[Cocci] hi all, how to match the args in function

[Cocci] hi all, how to match the args in function

[林嘉(程二 福州)]

I just began to learn cocci recently.
How to write a patch match the following?

Void func1(int s, bri_vlan_set_t t, char x)  yes
Void func2(int s,  char x)  no
Void func3(int s, bri_vlan_set_t t) yes

bri_vlan_set_t is defined type

I wrote:

@@
identifier fn, vp;
typedef bri_vlan_set_t;
@@
-fn(...bri_vlan_set_t vp...)

But it reports

Fatal error: exception Failure("minus: parse error:
 = File "r.cocci", line 5, column 7,  charpos = 60
    around = 'bri_vlan_set_t', whole content = -fn(...bri_vlan_set_t vp...)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://systeme.lip6.fr/pipermail/cocci/attachments/20140102/cfded7ef/attachment.html>

[Cocci] 答复: hi all, how to match the args in function

[Julia Lawall]

On Thu, 2 Jan 2014, ??(?? ??) wrote:

> Well, now I wrote as following:
> 
> @@
> identifier fn, vp;
> typedef bri_vlan_set_t;
> @@
> fn(...,bri_vlan_set_t vp,...);

Oops sorry.  I you want to match only the prototype, you do need to put 
the type out front.  So it would be

 @@
 identifier fn, vp;
 typedef bri_vlan_set_t;
 type T;
 @@
 T fn(...,bri_vlan_set_t vp,...);

julia

[Cocci] 答复: 答复: hi all, how to match the args in function

[Julia Lawall]

On Thu, 2 Jan 2014, ??(?? ??) wrote:

> Thank you very much, and
> I wrote as following which works now:
> 
> ------------
> @@
> identifier fn, vp;
> typedef bri_vlan_set_t;
> @@
> -fn(...,bri_vlan_set_t *vp,...)
> -{
> -...
> -}
> -------------
> 
> Now, here comes another problem, I want to find out those functions matching the prototype and memset vp in this function, for example
> 
> void vl_conbitmap(bri_vlan_set_t *vlanset) <-- condition 1
> {
>     if (action == BITMAP_EXCEPT) {
>         memset(vlanset, 0xff, sizeof(bri_vlan_set_t)); <-- condition2
>     }
> }
> 
> 
> How to write the patch?
> 
> I tried this:
> @@
> identifier fn, vp;
> typedef bri_vlan_set_t;
> expression e1,e2;
> @@
> -fn(...,bri_vlan_set_t *vp,...)
> -{
> -...
> -memset(vp, e1, e2);
> -...
> -}
> 
> But it seems failed, nothing output.

This means that every execution path has exactly one call to memset with 
vp as the first argument.  If you just mean that there should be at least 
one on every (non error) execution path, then you can put the following:

@@
identifier fn, vp;
typedef bri_vlan_set_t;
expression e1,e2;
@@
-fn(...,bri_vlan_set_t *vp,...)
-{
-<+...
-memset(vp, e1, e2);
-...+>
-}

If you want that there is at least one somewhere, with no constraints on 
the execution path, and you just want to find out about this, then you can 
put the following:

@@
identifier fn, vp;
typedef bri_vlan_set_t;
expression e1,e2;
@@
fn(...,bri_vlan_set_t *vp,...)
{
<+...
*memset(vp, e1, e2);
...+>
}  

* implicitly checks for the existence of such an execution path with the 
property, rather than requiring that all execution paths have the 
property.

If you really want to completely remove such functions, it is a bit more 
complicated.  First, you need the following rule, which explicitly uses 
exists, to find a fuction of interest:

@r exists@                    <----- new
identifier fn, vp;
typedef bri_vlan_set_t;
expression e1,e2;
position p;                   <----- new
@@
fn@p(...,bri_vlan_set_t *vp,...)  <----- new
{
<+...
memset(vp, e1, e2);
...+>
}  

And then the following rule, which considers all execution paths to 
remove the whole function:

@@
identifier r.fn;
position r.p;
@@

- fn at p(...) { ... }

The position variable is used to be sure that the second rule is matching 
the same function as the first one.  With ifdefs, you could have two 
definitions of functions with the same name.

julia

PS, please keep the mailing list in CC, as others could be interested in 
your question.

[Cocci] 答复: 答复: 答复: hi all, how to match the args in function

[林嘉(程二 福州)]

Thank you very much, it works now!

And do u know where can I get more document about SmPL grammer?  I only have a  < The SmPL Grammar (version 0.1.4)Research group on Coccinelle June 5, 2009>
Which describe some simply, and there are not enough examples,


Btw, I will cc maillist, :)
Thank you.

-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?1?2? 17:33
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: ??: ??: [Cocci] hi all, how to match the args in function

On Thu, 2 Jan 2014, ??(?? ??) wrote:

> Thank you very much, and
> I wrote as following which works now:
> 
> ------------
> @@
> identifier fn, vp;
> typedef bri_vlan_set_t;
> @@
> -fn(...,bri_vlan_set_t *vp,...)
> -{
> -...
> -}
> -------------
> 
> Now, here comes another problem, I want to find out those functions 
> matching the prototype and memset vp in this function, for example
> 
> void vl_conbitmap(bri_vlan_set_t *vlanset) <-- condition 1 {
>     if (action == BITMAP_EXCEPT) {
>         memset(vlanset, 0xff, sizeof(bri_vlan_set_t)); <-- condition2
>     }
> }
> 
> 
> How to write the patch?
> 
> I tried this:
> @@
> identifier fn, vp;
> typedef bri_vlan_set_t;
> expression e1,e2;
> @@
> -fn(...,bri_vlan_set_t *vp,...)
> -{
> -...
> -memset(vp, e1, e2);
> -...
> -}
> 
> But it seems failed, nothing output.

This means that every execution path has exactly one call to memset with vp as the first argument.  If you just mean that there should be at least one on every (non error) execution path, then you can put the following:

@@
identifier fn, vp;
typedef bri_vlan_set_t;
expression e1,e2;
@@
-fn(...,bri_vlan_set_t *vp,...)
-{
-<+...
-memset(vp, e1, e2);
-...+>
-}

If you want that there is at least one somewhere, with no constraints on the execution path, and you just want to find out about this, then you can put the following:

@@
identifier fn, vp;
typedef bri_vlan_set_t;
expression e1,e2;
@@
fn(...,bri_vlan_set_t *vp,...)
{
<+...
*memset(vp, e1, e2);
...+>
}  

* implicitly checks for the existence of such an execution path with the property, rather than requiring that all execution paths have the property.

If you really want to completely remove such functions, it is a bit more complicated.  First, you need the following rule, which explicitly uses exists, to find a fuction of interest:

@r exists@                    <----- new
identifier fn, vp;
typedef bri_vlan_set_t;
expression e1,e2;
position p;                   <----- new
@@
fn@p(...,bri_vlan_set_t *vp,...)  <----- new { <+...
memset(vp, e1, e2);
...+>
}  

And then the following rule, which considers all execution paths to remove the whole function:

@@
identifier r.fn;
position r.p;
@@

- fn at p(...) { ... }

The position variable is used to be sure that the second rule is matching the same function as the first one.  With ifdefs, you could have two definitions of functions with the same name.

julia

PS, please keep the mailing list in CC, as others could be interested in your question.

[Cocci] 答复: 答复: 答复: hi all, how to match the args in function

[Julia Lawall]

On Thu, 2 Jan 2014, ??(?? ??) wrote:

> Thank you very much, it works now!
> 
> And do u know where can I get more document about SmPL grammer?  I only 
> have a < The SmPL Grammar (version 0.1.4)Research group on Coccinelle 
> June 5, 2009> Which describe some simply, and there are not enough 
> examples,

I wonder where you got that version?  The version on the Coccinelle web 
page is for version 1.0.0-rc18 and is available here:

http://coccinelle.lip6.fr/docs/main_grammar.pdf

It is probably not very useful for learning Coccinelle, though.

If you are using Coccinelle version 0.1.4 you will need to upgrade.

One good resource is the mailing list, which is archived on gmane:

http://blog.gmane.org/gmane.comp.version-control.coccinelle

There are some examples with explanations on the web page:

http://coccinelle.lip6.fr/rules/

There are many examples without explanations here:

https://github.com/coccinelle/coccinellery

There are also examples in the Linux kernel source tree, in the 
scripts/coccinelle directory.

julia

[Cocci] 答复: 答复: 答复: 答复: hi all, how to match the args in function

[林嘉(程二 福州)]

Cocc version? I guess it's got by -v?
# spatch -version
spatch version 0.2.3 with Python support
maybe it is up to date?



and old version of grammer pdf is downloaded from: http://www.emn.fr/z-info/coccinelle/docs/main_grammar.pdf
thank you tell me the new version of pdf 


-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?1?2? 18:18
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: ??: ??: ??: [Cocci] hi all, how to match the args in function

On Thu, 2 Jan 2014, ??(?? ??) wrote:

> Thank you very much, it works now!
> 
> And do u know where can I get more document about SmPL grammer?  I 
> only have a < The SmPL Grammar (version 0.1.4)Research group on 
> Coccinelle June 5, 2009> Which describe some simply, and there are not 
> enough examples,

I wonder where you got that version?  The version on the Coccinelle web page is for version 1.0.0-rc18 and is available here:

http://coccinelle.lip6.fr/docs/main_grammar.pdf

It is probably not very useful for learning Coccinelle, though.

If you are using Coccinelle version 0.1.4 you will need to upgrade.

One good resource is the mailing list, which is archived on gmane:

http://blog.gmane.org/gmane.comp.version-control.coccinelle

There are some examples with explanations on the web page:

http://coccinelle.lip6.fr/rules/

There are many examples without explanations here:

https://github.com/coccinelle/coccinellery

There are also examples in the Linux kernel source tree, in the scripts/coccinelle directory.

julia

[Cocci] 答复: 答复: 答复: 答复: hi all, how to match the args in function

[Julia Lawall]

On Thu, 2 Jan 2014, ??(?? ??) wrote:

> Cocc version? I guess it's got by -v?
> # spatch -version
> spatch version 0.2.3 with Python support
> maybe it is up to date?

No, it is quite old.  The current version is 1.0.0-rc19
If you like, you can get it from github:
https://github.com/coccinelle/coccinelle

> and old version of grammer pdf is downloaded from: http://www.emn.fr/z-info/coccinelle/docs/main_grammar.pdf
> thank you tell me the new version of pdf

OK, that website is quite out of date.  Use coccinelle.lip6.fr.

julia

[Cocci] 答复: 答复: 答复: 答复: 答复: hi all, how to match the args in function

[林嘉(程二 福州)]

well, I downloaded the debian install package from ' coccinelle.lip6.fr.'

http://coccinelle.lip6.fr/download.php
quote:
Packages exist for several distributions:
Debian, thanks to Eugeniy Meshcheryakov
http://packages.debian.org/squeeze/coccinelle, maybe it hasn?t updated for a long time??

-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?1?2? 21:20
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: ??: ??: ??: ??: [Cocci] hi all, how to match the args in function

On Thu, 2 Jan 2014, ??(?? ??) wrote:

> Cocc version? I guess it's got by -v?
> # spatch -version
> spatch version 0.2.3 with Python support maybe it is up to date?

No, it is quite old.  The current version is 1.0.0-rc19 If you like, you can get it from github:
https://github.com/coccinelle/coccinelle

> and old version of grammer pdf is downloaded from: 
> http://www.emn.fr/z-info/coccinelle/docs/main_grammar.pdf
> thank you tell me the new version of pdf

OK, that website is quite out of date.  Use coccinelle.lip6.fr.

julia

[Cocci] 答复: 答复: 答复: 答复: 答复: hi all, how to match the args in function

[Julia Lawall]

On Fri, 3 Jan 2014, ??(?? ??) wrote:

> well, I downloaded the debian install package from ' coccinelle.lip6.fr.'
> 
> http://coccinelle.lip6.fr/download.php
> quote:
> Packages exist for several distributions:
> Debian, thanks to Eugeniy Meshcheryakov
> http://packages.debian.org/squeeze/coccinelle, maybe it hasn?t updated for a long time??

The debian package is updated regularly, I believe.  But the link on the 
Coccinelle web page doesn't go anywhere.  I will try to update it, but if 
you are using debian, then you could take the version in your package 
manager?

I see that the version in Debian squeeze is indeed 0.2.3, but I think that 
version of Debian is out of date.  The current stable version is wheezy, 
which has 1.0.0~rc12.  The testing and unstable versions have 1.0.0-rc19.

julia

> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
> ????: 2014?1?2? 21:20
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: ??: ??: ??: ??: [Cocci] hi all, how to match the args in function
> 
> On Thu, 2 Jan 2014, ??(?? ??) wrote:
> 
> > Cocc version? I guess it's got by -v?
> > # spatch -version
> > spatch version 0.2.3 with Python support maybe it is up to date?
> 
> No, it is quite old.  The current version is 1.0.0-rc19 If you like, you can get it from github:
> https://github.com/coccinelle/coccinelle
> 
> > and old version of grammer pdf is downloaded from: 
> > http://www.emn.fr/z-info/coccinelle/docs/main_grammar.pdf
> > thank you tell me the new version of pdf
> 
> OK, that website is quite out of date.  Use coccinelle.lip6.fr.
> 
> julia
> 

[Cocci] how to match this case?

[林嘉(程二 福州)]

Following is a code with bug

-----------------------------------
void set_timeout_value(int iParam)
{
	 int i;
	 byte b[100];
	 i = Geti();
	 if (i > 100 || i < 0){
	 	  return ;
	 	}	 	
	 b[i] = iParam;	/*  range of I is 0-100?so may overrun array* /

} 
-----------------------------------
Now I write a patch to find it, like this

@@
identifier fn, vp;
identifier I;
type T;
expression C;
@@
-fn(...,int vp,...)
-{
-T I[C];
-<+...
-if(vp > C || ...){
-...
-}
-...+>
-}


But it doesn?t work, give me a log:
rule starting on line 1: position variables or mixed modifs interfere with comm_assoc isobool (bool (int vp > int C) || ...)

can anyone tell why, thanks

[Cocci] how to match this case?

[Julia Lawall]

On Tue, 21 Jan 2014, ??(?? ??) wrote:

> Following is a code with bug
> 
> -----------------------------------
> void set_timeout_value(int iParam)
> {
> 	 int i;
> 	 byte b[100];
> 	 i = Geti();
> 	 if (i > 100 || i < 0){
> 	 	  return ;
> 	 	}	 	
> 	 b[i] = iParam;	/*  range of I is 0-100?so may overrun array* /
> 
> } 
> -----------------------------------
> Now I write a patch to find it, like this
> 
> @@
> identifier fn, vp;
> identifier I;
> type T;
> expression C;
> @@
> -fn(...,int vp,...)
> -{
> -T I[C];
> -<+...
> -if(vp > C || ...){
> -...
> -}
> -...+>
> -}
> 
> 
> But it doesn?t work, give me a log:
> rule starting on line 1: position variables or mixed modifs interfere with comm_assoc isobool (bool (int vp > int C) || ...)

I will look into it, but you could try putting disable comm_assoc in 
between the initial @@, if you just want to see if your rule is working a 
little bit.

julia


> 
> can anyone tell why, thanks
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
> 

[Cocci] 答复: how to match this case?

[林嘉(程二 福州)]

Like this?  @disable comm_assoc@

After I added, the warning dissapper, but nothing matched output,  it seems skip the statement I intended....?

init_defs_builtins: /usr/share/coccinelle/standard.h
HANDLING: single2.c

-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?1?21? 11:35
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: [Cocci] how to match this case?



On Tue, 21 Jan 2014, ??(?? ??) wrote:

> Following is a code with bug
> 
> -----------------------------------
> void set_timeout_value(int iParam)
> {
> 	 int i;
> 	 byte b[100];
> 	 i = Geti();
> 	 if (i > 100 || i < 0){
> 	 	  return ;
> 	 	}	 	
> 	 b[i] = iParam;	/*  range of I is 0-100?so may overrun array* /
> 
> }
> -----------------------------------
> Now I write a patch to find it, like this
> 
> @@
> identifier fn, vp;
> identifier I;
> type T;
> expression C;
> @@
> -fn(...,int vp,...)
> -{
> -T I[C];
> -<+...
> -if(vp > C || ...){
> -...
> -}
> -...+>
> -}
> 
> 
> But it doesn?t work, give me a log:
> rule starting on line 1: position variables or mixed modifs interfere 
> with comm_assoc isobool (bool (int vp > int C) || ...)

I will look into it, but you could try putting disable comm_assoc in between the initial @@, if you just want to see if your rule is working a little bit.

julia


> 
> can anyone tell why, thanks
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
> 

[Cocci] 答复: how to match this case?

[林嘉(程二 福州)]

Sorry, I made a mistake , if the checkpoint is local variable, I should write as

-int vp;
-T I[C];
-<+...
-if(vp > C || ...){

-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?1?21? 11:35
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: [Cocci] how to match this case?



On Tue, 21 Jan 2014, ??(?? ??) wrote:

> Following is a code with bug
> 
> -----------------------------------
> void set_timeout_value(int iParam)
> {
> 	 int i;
> 	 byte b[100];
> 	 i = Geti();
> 	 if (i > 100 || i < 0){
> 	 	  return ;
> 	 	}	 	
> 	 b[i] = iParam;	/*  range of I is 0-100?so may overrun array* /
> 
> }
> -----------------------------------
> Now I write a patch to find it, like this
> 
> @@
> identifier fn, vp;
> identifier I;
> type T;
> expression C;
> @@
> -fn(...,int vp,...)
> -{
> -T I[C];
> -<+...
> -if(vp > C || ...){
> -...
> -}
> -...+>
> -}
> 
> 
> But it doesn?t work, give me a log:
> rule starting on line 1: position variables or mixed modifs interfere 
> with comm_assoc isobool (bool (int vp > int C) || ...)

I will look into it, but you could try putting disable comm_assoc in between the initial @@, if you just want to see if your rule is working a little bit.

julia


> 
> can anyone tell why, thanks
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
> 

[Cocci] 答复: how to match this case?

[Julia Lawall]

On Tue, 21 Jan 2014, ??(?? ??) wrote:

> Sorry, I made a mistake , if the checkpoint is local variable, I should write as
> 
> -int vp;
  ...
> -T I[C];

Also, you don't know if the array is declared before or after vp.  So you 
would need 

(
- int vp;
  ...
- T I[C];
|
- T I[C];
  ...
- int vp;
)

julia

> -<+...
> -if(vp > C || ...){
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
> ????: 2014?1?21? 11:35
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] how to match this case?
> 
> 
> 
> On Tue, 21 Jan 2014, ??(?? ??) wrote:
> 
> > Following is a code with bug
> > 
> > -----------------------------------
> > void set_timeout_value(int iParam)
> > {
> > 	 int i;
> > 	 byte b[100];
> > 	 i = Geti();
> > 	 if (i > 100 || i < 0){
> > 	 	  return ;
> > 	 	}	 	
> > 	 b[i] = iParam;	/*  range of I is 0-100?so may overrun array* /
> > 
> > }
> > -----------------------------------
> > Now I write a patch to find it, like this
> > 
> > @@
> > identifier fn, vp;
> > identifier I;
> > type T;
> > expression C;
> > @@
> > -fn(...,int vp,...)
> > -{
> > -T I[C];
> > -<+...
> > -if(vp > C || ...){
> > -...
> > -}
> > -...+>
> > -}
> > 
> > 
> > But it doesn?t work, give me a log:
> > rule starting on line 1: position variables or mixed modifs interfere 
> > with comm_assoc isobool (bool (int vp > int C) || ...)
> 
> I will look into it, but you could try putting disable comm_assoc in between the initial @@, if you just want to see if your rule is working a little bit.
> 
> julia
> 
> 
> > 
> > can anyone tell why, thanks
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> > 
> 

[Cocci] 答复: 答复: how to match this case?

[林嘉(程二 福州)]

Thanks for remind.
But to the "|" operator , I still don?t know how to use, for example, I wrote


@disable comm_assoc@
identifier fn, p;
identifier I;
type T,T1;
expression C;
@@
-fn(...)
-{
-int p;
-...
-T I[C];
| 
-T I[C];
-...
-int p;
-<+...
-if(p > C){
-...
-}
-...+>
-}



It reports: init_defs_builtins: /usr/share/coccinelle/standard.h
126 127
Fatal error: exception Failure("minus: parse error: 
 = File "os.cocci", line 12, column 0,  charpos = 126
    around = '|', whole content = |
")

-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?1?21? 14:32
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: ??: [Cocci] how to match this case?

On Tue, 21 Jan 2014, ??(?? ??) wrote:

> Sorry, I made a mistake , if the checkpoint is local variable, I 
> should write as
> 
> -int vp;
  ...
> -T I[C];

Also, you don't know if the array is declared before or after vp.  So you would need 

(
- int vp;
  ...
- T I[C];
|
- T I[C];
  ...
- int vp;
)

julia

> -<+...
> -if(vp > C || ...){
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> ????: 2014?1?21? 11:35
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] how to match this case?
> 
> 
> 
> On Tue, 21 Jan 2014, ??(?? ??) wrote:
> 
> > Following is a code with bug
> > 
> > -----------------------------------
> > void set_timeout_value(int iParam)
> > {
> > 	 int i;
> > 	 byte b[100];
> > 	 i = Geti();
> > 	 if (i > 100 || i < 0){
> > 	 	  return ;
> > 	 	}	 	
> > 	 b[i] = iParam;	/*  range of I is 0-100?so may overrun array* /
> > 
> > }
> > -----------------------------------
> > Now I write a patch to find it, like this
> > 
> > @@
> > identifier fn, vp;
> > identifier I;
> > type T;
> > expression C;
> > @@
> > -fn(...,int vp,...)
> > -{
> > -T I[C];
> > -<+...
> > -if(vp > C || ...){
> > -...
> > -}
> > -...+>
> > -}
> > 
> > 
> > But it doesn?t work, give me a log:
> > rule starting on line 1: position variables or mixed modifs 
> > interfere with comm_assoc isobool (bool (int vp > int C) || ...)
> 
> I will look into it, but you could try putting disable comm_assoc in between the initial @@, if you just want to see if your rule is working a little bit.
> 
> julia
> 
> 
> > 
> > can anyone tell why, thanks
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> > 
> 

[Cocci] 答复: 答复: 答复: how to match this case?

[林嘉(程二 福州)]

I see, I should add '()' if I use '|'


-----????-----
???: cocci-bounces at systeme.lip6.fr [mailto:cocci-bounces at systeme.lip6.fr] ?? ??(?? ??)
????: 2014?1?21? 14:44
???: Julia Lawall
??: cocci at systeme.lip6.fr
??: [Cocci] ??: ??: how to match this case?

Thanks for remind.
But to the "|" operator , I still don?t know how to use, for example, I wrote


@disable comm_assoc@
identifier fn, p;
identifier I;
type T,T1;
expression C;
@@
-fn(...)
-{
-int p;
-...
-T I[C];
| 
-T I[C];
-...
-int p;
-<+...
-if(p > C){
-...
-}
-...+>
-}



It reports: init_defs_builtins: /usr/share/coccinelle/standard.h
126 127
Fatal error: exception Failure("minus: parse error: 
 = File "os.cocci", line 12, column 0,  charpos = 126
    around = '|', whole content = |
")

-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr]
????: 2014?1?21? 14:32
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: ??: [Cocci] how to match this case?

On Tue, 21 Jan 2014, ??(?? ??) wrote:

> Sorry, I made a mistake , if the checkpoint is local variable, I 
> should write as
> 
> -int vp;
  ...
> -T I[C];

Also, you don't know if the array is declared before or after vp.  So you would need 

(
- int vp;
  ...
- T I[C];
|
- T I[C];
  ...
- int vp;
)

julia

> -<+...
> -if(vp > C || ...){
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> ????: 2014?1?21? 11:35
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] how to match this case?
> 
> 
> 
> On Tue, 21 Jan 2014, ??(?? ??) wrote:
> 
> > Following is a code with bug
> > 
> > -----------------------------------
> > void set_timeout_value(int iParam)
> > {
> > 	 int i;
> > 	 byte b[100];
> > 	 i = Geti();
> > 	 if (i > 100 || i < 0){
> > 	 	  return ;
> > 	 	}	 	
> > 	 b[i] = iParam;	/*  range of I is 0-100?so may overrun array* /
> > 
> > }
> > -----------------------------------
> > Now I write a patch to find it, like this
> > 
> > @@
> > identifier fn, vp;
> > identifier I;
> > type T;
> > expression C;
> > @@
> > -fn(...,int vp,...)
> > -{
> > -T I[C];
> > -<+...
> > -if(vp > C || ...){
> > -...
> > -}
> > -...+>
> > -}
> > 
> > 
> > But it doesn?t work, give me a log:
> > rule starting on line 1: position variables or mixed modifs 
> > interfere with comm_assoc isobool (bool (int vp > int C) || ...)
> 
> I will look into it, but you could try putting disable comm_assoc in between the initial @@, if you just want to see if your rule is working a little bit.
> 
> julia
> 
> 
> > 
> > can anyone tell why, thanks
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> > 
> 
_______________________________________________
Cocci mailing list
Cocci at systeme.lip6.fr
https://systeme.lip6.fr/mailman/listinfo/cocci

[Cocci] 答复: 答复: how to match this case?

[Julia Lawall]

On Tue, 21 Jan 2014, ??(?? ??) wrote:

> Thanks for remind.
> But to the "|" operator , I still don?t know how to use, for example, I wrote

You need the () at the beginning and end, like I wrote.

julia

> 
> 
> @disable comm_assoc@
> identifier fn, p;
> identifier I;
> type T,T1;
> expression C;
> @@
> -fn(...)
> -{
> -int p;
> -...
> -T I[C];
> | 
> -T I[C];
> -...
> -int p;
> -<+...
> -if(p > C){
> -...
> -}
> -...+>
> -}
> 
> 
> 
> It reports: init_defs_builtins: /usr/share/coccinelle/standard.h
> 126 127
> Fatal error: exception Failure("minus: parse error: 
>  = File "os.cocci", line 12, column 0,  charpos = 126
>     around = '|', whole content = |
> ")
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
> ????: 2014?1?21? 14:32
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: ??: [Cocci] how to match this case?
> 
> On Tue, 21 Jan 2014, ??(?? ??) wrote:
> 
> > Sorry, I made a mistake , if the checkpoint is local variable, I 
> > should write as
> > 
> > -int vp;
>   ...
> > -T I[C];
> 
> Also, you don't know if the array is declared before or after vp.  So you would need 
> 
> (
> - int vp;
>   ...
> - T I[C];
> |
> - T I[C];
>   ...
> - int vp;
> )
> 
> julia
> 
> > -<+...
> > -if(vp > C || ...){
> > 
> > -----????-----
> > ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> > ????: 2014?1?21? 11:35
> > ???: ??(?? ??)
> > ??: cocci at systeme.lip6.fr
> > ??: Re: [Cocci] how to match this case?
> > 
> > 
> > 
> > On Tue, 21 Jan 2014, ??(?? ??) wrote:
> > 
> > > Following is a code with bug
> > > 
> > > -----------------------------------
> > > void set_timeout_value(int iParam)
> > > {
> > > 	 int i;
> > > 	 byte b[100];
> > > 	 i = Geti();
> > > 	 if (i > 100 || i < 0){
> > > 	 	  return ;
> > > 	 	}	 	
> > > 	 b[i] = iParam;	/*  range of I is 0-100?so may overrun array* /
> > > 
> > > }
> > > -----------------------------------
> > > Now I write a patch to find it, like this
> > > 
> > > @@
> > > identifier fn, vp;
> > > identifier I;
> > > type T;
> > > expression C;
> > > @@
> > > -fn(...,int vp,...)
> > > -{
> > > -T I[C];
> > > -<+...
> > > -if(vp > C || ...){
> > > -...
> > > -}
> > > -...+>
> > > -}
> > > 
> > > 
> > > But it doesn?t work, give me a log:
> > > rule starting on line 1: position variables or mixed modifs 
> > > interfere with comm_assoc isobool (bool (int vp > int C) || ...)
> > 
> > I will look into it, but you could try putting disable comm_assoc in between the initial @@, if you just want to see if your rule is working a little bit.
> > 
> > julia
> > 
> > 
> > > 
> > > can anyone tell why, thanks
> > > _______________________________________________
> > > Cocci mailing list
> > > Cocci at systeme.lip6.fr
> > > https://systeme.lip6.fr/mailman/listinfo/cocci
> > > 
> > 
> 

[Cocci] 答复: 答复: 答复: how to match this case?

[Julia Lawall]

On Tue, 21 Jan 2014, ??(?? ??) wrote:

> I see, I should add '()' if I use '|'

Yes :)

julia

> 
> 
> -----????-----
> ???: cocci-bounces at systeme.lip6.fr [mailto:cocci-bounces at systeme.lip6.fr] ?? ??(?? ??)
> ????: 2014?1?21? 14:44
> ???: Julia Lawall
> ??: cocci at systeme.lip6.fr
> ??: [Cocci] ??: ??: how to match this case?
> 
> Thanks for remind.
> But to the "|" operator , I still don?t know how to use, for example, I wrote
> 
> 
> @disable comm_assoc@
> identifier fn, p;
> identifier I;
> type T,T1;
> expression C;
> @@
> -fn(...)
> -{
> -int p;
> -...
> -T I[C];
> | 
> -T I[C];
> -...
> -int p;
> -<+...
> -if(p > C){
> -...
> -}
> -...+>
> -}
> 
> 
> 
> It reports: init_defs_builtins: /usr/share/coccinelle/standard.h
> 126 127
> Fatal error: exception Failure("minus: parse error: 
>  = File "os.cocci", line 12, column 0,  charpos = 126
>     around = '|', whole content = |
> ")
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> ????: 2014?1?21? 14:32
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: ??: [Cocci] how to match this case?
> 
> On Tue, 21 Jan 2014, ??(?? ??) wrote:
> 
> > Sorry, I made a mistake , if the checkpoint is local variable, I 
> > should write as
> > 
> > -int vp;
>   ...
> > -T I[C];
> 
> Also, you don't know if the array is declared before or after vp.  So you would need 
> 
> (
> - int vp;
>   ...
> - T I[C];
> |
> - T I[C];
>   ...
> - int vp;
> )
> 
> julia
> 
> > -<+...
> > -if(vp > C || ...){
> > 
> > -----????-----
> > ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> > ????: 2014?1?21? 11:35
> > ???: ??(?? ??)
> > ??: cocci at systeme.lip6.fr
> > ??: Re: [Cocci] how to match this case?
> > 
> > 
> > 
> > On Tue, 21 Jan 2014, ??(?? ??) wrote:
> > 
> > > Following is a code with bug
> > > 
> > > -----------------------------------
> > > void set_timeout_value(int iParam)
> > > {
> > > 	 int i;
> > > 	 byte b[100];
> > > 	 i = Geti();
> > > 	 if (i > 100 || i < 0){
> > > 	 	  return ;
> > > 	 	}	 	
> > > 	 b[i] = iParam;	/*  range of I is 0-100?so may overrun array* /
> > > 
> > > }
> > > -----------------------------------
> > > Now I write a patch to find it, like this
> > > 
> > > @@
> > > identifier fn, vp;
> > > identifier I;
> > > type T;
> > > expression C;
> > > @@
> > > -fn(...,int vp,...)
> > > -{
> > > -T I[C];
> > > -<+...
> > > -if(vp > C || ...){
> > > -...
> > > -}
> > > -...+>
> > > -}
> > > 
> > > 
> > > But it doesn?t work, give me a log:
> > > rule starting on line 1: position variables or mixed modifs 
> > > interfere with comm_assoc isobool (bool (int vp > int C) || ...)
> > 
> > I will look into it, but you could try putting disable comm_assoc in between the initial @@, if you just want to see if your rule is working a little bit.
> > 
> > julia
> > 
> > 
> > > 
> > > can anyone tell why, thanks
> > > _______________________________________________
> > > Cocci mailing list
> > > Cocci at systeme.lip6.fr
> > > https://systeme.lip6.fr/mailman/listinfo/cocci
> > > 
> > 
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
> 

[Cocci] infinitly loop in spatch?

[林嘉(程二 福州)]

When I use spatch to scan some file, I found it may loop sometimes, did anyone encounter too?

If using -debug,  it show pause at 'bind' step 

-----------------------------------------------------------------------
HANDLING: /root/windows/ok_file.c
-----------------------------------------------------------------------
let's go
-----------------------------------------------------------------------
-----------------------------------------------------------------------
-----------------------------------------------------------------------
rule starting on line 1 = 
-----------------------------------------------------------------------
dependencies for rule rule starting on line 1 satisfied:
binding in = []
binding relevant in = []
-----------------------------------------------------------------------
proto for rule starting on line 1 = 
-----------------------------------------------------------------------
dependencies for rule proto for rule starting on line 1 not satisfied:
rule starting on line 1 not satisfied
binding in environment = []
-----------------------------------------------------------------------
Finished
-----------------------------------------------------------------------
HANDLING: /root/loop_file.c
-----------------------------------------------------------------------
let's go
-----------------------------------------------------------------------
-----------------------------------------------------------------------
-----------------------------------------------------------------------
rule starting on line 1 = 
-----------------------------------------------------------------------
dependencies for rule rule starting on line 1 satisfied:
binding in = []
binding relevant in = []  <---  step here, then nothing more output.....     :(

[Cocci] infinitly loop in spatch?

[Julia Lawall]

On Tue, 21 Jan 2014, ??(?? ??) wrote:

> When I use spatch to scan some file, I found it may loop sometimes, did anyone encounter too?

It can run for a very long time, especially when there are nested loops,
and these loops contain many conditionals.  Basically it is trying all of
the control-flow paths.  One option is to tell it to ignore loops, with
the --no-loops option.  This can be unsafe, because it could miss some
matches from the bottom of the loop back to the top, but that might not be
relevant to your pattern.  Another option is to put a timeout --timeout
120, for example (120 seconds).  If it exceeds that time, it will do
nothing for the file.

If you want me to look at the problem in more detail, please send the
semantic patch and some problematic code.

julia


>
> If using -debug,  it show pause at 'bind' step
>
> -----------------------------------------------------------------------
> HANDLING: /root/windows/ok_file.c
> -----------------------------------------------------------------------
> let's go
> -----------------------------------------------------------------------
> -----------------------------------------------------------------------
> -----------------------------------------------------------------------
> rule starting on line 1 =
> -----------------------------------------------------------------------
> dependencies for rule rule starting on line 1 satisfied:
> binding in = []
> binding relevant in = []
> -----------------------------------------------------------------------
> proto for rule starting on line 1 =
> -----------------------------------------------------------------------
> dependencies for rule proto for rule starting on line 1 not satisfied:
> rule starting on line 1 not satisfied
> binding in environment = []
> -----------------------------------------------------------------------
> Finished
> -----------------------------------------------------------------------
> HANDLING: /root/loop_file.c
> -----------------------------------------------------------------------
> let's go
> -----------------------------------------------------------------------
> -----------------------------------------------------------------------
> -----------------------------------------------------------------------
> rule starting on line 1 =
> -----------------------------------------------------------------------
> dependencies for rule rule starting on line 1 satisfied:
> binding in = []
> binding relevant in = []  <---  step here, then nothing more output.....     :(
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
>

[Cocci] how to match this case?

[Julia Lawall]

On Tue, 21 Jan 2014, ??(?? ??) wrote:

> Following is a code with bug
>
> -----------------------------------
> void set_timeout_value(int iParam)
> {
> 	 int i;
> 	 byte b[100];
> 	 i = Geti();
> 	 if (i > 100 || i < 0){
> 	 	  return ;
> 	 	}
> 	 b[i] = iParam;	/*  range of I is 0-100?so may overrun array* /
>
> }
> -----------------------------------
> Now I write a patch to find it, like this
>
> @@
> identifier fn, vp;
> identifier I;
> type T;
> expression C;
> @@
> -fn(...,int vp,...)
> -{
> -T I[C];
> -<+...
> -if(vp > C || ...){
> -...
> -}
> -...+>
> -}
>
>
> But it doesn?t work, give me a log:
> rule starting on line 1: position variables or mixed modifs interfere with comm_assoc isobool (bool (int vp > int C) || ...)
>
> can anyone tell why, thanks

I tried the above code exactly, and I didn't get the problem.

I also tried removing int i; and changing i > 100 to iParam > 100, and
then it was able to remove the function definition.

julia

[Cocci] how to write such matching case?

[林嘉(程二 福州)]

In my project, it's required that if function A is called, and function B must be called following, just like this

Foo(...)
{
   If ... {
       A(....);
       ...
       B(....);
   } 
}

Or

Foo1(...)
{
        A(....);
        B(...);           
 }


How to write a patch to find the missing case? Such as 

Foo(...)
{
   If ... {
       A(....);
       ...
   } 
}

Or

Foo1(...)
{
       A(....);
       ...
}

[Cocci] how to write such matching case?

[Julia Lawall]

On Sun, 26 Jan 2014, ??(?? ??) wrote:

> In my project, it's required that if function A is called, and function B must be called following, just like this
>
> Foo(...)
> {
>    If ... {
>        A(....);
>        ...
>        B(....);
>    }
> }
>
> Or
>
> Foo1(...)
> {
>         A(....);
>         B(...);
>  }
>
>
> How to write a patch to find the missing case? Such as
>
> Foo(...)
> {
>    If ... {
>        A(....);
>        ...
>    }
> }
>
> Or
>
> Foo1(...)
> {
>        A(....);
>        ...
> }

A(...)
... When != B(...)

julia

[Cocci] 答复: how to write such matching case?

[林嘉(程二 福州)]

In SMPL,   'When != something'  means 'something' doesn?t exist ?



-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?1?26? 19:19
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: [Cocci] how to write such matching case?

On Sun, 26 Jan 2014, ??(?? ??) wrote:

> In my project, it's required that if function A is called, and function B must be called following, just like this
>
> Foo(...)
> {
>    If ... {
>        A(....);
>        ...
>        B(....);
>    }
> }
>
> Or
>
> Foo1(...)
> {
>         A(....);
>         B(...);
>  }
>
>
> How to write a patch to find the missing case? Such as
>
> Foo(...)
> {
>    If ... {
>        A(....);
>        ...
>    }
> }
>
> Or
>
> Foo1(...)
> {
>        A(....);
>        ...
> }

A(...)
... When != B(...)

julia

[Cocci] 答复: how to write such matching case?

[林嘉(程二 福州)]

I wrote this 

@@
@@
cli_printf(...)
... when != cli_client_execmd_fail(...)


it reports:
Fatal error: exception Failure("False should not be in the final result.  Perhaps your rule doesn't contain any +/-/* code, or you have a failed dependency.")


But i found there is not 'minus' in an example.cocci either....


//
//  Add missing pci_dev_put
//
// Target: Linux
// Copyright:  2012 - LIP6/INRIA
// License:  Licensed under ISC. See LICENSE or http://www.isc.org/software/license
// Author: Julia Lawall <Julia.Lawall@lip6.fr>
// URL: http://coccinelle.lip6.fr/ 
// URL: http://coccinellery.org/ 

@@
local idexpression x;
expression e;
@@

*x = pci_get_slot(...)
... when != true x == NULL
    when != pci_dev_put(x)
    when != e = x
    when != if (x != NULL) {<+... pci_dev_put(x); ...+>}
*return ...;





-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?1?26? 19:19
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: [Cocci] how to write such matching case?

On Sun, 26 Jan 2014, ??(?? ??) wrote:

> In my project, it's required that if function A is called, and function B must be called following, just like this
>
> Foo(...)
> {
>    If ... {
>        A(....);
>        ...
>        B(....);
>    }
> }
>
> Or
>
> Foo1(...)
> {
>         A(....);
>         B(...);
>  }
>
>
> How to write a patch to find the missing case? Such as
>
> Foo(...)
> {
>    If ... {
>        A(....);
>        ...
>    }
> }
>
> Or
>
> Foo1(...)
> {
>        A(....);
>        ...
> }

A(...)
... When != B(...)

julia

[Cocci] 答复: how to write such matching case?

[Julia Lawall]

On Mon, 27 Jan 2014, ??(?? ??) wrote:

> I wrote this 
> 
> @@
> @@
> cli_printf(...)
> ... when != cli_client_execmd_fail(...)
> 
> 
> it reports:
> Fatal error: exception Failure("False should not be in the final result.  Perhaps your rule doesn't contain any +/-/* code, or you have a failed dependency.")
> 
> 
> But i found there is not 'minus' in an example.cocci either....

There is a * (indicating a line of interest).  You can use that.  If you 
use that, you will get an exists semantics for the ... (there exists an 
execution path that satisfies the property).  Otherwise, you get a foral 
semantics (all execution paths have to satisfy the property).

julia

> 
> 
> //
> //  Add missing pci_dev_put
> //
> // Target: Linux
> // Copyright:  2012 - LIP6/INRIA
> // License:  Licensed under ISC. See LICENSE or http://www.isc.org/software/license
> // Author: Julia Lawall <Julia.Lawall@lip6.fr>
> // URL: http://coccinelle.lip6.fr/ 
> // URL: http://coccinellery.org/ 
> 
> @@
> local idexpression x;
> expression e;
> @@
> 
> *x = pci_get_slot(...)
> ... when != true x == NULL
>     when != pci_dev_put(x)
>     when != e = x
>     when != if (x != NULL) {<+... pci_dev_put(x); ...+>}
> *return ...;
> 
> 
> 
> 
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
> ????: 2014?1?26? 19:19
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] how to write such matching case?
> 
> On Sun, 26 Jan 2014, ??(?? ??) wrote:
> 
> > In my project, it's required that if function A is called, and function B must be called following, just like this
> >
> > Foo(...)
> > {
> >    If ... {
> >        A(....);
> >        ...
> >        B(....);
> >    }
> > }
> >
> > Or
> >
> > Foo1(...)
> > {
> >         A(....);
> >         B(...);
> >  }
> >
> >
> > How to write a patch to find the missing case? Such as
> >
> > Foo(...)
> > {
> >    If ... {
> >        A(....);
> >        ...
> >    }
> > }
> >
> > Or
> >
> > Foo1(...)
> > {
> >        A(....);
> >        ...
> > }
> 
> A(...)
> ... When != B(...)
> 
> julia
> 

[Cocci] 答复: how to write such matching case?

[Julia Lawall]

On Mon, 27 Jan 2014, ??(?? ??) wrote:

> In SMPL,   'When != something'  means 'something' doesn?t exist ?

Yes, within the execution path(s) matched by ...

julia

> 
> 
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
> ????: 2014?1?26? 19:19
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] how to write such matching case?
> 
> On Sun, 26 Jan 2014, ??(?? ??) wrote:
> 
> > In my project, it's required that if function A is called, and function B must be called following, just like this
> >
> > Foo(...)
> > {
> >    If ... {
> >        A(....);
> >        ...
> >        B(....);
> >    }
> > }
> >
> > Or
> >
> > Foo1(...)
> > {
> >         A(....);
> >         B(...);
> >  }
> >
> >
> > How to write a patch to find the missing case? Such as
> >
> > Foo(...)
> > {
> >    If ... {
> >        A(....);
> >        ...
> >    }
> > }
> >
> > Or
> >
> > Foo1(...)
> > {
> >        A(....);
> >        ...
> > }
> 
> A(...)
> ... When != B(...)
> 
> julia
> 

[Cocci] 答复: 答复: how to write such matching case?

[林嘉(程二 福州)]

Thanks for explain, but if I want to write a simple match, that is function foo must be called in main()
e.g.

int main(...)
{
    ...
    Foo(...);
    ...
}

I write it as

@@
@@
main(...) {
  <...
    when != reg_process_exit_action(...)
  ...>
}

Reports: 53 57
Fatal error: exception Failure("minus: parse error: 
 = File "rgos.cocci", line 6, column 2,  charpos = 53
    around = '...>', whole content =   ...>
")

What does it mean? 

By the way, I still don?t know when I should use '<'and'>' and when need not?

-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?1?27? 20:34
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: ??: [Cocci] how to write such matching case?

On Mon, 27 Jan 2014, ??(?? ??) wrote:

> I wrote this
> 
> @@
> @@
> cli_printf(...)
> ... when != cli_client_execmd_fail(...)
> 
> 
> it reports:
> Fatal error: exception Failure("False should not be in the final 
> result.  Perhaps your rule doesn't contain any +/-/* code, or you have 
> a failed dependency.")
> 
> 
> But i found there is not 'minus' in an example.cocci either....

There is a * (indicating a line of interest).  You can use that.  If you use that, you will get an exists semantics for the ... (there exists an execution path that satisfies the property).  Otherwise, you get a foral semantics (all execution paths have to satisfy the property).

julia

> 
> 
> //
> //  Add missing pci_dev_put
> //
> // Target: Linux
> // Copyright:  2012 - LIP6/INRIA
> // License:  Licensed under ISC. See LICENSE or 
> http://www.isc.org/software/license
> // Author: Julia Lawall <Julia.Lawall@lip6.fr> // URL: 
> http://coccinelle.lip6.fr/ // URL: http://coccinellery.org/
> 
> @@
> local idexpression x;
> expression e;
> @@
> 
> *x = pci_get_slot(...)
> ... when != true x == NULL
>     when != pci_dev_put(x)
>     when != e = x
>     when != if (x != NULL) {<+... pci_dev_put(x); ...+>} *return ...;
> 
> 
> 
> 
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> ????: 2014?1?26? 19:19
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] how to write such matching case?
> 
> On Sun, 26 Jan 2014, ??(?? ??) wrote:
> 
> > In my project, it's required that if function A is called, and 
> > function B must be called following, just like this
> >
> > Foo(...)
> > {
> >    If ... {
> >        A(....);
> >        ...
> >        B(....);
> >    }
> > }
> >
> > Or
> >
> > Foo1(...)
> > {
> >         A(....);
> >         B(...);
> >  }
> >
> >
> > How to write a patch to find the missing case? Such as
> >
> > Foo(...)
> > {
> >    If ... {
> >        A(....);
> >        ...
> >    }
> > }
> >
> > Or
> >
> > Foo1(...)
> > {
> >        A(....);
> >        ...
> > }
> 
> A(...)
> ... When != B(...)
> 
> julia
> 

[Cocci] 答复: 答复: how to write such matching case?

[Peter Senna Tschudin]

You can try:

@@
@@
* main () {
   ...
   when != Foo(...)
}

This semantic patch will match functions named main that do not
contain call to function Foo;

On Mon, Jan 27, 2014 at 2:55 PM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> Thanks for explain, but if I want to write a simple match, that is function foo must be called in main()
> e.g.
>
> int main(...)
> {
>     ...
>     Foo(...);
>     ...
> }
>
> I write it as
>
> @@
> @@
> main(...) {
>   <...
>     when != reg_process_exit_action(...)
>   ...>
> }
>
> Reports: 53 57
> Fatal error: exception Failure("minus: parse error:
>  = File "rgos.cocci", line 6, column 2,  charpos = 53
>     around = '...>', whole content =   ...>
> ")
>
> What does it mean?
>
> By the way, I still don?t know when I should use '<'and'>' and when need not?
>
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> ????: 2014?1?27? 20:34
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: ??: [Cocci] how to write such matching case?
>
> On Mon, 27 Jan 2014, ??(?? ??) wrote:
>
>> I wrote this
>>
>> @@
>> @@
>> cli_printf(...)
>> ... when != cli_client_execmd_fail(...)
>>
>>
>> it reports:
>> Fatal error: exception Failure("False should not be in the final
>> result.  Perhaps your rule doesn't contain any +/-/* code, or you have
>> a failed dependency.")
>>
>>
>> But i found there is not 'minus' in an example.cocci either....
>
> There is a * (indicating a line of interest).  You can use that.  If you use that, you will get an exists semantics for the ... (there exists an execution path that satisfies the property).  Otherwise, you get a foral semantics (all execution paths have to satisfy the property).
>
> julia
>
>>
>>
>> //
>> //  Add missing pci_dev_put
>> //
>> // Target: Linux
>> // Copyright:  2012 - LIP6/INRIA
>> // License:  Licensed under ISC. See LICENSE or
>> http://www.isc.org/software/license
>> // Author: Julia Lawall <Julia.Lawall@lip6.fr> // URL:
>> http://coccinelle.lip6.fr/ // URL: http://coccinellery.org/
>>
>> @@
>> local idexpression x;
>> expression e;
>> @@
>>
>> *x = pci_get_slot(...)
>> ... when != true x == NULL
>>     when != pci_dev_put(x)
>>     when != e = x
>>     when != if (x != NULL) {<+... pci_dev_put(x); ...+>} *return ...;
>>
>>
>>
>>
>>
>> -----????-----
>> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
>> ????: 2014?1?26? 19:19
>> ???: ??(?? ??)
>> ??: cocci at systeme.lip6.fr
>> ??: Re: [Cocci] how to write such matching case?
>>
>> On Sun, 26 Jan 2014, ??(?? ??) wrote:
>>
>> > In my project, it's required that if function A is called, and
>> > function B must be called following, just like this
>> >
>> > Foo(...)
>> > {
>> >    If ... {
>> >        A(....);
>> >        ...
>> >        B(....);
>> >    }
>> > }
>> >
>> > Or
>> >
>> > Foo1(...)
>> > {
>> >         A(....);
>> >         B(...);
>> >  }
>> >
>> >
>> > How to write a patch to find the missing case? Such as
>> >
>> > Foo(...)
>> > {
>> >    If ... {
>> >        A(....);
>> >        ...
>> >    }
>> > }
>> >
>> > Or
>> >
>> > Foo1(...)
>> > {
>> >        A(....);
>> >        ...
>> > }
>>
>> A(...)
>> ... When != B(...)
>>
>> julia
>>
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci



-- 
Peter

[Cocci] 答复: 答复: how to write such matching case?

[Julia Lawall]

On Mon, 27 Jan 2014, Peter Senna Tschudin wrote:

> You can try:
> 
> @@
> @@
> * main () {
>    ...
>    when != Foo(...)
> }

Normally the first when should be on the same line as the ...

julia

> This semantic patch will match functions named main that do not
> contain call to function Foo;
> 
> On Mon, Jan 27, 2014 at 2:55 PM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > Thanks for explain, but if I want to write a simple match, that is function foo must be called in main()
> > e.g.
> >
> > int main(...)
> > {
> >     ...
> >     Foo(...);
> >     ...
> > }
> >
> > I write it as
> >
> > @@
> > @@
> > main(...) {
> >   <...
> >     when != reg_process_exit_action(...)
> >   ...>
> > }
> >
> > Reports: 53 57
> > Fatal error: exception Failure("minus: parse error:
> >  = File "rgos.cocci", line 6, column 2,  charpos = 53
> >     around = '...>', whole content =   ...>
> > ")
> >
> > What does it mean?
> >
> > By the way, I still don?t know when I should use '<'and'>' and when need not?
> >
> > -----????-----
> > ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> > ????: 2014?1?27? 20:34
> > ???: ??(?? ??)
> > ??: cocci at systeme.lip6.fr
> > ??: Re: ??: [Cocci] how to write such matching case?
> >
> > On Mon, 27 Jan 2014, ??(?? ??) wrote:
> >
> >> I wrote this
> >>
> >> @@
> >> @@
> >> cli_printf(...)
> >> ... when != cli_client_execmd_fail(...)
> >>
> >>
> >> it reports:
> >> Fatal error: exception Failure("False should not be in the final
> >> result.  Perhaps your rule doesn't contain any +/-/* code, or you have
> >> a failed dependency.")
> >>
> >>
> >> But i found there is not 'minus' in an example.cocci either....
> >
> > There is a * (indicating a line of interest).  You can use that.  If you use that, you will get an exists semantics for the ... (there exists an execution path that satisfies the property).  Otherwise, you get a foral semantics (all execution paths have to satisfy the property).
> >
> > julia
> >
> >>
> >>
> >> //
> >> //  Add missing pci_dev_put
> >> //
> >> // Target: Linux
> >> // Copyright:  2012 - LIP6/INRIA
> >> // License:  Licensed under ISC. See LICENSE or
> >> http://www.isc.org/software/license
> >> // Author: Julia Lawall <Julia.Lawall@lip6.fr> // URL:
> >> http://coccinelle.lip6.fr/ // URL: http://coccinellery.org/
> >>
> >> @@
> >> local idexpression x;
> >> expression e;
> >> @@
> >>
> >> *x = pci_get_slot(...)
> >> ... when != true x == NULL
> >>     when != pci_dev_put(x)
> >>     when != e = x
> >>     when != if (x != NULL) {<+... pci_dev_put(x); ...+>} *return ...;
> >>
> >>
> >>
> >>
> >>
> >> -----????-----
> >> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> >> ????: 2014?1?26? 19:19
> >> ???: ??(?? ??)
> >> ??: cocci at systeme.lip6.fr
> >> ??: Re: [Cocci] how to write such matching case?
> >>
> >> On Sun, 26 Jan 2014, ??(?? ??) wrote:
> >>
> >> > In my project, it's required that if function A is called, and
> >> > function B must be called following, just like this
> >> >
> >> > Foo(...)
> >> > {
> >> >    If ... {
> >> >        A(....);
> >> >        ...
> >> >        B(....);
> >> >    }
> >> > }
> >> >
> >> > Or
> >> >
> >> > Foo1(...)
> >> > {
> >> >         A(....);
> >> >         B(...);
> >> >  }
> >> >
> >> >
> >> > How to write a patch to find the missing case? Such as
> >> >
> >> > Foo(...)
> >> > {
> >> >    If ... {
> >> >        A(....);
> >> >        ...
> >> >    }
> >> > }
> >> >
> >> > Or
> >> >
> >> > Foo1(...)
> >> > {
> >> >        A(....);
> >> >        ...
> >> > }
> >>
> >> A(...)
> >> ... When != B(...)
> >>
> >> julia
> >>
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> 
> 
> 
> -- 
> Peter
> 

[Cocci] 答复: 答复: 答复: how to write such matching case?

[林嘉(程二 福州)]

Just like this?
@@
@@
* main(...){
   ...when != reg_process_exit_action(...)
}


I tried, but it seems does not work...

I make the function reg...() called in main(), but it still report main()

-----????-----
???: Peter Senna Tschudin [mailto:peter.senna at gmail.com] 
????: 2014?1?27? 22:12
???: ??(?? ??)
??: Julia Lawall; cocci at systeme.lip6.fr
??: Re: [Cocci] ??: ??: how to write such matching case?

You can try:

@@
@@
* main () {
   ...
   when != Foo(...)
}

This semantic patch will match functions named main that do not contain call to function Foo;

On Mon, Jan 27, 2014 at 2:55 PM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> Thanks for explain, but if I want to write a simple match, that is 
> function foo must be called in main() e.g.
>
> int main(...)
> {
>     ...
>     Foo(...);
>     ...
> }
>
> I write it as
>
> @@
> @@
> main(...) {
>   <...
>     when != reg_process_exit_action(...)
>   ...>
> }
>
> Reports: 53 57
> Fatal error: exception Failure("minus: parse error:
>  = File "rgos.cocci", line 6, column 2,  charpos = 53
>     around = '...>', whole content =   ...>
> ")
>
> What does it mean?
>
> By the way, I still don?t know when I should use '<'and'>' and when need not?
>
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> ????: 2014?1?27? 20:34
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: ??: [Cocci] how to write such matching case?
>
> On Mon, 27 Jan 2014, ??(?? ??) wrote:
>
>> I wrote this
>>
>> @@
>> @@
>> cli_printf(...)
>> ... when != cli_client_execmd_fail(...)
>>
>>
>> it reports:
>> Fatal error: exception Failure("False should not be in the final 
>> result.  Perhaps your rule doesn't contain any +/-/* code, or you 
>> have a failed dependency.")
>>
>>
>> But i found there is not 'minus' in an example.cocci either....
>
> There is a * (indicating a line of interest).  You can use that.  If you use that, you will get an exists semantics for the ... (there exists an execution path that satisfies the property).  Otherwise, you get a foral semantics (all execution paths have to satisfy the property).
>
> julia
>
>>
>>
>> //
>> //  Add missing pci_dev_put
>> //
>> // Target: Linux
>> // Copyright:  2012 - LIP6/INRIA
>> // License:  Licensed under ISC. See LICENSE or 
>> http://www.isc.org/software/license
>> // Author: Julia Lawall <Julia.Lawall@lip6.fr> // URL:
>> http://coccinelle.lip6.fr/ // URL: http://coccinellery.org/
>>
>> @@
>> local idexpression x;
>> expression e;
>> @@
>>
>> *x = pci_get_slot(...)
>> ... when != true x == NULL
>>     when != pci_dev_put(x)
>>     when != e = x
>>     when != if (x != NULL) {<+... pci_dev_put(x); ...+>} *return ...;
>>
>>
>>
>>
>>
>> -----????-----
>> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
>> ????: 2014?1?26? 19:19
>> ???: ??(?? ??)
>> ??: cocci at systeme.lip6.fr
>> ??: Re: [Cocci] how to write such matching case?
>>
>> On Sun, 26 Jan 2014, ??(?? ??) wrote:
>>
>> > In my project, it's required that if function A is called, and 
>> > function B must be called following, just like this
>> >
>> > Foo(...)
>> > {
>> >    If ... {
>> >        A(....);
>> >        ...
>> >        B(....);
>> >    }
>> > }
>> >
>> > Or
>> >
>> > Foo1(...)
>> > {
>> >         A(....);
>> >         B(...);
>> >  }
>> >
>> >
>> > How to write a patch to find the missing case? Such as
>> >
>> > Foo(...)
>> > {
>> >    If ... {
>> >        A(....);
>> >        ...
>> >    }
>> > }
>> >
>> > Or
>> >
>> > Foo1(...)
>> > {
>> >        A(....);
>> >        ...
>> > }
>>
>> A(...)
>> ... When != B(...)
>>
>> julia
>>
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci



--
Peter

[Cocci] 答复: 答复: 答复: how to write such matching case?

[Julia Lawall]

On Tue, 28 Jan 2014, ??(?? ??) wrote:

> Just like this?
> @@
> @@
> * main(...){
>    ...when != reg_process_exit_action(...)
> }
> 
> 
> I tried, but it seems does not work...
> 
> I make the function reg...() called in main(), but it still report main()

Probably because there exists a control flow path in which it is not 
called.  If you want the forall semantics, put forall in between the 
initial @@

julia


> 
> -----????-----
> ???: Peter Senna Tschudin [mailto:peter.senna at gmail.com] 
> ????: 2014?1?27? 22:12
> ???: ??(?? ??)
> ??: Julia Lawall; cocci at systeme.lip6.fr
> ??: Re: [Cocci] ??: ??: how to write such matching case?
> 
> You can try:
> 
> @@
> @@
> * main () {
>    ...
>    when != Foo(...)
> }
> 
> This semantic patch will match functions named main that do not contain call to function Foo;
> 
> On Mon, Jan 27, 2014 at 2:55 PM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > Thanks for explain, but if I want to write a simple match, that is 
> > function foo must be called in main() e.g.
> >
> > int main(...)
> > {
> >     ...
> >     Foo(...);
> >     ...
> > }
> >
> > I write it as
> >
> > @@
> > @@
> > main(...) {
> >   <...
> >     when != reg_process_exit_action(...)
> >   ...>
> > }
> >
> > Reports: 53 57
> > Fatal error: exception Failure("minus: parse error:
> >  = File "rgos.cocci", line 6, column 2,  charpos = 53
> >     around = '...>', whole content =   ...>
> > ")
> >
> > What does it mean?
> >
> > By the way, I still don?t know when I should use '<'and'>' and when need not?
> >
> > -----????-----
> > ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> > ????: 2014?1?27? 20:34
> > ???: ??(?? ??)
> > ??: cocci at systeme.lip6.fr
> > ??: Re: ??: [Cocci] how to write such matching case?
> >
> > On Mon, 27 Jan 2014, ??(?? ??) wrote:
> >
> >> I wrote this
> >>
> >> @@
> >> @@
> >> cli_printf(...)
> >> ... when != cli_client_execmd_fail(...)
> >>
> >>
> >> it reports:
> >> Fatal error: exception Failure("False should not be in the final 
> >> result.  Perhaps your rule doesn't contain any +/-/* code, or you 
> >> have a failed dependency.")
> >>
> >>
> >> But i found there is not 'minus' in an example.cocci either....
> >
> > There is a * (indicating a line of interest).  You can use that.  If you use that, you will get an exists semantics for the ... (there exists an execution path that satisfies the property).  Otherwise, you get a foral semantics (all execution paths have to satisfy the property).
> >
> > julia
> >
> >>
> >>
> >> //
> >> //  Add missing pci_dev_put
> >> //
> >> // Target: Linux
> >> // Copyright:  2012 - LIP6/INRIA
> >> // License:  Licensed under ISC. See LICENSE or 
> >> http://www.isc.org/software/license
> >> // Author: Julia Lawall <Julia.Lawall@lip6.fr> // URL:
> >> http://coccinelle.lip6.fr/ // URL: http://coccinellery.org/
> >>
> >> @@
> >> local idexpression x;
> >> expression e;
> >> @@
> >>
> >> *x = pci_get_slot(...)
> >> ... when != true x == NULL
> >>     when != pci_dev_put(x)
> >>     when != e = x
> >>     when != if (x != NULL) {<+... pci_dev_put(x); ...+>} *return ...;
> >>
> >>
> >>
> >>
> >>
> >> -----????-----
> >> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> >> ????: 2014?1?26? 19:19
> >> ???: ??(?? ??)
> >> ??: cocci at systeme.lip6.fr
> >> ??: Re: [Cocci] how to write such matching case?
> >>
> >> On Sun, 26 Jan 2014, ??(?? ??) wrote:
> >>
> >> > In my project, it's required that if function A is called, and 
> >> > function B must be called following, just like this
> >> >
> >> > Foo(...)
> >> > {
> >> >    If ... {
> >> >        A(....);
> >> >        ...
> >> >        B(....);
> >> >    }
> >> > }
> >> >
> >> > Or
> >> >
> >> > Foo1(...)
> >> > {
> >> >         A(....);
> >> >         B(...);
> >> >  }
> >> >
> >> >
> >> > How to write a patch to find the missing case? Such as
> >> >
> >> > Foo(...)
> >> > {
> >> >    If ... {
> >> >        A(....);
> >> >        ...
> >> >    }
> >> > }
> >> >
> >> > Or
> >> >
> >> > Foo1(...)
> >> > {
> >> >        A(....);
> >> >        ...
> >> > }
> >>
> >> A(...)
> >> ... When != B(...)
> >>
> >> julia
> >>
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> 
> 
> 
> --
> Peter
> 

[Cocci] 答复: 答复: 答复: 答复: how to write such matching case?

[林嘉(程二 福州)]

You mean this?

@forall@
@@
* main(){
   ...when != reg_process_exit_action(...)
}

And my 'main' function is straightforward without flow control

int main(void)
{
    struct AAAA foo;
    struct AAA foo2;
    struct BBB foo3, *p;
    printf("%d\n", sizeof(((struct AAAA*)0)->aa));
    printf("%d %d %d\n", foo.aa.c, foo2.c, foo3.value); 
    memset(p, 0, sizeof(*p));    
    reg_process_exit_action(0, RESTART);    
    return 0;
}



-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?1?28? 15:25
???: ??(?? ??)
??: Peter Senna Tschudin; cocci at systeme.lip6.fr
??: Re: ??: [Cocci] ??: ??: how to write such matching case?



On Tue, 28 Jan 2014, ??(?? ??) wrote:

> Just like this?
> @@
> @@
> * main(...){
>    ...when != reg_process_exit_action(...) }
> 
> 
> I tried, but it seems does not work...
> 
> I make the function reg...() called in main(), but it still report 
> main()

Probably because there exists a control flow path in which it is not called.  If you want the forall semantics, put forall in between the initial @@

julia


> 
> -----????-----
> ???: Peter Senna Tschudin [mailto:peter.senna at gmail.com]
> ????: 2014?1?27? 22:12
> ???: ??(?? ??)
> ??: Julia Lawall; cocci at systeme.lip6.fr
> ??: Re: [Cocci] ??: ??: how to write such matching case?
> 
> You can try:
> 
> @@
> @@
> * main () {
>    ...
>    when != Foo(...)
> }
> 
> This semantic patch will match functions named main that do not 
> contain call to function Foo;
> 
> On Mon, Jan 27, 2014 at 2:55 PM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > Thanks for explain, but if I want to write a simple match, that is 
> > function foo must be called in main() e.g.
> >
> > int main(...)
> > {
> >     ...
> >     Foo(...);
> >     ...
> > }
> >
> > I write it as
> >
> > @@
> > @@
> > main(...) {
> >   <...
> >     when != reg_process_exit_action(...)
> >   ...>
> > }
> >
> > Reports: 53 57
> > Fatal error: exception Failure("minus: parse error:
> >  = File "rgos.cocci", line 6, column 2,  charpos = 53
> >     around = '...>', whole content =   ...>
> > ")
> >
> > What does it mean?
> >
> > By the way, I still don?t know when I should use '<'and'>' and when need not?
> >
> > -----????-----
> > ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> > ????: 2014?1?27? 20:34
> > ???: ??(?? ??)
> > ??: cocci at systeme.lip6.fr
> > ??: Re: ??: [Cocci] how to write such matching case?
> >
> > On Mon, 27 Jan 2014, ??(?? ??) wrote:
> >
> >> I wrote this
> >>
> >> @@
> >> @@
> >> cli_printf(...)
> >> ... when != cli_client_execmd_fail(...)
> >>
> >>
> >> it reports:
> >> Fatal error: exception Failure("False should not be in the final 
> >> result.  Perhaps your rule doesn't contain any +/-/* code, or you 
> >> have a failed dependency.")
> >>
> >>
> >> But i found there is not 'minus' in an example.cocci either....
> >
> > There is a * (indicating a line of interest).  You can use that.  If you use that, you will get an exists semantics for the ... (there exists an execution path that satisfies the property).  Otherwise, you get a foral semantics (all execution paths have to satisfy the property).
> >
> > julia
> >
> >>
> >>
> >> //
> >> //  Add missing pci_dev_put
> >> //
> >> // Target: Linux
> >> // Copyright:  2012 - LIP6/INRIA
> >> // License:  Licensed under ISC. See LICENSE or 
> >> http://www.isc.org/software/license
> >> // Author: Julia Lawall <Julia.Lawall@lip6.fr> // URL:
> >> http://coccinelle.lip6.fr/ // URL: http://coccinellery.org/
> >>
> >> @@
> >> local idexpression x;
> >> expression e;
> >> @@
> >>
> >> *x = pci_get_slot(...)
> >> ... when != true x == NULL
> >>     when != pci_dev_put(x)
> >>     when != e = x
> >>     when != if (x != NULL) {<+... pci_dev_put(x); ...+>} *return 
> >> ...;
> >>
> >>
> >>
> >>
> >>
> >> -----????-----
> >> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> >> ????: 2014?1?26? 19:19
> >> ???: ??(?? ??)
> >> ??: cocci at systeme.lip6.fr
> >> ??: Re: [Cocci] how to write such matching case?
> >>
> >> On Sun, 26 Jan 2014, ??(?? ??) wrote:
> >>
> >> > In my project, it's required that if function A is called, and 
> >> > function B must be called following, just like this
> >> >
> >> > Foo(...)
> >> > {
> >> >    If ... {
> >> >        A(....);
> >> >        ...
> >> >        B(....);
> >> >    }
> >> > }
> >> >
> >> > Or
> >> >
> >> > Foo1(...)
> >> > {
> >> >         A(....);
> >> >         B(...);
> >> >  }
> >> >
> >> >
> >> > How to write a patch to find the missing case? Such as
> >> >
> >> > Foo(...)
> >> > {
> >> >    If ... {
> >> >        A(....);
> >> >        ...
> >> >    }
> >> > }
> >> >
> >> > Or
> >> >
> >> > Foo1(...)
> >> > {
> >> >        A(....);
> >> >        ...
> >> > }
> >>
> >> A(...)
> >> ... When != B(...)
> >>
> >> julia
> >>
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> 
> 
> 
> --
> Peter
> 

[Cocci] 答复: 答复: 答复: 答复: how to write such matching case?

[Julia Lawall]

On Wed, 12 Feb 2014, ??(?? ??) wrote:

> You mean this?
> 
> @forall@
> @@
> * main(){

I've lost track of the issue, but here you have () as your parameter list 
and in the code there is a void.  If you want void, you need to put void.  
If you don't care about the contents of the parameter list, you should put 
...

julia

>    ...when != reg_process_exit_action(...)
> }
> 
> And my 'main' function is straightforward without flow control
> 
> int main(void)
> {
>     struct AAAA foo;
>     struct AAA foo2;
>     struct BBB foo3, *p;
>     printf("%d\n", sizeof(((struct AAAA*)0)->aa));
>     printf("%d %d %d\n", foo.aa.c, foo2.c, foo3.value); 
>     memset(p, 0, sizeof(*p));    
>     reg_process_exit_action(0, RESTART);    
>     return 0;
> }
> 
> 
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
> ????: 2014?1?28? 15:25
> ???: ??(?? ??)
> ??: Peter Senna Tschudin; cocci at systeme.lip6.fr
> ??: Re: ??: [Cocci] ??: ??: how to write such matching case?
> 
> 
> 
> On Tue, 28 Jan 2014, ??(?? ??) wrote:
> 
> > Just like this?
> > @@
> > @@
> > * main(...){
> >    ...when != reg_process_exit_action(...) }
> > 
> > 
> > I tried, but it seems does not work...
> > 
> > I make the function reg...() called in main(), but it still report 
> > main()
> 
> Probably because there exists a control flow path in which it is not called.  If you want the forall semantics, put forall in between the initial @@
> 
> julia
> 
> 
> > 
> > -----????-----
> > ???: Peter Senna Tschudin [mailto:peter.senna at gmail.com]
> > ????: 2014?1?27? 22:12
> > ???: ??(?? ??)
> > ??: Julia Lawall; cocci at systeme.lip6.fr
> > ??: Re: [Cocci] ??: ??: how to write such matching case?
> > 
> > You can try:
> > 
> > @@
> > @@
> > * main () {
> >    ...
> >    when != Foo(...)
> > }
> > 
> > This semantic patch will match functions named main that do not 
> > contain call to function Foo;
> > 
> > On Mon, Jan 27, 2014 at 2:55 PM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > > Thanks for explain, but if I want to write a simple match, that is 
> > > function foo must be called in main() e.g.
> > >
> > > int main(...)
> > > {
> > >     ...
> > >     Foo(...);
> > >     ...
> > > }
> > >
> > > I write it as
> > >
> > > @@
> > > @@
> > > main(...) {
> > >   <...
> > >     when != reg_process_exit_action(...)
> > >   ...>
> > > }
> > >
> > > Reports: 53 57
> > > Fatal error: exception Failure("minus: parse error:
> > >  = File "rgos.cocci", line 6, column 2,  charpos = 53
> > >     around = '...>', whole content =   ...>
> > > ")
> > >
> > > What does it mean?
> > >
> > > By the way, I still don?t know when I should use '<'and'>' and when need not?
> > >
> > > -----????-----
> > > ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> > > ????: 2014?1?27? 20:34
> > > ???: ??(?? ??)
> > > ??: cocci at systeme.lip6.fr
> > > ??: Re: ??: [Cocci] how to write such matching case?
> > >
> > > On Mon, 27 Jan 2014, ??(?? ??) wrote:
> > >
> > >> I wrote this
> > >>
> > >> @@
> > >> @@
> > >> cli_printf(...)
> > >> ... when != cli_client_execmd_fail(...)
> > >>
> > >>
> > >> it reports:
> > >> Fatal error: exception Failure("False should not be in the final 
> > >> result.  Perhaps your rule doesn't contain any +/-/* code, or you 
> > >> have a failed dependency.")
> > >>
> > >>
> > >> But i found there is not 'minus' in an example.cocci either....
> > >
> > > There is a * (indicating a line of interest).  You can use that.  If you use that, you will get an exists semantics for the ... (there exists an execution path that satisfies the property).  Otherwise, you get a foral semantics (all execution paths have to satisfy the property).
> > >
> > > julia
> > >
> > >>
> > >>
> > >> //
> > >> //  Add missing pci_dev_put
> > >> //
> > >> // Target: Linux
> > >> // Copyright:  2012 - LIP6/INRIA
> > >> // License:  Licensed under ISC. See LICENSE or 
> > >> http://www.isc.org/software/license
> > >> // Author: Julia Lawall <Julia.Lawall@lip6.fr> // URL:
> > >> http://coccinelle.lip6.fr/ // URL: http://coccinellery.org/
> > >>
> > >> @@
> > >> local idexpression x;
> > >> expression e;
> > >> @@
> > >>
> > >> *x = pci_get_slot(...)
> > >> ... when != true x == NULL
> > >>     when != pci_dev_put(x)
> > >>     when != e = x
> > >>     when != if (x != NULL) {<+... pci_dev_put(x); ...+>} *return 
> > >> ...;
> > >>
> > >>
> > >>
> > >>
> > >>
> > >> -----????-----
> > >> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> > >> ????: 2014?1?26? 19:19
> > >> ???: ??(?? ??)
> > >> ??: cocci at systeme.lip6.fr
> > >> ??: Re: [Cocci] how to write such matching case?
> > >>
> > >> On Sun, 26 Jan 2014, ??(?? ??) wrote:
> > >>
> > >> > In my project, it's required that if function A is called, and 
> > >> > function B must be called following, just like this
> > >> >
> > >> > Foo(...)
> > >> > {
> > >> >    If ... {
> > >> >        A(....);
> > >> >        ...
> > >> >        B(....);
> > >> >    }
> > >> > }
> > >> >
> > >> > Or
> > >> >
> > >> > Foo1(...)
> > >> > {
> > >> >         A(....);
> > >> >         B(...);
> > >> >  }
> > >> >
> > >> >
> > >> > How to write a patch to find the missing case? Such as
> > >> >
> > >> > Foo(...)
> > >> > {
> > >> >    If ... {
> > >> >        A(....);
> > >> >        ...
> > >> >    }
> > >> > }
> > >> >
> > >> > Or
> > >> >
> > >> > Foo1(...)
> > >> > {
> > >> >        A(....);
> > >> >        ...
> > >> > }
> > >>
> > >> A(...)
> > >> ... When != B(...)
> > >>
> > >> julia
> > >>
> > > _______________________________________________
> > > Cocci mailing list
> > > Cocci at systeme.lip6.fr
> > > https://systeme.lip6.fr/mailman/listinfo/cocci
> > 
> > 
> > 
> > --
> > Peter
> > 
> 

[Cocci] Fatal error: exception Failure("no python")

[林嘉(程二 福州)]

I did some 'apt-get install update' and 'apt-get install upgrade' today

Then I found cocc can not work, it reports: Fatal error: exception Failure("no python")

But I run python -v is ok

# python -V
Python 2.7.3

Why? How can I fix this? 

[Cocci] Fatal error: exception Failure("no python")

[Julia Lawall]

On Fri, 14 Feb 2014, ??(?? ??) wrote:

> I did some 'apt-get install update' and 'apt-get install upgrade' today
>
> Then I found cocc can not work, it reports: Fatal error: exception Failure("no python")
>
> But I run python -v is ok
>
> # python -V
> Python 2.7.3
>
> Why? How can I fix this?

Perhaps your python is not in a standard place?

What distribution of Linux are you using?  What version of Coccinelle do
you have?

julia

[Cocci] Fatal error: exception Failure("no python")

[Nicolas Palix]

Hi,

On Fri, Feb 14, 2014 at 8:57 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> Then I found cocc can not work, it reports: Fatal error: exception Failure("no python")

You get is error when spatch has not been compiled with Python support
but you are still using cocci file with Python snippet.

Does ' spatch --version' report python support ?
Such as
     spatch version 1.0.0-rc15 with Python support and with PCRE support

Regards,
-- 
Nicolas Palix
Tel: +33 4 76 51 46 27
http://lig-membres.imag.fr/palix/

[Cocci] 答复: Fatal error: exception Failure("no python")

[林嘉(程二 福州)]

-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?2?14? 16:56
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: [Cocci] Fatal error: exception Failure("no python")

On Fri, 14 Feb 2014, ??(?? ??) wrote:

> I did some 'apt-get install update' and 'apt-get install upgrade' 
> today
>
> Then I found cocc can not work, it reports: Fatal error: exception 
> Failure("no python")
>
> But I run python -v is ok
>
> # python -V
> Python 2.7.3
>
> Why? How can I fix this?

Perhaps your python is not in a standard place?

I don?t know how python is placed..but the result of whereis is

# whereis python
python: /usr/bin/python2.6 /usr/bin/python2.7 /usr/bin/python2.7-config /usr/bin/python /etc/python2.6 /etc/python2.7 /etc/python /usr/lib/python2.6 /usr/lib/python2.7 /usr/bin/X11/python2.6 /usr/bin/X11/python2.7 /usr/bin/X11/python2.7-config /usr/bin/X11/python /usr/local/lib/python2.6 /usr/local/lib/python2.7 /usr/include/python2.6 /usr/include/python2.7 /usr/include/python2.7_d /usr/include/python2.6_d /usr/share/python /usr/share/man/man1/python.1.gz



What distribution of Linux are you using?  What version of Coccinelle do you have?
~~~~~~~~~~~ uname -ar
Linux debian 2.6.33.7-co-0.7.9 #1 PREEMPT Sat Apr 9 20:30:51 UTC 2011 i686 GNU/Linux

I installed 0.2.3.firstly,  then installed 1.0.0-rc19
julia

[Cocci] 答复: Fatal error: exception Failure("no python")

[林嘉(程二 福州)]

# spatch --version
spatch version 1.0.0-rc19 without Python support and with Str regexp support

but I didn't make and make install with option '--without-python '
that is I pressed  ./configure --enable-release, just it


and now when I run "/usr/bin/spatch -sp_file ....", all is ok, it seems version 0.2.3 runs
but "spatch -sp_file " reports fail, seems version 1.0 runs?

-----????-----
???: npalix.work at gmail.com [mailto:npalix.work at gmail.com] ?? Nicolas Palix
????: 2014?2?14? 17:51
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: [Cocci] Fatal error: exception Failure("no python")

Hi,

On Fri, Feb 14, 2014 at 8:57 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> Then I found cocc can not work, it reports: Fatal error: exception 
> Failure("no python")

You get is error when spatch has not been compiled with Python support but you are still using cocci file with Python snippet.

Does ' spatch --version' report python support ?
Such as
     spatch version 1.0.0-rc15 with Python support and with PCRE support

Regards,
--
Nicolas Palix
Tel: +33 4 76 51 46 27
http://lig-membres.imag.fr/palix/

[Cocci] 答复: Fatal error: exception Failure("no python")

[Julia Lawall]

On Fri, 14 Feb 2014, ??(?? ??) wrote:

> # spatch --version
> spatch version 1.0.0-rc19 without Python support and with Str regexp support
>
> but I didn't make and make install with option '--without-python '
> that is I pressed  ./configure --enable-release, just it
>
>
> and now when I run "/usr/bin/spatch -sp_file ....", all is ok, it seems version 0.2.3 runs
> but "spatch -sp_file " reports fail, seems version 1.0 runs?

I'm not sure to understand what you are doing.  Are you compiling
coccinelle from the source code, or are you using one from a distribution.
If you are compiling from source code, what does it say about python when
you run ./configure?

I think that you are not the first to have this problem.  Maybe search for
python on the gmane archives of the mailing list:

http://blog.gmane.org/gmane.comp.version-control.coccinelle

julia


>
> -----????-----
> ???: npalix.work at gmail.com [mailto:npalix.work at gmail.com] ?? Nicolas Palix
> ????: 2014?2?14? 17:51
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] Fatal error: exception Failure("no python")
>
> Hi,
>
> On Fri, Feb 14, 2014 at 8:57 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > Then I found cocc can not work, it reports: Fatal error: exception
> > Failure("no python")
>
> You get is error when spatch has not been compiled with Python support but you are still using cocci file with Python snippet.
>
> Does ' spatch --version' report python support ?
> Such as
>      spatch version 1.0.0-rc15 with Python support and with PCRE support
>
> Regards,
> --
> Nicolas Palix
> Tel: +33 4 76 51 46 27
> http://lig-membres.imag.fr/palix/
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
>

[Cocci] 答复: 答复: Fatal error: exception Failure("no python")

[林嘉(程二 福州)]

Sorry , I didn?t describe clearly.  Well,  The step is following:

1  I installed cocc by using deb package 'coccinelle_0.2.3.deb-1_i386.deb' at first.   All is ok.

2  Then I think its version may be too low. So I downloaded the source of version 1.0 and make install.  All is ok.

3  I made some 'apt-get install upgrade and update' for another reason, I didn?t think it has something to do with cocc.

4  the problem occur when I use cocc. 

5  I found I can run cocc by "/usr/bin/spatch " , but not by "spatch ",  I think there is some chaos in the cocc config now, see following...

root at debian:~/windows/ctest/cocc# /usr/bin/spatch -version 
spatch version 0.2.3 with Python support
root at debian:~/windows/ctest/cocc# spatch -version
spatch version 1.0.0-rc19 without Python support and with Str regexp support 
root at debian:~/windows/ctest/cocc#



-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?2?14? 20:58
???: ??(?? ??)
??: Nicolas Palix; cocci at systeme.lip6.fr
??: Re: [Cocci] ??: Fatal error: exception Failure("no python")

On Fri, 14 Feb 2014, ??(?? ??) wrote:

> # spatch --version
> spatch version 1.0.0-rc19 without Python support and with Str regexp 
> support
>
> but I didn't make and make install with option '--without-python '
> that is I pressed  ./configure --enable-release, just it
>
>
> and now when I run "/usr/bin/spatch -sp_file ....", all is ok, it 
> seems version 0.2.3 runs but "spatch -sp_file " reports fail, seems version 1.0 runs?

I'm not sure to understand what you are doing.  Are you compiling coccinelle from the source code, or are you using one from a distribution.
If you are compiling from source code, what does it say about python when you run ./configure?

I think that you are not the first to have this problem.  Maybe search for python on the gmane archives of the mailing list:

http://blog.gmane.org/gmane.comp.version-control.coccinelle

julia


>
> -----????-----
> ???: npalix.work at gmail.com [mailto:npalix.work at gmail.com] ?? Nicolas 
> Palix
> ????: 2014?2?14? 17:51
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] Fatal error: exception Failure("no python")
>
> Hi,
>
> On Fri, Feb 14, 2014 at 8:57 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > Then I found cocc can not work, it reports: Fatal error: exception 
> > Failure("no python")
>
> You get is error when spatch has not been compiled with Python support but you are still using cocci file with Python snippet.
>
> Does ' spatch --version' report python support ?
> Such as
>      spatch version 1.0.0-rc15 with Python support and with PCRE 
> support
>
> Regards,
> --
> Nicolas Palix
> Tel: +33 4 76 51 46 27
> http://lig-membres.imag.fr/palix/
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
>

[Cocci] 答复: 答复: Fatal error: exception Failure("no python")

[Julia Lawall]

On Sat, 15 Feb 2014, ??(?? ??) wrote:

> Sorry , I didn?t describe clearly.  Well,  The step is following:
> 
> 1  I installed cocc by using deb package 'coccinelle_0.2.3.deb-1_i386.deb' at first.   All is ok.

This is not good at all.  But I am not sure why you are using that 
version.  That is from the old stable version of Debian.  The current 
stable version is 1.0.0-rc12.

> 2 Then I think its version may be too low. So I downloaded the source of 
> version 1.0 and make install.  All is ok.

Did you run ./configure?  I'm surprised that it owuld let you run make 
without doing that.  ./configure will normally succeed, but I believe it 
says at the end what features it managed to find.

> 3 I made some 'apt-get install upgrade and update' for another reason, I 
> didn?t think it has something to do with cocc.
> 
> 4  the problem occur when I use cocc. 
> 
> 5 I found I can run cocc by "/usr/bin/spatch " , but not by "spatch ", I 
> think there is some chaos in the cocc config now, see following...

Coccinelle goes in /usr/local/bin

> root at debian:~/windows/ctest/cocc# /usr/bin/spatch -version 
> spatch version 0.2.3 with Python support
> root at debian:~/windows/ctest/cocc# spatch -version
> spatch version 1.0.0-rc19 without Python support and with Str regexp support 
> root at debian:~/windows/ctest/cocc#

Clearly it has not found python during the configure operation.  Maybe if 
you rerun configure and save the output in a file, you will find the 
problem.

julia

> 
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
> ????: 2014?2?14? 20:58
> ???: ??(?? ??)
> ??: Nicolas Palix; cocci at systeme.lip6.fr
> ??: Re: [Cocci] ??: Fatal error: exception Failure("no python")
> 
> On Fri, 14 Feb 2014, ??(?? ??) wrote:
> 
> > # spatch --version
> > spatch version 1.0.0-rc19 without Python support and with Str regexp 
> > support
> >
> > but I didn't make and make install with option '--without-python '
> > that is I pressed  ./configure --enable-release, just it
> >
> >
> > and now when I run "/usr/bin/spatch -sp_file ....", all is ok, it 
> > seems version 0.2.3 runs but "spatch -sp_file " reports fail, seems version 1.0 runs?
> 
> I'm not sure to understand what you are doing.  Are you compiling coccinelle from the source code, or are you using one from a distribution.
> If you are compiling from source code, what does it say about python when you run ./configure?
> 
> I think that you are not the first to have this problem.  Maybe search for python on the gmane archives of the mailing list:
> 
> http://blog.gmane.org/gmane.comp.version-control.coccinelle
> 
> julia
> 
> 
> >
> > -----????-----
> > ???: npalix.work at gmail.com [mailto:npalix.work at gmail.com] ?? Nicolas 
> > Palix
> > ????: 2014?2?14? 17:51
> > ???: ??(?? ??)
> > ??: cocci at systeme.lip6.fr
> > ??: Re: [Cocci] Fatal error: exception Failure("no python")
> >
> > Hi,
> >
> > On Fri, Feb 14, 2014 at 8:57 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > > Then I found cocc can not work, it reports: Fatal error: exception 
> > > Failure("no python")
> >
> > You get is error when spatch has not been compiled with Python support but you are still using cocci file with Python snippet.
> >
> > Does ' spatch --version' report python support ?
> > Such as
> >      spatch version 1.0.0-rc15 with Python support and with PCRE 
> > support
> >
> > Regards,
> > --
> > Nicolas Palix
> > Tel: +33 4 76 51 46 27
> > http://lig-membres.imag.fr/palix/
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> >
> 

[Cocci] 答复: 答复: 答复: Fatal error: exception Failure("no python")

[林嘉(程二 福州)]

Yes, I rerun the configure with enable python, and found lack of the pkg-config python lib 


./configure --enable-release --enable-ocaml --enable-python --enable-pcre-syntax
reports?
./configure: line 15220: /root/windows/install/coccinelle-master/coccinelle-master/setup/fake-subst.sh: Permission denied
configure: error: the python scripts feature is enabled but the pkg-config python library is not found



finally, I directly download the 1.0 version Debian package from https://packages.debian.org/wheezy/i386/coccinelle/download


-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?2?15? 14:18
???: ??(?? ??)
??: Nicolas Palix; cocci at systeme.lip6.fr
??: Re: ??: [Cocci] ??: Fatal error: exception Failure("no python")

On Sat, 15 Feb 2014, ??(?? ??) wrote:

> Sorry , I didn?t describe clearly.  Well,  The step is following:
> 
> 1  I installed cocc by using deb package 'coccinelle_0.2.3.deb-1_i386.deb' at first.   All is ok.

This is not good at all.  But I am not sure why you are using that version.  That is from the old stable version of Debian.  The current stable version is 1.0.0-rc12.

> 2 Then I think its version may be too low. So I downloaded the source 
> of version 1.0 and make install.  All is ok.

Did you run ./configure?  I'm surprised that it owuld let you run make without doing that.  ./configure will normally succeed, but I believe it says at the end what features it managed to find.

> 3 I made some 'apt-get install upgrade and update' for another reason, 
> I didn?t think it has something to do with cocc.
> 
> 4  the problem occur when I use cocc. 
> 
> 5 I found I can run cocc by "/usr/bin/spatch " , but not by "spatch ", 
> I think there is some chaos in the cocc config now, see following...

Coccinelle goes in /usr/local/bin

> root at debian:~/windows/ctest/cocc# /usr/bin/spatch -version spatch 
> version 0.2.3 with Python support root at debian:~/windows/ctest/cocc# 
> spatch -version spatch version 1.0.0-rc19 without Python support and 
> with Str regexp support root at debian:~/windows/ctest/cocc#

Clearly it has not found python during the configure operation.  Maybe if you rerun configure and save the output in a file, you will find the problem.

julia

> 
> 
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> ????: 2014?2?14? 20:58
> ???: ??(?? ??)
> ??: Nicolas Palix; cocci at systeme.lip6.fr
> ??: Re: [Cocci] ??: Fatal error: exception Failure("no python")
> 
> On Fri, 14 Feb 2014, ??(?? ??) wrote:
> 
> > # spatch --version
> > spatch version 1.0.0-rc19 without Python support and with Str regexp 
> > support
> >
> > but I didn't make and make install with option '--without-python '
> > that is I pressed  ./configure --enable-release, just it
> >
> >
> > and now when I run "/usr/bin/spatch -sp_file ....", all is ok, it 
> > seems version 0.2.3 runs but "spatch -sp_file " reports fail, seems version 1.0 runs?
> 
> I'm not sure to understand what you are doing.  Are you compiling coccinelle from the source code, or are you using one from a distribution.
> If you are compiling from source code, what does it say about python when you run ./configure?
> 
> I think that you are not the first to have this problem.  Maybe search for python on the gmane archives of the mailing list:
> 
> http://blog.gmane.org/gmane.comp.version-control.coccinelle
> 
> julia
> 
> 
> >
> > -----????-----
> > ???: npalix.work at gmail.com [mailto:npalix.work at gmail.com] ?? Nicolas 
> > Palix
> > ????: 2014?2?14? 17:51
> > ???: ??(?? ??)
> > ??: cocci at systeme.lip6.fr
> > ??: Re: [Cocci] Fatal error: exception Failure("no python")
> >
> > Hi,
> >
> > On Fri, Feb 14, 2014 at 8:57 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > > Then I found cocc can not work, it reports: Fatal error: exception 
> > > Failure("no python")
> >
> > You get is error when spatch has not been compiled with Python support but you are still using cocci file with Python snippet.
> >
> > Does ' spatch --version' report python support ?
> > Such as
> >      spatch version 1.0.0-rc15 with Python support and with PCRE 
> > support
> >
> > Regards,
> > --
> > Nicolas Palix
> > Tel: +33 4 76 51 46 27
> > http://lig-membres.imag.fr/palix/
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> >
> 

[Cocci] what is the function of 'position' in SMPL

[林嘉(程二 福州)]

e.g. in find_unsigned.cocci

@u@ type T; unsigned T i; position p; @@
 i at p < 0


what is the difference with

@u@ type T; unsigned T i; @@
 i < 0


and what is the meaning of operator '@' ahead 'p'?
 

[Cocci] 答复: 答复: 答复: Fatal error: exception Failure("no python")

[Julia Lawall]

In the mailing list archives there is some discussion of Python:

http://blog.gmane.org/gmane.comp.version-control.coccinelle/day=20130823
http://blog.gmane.org/gmane.comp.version-control.coccinelle/day=20130824

julia

On Mon, 17 Feb 2014, ??(?? ??) wrote:

> Yes, I rerun the configure with enable python, and found lack of the pkg-config python lib
>
>
> ./configure --enable-release --enable-ocaml --enable-python --enable-pcre-syntax
> reports?
> ./configure: line 15220: /root/windows/install/coccinelle-master/coccinelle-master/setup/fake-subst.sh: Permission denied
> configure: error: the python scripts feature is enabled but the pkg-config python library is not found
>
>
>
> finally, I directly download the 1.0 version Debian package from https://packages.debian.org/wheezy/i386/coccinelle/download
>
>
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> ????: 2014?2?15? 14:18
> ???: ??(?? ??)
> ??: Nicolas Palix; cocci at systeme.lip6.fr
> ??: Re: ??: [Cocci] ??: Fatal error: exception Failure("no python")
>
> On Sat, 15 Feb 2014, ??(?? ??) wrote:
>
> > Sorry , I didn?t describe clearly.  Well,  The step is following:
> >
> > 1  I installed cocc by using deb package 'coccinelle_0.2.3.deb-1_i386.deb' at first.   All is ok.
>
> This is not good at all.  But I am not sure why you are using that version.  That is from the old stable version of Debian.  The current stable version is 1.0.0-rc12.
>
> > 2 Then I think its version may be too low. So I downloaded the source
> > of version 1.0 and make install.  All is ok.
>
> Did you run ./configure?  I'm surprised that it owuld let you run make without doing that.  ./configure will normally succeed, but I believe it says at the end what features it managed to find.
>
> > 3 I made some 'apt-get install upgrade and update' for another reason,
> > I didn?t think it has something to do with cocc.
> >
> > 4  the problem occur when I use cocc.
> >
> > 5 I found I can run cocc by "/usr/bin/spatch " , but not by "spatch ",
> > I think there is some chaos in the cocc config now, see following...
>
> Coccinelle goes in /usr/local/bin
>
> > root at debian:~/windows/ctest/cocc# /usr/bin/spatch -version spatch
> > version 0.2.3 with Python support root at debian:~/windows/ctest/cocc#
> > spatch -version spatch version 1.0.0-rc19 without Python support and
> > with Str regexp support root at debian:~/windows/ctest/cocc#
>
> Clearly it has not found python during the configure operation.  Maybe if you rerun configure and save the output in a file, you will find the problem.
>
> julia
>
> >
> >
> > -----????-----
> > ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> > ????: 2014?2?14? 20:58
> > ???: ??(?? ??)
> > ??: Nicolas Palix; cocci at systeme.lip6.fr
> > ??: Re: [Cocci] ??: Fatal error: exception Failure("no python")
> >
> > On Fri, 14 Feb 2014, ??(?? ??) wrote:
> >
> > > # spatch --version
> > > spatch version 1.0.0-rc19 without Python support and with Str regexp
> > > support
> > >
> > > but I didn't make and make install with option '--without-python '
> > > that is I pressed  ./configure --enable-release, just it
> > >
> > >
> > > and now when I run "/usr/bin/spatch -sp_file ....", all is ok, it
> > > seems version 0.2.3 runs but "spatch -sp_file " reports fail, seems version 1.0 runs?
> >
> > I'm not sure to understand what you are doing.  Are you compiling coccinelle from the source code, or are you using one from a distribution.
> > If you are compiling from source code, what does it say about python when you run ./configure?
> >
> > I think that you are not the first to have this problem.  Maybe search for python on the gmane archives of the mailing list:
> >
> > http://blog.gmane.org/gmane.comp.version-control.coccinelle
> >
> > julia
> >
> >
> > >
> > > -----????-----
> > > ???: npalix.work at gmail.com [mailto:npalix.work at gmail.com] ?? Nicolas
> > > Palix
> > > ????: 2014?2?14? 17:51
> > > ???: ??(?? ??)
> > > ??: cocci at systeme.lip6.fr
> > > ??: Re: [Cocci] Fatal error: exception Failure("no python")
> > >
> > > Hi,
> > >
> > > On Fri, Feb 14, 2014 at 8:57 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > > > Then I found cocc can not work, it reports: Fatal error: exception
> > > > Failure("no python")
> > >
> > > You get is error when spatch has not been compiled with Python support but you are still using cocci file with Python snippet.
> > >
> > > Does ' spatch --version' report python support ?
> > > Such as
> > >      spatch version 1.0.0-rc15 with Python support and with PCRE
> > > support
> > >
> > > Regards,
> > > --
> > > Nicolas Palix
> > > Tel: +33 4 76 51 46 27
> > > http://lig-membres.imag.fr/palix/
> > > _______________________________________________
> > > Cocci mailing list
> > > Cocci at systeme.lip6.fr
> > > https://systeme.lip6.fr/mailman/listinfo/cocci
> > >
> >
>

[Cocci] 答复: 答复: 答复: Fatal error: exception Failure("no python")

[Arie Middelkoop]

Check this line:

./configure: line 15220: /root/windows/install/
coccinelle-master/coccinelle-master/setup/fake-subst.sh: Permission denied

That script is executed to find python when pkg-config was not found.
Check whether the executable bit is set.

Arie

2014-02-17 2:27 GMT+01:00 ??(?? ??) <linjia@ruijie.com.cn>:
> Yes, I rerun the configure with enable python, and found lack of the pkg-config python lib
>
>
> ./configure --enable-release --enable-ocaml --enable-python --enable-pcre-syntax
> reports?
> ./configure: line 15220: /root/windows/install/coccinelle-master/coccinelle-master/setup/fake-subst.sh: Permission denied
> configure: error: the python scripts feature is enabled but the pkg-config python library is not found
>
>
>
> finally, I directly download the 1.0 version Debian package from https://packages.debian.org/wheezy/i386/coccinelle/download
>
>
> -----????-----
> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> ????: 2014?2?15? 14:18
> ???: ??(?? ??)
> ??: Nicolas Palix; cocci at systeme.lip6.fr
> ??: Re: ??: [Cocci] ??: Fatal error: exception Failure("no python")
>
> On Sat, 15 Feb 2014, ??(?? ??) wrote:
>
>> Sorry , I didn?t describe clearly.  Well,  The step is following:
>>
>> 1  I installed cocc by using deb package 'coccinelle_0.2.3.deb-1_i386.deb' at first.   All is ok.
>
> This is not good at all.  But I am not sure why you are using that version.  That is from the old stable version of Debian.  The current stable version is 1.0.0-rc12.
>
>> 2 Then I think its version may be too low. So I downloaded the source
>> of version 1.0 and make install.  All is ok.
>
> Did you run ./configure?  I'm surprised that it owuld let you run make without doing that.  ./configure will normally succeed, but I believe it says at the end what features it managed to find.
>
>> 3 I made some 'apt-get install upgrade and update' for another reason,
>> I didn?t think it has something to do with cocc.
>>
>> 4  the problem occur when I use cocc.
>>
>> 5 I found I can run cocc by "/usr/bin/spatch " , but not by "spatch ",
>> I think there is some chaos in the cocc config now, see following...
>
> Coccinelle goes in /usr/local/bin
>
>> root at debian:~/windows/ctest/cocc# /usr/bin/spatch -version spatch
>> version 0.2.3 with Python support root at debian:~/windows/ctest/cocc#
>> spatch -version spatch version 1.0.0-rc19 without Python support and
>> with Str regexp support root at debian:~/windows/ctest/cocc#
>
> Clearly it has not found python during the configure operation.  Maybe if you rerun configure and save the output in a file, you will find the problem.
>
> julia
>
>>
>>
>> -----????-----
>> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
>> ????: 2014?2?14? 20:58
>> ???: ??(?? ??)
>> ??: Nicolas Palix; cocci at systeme.lip6.fr
>> ??: Re: [Cocci] ??: Fatal error: exception Failure("no python")
>>
>> On Fri, 14 Feb 2014, ??(?? ??) wrote:
>>
>> > # spatch --version
>> > spatch version 1.0.0-rc19 without Python support and with Str regexp
>> > support
>> >
>> > but I didn't make and make install with option '--without-python '
>> > that is I pressed  ./configure --enable-release, just it
>> >
>> >
>> > and now when I run "/usr/bin/spatch -sp_file ....", all is ok, it
>> > seems version 0.2.3 runs but "spatch -sp_file " reports fail, seems version 1.0 runs?
>>
>> I'm not sure to understand what you are doing.  Are you compiling coccinelle from the source code, or are you using one from a distribution.
>> If you are compiling from source code, what does it say about python when you run ./configure?
>>
>> I think that you are not the first to have this problem.  Maybe search for python on the gmane archives of the mailing list:
>>
>> http://blog.gmane.org/gmane.comp.version-control.coccinelle
>>
>> julia
>>
>>
>> >
>> > -----????-----
>> > ???: npalix.work at gmail.com [mailto:npalix.work at gmail.com] ?? Nicolas
>> > Palix
>> > ????: 2014?2?14? 17:51
>> > ???: ??(?? ??)
>> > ??: cocci at systeme.lip6.fr
>> > ??: Re: [Cocci] Fatal error: exception Failure("no python")
>> >
>> > Hi,
>> >
>> > On Fri, Feb 14, 2014 at 8:57 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
>> > > Then I found cocc can not work, it reports: Fatal error: exception
>> > > Failure("no python")
>> >
>> > You get is error when spatch has not been compiled with Python support but you are still using cocci file with Python snippet.
>> >
>> > Does ' spatch --version' report python support ?
>> > Such as
>> >      spatch version 1.0.0-rc15 with Python support and with PCRE
>> > support
>> >
>> > Regards,
>> > --
>> > Nicolas Palix
>> > Tel: +33 4 76 51 46 27
>> > http://lig-membres.imag.fr/palix/
>> > _______________________________________________
>> > Cocci mailing list
>> > Cocci at systeme.lip6.fr
>> > https://systeme.lip6.fr/mailman/listinfo/cocci
>> >
>>
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci

[Cocci] 答复: 答复: 答复: Fatal error: exception Failure("no python")

[Julia Lawall]

On Mon, 17 Feb 2014, Arie Middelkoop wrote:

> Check this line:
>
> ./configure: line 15220: /root/windows/install/
> coccinelle-master/coccinelle-master/setup/fake-subst.sh: Permission denied
>
> That script is executed to find python when pkg-config was not found.
> Check whether the executable bit is set.

Thanks for the suggestion.  For me it is, but perhaps it changes when
someone downloads the code?

julia

>
> Arie
>
> 2014-02-17 2:27 GMT+01:00 ??(?? ??) <linjia@ruijie.com.cn>:
> > Yes, I rerun the configure with enable python, and found lack of the pkg-config python lib
> >
> >
> > ./configure --enable-release --enable-ocaml --enable-python --enable-pcre-syntax
> > reports?
> > ./configure: line 15220: /root/windows/install/coccinelle-master/coccinelle-master/setup/fake-subst.sh: Permission denied
> > configure: error: the python scripts feature is enabled but the pkg-config python library is not found
> >
> >
> >
> > finally, I directly download the 1.0 version Debian package from https://packages.debian.org/wheezy/i386/coccinelle/download
> >
> >
> > -----????-----
> > ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> > ????: 2014?2?15? 14:18
> > ???: ??(?? ??)
> > ??: Nicolas Palix; cocci at systeme.lip6.fr
> > ??: Re: ??: [Cocci] ??: Fatal error: exception Failure("no python")
> >
> > On Sat, 15 Feb 2014, ??(?? ??) wrote:
> >
> >> Sorry , I didn?t describe clearly.  Well,  The step is following:
> >>
> >> 1  I installed cocc by using deb package 'coccinelle_0.2.3.deb-1_i386.deb' at first.   All is ok.
> >
> > This is not good at all.  But I am not sure why you are using that version.  That is from the old stable version of Debian.  The current stable version is 1.0.0-rc12.
> >
> >> 2 Then I think its version may be too low. So I downloaded the source
> >> of version 1.0 and make install.  All is ok.
> >
> > Did you run ./configure?  I'm surprised that it owuld let you run make without doing that.  ./configure will normally succeed, but I believe it says at the end what features it managed to find.
> >
> >> 3 I made some 'apt-get install upgrade and update' for another reason,
> >> I didn?t think it has something to do with cocc.
> >>
> >> 4  the problem occur when I use cocc.
> >>
> >> 5 I found I can run cocc by "/usr/bin/spatch " , but not by "spatch ",
> >> I think there is some chaos in the cocc config now, see following...
> >
> > Coccinelle goes in /usr/local/bin
> >
> >> root at debian:~/windows/ctest/cocc# /usr/bin/spatch -version spatch
> >> version 0.2.3 with Python support root at debian:~/windows/ctest/cocc#
> >> spatch -version spatch version 1.0.0-rc19 without Python support and
> >> with Str regexp support root at debian:~/windows/ctest/cocc#
> >
> > Clearly it has not found python during the configure operation.  Maybe if you rerun configure and save the output in a file, you will find the problem.
> >
> > julia
> >
> >>
> >>
> >> -----????-----
> >> ???: Julia Lawall [mailto:julia.lawall at lip6.fr]
> >> ????: 2014?2?14? 20:58
> >> ???: ??(?? ??)
> >> ??: Nicolas Palix; cocci at systeme.lip6.fr
> >> ??: Re: [Cocci] ??: Fatal error: exception Failure("no python")
> >>
> >> On Fri, 14 Feb 2014, ??(?? ??) wrote:
> >>
> >> > # spatch --version
> >> > spatch version 1.0.0-rc19 without Python support and with Str regexp
> >> > support
> >> >
> >> > but I didn't make and make install with option '--without-python '
> >> > that is I pressed  ./configure --enable-release, just it
> >> >
> >> >
> >> > and now when I run "/usr/bin/spatch -sp_file ....", all is ok, it
> >> > seems version 0.2.3 runs but "spatch -sp_file " reports fail, seems version 1.0 runs?
> >>
> >> I'm not sure to understand what you are doing.  Are you compiling coccinelle from the source code, or are you using one from a distribution.
> >> If you are compiling from source code, what does it say about python when you run ./configure?
> >>
> >> I think that you are not the first to have this problem.  Maybe search for python on the gmane archives of the mailing list:
> >>
> >> http://blog.gmane.org/gmane.comp.version-control.coccinelle
> >>
> >> julia
> >>
> >>
> >> >
> >> > -----????-----
> >> > ???: npalix.work at gmail.com [mailto:npalix.work at gmail.com] ?? Nicolas
> >> > Palix
> >> > ????: 2014?2?14? 17:51
> >> > ???: ??(?? ??)
> >> > ??: cocci at systeme.lip6.fr
> >> > ??: Re: [Cocci] Fatal error: exception Failure("no python")
> >> >
> >> > Hi,
> >> >
> >> > On Fri, Feb 14, 2014 at 8:57 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> >> > > Then I found cocc can not work, it reports: Fatal error: exception
> >> > > Failure("no python")
> >> >
> >> > You get is error when spatch has not been compiled with Python support but you are still using cocci file with Python snippet.
> >> >
> >> > Does ' spatch --version' report python support ?
> >> > Such as
> >> >      spatch version 1.0.0-rc15 with Python support and with PCRE
> >> > support
> >> >
> >> > Regards,
> >> > --
> >> > Nicolas Palix
> >> > Tel: +33 4 76 51 46 27
> >> > http://lig-membres.imag.fr/palix/
> >> > _______________________________________________
> >> > Cocci mailing list
> >> > Cocci at systeme.lip6.fr
> >> > https://systeme.lip6.fr/mailman/listinfo/cocci
> >> >
> >>
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
>

[Cocci] what is the function of 'position' in SMPL

[Peter Senna Tschudin]

p is a position metavariable. It is useful when you want to print line
and column numbers of interesting parts.

The @p means that the position of what matches to 'i' will be saved into p.

There is some information available at:
http://coccinelle.lip6.fr/docs/main_grammar.pdf

On Mon, Feb 17, 2014 at 6:45 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> e.g. in find_unsigned.cocci
>
> @u@ type T; unsigned T i; position p; @@
>  i at p < 0
>
>
> what is the difference with
>
> @u@ type T; unsigned T i; @@
>  i < 0
>
>
> and what is the meaning of operator '@' ahead 'p'?
>
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci



-- 
Peter

[Cocci] 答复: what is the function of 'position' in SMPL

[林嘉(程二 福州)]

Thank to all 

What the differentce among 'Identifier declaration statement' in metavariable portion?
I  just guess they all can be used to express a name of variable or function

For example:

@@
Identifier a; or declaration a;?
@@
...........



-----????-----
???: Peter Senna Tschudin [mailto:peter.senna at gmail.com] 
????: 2014?2?17? 21:35
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: [Cocci] what is the function of 'position' in SMPL

p is a position metavariable. It is useful when you want to print line and column numbers of interesting parts.

The @p means that the position of what matches to 'i' will be saved into p.

There is some information available at:
http://coccinelle.lip6.fr/docs/main_grammar.pdf

On Mon, Feb 17, 2014 at 6:45 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> e.g. in find_unsigned.cocci
>
> @u@ type T; unsigned T i; position p; @@  i at p < 0
>
>
> what is the difference with
>
> @u@ type T; unsigned T i; @@
>  i < 0
>
>
> and what is the meaning of operator '@' ahead 'p'?
>
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci



-- 
Peter

[Cocci] 答复: what is the function of 'position' in SMPL

[Julia Lawall]

On Tue, 18 Feb 2014, ??(?? ??) wrote:

> Thank to all 
> 
> What the differentce among 'Identifier declaration statement' in metavariable portion?
> I  just guess they all can be used to express a name of variable or function

Identifier is just the name a.  A declaration has a type and a semicolon.
For example:

int x;

is a declaration, while

x

is an identifier.

julia

> For example:
> 
> @@
> Identifier a; or declaration a;?
> @@
> ...........
> 
> 
> 
> -----????-----
> ???: Peter Senna Tschudin [mailto:peter.senna at gmail.com] 
> ????: 2014?2?17? 21:35
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] what is the function of 'position' in SMPL
> 
> p is a position metavariable. It is useful when you want to print line and column numbers of interesting parts.
> 
> The @p means that the position of what matches to 'i' will be saved into p.
> 
> There is some information available at:
> http://coccinelle.lip6.fr/docs/main_grammar.pdf
> 
> On Mon, Feb 17, 2014 at 6:45 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > e.g. in find_unsigned.cocci
> >
> > @u@ type T; unsigned T i; position p; @@  i at p < 0
> >
> >
> > what is the difference with
> >
> > @u@ type T; unsigned T i; @@
> >  i < 0
> >
> >
> > and what is the meaning of operator '@' ahead 'p'?
> >
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> 
> 
> 
> -- 
> Peter
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
> 

[Cocci] 答复: what is the function of 'position' in SMPL

[Julia Lawall]

On Tue, 18 Feb 2014, ??(?? ??) wrote:

> Thank to all 
> 
> What the differentce among 'Identifier declaration statement' in metavariable portion?
> I  just guess they all can be used to express a name of variable or function

An identifier can be used to express the name of a variable, a function, a 
field, the name of a structure, the name of an enum, the name of a 
#define, etc.

julia

> 
> For example:
> 
> @@
> Identifier a; or declaration a;?
> @@
> ...........
> 
> 
> 
> -----????-----
> ???: Peter Senna Tschudin [mailto:peter.senna at gmail.com] 
> ????: 2014?2?17? 21:35
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] what is the function of 'position' in SMPL
> 
> p is a position metavariable. It is useful when you want to print line and column numbers of interesting parts.
> 
> The @p means that the position of what matches to 'i' will be saved into p.
> 
> There is some information available at:
> http://coccinelle.lip6.fr/docs/main_grammar.pdf
> 
> On Mon, Feb 17, 2014 at 6:45 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > e.g. in find_unsigned.cocci
> >
> > @u@ type T; unsigned T i; position p; @@  i at p < 0
> >
> >
> > what is the difference with
> >
> > @u@ type T; unsigned T i; @@
> >  i < 0
> >
> >
> > and what is the meaning of operator '@' ahead 'p'?
> >
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> 
> 
> 
> -- 
> Peter
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
> 

[Cocci] 答复: 答复: what is the function of 'position' in SMPL

[林嘉(程二 福州)]

Thanks,  So, what's the difference between 'statement' and 'expression'?
I guess statement means a complete line in c code, such as  's = socket(...);' while expression means an incomplete part, e.g. right hand of operator, 'socket(inet, 0, 0, 0)' ?



-----????-----
???: Julia Lawall [mailto:julia.lawall at lip6.fr] 
????: 2014?2?18? 15:35
???: ??(?? ??)
??: cocci at systeme.lip6.fr
??: Re: [Cocci] ??: what is the function of 'position' in SMPL

On Tue, 18 Feb 2014, ??(?? ??) wrote:

> Thank to all
> 
> What the differentce among 'Identifier declaration statement' in 
> metavariable portion?
> I  just guess they all can be used to express a name of variable or 
> function

An identifier can be used to express the name of a variable, a function, a field, the name of a structure, the name of an enum, the name of a #define, etc.

julia

> 
> For example:
> 
> @@
> Identifier a; or declaration a;?
> @@
> ...........
> 
> 
> 
> -----????-----
> ???: Peter Senna Tschudin [mailto:peter.senna at gmail.com]
> ????: 2014?2?17? 21:35
> ???: ??(?? ??)
> ??: cocci at systeme.lip6.fr
> ??: Re: [Cocci] what is the function of 'position' in SMPL
> 
> p is a position metavariable. It is useful when you want to print line and column numbers of interesting parts.
> 
> The @p means that the position of what matches to 'i' will be saved into p.
> 
> There is some information available at:
> http://coccinelle.lip6.fr/docs/main_grammar.pdf
> 
> On Mon, Feb 17, 2014 at 6:45 AM, ??(?? ??) <linjia@ruijie.com.cn> wrote:
> > e.g. in find_unsigned.cocci
> >
> > @u@ type T; unsigned T i; position p; @@  i at p < 0
> >
> >
> > what is the difference with
> >
> > @u@ type T; unsigned T i; @@
> >  i < 0
> >
> >
> > and what is the meaning of operator '@' ahead 'p'?
> >
> > _______________________________________________
> > Cocci mailing list
> > Cocci at systeme.lip6.fr
> > https://systeme.lip6.fr/mailman/listinfo/cocci
> 
> 
> 
> --
> Peter
> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
> 

[Cocci] 答复: 答复: what is the function of 'position' in SMPL

[Julia Lawall]

On Tue, 18 Feb 2014, ??(?? ??) wrote:

> Thanks,  So, what's the difference between 'statement' and 'expression'?
> I guess statement means a complete line in c code, such as  's = socket(...);' while expression means an incomplete part, e.g. right hand of operator, 'socket(inet, 0, 0, 0)' ?

A statement has no value.  So like s = socket(...); but also if (...)
return;, etc.

An expression has a value.  So like 3 + 4, f(), etc.  x = y without a
trailing semicolon is also an expression.

There is a simple grammar for a C-like language here:

http://marvin.cs.uidaho.edu/Teaching/CS445/c-Grammar.pdf

This is not the grammar that we follow, but it can give the idea.

julia