +gs  clippath
> +2213 3140 m 2311 3255 l 2357 3216 l 2258 3101 l 2258 3101 l 2314 3212 l 2213 3140 l cp
> +eoclip
> +n 1875 2700 m
> + 2325 3225 l gs col0 s gr gr
> +
> +% arrowhead
> +n 2213 3140 m 2314 3212 l 2258 3101 l  col0 s
> +% Polyline
> +gs  clippath
> +1585 2406 m 1734 2432 l 1744 2372 l 1595 2347 l 1595 2347 l 1709 2397 l 1585 2406 l cp
> +eoclip
> +n 1275 2325 m
> + 1725 2400 l gs col0 s gr gr
> +
> +% arrowhead
> +n 1585 2406 m 1709 2397 l 1595 2347 l  col0 s
>  /NimbusSanL-Regu ff 150.00 scf sf
>  2550 1200 m
>  gs 1 -1 sc (Context Switch) col0 sh gr
> @@ -227,6 +254,15 @@ gs 1 -1 sc  90.0 rot (list_del\(\)) col0 sh gr
>  /NimbusMonL-Regu-Bold ff 150.00 scf sf
>  675 3750 m
>  gs 1 -1 sc  90.0 rot (free\(\)) dup sw pop neg 0 rm  col0 sh gr
> +/NimbusSanL-Regu ff 100.00 scf sf
> +1275 1500 m
> +gs 1 -1 sc (run_on\(1\)) col0 sh gr
> +/NimbusSanL-Regu ff 100.00 scf sf
> +1350 2325 m
> +gs 1 -1 sc (run_on\(2\)) col0 sh gr
> +/NimbusSanL-Regu ff 100.00 scf sf
> +1950 2700 m
> +gs 1 -1 sc (run_on\(3\)) col0 sh gr
>  % Ellipse
>  n 2400 3375 75 75 0 360 DrawEllipse gs col28 1.00 shd ef gr gs col0 s gr
>  
> @@ -242,12 +278,12 @@ n 1800 2550 75 75 0 360 DrawEllipse gs col28 1.00 shd ef gr gs col0 s gr
>  % Polyline
>  15.000 slw
>   [60] 0 sd
> -n 600 2100 m
> - 2625 2100 l gs col28 s gr  [] 0 sd
> -% Polyline
> - [60] 0 sd
>  n 600 3450 m
>   2625 3450 l gs col2 s gr  [] 0 sd
> +% Polyline
> + [60] 0 sd
> +n 600 2100 m
> + 2625 2100 l gs col28 s gr  [] 0 sd
>  % here ends figure;
>  pagefooter
>  showpage
> diff --git a/defer/QSBRGracePeriod.fig b/defer/QSBRGracePeriod.fig
> index 3bee0de..77b232a 100644
> --- a/defer/QSBRGracePeriod.fig
> +++ b/defer/QSBRGracePeriod.fig
> @@ -1,4 +1,4 @@
> -#FIG 3.2  Produced by xfig version 3.2.5b
> +#FIG 3.2  Produced by xfig version 3.2.5c
>  Landscape
>  Center
>  Inches
> @@ -28,8 +28,6 @@ Single
>  	 1200 900 1200 3900
>  2 2 0 1 0 3 50 -1 20 0.000 0 0 -1 0 0 5
>  	 1200 2550 1275 2550 1275 2850 1200 2850 1200 2550
> -2 1 1 2 28 7 49 -1 -1 4.000 0 0 -1 0 0 2
> -	 600 2100 2625 2100
>  2 1 1 2 2 7 49 -1 -1 4.000 0 0 -1 0 0 2
>  	 600 3450 2625 3450
>  2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 1 1 2
> @@ -49,12 +47,26 @@ Single
>  2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 1 0 2
>  	0 0 1.00 60.00 120.00
>  	 1200 3450 750 3750
> -4 0 0 50 -1 16 9 0.0000 4 120 1110 2550 1200 Context Switch\001
> -4 0 0 50 -1 16 9 0.0000 4 120 480 2550 1725 Reader\001
> -4 1 0 50 -1 16 9 1.5708 4 120 900 750 2700 Grace Period\001
> -4 1 0 50 -1 16 9 0.0000 4 120 480 1200 825 CPU 1\001
> -4 1 0 50 -1 16 9 0.0000 4 120 480 1800 825 CPU 2\001
> -4 1 0 50 -1 16 9 0.0000 4 120 480 2400 825 CPU 3\001
> -4 0 0 50 -1 14 9 1.5708 4 105 1440 150 1650 wait for readers\001
> -4 0 0 50 -1 14 9 1.5708 4 135 900 525 1350 list_del()\001
> -4 2 0 50 -1 14 9 1.5708 4 120 540 675 3750 free()\001
> +2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 1 0 2
> +	0 0 1.00 60.00 120.00
> +	 2250 1275 1275 1950
> +2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 1 0 2
> +	0 0 1.00 60.00 120.00
> +	 1875 2700 2325 3225
> +2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 1 0 2
> +	0 0 1.00 60.00 120.00
> +	 1275 2325 1725 2400
> +2 1 1 2 28 7 49 -1 -1 4.000 0 0 -1 0 0 2
> +	 600 2100 2625 2100
> +4 0 0 50 -1 16 9 0.0000 4 135 1260 2550 1200 Context Switch\001
> +4 0 0 50 -1 16 9 0.0000 4 135 540 2550 1725 Reader\001
> +4 1 0 50 -1 16 9 1.5708 4 135 1080 750 2700 Grace Period\001
> +4 1 0 50 -1 16 9 0.0000 4 135 450 1200 825 CPU 1\001
> +4 1 0 50 -1 16 9 0.0000 4 135 450 1800 825 CPU 2\001
> +4 1 0 50 -1 16 9 0.0000 4 135 450 2400 825 CPU 3\001
> +4 0 0 50 -1 14 9 1.5708 4 135 1440 150 1650 wait for readers\001
> +4 0 0 50 -1 14 9 1.5708 4 165 900 525 1350 list_del()\001
> +4 2 0 50 -1 14 9 1.5708 4 165 540 675 3750 free()\001
> +4 0 0 50 -1 0 6 0.0000 4 165 810 1275 1500 run_on(1)\001
> +4 0 0 50 -1 0 6 0.0000 4 165 810 1350 2325 run_on(2)\001
> +4 0 0 50 -1 0 6 0.0000 4 165 810 1950 2700 run_on(3)\001
> -- 
> 1.9.1
>