[Bug tools/24509] New: eu-readelf does not know how to dissect DW_AT_discr_list

[Bug tools/24509] New: eu-readelf does not know how to dissect DW_AT_discr_list

[tromey at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=24509

            Bug ID: 24509
           Summary: eu-readelf does not know how to dissect
                    DW_AT_discr_list
           Product: elfutils
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: tools
          Assignee: unassigned at sourceware dot org
          Reporter: tromey at sourceware dot org
                CC: elfutils-devel at sourceware dot org
  Target Milestone: ---

Consider the Ada test case in
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83935#c2.

Compiling to a .o and then using eu-readelf will yield something
like:

...
 [    6b]        variant              abbrev: 5
                 discr_list           (block1) 7 byte block: 01 01 ff ff ff ff
07

discr_list has a specified structure, and it would be nice
if eu-readelf could dump it more verbosely.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug tools/24509] eu-readelf does not know how to dissect DW_AT_discr_list

[mark at klomp dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=24509

Mark Wielaard <mark at klomp dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mark at klomp dot org

--- Comment #1 from Mark Wielaard <mark at klomp dot org> ---
Created attachment 11760
  --> https://sourceware.org/bugzilla/attachment.cgi?id=11760&action=edit
quick DW_AT_discr_list prototype

Here is a quick hack to show something more intelligent. It doesn't yet handle
signed values.

Could you provide a few different examples that we could use as testcases?

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug tools/24509] eu-readelf does not know how to dissect DW_AT_discr_list

[tromey at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=24509

--- Comment #2 from Tom Tromey <tromey at sourceware dot org> ---
Here's one that uses both ranges and discrete elements:

package Rng is

   type Rec (I : Integer) is record
      case I is
         when Positive =>
            C : Character;
            case I is
               when 1..15 | 17 | 23 =>
                  null;
               when others =>
                  N : Natural;
            end case;
         when others =>
            S : String (1 .. 10);
      end case;
   end record;

   R : Rec (1);

end Rng;



And here's one that uses unsigned types:

package Urng is

   type Unsigned is mod 65536;
   type Rec (I : Unsigned) is record
      case I is
         when 17 | 23 | 32768..65535 =>
            null;
         when others =>
            S : String (1 .. 10);
      end case;
   end record;

   R : Rec (1);

end Urng;

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug tools/24509] eu-readelf does not know how to dissect DW_AT_discr_list

[tromey at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=24509

--- Comment #3 from Tom Tromey <tromey at sourceware dot org> ---
To check for negative values, I think you can add another
clause, here's what I did:

package Rng is

   type Rec (I : Integer) is record
      case I is
         when Positive =>
            C : Character;
            case I is
               when 1..15 | 17 | 23 =>
                  null;
               when others =>
                  N : Natural;
            end case;
         when -52..-1 =>
            Q: Natural;
         when others =>
            S : String (1 .. 10);
      end case;
   end record;

   R : Rec (1);

end Rng;


The proposed patch does not handle this properly:

 [    9f]        variant              abbrev: 5
                 discr_list           (block1) range: 76-127

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug tools/24509] eu-readelf does not know how to dissect DW_AT_discr_list

[mark at klomp dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=24509

Mark Wielaard <mark at klomp dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #11760|0                           |1
        is obsolete|                            |

--- Comment #4 from Mark Wielaard <mark at klomp dot org> ---
Created attachment 11762
  --> https://sourceware.org/bugzilla/attachment.cgi?id=11762&action=edit
discr_list patch that handles signed/unsigned variant_part types

Updated patch that handles signed/unsigned discriminators.

So this is what I used as testcase to have a signed (Integer) and unsigned
(Character) discr.

package Rng is

   type Rec (I : Integer) is record
      case I is
         when Positive =>
            C : Character;
            case I is
               when 1..15 | 17 | 23..255 =>
                  null;
               when others =>
                  N : Natural;
            end case;
         when -52..-1 =>
            Q: Natural;
         when -53 =>
            R: Character;
         when others =>
            S : String (1 .. 10);
      end case;
   end record;

   Subtype Uppercase is Character Range 'A'..'Z';
   Subtype Lowercase is Character Range 'a'..'z';
   Subtype Numbers is Character Range '0'..'9';

   type RecC (C : Character) is record
      case C is
         when '@' =>
            B : Boolean;
         when Lowercase | Uppercase | Numbers =>
            D : Character;
         when others =>
            null;
      end case;
   end record;

   R : Rec (1);
   RC : RecC ('$');

end Rng;

Does the output look like what you would expect?

BTW. The byte_size expressions are impressive!

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug tools/24509] eu-readelf does not know how to dissect DW_AT_discr_list

[tromey at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=24509

--- Comment #5 from Tom Tromey <tromey at sourceware dot org> ---
FWIW the output looks reasonable to me now.  Thanks for doing this!

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug tools/24509] eu-readelf does not know how to dissect DW_AT_discr_list

[mark at klomp dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=24509

Mark Wielaard <mark at klomp dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED

--- Comment #6 from Mark Wielaard <mark at klomp dot org> ---
Proposed patch, added default/empty block handling, data bounds checks and
testcases: https://sourceware.org/ml/elfutils-devel/2019-q2/msg00072.html

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug tools/24509] eu-readelf does not know how to dissect DW_AT_discr_list

[mark at klomp dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=24509

Mark Wielaard <mark at klomp dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED

--- Comment #7 from Mark Wielaard <mark at klomp dot org> ---
commit 643cbb275d65533472c0f53391f9fc1d5d9a2efc
Author: Mark Wielaard <mark@klomp.org>
Date:   Sun May 5 23:18:36 2019 +0200

    readelf: Decode DW_AT_discr_list block attributes.

    Decode DW_AT_descr_list blocks using the DW_DSC values.
    This requires knowing the signedness of the discriminant.
    Which means the attr_callback function needs access to the
    parent DIE. Pass the whole DIE path, plus the current level.
    That way the type of the discriminant can be looked up in
    the variant_part (parent) DIE of the variant DIE (which has
    the discr_list attribute).

    Add a testcase using both signed and unsigned discriminants.

    https://sourceware.org/bugzilla/show_bug.cgi?id=24509

    Signed-off-by: Mark Wielaard <mark@klomp.org>

-- 
You are receiving this mail because:
You are on the CC list for the bug.