Re: [PATCH] [gdb/exp] Fix cast handling for indirection

[Tom de Vries <tdevries@suse.de>]

On 5/3/24 04:31, Kevin Buettner wrote:
> On Thu,  2 May 2024 17:49:02 +0200
> Tom de Vries <tdevries@suse.de> wrote:
> 
>> Consider a test-case compiled without debug info, containing:
>> ...
>> char a = 'a';
>>
>> char *
>> a_loc (void)
>> {
>>    return &a;
>> }
>> ...
>>
>> We get:
>> ...
>> (gdb) p (char)*a_loc ()
>> Cannot access memory at address 0x10
>> ...
>>
>> There's a bug in unop_ind_base_operation::evaluate that evaluates
>> "(char)*a_loc ()" the same as:
>> ...
>> (gdb) p (char)*(char)a_loc ()
> 

Hi Kevin,

thanks for the review.

> This surprised me.  I would have thought that the evaluation would
> have been:
> 
>      (char)*(int)a_loc ()
> 
> ...due to the fact that functions lacking an explicit return type
> return 'int' in traditional C.
> 

Default-to-int was my first guess, but after debugging and investigating 
the code in unop_ind_base_operation::evaluate, I realized that the cast 
(passed in expect_type) is passed unmodified to the dereferenced expression.

I used this in the test-case, by changing the cast from int to char, to 
maximize the changes that the PR will produce an incorrect pointer.

>> Cannot access memory at address 0x10
>> ...
>>
>> Fix this by instead evaluating it the same as:
>> ...
>> (gdb) p (char)*(char *)a_loc ()
>> $1 = 97 'a'
>> ...
>>
>> Tested on x86_64-linux.
>>
>> PR exp/31693
>> Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=31693
> 
> Aside from the (possible) nit in the commit log that I mention
> above, the patch and test case look good to me.
> 
> Approved-by: Kevin Buettner <kevinb@redhat.com>
> 

Thanks, pushed unmodified.
- Tom

> Kevin
>