[bug #45833] support load/save of classdef objects

Please use the bug tracker to post updates to a bug report.  The mailing list is intended as a read-only notification stream.  Info posted to this mailing list address won't appear in the tracker database where it is most useful.

Follow-up Comment #95, bug #45833 (group octave):

There are a couple of compiler warnings when building Octave for a 32-bit
target.
E.g.,
https://github.com/gnu-octave/octave/actions/runs/21731757111/job/62687797007#step:13:4312

../libinterp/corefcn/ls-mat-subsys.cc: In function 'octave_value
octave::load_mcos_object(const octave_value&, bool)':
../libinterp/corefcn/ls-mat-subsys.cc:174:33: warning: comparison of integer
expressions of different signedness: 'octave_idx_type' {aka 'int'} and
'uint32_t' {aka 'unsigned int'} [-Wsign-compare]
  174 |   for (octave_idx_type i = 0; i < objdims; i++)
      |                               ~~^~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc: In member function 'bool
octave::subsystem_handler::read_filewrapper(const Cell&, bool)':
../libinterp/corefcn/ls-mat-subsys.cc:326:22: warning: cast from 'const
octave_int<unsigned char>*' to 'const uint32_t*' {aka 'const unsigned int*'}
increases required alignment of target type [-Wcast-align]
  326 |   uint32_t version = reinterpret_cast<const uint32_t&>
(fwrap_metadata_array(0));
      |
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc:335:17: warning: cast from 'const
octave_int<unsigned char>*' to 'const uint32_t*' {aka 'const unsigned int*'}
increases required alignment of target type [-Wcast-align]
  335 |   m_num_names = reinterpret_cast<const uint32_t&>
(fwrap_metadata_array(4));
      |
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc:342:29: warning: cast from 'const
octave_int<unsigned char>*' to 'const uint32_t*' {aka 'const unsigned int*'}
increases required alignment of target type [-Wcast-align]
  342 |       m_region_offsets[i] = reinterpret_cast<const uint32_t&>
      |                             ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  343 |                               (fwrap_metadata_array(8 + i * 4));
      |                               ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc:388:22: warning: cast from 'const
octave_int<unsigned char>*' to 'const uint32_t*' {aka 'const unsigned int*'}
increases required alignment of target type [-Wcast-align]
  388 |       uint32_t val = reinterpret_cast<const uint32_t&>
      |                      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  389 |                        (fwrap_metadata_array(m_region_offsets[0] + i *
4));
      |
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc:401:22: warning: cast from 'const
octave_int<unsigned char>*' to 'const uint32_t*' {aka 'const unsigned int*'}
increases required alignment of target type [-Wcast-align]
  401 |       uint32_t val = reinterpret_cast<const uint32_t&>
      |                      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  402 |                        (fwrap_metadata_array(m_region_offsets[2] + i *
4));
      |
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc:414:22: warning: cast from 'const
octave_int<unsigned char>*' to 'const uint32_t*' {aka 'const unsigned int*'}
increases required alignment of target type [-Wcast-align]
  414 |       uint32_t val = reinterpret_cast<const uint32_t&>
      |                      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  415 |                        (fwrap_metadata_array(m_region_offsets[3] + i *
4));
      |
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc:427:22: warning: cast from 'const
octave_int<unsigned char>*' to 'const uint32_t*' {aka 'const unsigned int*'}
increases required alignment of target type [-Wcast-align]
  427 |       uint32_t val = reinterpret_cast<const uint32_t&>
      |                      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  428 |                        (fwrap_metadata_array(m_region_offsets[1] + i *
4));
      |
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc:440:22: warning: cast from 'const
octave_int<unsigned char>*' to 'const uint32_t*' {aka 'const unsigned int*'}
increases required alignment of target type [-Wcast-align]
  440 |       uint32_t val = reinterpret_cast<const uint32_t&>
      |                      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  441 |                        (fwrap_metadata_array(m_region_offsets[4] + i *
4));
      |
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc: In member function 'octave_map
octave::subsystem_handler::get_object_properties(uint32_t, uint32_t, bool,
bool)':
../libinterp/corefcn/ls-mat-subsys.cc:503:51: warning: comparison of integer
expressions of different signedness: 'octave_idx_type' {aka 'int'} and 'const
uint32_t' {aka 'const unsigned int'} [-Wsign-compare]
  503 |   for (octave_idx_type current_id = 0; current_id < obj_type_id;
current_id++)
      |                                        ~~~~~~~~~~~^~~~~~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc:521:33: warning: comparison of integer
expressions of different signedness: 'octave_idx_type' {aka 'int'} and
'uint32_t' {aka 'unsigned int'} [-Wsign-compare]
  521 |   for (octave_idx_type i = 0; i < nprops && ptr + 2 < end_ptr; i++)
      |                               ~~^~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc: In member function 'uint32_t
octave::subsystem_handler::get_name_index(const std::string&)':
../libinterp/corefcn/ls-mat-subsys.cc:622:26: warning: comparison of integer
expressions of different signedness: 'uint32_t' {aka 'unsigned int'} and
'octave_idx_type' {aka 'int'} [-Wsign-compare]
  622 |   for (uint32_t i = 0; i < m_prop_class_names.numel (); i++)
      |                        ~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../libinterp/corefcn/ls-mat-subsys.cc: In member function 'void
octave::subsystem_handler::process_object_properties(const
std::vector<std::tuple<octave_map, unsigned int, bool> >&, const
std::vector<bool>&, uint32_t)':
../libinterp/corefcn/ls-mat-subsys.cc:750:55: warning: comparison of integer
expressions of different signedness: 'octave_idx_type' {aka 'int'} and
'uint32_t' {aka 'unsigned int'} [-Wsign-compare]
  750 |           for (octave_idx_type prop_idx = 0; prop_idx < num_props;
prop_idx++)
      |                                              ~~~~~~~~~^~~~~~~~~~~


When building for targets with 32-bit pointers, `octave_idx_type` is a signed
32-bit integer. Casting `octave_idx_type` to or from 32-bit unsigned integers
can lead to integer overflow errors.

Are the integers in the -v7 files really unsigned? I seem to recall that
saving in that file format is limited to 2 GB (not 4 GB).



    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?45833>

_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/

-----BEGIN PGP SIGNATURE-----

iHUEABYIAB0WIQQk97aszIMMAvLLwm6qLAuaBUf3TgUCaYYSPQAKCRCqLAuaBUf3
TvYhAQC5XG/p40SBe+J7VcmM7+owWKhyfq3gEqCfPhnGoPXs2AD8CJ/Y0sj5sGGW
qZUsdGp+qPIM6oSNjeCHn2lCQ4OLaQA=
=8iCh
-----END PGP SIGNATURE-----