[GRASS-dev] seg fault with temporal modules

Hi devs,

I get ‘Segmentation fault (core dumped)’ after running temporal commands. Commands work, but they finish with segfault. I started observing this yesterday evening that I recompiled trunk to get lattest temporal changes. I was working with temporal modules the day before that with no issues.

Luca suggested off-list that my DB must be corrupted somehow because others do not have that issue and neither myself in a different machine with freshly compiled trunk. So, I’m lost…

To test further I did 2 things:
1). remove my nc_basic location and download it again in my grassdata:

GRASS 7.7.svn (nc_basic_spm_grass7):~ > t.connect -d
Default driver / database set to:
driver: sqlite
database: $GISDBASE/$LOCATION_NAME/$MAPSET/tgis/sqlite.db

GRASS 7.7.svn (nc_basic_spm_grass7):~ > t.list
WARNING: Temporal database connection defined as:
/home/veroandreo/grassdata/nc_basic_spm_grass7/PERMANENT/tgis/sqlite.db
But database file does not exist.
Creating temporal database:
/home/veroandreo/grassdata/nc_basic_spm_grass7/PERMANENT/tgis/sqlite.db

Segmentation fault (core dumped)

2). create a new grassdata and download a different location

GRASS 7.7.svn (worldlocation):~ > t.connect -d
Default driver / database set to:
driver: sqlite
database: $GISDBASE/$LOCATION_NAME/$MAPSET/tgis/sqlite.db

GRASS 7.7.svn (worldlocation):~ > t.list
WARNING: Temporal database connection defined as:
/home/veroandreo/grassdata_test/worldlocation/PERMANENT/tgis/sqlite.db
But database file does not exist.
Creating temporal database:
/home/veroandreo/grassdata_test/worldlocation/PERMANENT/tgis/sqlite.db

Segmentation fault (core dumped)

same result… any other ideas??

Thanks much in advance!
Vero

On Fri, Sep 28, 2018 at 3:18 PM Veronica Andreo <veroandreo@gmail.com> wrote:

Hi devs,

I get 'Segmentation fault (core dumped)' after running temporal commands. Commands work, but they finish with segfault.

We has a similar (?) problem here:
    On Sat, Jul 28, 2018 at 11:40 PM Veronica Andreo wrote...

The reason was a broken external library. Maybe that problem came back
in a system update?

You may try t.info with strace like the last time.

Best
Markus

On Fri, Sep 28, 2018 at 4:03 PM Markus Neteler <neteler@osgeo.org> wrote:

On Fri, Sep 28, 2018 at 3:18 PM Veronica Andreo <veroandreo@gmail.com> wrote:
> I get 'Segmentation fault (core dumped)' after running temporal commands. Commands work, but they finish with segfault.

We has a similar (?) problem here:
    On Sat, Jul 28, 2018 at 11:40 PM Veronica Andreo wrote...

The reason was a broken external library. Maybe that problem came back
in a system update?

You may try t.info with strace like the last time.

Seems that
https://bugzilla.redhat.com/show_bug.cgi?id=1605231
may be back, i.e. bug reintroduced in openblas-0.3.3 (the update just
came in through the Fedora update channel)...

I'll check later on my system.

Best
Markus

On Fri, Sep 28, 2018 at 5:37 PM Markus Neteler <neteler@osgeo.org> wrote:

On Fri, Sep 28, 2018 at 4:03 PM Markus Neteler <neteler@osgeo.org> wrote:
> On Fri, Sep 28, 2018 at 3:18 PM Veronica Andreo <veroandreo@gmail.com> wrote:
> > I get 'Segmentation fault (core dumped)' after running temporal commands. Commands work, but they finish with segfault.

...

> You may try t.info with strace like the last time.

Seems that
https://bugzilla.redhat.com/show_bug.cgi?id=1605231
may be back, i.e. bug reintroduced in openblas-0.3.3 (the update just
came in through the Fedora update channel)...

Reported upstream as
https://bugzilla.redhat.com/show_bug.cgi?id=1634060

Markus

Markus Neteler <neteler@osgeo.org> schrieb am Fr., 28. Sep. 2018, 17:46:

On Fri, Sep 28, 2018 at 5:37 PM Markus Neteler <neteler@osgeo.org> wrote:

On Fri, Sep 28, 2018 at 4:03 PM Markus Neteler <neteler@osgeo.org> wrote:

On Fri, Sep 28, 2018 at 3:18 PM Veronica Andreo <veroandreo@gmail.com> wrote:

I get ‘Segmentation fault (core dumped)’ after running temporal commands. Commands work, but they finish with segfault.

You may try t.info with strace like the last time.

Seems that
https://bugzilla.redhat.com/show_bug.cgi?id=1605231
may be back, i.e. bug reintroduced in openblas-0.3.3 (the update just
came in through the Fedora update channel)…

Reported upstream as
https://bugzilla.redhat.com/show_bug.cgi?id=1634060

Fixed upstream:

<bugzilla@redhat.com> schrieb am So., 30. Sep. 2018, 15:44:
— Comment #2 from Fedora Update System <updates@fedoraproject.org> —
openblas-0.3.3-2.fc28 has been submitted as an update to Fedora 28.
https://bodhi.fedoraproject.org/updates/FEDORA-2018-dcb9bcdb5c

It may take a day to propagate through the update channel.

Other distros will follow or still have an older openblas package version…

Cheers,
Markus

Hi Markus,

Thanks much for reporting the issue and then reporting back here. I had completely forgotten that we had the same issue shortly ago… eheh

Cheers,
Vero

El dom., 30 sep. 2018 13:04, Markus Neteler <neteler@osgeo.org> escribió:

Markus Neteler <neteler@osgeo.org> schrieb am Fr., 28. Sep. 2018, 17:46:

On Fri, Sep 28, 2018 at 5:37 PM Markus Neteler <neteler@osgeo.org> wrote:

On Fri, Sep 28, 2018 at 4:03 PM Markus Neteler <neteler@osgeo.org> wrote:

On Fri, Sep 28, 2018 at 3:18 PM Veronica Andreo <veroandreo@gmail.com> wrote:

I get ‘Segmentation fault (core dumped)’ after running temporal commands. Commands work, but they finish with segfault.

You may try t.info with strace like the last time.

Seems that
https://bugzilla.redhat.com/show_bug.cgi?id=1605231
may be back, i.e. bug reintroduced in openblas-0.3.3 (the update just
came in through the Fedora update channel)…

Reported upstream as
https://bugzilla.redhat.com/show_bug.cgi?id=1634060

Fixed upstream:

<bugzilla@redhat.com> schrieb am So., 30. Sep. 2018, 15:44:
— Comment #2 from Fedora Update System <updates@fedoraproject.org> —
openblas-0.3.3-2.fc28 has been submitted as an update to Fedora 28.
https://bodhi.fedoraproject.org/updates/FEDORA-2018-dcb9bcdb5c

It may take a day to propagate through the update channel.

Other distros will follow or still have an older openblas package version…

Cheers,
Markus

On Sun, Sep 30, 2018 at 9:08 PM Veronica Andreo <veroandreo@gmail.com> wrote:

Hi Markus,

Thanks much for reporting the issue and then reporting back here. I had completely forgotten that we had the same issue shortly ago... eheh

I just upgraded openblas-0.3.3-2 from the Fedora testing server, it is
functional again and t.list works as expected.

Cheers,
Markus