[Bug default/19708] New: See who runs out of patience first?

[Bug default/19708] New: See who runs out of patience first?

[michi.henning at canonical dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=19708

            Bug ID: 19708
           Summary: See who runs out of patience first?
           Product: libabigail
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: default
          Assignee: dodji at redhat dot com
          Reporter: michi.henning at canonical dot com
                CC: libabigail at sourceware dot org
  Target Milestone: ---

Referring to bug 19706 again.

I start out with the following suppressions:

[suppress_function]
    name_regexp = boost::.*
    change_kind = all
    allow_other_aliases = false

[suppress_function]
    name_regexp = std::.*
    change_kind = all
    allow_other_aliases = false

[suppress_function]
    name_regexp = unity::scopes::testing::.*
    change_kind = all
    allow_other_aliases = false

[suppress_function]
    name_regexp = unity::scopes::internal::.*
    change_kind = all
    allow_other_aliases = false

Now I run abidiff. I get (trailing lines for run_scope suppressed):

2 functions with some indirect sub-type change:

  [C]'method unity::scopes::ActivationQueryBase::ActivationQueryBase(const
unity::scopes::Result&, const unity::scopes::ActionMetadata&)' at
ActivationQueryBase.cpp:29:1 has some indirect sub-type changes:
    parameter 1 of type 'const unity::scopes::Result&' has sub-type changes:
      in referenced type 'const unity::scopes::Result':
        in unqualified underlying type 'class unity::scopes::Result' at
Result.h:50:1:
          no data member change (1 filtered);

  [C]'method void unity::scopes::Runtime::run_scope(unity::scopes::ScopeBase*,
const std::__cxx11::string&)' at Runtime.h:120:1 has some indirect sub-type
changes:
    parameter 1 of type 'unity::scopes::ScopeBase*' has sub-type changes:
      in pointed to type 'class unity::scopes::ScopeBase' at ScopeBase.h:133:1:
        no member function changes (6 filtered);

        1 data member change:
         type of 'std::unique_ptr<unity::scopes::internal::ScopeBaseImpl,
std::default_delete<unity::scopes::internal::ScopeBaseImpl> >
unity::scopes::ScopeBase::p' changed:
          ...

OK, so I add the following suppressions:

[suppress_function]
    name = unity::scopes::Runtime::run_scope
    change_kind = function-subtype-change
    allow_other_aliases = false

[suppress_function]
    name_regexp = unity::scopes::ActivationQueryBase::ActivationQueryBase
    change_kind = function-subtype-change
    allow_other_aliases = false

This gets rid of the complaint about run_scope and the constructor, but now I
get this:

1 function with some indirect sub-type change:

  [C]'method unity::scopes::internal::ActivationQueryBaseImpl*
unity::scopes::ActivationQueryBase::fwd()' at ActivationQueryBase.cpp:80:1 has
some indirect sub-type changes: 
    return type changed:
      in pointed to type 'class
unity::scopes::internal::ActivationQueryBaseImpl' at
ActivationQueryBaseImpl.h:36:1:
        1 data member change:
         type of 'unity::scopes::Result
unity::scopes::internal::ActivationQueryBaseImpl::result_' changed:
           no data member change (1 filtered);

OK, fair enough, I edit the previous suppression to look like this instead and
try again.

[suppress_function]
    name_regexp = unity::scopes::ActivationQueryBase::.*
    change_kind = function-subtype-change
    allow_other_aliases = false

I no longer get the complaints about ActivationQueryBase, but:

1 function with some indirect sub-type change:

  [C]'method unity::scopes::ActivationResponse::ActivationResponse(const
unity::scopes::Result&)' at ActivationResponse.h:80:1 has some indirect
sub-type changes:
    parameter 1 of type 'const unity::scopes::Result&' has sub-type changes:
      in referenced type 'const unity::scopes::Result':
        in unqualified underlying type 'class unity::scopes::Result' at
Result.h:50:1:
          no data member change (1 filtered);

Fine, I'm getting the hang of it now and add:

[suppress_function]
    name_regexp = unity::scopes::ActivationResponse::.*
    change_kind = function-subtype-change
    allow_other_aliases = false

Now I get:

1 function with some indirect sub-type change:

  [C]'method
unity::scopes::CategorisedResult::CategorisedResult(unity::scopes::CategorisedResult&&)'
at CategorisedResult.h:41:1 has some indirect sub-type changes:
    parameter 1 of type 'unity::scopes::CategorisedResult&&' has sub-type
changes:
      in referenced type 'class unity::scopes::CategorisedResult' at
CategorisedResult.h:41:1:
        no base class change (1 filtered);

At that point, I lost patience.

It seems like a game of whack-a-mole: every time I add a suppression, a new
complaint pops up and, when I suppress that complaint, I get yet another
different complaint.

This isn't workable as is. Is there any way around this? (If possible, with
1.0.rc2?)

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

[Bug default/19708] See who runs out of patience first?

[michi.henning at canonical dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=19708

--- Comment #1 from Michi Henning <michi.henning at canonical dot com> ---
OK, so it looks like what I really need is this:

[suppress_function]
    name_regexp = unity::scopes::internal::.*
    change_kind = all
    allow_other_aliases = false

[suppress_type]
    name_regexp = unity::scopes::internal::.*

With this, I get the silence I was hoping for.

But I find the incremental complaints I described in the OP surprising. Is this
intentional?

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