[v2,02/15] Fix calling ifunc functions when resolver has debug info and different name

  Currently, on Fedora 27 (glibc 2.26), if you try to call strlen in the
inferior you get:

 (gdb) p strlen ("hello")
 $1 = (size_t (*)(const char *)) 0x7ffff554aac0 <__strlen_avx2>

strlen is an ifunc function, and what we see above is the result of
calling the ifunc resolver in the inferior.  That returns a pointer to
the actual target function that implements strlen on my machine.  GDB
should have turned around and called the resolver automatically
without the user noticing.

This is was caused by commit:

  commit bf223d3e808e6fec9ee165d3d48beb74837796de
  Date: Mon Aug 21 11:34:32 2017 +0100

      Handle function aliases better (PR gdb/19487, errno printing)

which added the find_function_alias_target call to c-exp.y, to try to
find an alias with debug info for a minsym.  For ifunc symbols, that
finds the ifunc's resolver if it has debug info (in the example it's
called "strlen_ifunc"), with the result that GDB calls that as a
regular function.

After this commit, we get now get:

  (top-gdb) p strlen ("hello")
  '__strlen_avx2' has unknown return type; cast the call to its declared return type

Which is correct, because __strlen_avx2 is written in assembly.
That'll be improved in a following patch, though.

gdb/ChangeLog:
yyyy-mm-dd  Pedro Alves  <palves@redhat.com>

	* c-exp.y (variable production): Skip finding an alias for ifunc
	symbols.
---
 gdb/c-exp.y | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
  

On 2018-03-25 03:19 PM, Pedro Alves wrote:
> Currently, on Fedora 27 (glibc 2.26), if you try to call strlen in the
> inferior you get:
> 
>  (gdb) p strlen ("hello")
>  $1 = (size_t (*)(const char *)) 0x7ffff554aac0 <__strlen_avx2>
> 
> strlen is an ifunc function, and what we see above is the result of
> calling the ifunc resolver in the inferior.  That returns a pointer to
> the actual target function that implements strlen on my machine.  GDB
> should have turned around and called the resolver automatically
> without the user noticing.
> 
> This is was caused by commit:
> 
>   commit bf223d3e808e6fec9ee165d3d48beb74837796de
>   Date: Mon Aug 21 11:34:32 2017 +0100
> 
>       Handle function aliases better (PR gdb/19487, errno printing)
> 
> which added the find_function_alias_target call to c-exp.y, to try to
> find an alias with debug info for a minsym.  For ifunc symbols, that
> finds the ifunc's resolver if it has debug info (in the example it's
> called "strlen_ifunc"), with the result that GDB calls that as a
> regular function.
> 
> After this commit, we get now get:
> 
>   (top-gdb) p strlen ("hello")
>   '__strlen_avx2' has unknown return type; cast the call to its declared return type
> 
> Which is correct, because __strlen_avx2 is written in assembly.
> That'll be improved in a following patch, though.
> 
> gdb/ChangeLog:
> yyyy-mm-dd  Pedro Alves  <palves@redhat.com>
> 
> 	* c-exp.y (variable production): Skip finding an alias for ifunc
> 	symbols.
> ---
>  gdb/c-exp.y | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/gdb/c-exp.y b/gdb/c-exp.y
> index 8dc3c068a5e..e2ea07cd792 100644
> --- a/gdb/c-exp.y
> +++ b/gdb/c-exp.y
> @@ -1081,7 +1081,9 @@ variable:	name_not_typename
>  				 is important for example for "p
>  				 *__errno_location()".  */
>  			      symbol *alias_target
> -				= find_function_alias_target (msymbol);
> +				= (msymbol.minsym->type != mst_text_gnu_ifunc
> +				   ? find_function_alias_target (msymbol)
> +				   : NULL);
>  			      if (alias_target != NULL)
>  				{
>  				  write_exp_elt_opcode (pstate, OP_VAR_VALUE);
> 

Just one question, to which I really don't have a preconceived answer:

Would it make sense to add that filtering to find_function_alias_target or
another function deeper in the call chain instead?  In other words, would
it ever make sense for find_function_alias_target to return an non-NULL
result for an mst_text_gnu_ifunc minimal symbol?

Simon

Simon
  

Hi Simon,

On 04/01/2018 04:44 AM, Simon Marchi wrote:
>> diff --git a/gdb/c-exp.y b/gdb/c-exp.y
>> index 8dc3c068a5e..e2ea07cd792 100644
>> --- a/gdb/c-exp.y
>> +++ b/gdb/c-exp.y
>> @@ -1081,7 +1081,9 @@ variable:	name_not_typename
>>  				 is important for example for "p
>>  				 *__errno_location()".  */
>>  			      symbol *alias_target
>> -				= find_function_alias_target (msymbol);
>> +				= (msymbol.minsym->type != mst_text_gnu_ifunc
>> +				   ? find_function_alias_target (msymbol)
>> +				   : NULL);
>>  			      if (alias_target != NULL)
>>  				{
>>  				  write_exp_elt_opcode (pstate, OP_VAR_VALUE);
>>
> 
> Just one question, to which I really don't have a preconceived answer:
> 
> Would it make sense to add that filtering to find_function_alias_target or
> another function deeper in the call chain instead?  In other words, would
> it ever make sense for find_function_alias_target to return an non-NULL
> result for an mst_text_gnu_ifunc minimal symbol?

For ifuncs, find_function_alias_target will finds the debug symbol for
the resolver.  If that has a different name from the ifunc (it will
if you use gcc's attribute ifunc, then it'll work the same as any other
minsym, in the sense that it'll find an alias.

So from that angle, the function works as intended, and it
wasn't clear that we'll always want to treat ifuncs differently.
Note this is the only caller of find_function_alias_target currently.

Another reason for leaving the check here is that patch #4 adds
another case of "do this for ifuncs" just above this code, but for
the "sym.symbol" case.  It felt better to keep both of the
"for ifunc, do this" cases close together.

Thanks,
Pedro Alves