[PATCH] Stale FIXME

[PATCH] Stale FIXME

[David Deephanphongs]

There appears to be a stale FIXME in main.c - it warns about the
argument after the program-name (corearg):
  /* FIXME: The documentation says this can be a "ProcID". as well. */

GDB /does/ treat the corearg as a PID, though.
First it tries to open corearg as a core-file.  If that fails, it attempts
to use it as a PID (main.c:615-623).

----- Changelog Entry ------
2001-05-22  David Deephanphongs  <david@llamedos.org>

	* main.c: Removed stale FIXME - GDB /does/ try to use the second
	argument as a PID if the arg. isn't a core file.

----- End Changelog Entry -----

----- Patch -----
diff -c3p gdb/main.c gdb-new/main.c
*** gdb/main.c	Mon May  7 15:03:11 2001
--- gdb-new/main.c	Tue May 22 02:53:54 2001
*************** extern int gdbtk_test (char *);
*** 473,479 ****
  	  execarg = argv[optind];
  	  break;
  	case 2:
- 	  /* FIXME: The documentation says this can be a "ProcID". as well. */
  	  corearg = argv[optind];
  	  break;
  	case 3:
--- 473,478 ----
------ End Patch ------

-- 
Brother Preptil, the master of the music, had described Brutha's voice as
putting him in mind of a disappointed vulture arriving too late at the dead
donkey.
        -- (Terry Pratchett, Small Gods)

Re: [PATCH] Stale FIXME

[Fernando Nasser]

David Deephanphongs wrote:
> 
> There appears to be a stale FIXME in main.c - it warns about the
> argument after the program-name (corearg):
>   /* FIXME: The documentation says this can be a "ProcID". as well. */
> 
> GDB /does/ treat the corearg as a PID, though.
> First it tries to open corearg as a core-file.  If that fails, it attempts
> to use it as a PID (main.c:615-623).

Yes, and I use it all the time to debug gdb with gdb.  That was a badly worded FIXME text.   It was supposed to say that we should not be overloading variables and filling them with something that is not related to its name -- 4 bytes is a small price to pay for clarity IMO.

I still would like to keep a note there that saying that corearg may be a PID and not a core file name. Something like:

	  /* COREARG can be the name of a core file or
             the "ProcID" of a process we should attach to. */

Can you submit your patch with the replacement text?  I will ask permission to Elena and check it in.

Thanks for the catch.

-- 
Fernando Nasser
Red Hat - Toronto                       E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9

Re: [PATCH] Stale FIXME

[Elena Zannoni]

Fernando Nasser writes:
 > David Deephanphongs wrote:
 > > 
 > > There appears to be a stale FIXME in main.c - it warns about the
 > > argument after the program-name (corearg):
 > >   /* FIXME: The documentation says this can be a "ProcID". as well. */
 > > 
 > > GDB /does/ treat the corearg as a PID, though.
 > > First it tries to open corearg as a core-file.  If that fails, it attempts
 > > to use it as a PID (main.c:615-623).
 > 
 > Yes, and I use it all the time to debug gdb with gdb.  That was a badly worded FIXME text.   It was supposed to say that we should not be overloading variables and filling them with something that is not related to its name -- 4 bytes is a small price to pay for clarity IMO.
 > 
 > I still would like to keep a note there that saying that corearg may be a PID and not a core file name. Something like:
 > 
 > 	  /* COREARG can be the name of a core file or
 >              the "ProcID" of a process we should attach to. */
 > 
 > Can you submit your patch with the replacement text?  I will ask permission to Elena and check it in.
 > 

I see, you added the comment... :-) So I trust you know what you meant
to say. Please, modify the comment as you indicated above.

Thanks
Elena


 > Thanks for the catch.
 > 
 > -- 
 > Fernando Nasser
 > Red Hat - Toronto                       E-Mail:  fnasser@redhat.com
 > 2323 Yonge Street, Suite #300
 > Toronto, Ontario   M4P 2C9
 > 

Re: [PATCH] Stale FIXME

[Fernando Nasser]

Elena Zannoni wrote:
> 
> Fernando Nasser writes:
>  > David Deephanphongs wrote:
>  > >
>  > > There appears to be a stale FIXME in main.c - it warns about the
>  > > argument after the program-name (corearg):
>  > >   /* FIXME: The documentation says this can be a "ProcID". as well. */
>  > >
>  > > GDB /does/ treat the corearg as a PID, though.
>  > > First it tries to open corearg as a core-file.  If that fails, it attempts
>  > > to use it as a PID (main.c:615-623).
>  >
>  > Yes, and I use it all the time to debug gdb with gdb.  That was a badly worded FIXME text.   It was supposed to say that we should not be overloading variables and filling them with something that is not related to its name -- 4 bytes is a small price to pay for clarity IMO.
>  >
>  > I still would like to keep a note there that saying that corearg may be a PID and not a core file name. Something like:
>  >
>  >        /* COREARG can be the name of a core file or
>  >              the "ProcID" of a process we should attach to. */
>  >
>  > Can you submit your patch with the replacement text?  I will ask permission to Elena and check it in.
>  >
> 
> I see, you added the comment... :-) So I trust you know what you meant
> to say. Please, modify the comment as you indicated above.
> 
> Thanks
> Elena
> 

Thank you Elena.

There is also a patch from  David Deephanphongs  that proposes creating
a new variable in this same thread.

-- 
Fernando Nasser
Red Hat Canada Ltd.                     E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9

Re: [PATCH] Stale FIXME

[David Deephanphongs]

Thus spake Fernando Nasser (fnasser@redhat.com):

> Thank you Elena.
> 
> There is also a patch from  David Deephanphongs  that proposes creating
> a new variable in this same thread.
> 

FYI, I'm waiting for some base aye/nay from Elena before I deal with any
renaming-of-variables issues in the second patch.

Dave

-- 
"Well, 'scuse me. I was jus' tryin' to save the world."
        -- Gaspode the wonder dog
           (Terry Pratchett, Moving Pictures)