Problem report #20

Problem report #20

[Dan Nicolaescu]

CID: 20
Checker: FORWARD_NULL (help)
File: emacs/src/insdel.c
Function: prepare_to_modify_buffer
Description: Variable "preserve_ptr" tracked as NULL was passed to a function that dereferences it.

Event var_compare_op: Added "preserve_ptr" due to comparison "preserve_ptr != 0"
Also see events: [var_deref_model]
At conditional (1): "preserve_ptr != 0" taking false path

2044 	      if (preserve_ptr)
2045 		{
2046 		  Lisp_Object preserve_marker;
2047 		  struct gcpro gcpro1;
2048 		  preserve_marker = Fcopy_marker (make_number (*preserve_ptr), Qnil);
2049 		  GCPRO1 (preserve_marker);
2050 		  verify_interval_modification (current_buffer, start, end);
2051 		  *preserve_ptr = marker_position (preserve_marker);
2052 		  unchain_marker (XMARKER (preserve_marker));
2053 		  UNGCPRO;
2054 		}
2055 	      else
2056 		verify_interval_modification (current_buffer, start, end);
2057 	    }
2058 	
2059 	#ifdef CLASH_DETECTION

At conditional (2): "(current_buffer)->file_truename != Qnil" taking true path
At conditional (3): "(current_buffer)->filename != Qnil" taking true path
At conditional (4): "((current_buffer)->text)->save_modiff >= ((current_buffer)->text)->modiff" taking true path

2060 	  if (!NILP (current_buffer->file_truename)
2061 	      /* Make binding buffer-file-name to nil effective.  */
2062 	      && !NILP (current_buffer->filename)
2063 	      && SAVE_MODIFF >= MODIFF)
2064 	    lock_file (current_buffer->file_truename);
2065 	#else
2066 	  /* At least warn if this file has changed on disk since it was visited.  */
2067 	  if (!NILP (current_buffer->filename)
2068 	      && SAVE_MODIFF >= MODIFF
2069 	      && NILP (Fverify_visited_file_modtime (Fcurrent_buffer ()))
2070 	      && !NILP (Ffile_exists_p (current_buffer->filename)))
2071 	    call1 (intern ("ask-user-about-supersession-threat"),
2072 		   current_buffer->filename);
2073 	#endif /* not CLASH_DETECTION */
2074 	

Event var_deref_model: Variable "preserve_ptr" tracked as NULL was passed to a function that dereferences it. [model]
Also see events: [var_compare_op]

2075 	  signal_before_change (start, end, preserve_ptr);
2076 	

Re: Problem report #20

[Stuart D. Herring]

> Event var_deref_model: Variable "preserve_ptr" tracked as NULL was passed
> to a function that dereferences it. [model]
> Also see events: [var_compare_op]
>
> 2075 	  signal_before_change (start, end, preserve_ptr);

signal_before_change() doesn't even -use- preserve_ptr (is that a bug?),
much less dereference it.  No bug (here).

Davis

-- 
This product is sold by volume, not by mass.  If it appears too dense or
too sparse, it is because mass-energy conversion has occurred during
shipping.

Re: Problem report #20

[Stuart D. Herring]

I previously wrote:

>> Event var_deref_model: Variable "preserve_ptr" tracked as NULL was
>> passed
>> to a function that dereferences it. [model]
>> Also see events: [var_compare_op]
>>
>> 2075 	  signal_before_change (start, end, preserve_ptr);
>
> signal_before_change() doesn't even -use- preserve_ptr (is that a bug?),
> much less dereference it.  No bug (here).

Dan pointed out that signal_before_change does use preserve_ptr via
macros.    (Thanks.)  But the macros check the pointer before using it, so
signal_before_change(...,NULL) is safe, and so prepare_to_modify_buffer
(still) has no bug; it only uses preserve_ptr by passing it to
signal_before_change, and so the pointer is allowed to be NULL.

Davis

-- 
This product is sold by volume, not by mass.  If it appears too dense or
too sparse, it is because mass-energy conversion has occurred during
shipping.