[PATCH]: ptype of pointer to typedef

[PATCH]: ptype of pointer to typedef

[Andrew Burgess]

Given a test program like this:

typedef char (my_char_array) [4];
my_char_array array = {'a', 'b', 'c', 'd'};
my_char_array *array_p = &array;

typedef char my_char;
my_char array2 [4] = {'a', 'b', 'c', 'd'};
my_char (*array2_p)[4] = &array2;

int
main ( void )
{
   return 0;
}

Then in gdb:

(gdb) ptype array2_p
type = char (*)[4]
(gdb) ptype array_p
type = char *)[4]


Notice in the array_p case, the missing "(" character.

Patch below include a fix, and a test case.  The test case builds on an 
existing test that was marked as xfail in a few cases, I've extended the 
xfail to cover the new case I've added, but I have no way to check if 
this is the right thing to do or not, I'm happy to change the patch if 
anyone has an opinion.

Ok to commit?

Cheers,
Andrew

gdb/ChangeLog

2012-09-03  Andrew Bugess  <aburgess@broadcom.com>

	* c-typeprint.c (c_type_print_varspec_prefix): Pass through the
	passed_a_ptr flag when displaying typedef types.

diff --git a/gdb/c-typeprint.c b/gdb/c-typeprint.c
index c2a775a..b51ced8 100644
--- a/gdb/c-typeprint.c
+++ b/gdb/c-typeprint.c
@@ -306,7 +306,7 @@ c_type_print_varspec_prefix (struct type *type,

      case TYPE_CODE_TYPEDEF:
        c_type_print_varspec_prefix (TYPE_TARGET_TYPE (type),
-                                  stream, show, 0, 0);
+                                  stream, show, passed_a_ptr, 0);
        break;

      case TYPE_CODE_UNDEF:

gdb/testsuite/ChangeLog

2012-09-03  Andrew Burgess  <aburgess@broadcom.com>

	* gdb.base/ptype.exp: Test ptype on a pointer to a typedef.

diff --git a/gdb/testsuite/gdb.base/ptype.exp 
b/gdb/testsuite/gdb.base/ptype.exp
index c7bede2..0eef17d 100644
--- a/gdb/testsuite/gdb.base/ptype.exp
+++ b/gdb/testsuite/gdb.base/ptype.exp
@@ -370,6 +370,10 @@ if {!$gcc_compiled} then { setup_xfail "rs6000-*-*" 
"i*86-*
  if {$hp_aCC_compiler} {setup_xfail "hppa*-*-*"}
  gdb_test "ptype t_char_array" "type = (|unsigned )char \\\[0?\\\]"

+if {!$gcc_compiled} then { setup_xfail "rs6000-*-*" "i*86-*-sysv4*" }
+if {$hp_aCC_compiler} {setup_xfail "hppa*-*-*"}
+gdb_test "ptype pv_char_array" "type = (|unsigned )char 
\\(\\*\\)\\\[0?\\\]"
+
  #
  ##
  ## test ptype command with pointers

Re: [PATCH]: ptype of pointer to typedef

[Andrew Burgess]

On 03/09/2012 10:45 AM, Andrew Burgess wrote:

> 2012-09-03  Andrew Bugess  <aburgess@broadcom.com>
>

That's slightly embarrassing, apparently I can't spell my own name :(

Andrew

Re: [PATCH]: ptype of pointer to typedef

[Andrew Burgess]

Ping!

On 03/09/2012 10:45 AM, Andrew Burgess wrote:
> Given a test program like this:
>
> typedef char (my_char_array) [4];
> my_char_array array = {'a', 'b', 'c', 'd'};
> my_char_array *array_p = &array;
>
> typedef char my_char;
> my_char array2 [4] = {'a', 'b', 'c', 'd'};
> my_char (*array2_p)[4] = &array2;
>
> int
> main ( void )
> {
>    return 0;
> }
>
> Then in gdb:
>
> (gdb) ptype array2_p
> type = char (*)[4]
> (gdb) ptype array_p
> type = char *)[4]
>
>
> Notice in the array_p case, the missing "(" character.
>
> Patch below include a fix, and a test case.  The test case builds on an
> existing test that was marked as xfail in a few cases, I've extended the
> xfail to cover the new case I've added, but I have no way to check if
> this is the right thing to do or not, I'm happy to change the patch if
> anyone has an opinion.
>
> Ok to commit?
>
> Cheers,
> Andrew
>
> gdb/ChangeLog
>
> 2012-09-03  Andrew Bugess  <aburgess@broadcom.com>
>
>      * c-typeprint.c (c_type_print_varspec_prefix): Pass through the
>      passed_a_ptr flag when displaying typedef types.
>
> diff --git a/gdb/c-typeprint.c b/gdb/c-typeprint.c
> index c2a775a..b51ced8 100644
> --- a/gdb/c-typeprint.c
> +++ b/gdb/c-typeprint.c
> @@ -306,7 +306,7 @@ c_type_print_varspec_prefix (struct type *type,
>
>       case TYPE_CODE_TYPEDEF:
>         c_type_print_varspec_prefix (TYPE_TARGET_TYPE (type),
> -                                  stream, show, 0, 0);
> +                                  stream, show, passed_a_ptr, 0);
>         break;
>
>       case TYPE_CODE_UNDEF:
>
> gdb/testsuite/ChangeLog
>
> 2012-09-03  Andrew Burgess  <aburgess@broadcom.com>
>
>      * gdb.base/ptype.exp: Test ptype on a pointer to a typedef.
>
> diff --git a/gdb/testsuite/gdb.base/ptype.exp
> b/gdb/testsuite/gdb.base/ptype.exp
> index c7bede2..0eef17d 100644
> --- a/gdb/testsuite/gdb.base/ptype.exp
> +++ b/gdb/testsuite/gdb.base/ptype.exp
> @@ -370,6 +370,10 @@ if {!$gcc_compiled} then { setup_xfail "rs6000-*-*"
> "i*86-*
>   if {$hp_aCC_compiler} {setup_xfail "hppa*-*-*"}
>   gdb_test "ptype t_char_array" "type = (|unsigned )char \\\[0?\\\]"
>
> +if {!$gcc_compiled} then { setup_xfail "rs6000-*-*" "i*86-*-sysv4*" }
> +if {$hp_aCC_compiler} {setup_xfail "hppa*-*-*"}
> +gdb_test "ptype pv_char_array" "type = (|unsigned )char
> \\(\\*\\)\\\[0?\\\]"
> +
>   #
>   ##
>   ## test ptype command with pointers
>
>
>
>

Re: [PATCH]: ptype of pointer to typedef

[Tom Tromey]

>>>>> "Andrew" == Andrew Burgess <aburgess@broadcom.com> writes:

Andrew> The test case builds on an existing test that was marked as
Andrew> xfail in a few cases, I've extended the xfail to cover the new
Andrew> case I've added, but I have no way to check if this is the right
Andrew> thing to do or not, I'm happy to change the patch if anyone has
Andrew> an opinion.

I would just drop the xfails.
If they are needed, somebody with the affected environment can write them.

The patch is ok with that change.

Andrew> +gdb_test "ptype pv_char_array" "type = (|unsigned )char
Andrew> \\(\\*\\)\\\[0?\\\]"
Andrew> +

Your email got wrapped somehow.

Tom

Re: [PATCH]: ptype of pointer to typedef

[Andrew Burgess]

On 10/09/2012 4:44 PM, Tom Tromey wrote:
>>>>>> "Andrew" == Andrew Burgess <aburgess@broadcom.com> writes:
>
> Andrew> The test case builds on an existing test that was marked as
> Andrew> xfail in a few cases, I've extended the xfail to cover the new
> Andrew> case I've added, but I have no way to check if this is the right
> Andrew> thing to do or not, I'm happy to change the patch if anyone has
> Andrew> an opinion.
>
> I would just drop the xfails.
> If they are needed, somebody with the affected environment can write them.

Done.

>
> The patch is ok with that change.
>
> Andrew> +gdb_test "ptype pv_char_array" "type = (|unsigned )char
> Andrew> \\(\\*\\)\\\[0?\\\]"
> Andrew> +
>
> Your email got wrapped somehow.

Apologies.

Committed with the above change.

Thanks,
Andrew