Re: [PATCH] ctf-reader: Fix length in dynamic array definition

[Guillermo Martinez <guillermo.e.martinez@oracle.com>]

On Saturday, November 27, 2021 12:03:23 PM CST Jose E. Marchesi wrote:
> 
> > Hello libabigail team,
> >
> > This patch fix the `length' property of `subrange' node when
> > we define a dynamic length array,
> >
> > Please let me know your comments,
> >
> > Thanks in advanced!,
> > Guillermo
> >
> > Defining an array type with dynamic length, node `subrange'
> > in the abixml file doesn't write the accurate `length'
> > property `infinite', instead `1' is written:
> >     <subrange length='1' .../>
> >
> > So, member function `array_type_def::subrange_type::is_infinite'
> > is set when `upper_bound' value is equal to `0' or UINT64_MAX.
> >
> > 	* src/abg-ctf-reader.cc	(process_ctf_array_type):
> > 	set subrange_type::is_infinite when `upper_bound' value
> > 	is equal to `0' or UINT64_MAX.
> >
> > Signed-off-by: Guillermo E. Martinez <guillermo.e.martinez@oracle.com>
> > ---
> >  src/abg-ctf-reader.cc | 20 ++++++++++++++++++++
> >  1 file changed, 20 insertions(+)
> >
> > diff --git a/src/abg-ctf-reader.cc b/src/abg-ctf-reader.cc
> > index c571d825..687686d9 100644
> > --- a/src/abg-ctf-reader.cc
> > +++ b/src/abg-ctf-reader.cc
> > @@ -472,6 +472,7 @@ process_ctf_array_type(read_context *ctxt,
> >  {
> >    array_type_def_sptr result;
> >    ctf_arinfo_t ctf_ainfo;
> > +  bool is_infinite = false;
> >  
> >    /* First, get the information about the CTF array.  */
> >    if (static_cast<ctf_id_t>(ctf_array_info(ctf_dictionary,
> > @@ -507,6 +508,19 @@ process_ctf_array_type(read_context *ctxt,
> >    lower_bound.set_unsigned(0); /* CTF supports C only.  */
> >    upper_bound.set_unsigned(nelems > 0 ? nelems - 1 : 0U);
> >  
> > +  if (upper_bound.get_unsigned_value() == 0)
> > +    {
> > +      is_infinite = true;
> > +    }
> 
> These braces are superfluous.
OK, I'll will remove curly braces in next version.

> > +  if (UINT64_MAX == upper_bound.get_unsigned_value())
> > +    {
> > +      // If the upper_bound size is the max of the integer value, then
> > +      // it most certainly means infinite size.
> > +      is_infinite = true;
> > +      upper_bound.set_unsigned(0);
> > +    }
> 
> Is that specified by the CTF spec?
Ohh .. sorry is for DWARF, will be remove in next patch.
> > +
> >    subrange.reset(new array_type_def::subrange_type(ctxt->ir_env,
> >                                                     "",
> >                                                     lower_bound,
> > @@ -517,6 +531,12 @@ process_ctf_array_type(read_context *ctxt,
> >    if (!subrange)
> >      return result;
> >  
> > +  subrange->is_infinite(is_infinite);
> > +  ABG_ASSERT(subrange->is_infinite()
> > +             || (subrange->get_length() ==
> > +                 (uint64_t) (subrange->get_upper_bound()
> > +                      - subrange->get_lower_bound() + 1)));
> 
> Hm why this assert?
After previous comment it's not required.
> 
> > +
> >    add_decl_to_scope(subrange, tunit->get_global_scope());
> >    canonicalize(subrange);
> >    subranges.push_back(subrange);
> 
Thanks Jose!
guillermo