Re: [GRASS-user] v.net.iso - segmentation fault error

[Eduardo Corbelle Rico]

O Ven, 18-11-2016 ás 22:45 +0100, Markus Metz escribiu:
> 
> 
> On Fri, Nov 18, 2016 at 4:48 PM, Eduardo Corbelle Rico  l...@gmx.net> wrote:
> >
> >
> > O Xov, 17-11-2016 ás 08:52 +0100, Markus Metz escribiu:
> > >
> > > On Wed, Nov 16, 2016 at 7:14 PM, Eduardo Corbelle Rico
>  > > l...@gmx.net> wrote:
> > > >
> > > > O Mér, 16-11-2016 ás 15:22 +0100, Markus Metz escribiu:
> > > > >
> > > > > On Wed, Nov 16, 2016 at 11:44 AM, Eduardo Corbelle Rico
> > >  > > > > e...@gmx.net> wrote:
> > > > > >
> > > > > > O Xov, 27-10-2016 ás 14:56 +0200, Markus Metz escribiu:
> > > > > > > On Thu, Oct 27, 2016 at 11:16 AM, Eduardo Corbelle Rico
> > > > > > >  wrote:
> > > > > > > >
> > > > > > > > O Mér, 26-10-2016 ás 23:18 +0200, Markus Metz escribiu:
> > > > > > > > > On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle
> Rico
> > > > > > > > >  wrote:
> > > > > > > > > >
> > > > > > > > > > Dear all,
> > > > > > > > > >
> > > > > > > > > > I am unable to use v.net.iso because of a
> "segmentation
> > > > > fault"
> > > > > > > > > > error. I
> > > > > > > > > >  have a script that used to work correctly in GRASS
> 6.4
> > > (32
> > > > > > > > > > bit)
> > > > > > > > > > but
> > > > > > > > > > produces this error after I switched to GRASS 7.0.4
> (64
> > > > > bit).
> > > > > > > > > >
> > > > > > > > > > If I try to follow both examples shown in https://g
> rass
> > > .osg
> > > > > eo.o
> > > > > > > > > > rg/g
> > > > > > > > > > rass
> > > > > > > > > > 70/manuals/v.net.iso.html (using the Spearfish
> dataset
> > > for
> > > > > > > > > > GRASS 7)
> > > > > > > > > > the
> > > > > > > > > > same error appears:
> > > > > > > > > >
> > > > > > > > > > > Building graph...
> > > > > > > > > > > Registering arcs...
> > > > > > > > > > > Segmentation fault
> > > > > > > > >
> > > > > > > > > The segmentation fault must happen in the vector
> > > libraries. I
> > > > > > > > > tested
> > > > > > > > > on Linux and can not reproduce the segmentation
> fault.
> > > > > Valgrind
> > > > > > > > > also
> > > > > > > > > does not show anything that could cause a
> segmentation
> > > fault.
> > > > > Can
> > > > > > > > > you
> > > > > > > > > provide a gdb backtrace?
> > > > > > > > >
> > > > > > > > > Markus M
> > > > > > > >
> > > > > > > > Thank you Markus,
> > > > > > > >
> > > > > > > > My system is debian Stretch, and the following lines
> > > describe
> > > > > my
> > > > > > > > attempt at generating a gdb backtrace. Please let me
> know
> > > if I
> > > > > > > > should
> > > > > > > > do otherwise.
> > > > > > >
> > > > > > > So far so good. The module crashes in a library function
> that
> > > > > comes
> > > > > > > from libavl without modifications, but I can still not
> see
> > > why it
> > > > > > > could cause a segmentation fault. Can you recompile with
> the
> > > > > compiler
> > > > > > > debugging option -g? gdb will then be able to provide
> more
> > > > > > > information:
> > > > > > >
> > > > > > > make distclean
> > > > > > > CFLAGS="-g" ./configure 
> > > > > > > make
> > > > > > >
> > > > > > > optionally make install
> > > > > > >
> > > > > > > Then start GRASS and use gdb for v.net.iso as before.
> > > > > > >
> > > > > > > Markus M
> > > > > > >
> > > > > >
> > > > > > Dear Markus,
> > > > > >
> > > > > > Following those instructions does not seem to have changed
> the
> > > > > > resulting backtrace info (copied below). Maybe I missed
> > > something
> > > > > along
> > > > > > the way?
> > > > > >
> > > > > > -
> > > > >
> > > > >
> > > > > Is this something Debian specific that debugging symbols must
> be
> > > > > explicitly activated? Are you using GCC or another compiler?
> > > > >
> > > > > Alternatively, you can also try valgrind:
> > > > >
> > > > > CMD="v.net.iso input=myroads_net output=myroads_net_iso
> > > > > center_cats=1-10 costs=1000,2000,5000 --o"
> > > > > valgrind --tool=memcheck --log-file=v.net.iso.vg.log $CMD
> > > > >
> > > > > Markus M
> > > > >
> > > >
> > > > You were right: it was my fault not having grass-core-dbgsym
> and
> > > grass-
> > > > dev-dbgsym packages installed.
> > >
> > > You do not need these packages when you compile from source.
> > >
> > > > I copied below the results of the
> > > > debugging commands:
> > > >
> > > > 
> > > > GRASS 7.0.5 (spearfish60_grass7):~ > gdb `which v.net.iso`
> > > > GNU gdb (Debian 7.11.1-2+b1) 7.11.1
> > > > Copyright (C) 2016 Free Software Foundation, Inc.
> > > > License GPLv3+: GNU GPL version 3 or later  nses
> > > /gpl
> > > > .html>
> > > > This is free software: you are free to change and redistribute
> it.
> > > > There is NO WARRANTY, to the extent permitted by law.  Type
> "show
> > > > copying"
> > > > and "show warranty" for details.
> > > > This GDB was configured as "x86_64-linux-gnu".
> > > > Type "show configuration" for configuration details.
> > > > For bug reporting instructions, please see:

Re: [GRASS-user] v.net.iso - segmentation fault error

[Markus Metz]

On Fri, Nov 18, 2016 at 4:48 PM, Eduardo Corbelle Rico <
eduardo.corbe...@gmx.net> wrote:
>
>
> O Xov, 17-11-2016 ás 08:52 +0100, Markus Metz escribiu:
> >
> > On Wed, Nov 16, 2016 at 7:14 PM, Eduardo Corbelle Rico  > l...@gmx.net> wrote:
> > >
> > > O Mér, 16-11-2016 ás 15:22 +0100, Markus Metz escribiu:
> > > >
> > > > On Wed, Nov 16, 2016 at 11:44 AM, Eduardo Corbelle Rico
> >  > > > e...@gmx.net> wrote:
> > > > >
> > > > > O Xov, 27-10-2016 ás 14:56 +0200, Markus Metz escribiu:
> > > > > > On Thu, Oct 27, 2016 at 11:16 AM, Eduardo Corbelle Rico
> > > > > >  wrote:
> > > > > > >
> > > > > > > O Mér, 26-10-2016 ás 23:18 +0200, Markus Metz escribiu:
> > > > > > > > On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle Rico
> > > > > > > >  wrote:
> > > > > > > > >
> > > > > > > > > Dear all,
> > > > > > > > >
> > > > > > > > > I am unable to use v.net.iso because of a "segmentation
> > > > fault"
> > > > > > > > > error. I
> > > > > > > > >  have a script that used to work correctly in GRASS 6.4
> > (32
> > > > > > > > > bit)
> > > > > > > > > but
> > > > > > > > > produces this error after I switched to GRASS 7.0.4 (64
> > > > bit).
> > > > > > > > >
> > > > > > > > > If I try to follow both examples shown in https://grass
> > .osg
> > > > eo.o
> > > > > > > > > rg/g
> > > > > > > > > rass
> > > > > > > > > 70/manuals/v.net.iso.html (using the Spearfish dataset
> > for
> > > > > > > > > GRASS 7)
> > > > > > > > > the
> > > > > > > > > same error appears:
> > > > > > > > >
> > > > > > > > > > Building graph...
> > > > > > > > > > Registering arcs...
> > > > > > > > > > Segmentation fault
> > > > > > > >
> > > > > > > > The segmentation fault must happen in the vector
> > libraries. I
> > > > > > > > tested
> > > > > > > > on Linux and can not reproduce the segmentation fault.
> > > > Valgrind
> > > > > > > > also
> > > > > > > > does not show anything that could cause a segmentation
> > fault.
> > > > Can
> > > > > > > > you
> > > > > > > > provide a gdb backtrace?
> > > > > > > >
> > > > > > > > Markus M
> > > > > > >
> > > > > > > Thank you Markus,
> > > > > > >
> > > > > > > My system is debian Stretch, and the following lines
> > describe
> > > > my
> > > > > > > attempt at generating a gdb backtrace. Please let me know
> > if I
> > > > > > > should
> > > > > > > do otherwise.
> > > > > >
> > > > > > So far so good. The module crashes in a library function that
> > > > comes
> > > > > > from libavl without modifications, but I can still not see
> > why it
> > > > > > could cause a segmentation fault. Can you recompile with the
> > > > compiler
> > > > > > debugging option -g? gdb will then be able to provide more
> > > > > > information:
> > > > > >
> > > > > > make distclean
> > > > > > CFLAGS="-g" ./configure 
> > > > > > make
> > > > > >
> > > > > > optionally make install
> > > > > >
> > > > > > Then start GRASS and use gdb for v.net.iso as before.
> > > > > >
> > > > > > Markus M
> > > > > >
> > > > >
> > > > > Dear Markus,
> > > > >
> > > > > Following those instructions does not seem to have changed the
> > > > > resulting backtrace info (copied below). Maybe I missed
> > something
> > > > along
> > > > > the way?
> > > > >
> > > > > -
> > > >
> > > >
> > > > Is this something Debian specific that debugging symbols must be
> > > > explicitly activated? Are you using GCC or another compiler?
> > > >
> > > > Alternatively, you can also try valgrind:
> > > >
> > > > CMD="v.net.iso input=myroads_net output=myroads_net_iso
> > > > center_cats=1-10 costs=1000,2000,5000 --o"
> > > > valgrind --tool=memcheck --log-file=v.net.iso.vg.log $CMD
> > > >
> > > > Markus M
> > > >
> > >
> > > You were right: it was my fault not having grass-core-dbgsym and
> > grass-
> > > dev-dbgsym packages installed.
> >
> > You do not need these packages when you compile from source.
> >
> > > I copied below the results of the
> > > debugging commands:
> > >
> > > 
> > > GRASS 7.0.5 (spearfish60_grass7):~ > gdb `which v.net.iso`
> > > GNU gdb (Debian 7.11.1-2+b1) 7.11.1
> > > Copyright (C) 2016 Free Software Foundation, Inc.
> > > License GPLv3+: GNU GPL version 3 or later  > /gpl
> > > .html>
> > > This is free software: you are free to change and redistribute it.
> > > There is NO WARRANTY, to the extent permitted by law.  Type "show
> > > copying"
> > > and "show warranty" for details.
> > > This GDB was configured as "x86_64-linux-gnu".
> > > Type "show configuration" for configuration details.
> > > For bug reporting instructions, please see:
> > > .
> > > Find the GDB manual and other documentation resources online at:
> > > .
> > > For help, type "help".
> > > Type "apropos word" to search for commands related to "word"...
> > > Reading symbols from /usr/lib/grass70/bin/v.net.iso...Reading
> > symbols
> > > from 

Re: [GRASS-user] v.net.iso - segmentation fault error

[Eduardo Corbelle Rico]

O Xov, 17-11-2016 ás 08:52 +0100, Markus Metz escribiu:
> 
> On Wed, Nov 16, 2016 at 7:14 PM, Eduardo Corbelle Rico  l...@gmx.net> wrote:
> >
> > O Mér, 16-11-2016 ás 15:22 +0100, Markus Metz escribiu:
> > >
> > > On Wed, Nov 16, 2016 at 11:44 AM, Eduardo Corbelle Rico
>  > > e...@gmx.net> wrote:
> > > >
> > > > O Xov, 27-10-2016 ás 14:56 +0200, Markus Metz escribiu:
> > > > > On Thu, Oct 27, 2016 at 11:16 AM, Eduardo Corbelle Rico
> > > > >  wrote:
> > > > > >
> > > > > > O Mér, 26-10-2016 ás 23:18 +0200, Markus Metz escribiu:
> > > > > > > On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle Rico
> > > > > > >  wrote:
> > > > > > > >
> > > > > > > > Dear all,
> > > > > > > >
> > > > > > > > I am unable to use v.net.iso because of a "segmentation
> > > fault"
> > > > > > > > error. I
> > > > > > > >  have a script that used to work correctly in GRASS 6.4
> (32
> > > > > > > > bit)
> > > > > > > > but
> > > > > > > > produces this error after I switched to GRASS 7.0.4 (64
> > > bit).
> > > > > > > >
> > > > > > > > If I try to follow both examples shown in https://grass
> .osg
> > > eo.o
> > > > > > > > rg/g
> > > > > > > > rass
> > > > > > > > 70/manuals/v.net.iso.html (using the Spearfish dataset
> for
> > > > > > > > GRASS 7)
> > > > > > > > the
> > > > > > > > same error appears:
> > > > > > > >
> > > > > > > > > Building graph...
> > > > > > > > > Registering arcs...
> > > > > > > > > Segmentation fault
> > > > > > >
> > > > > > > The segmentation fault must happen in the vector
> libraries. I
> > > > > > > tested
> > > > > > > on Linux and can not reproduce the segmentation fault.
> > > Valgrind
> > > > > > > also
> > > > > > > does not show anything that could cause a segmentation
> fault.
> > > Can
> > > > > > > you
> > > > > > > provide a gdb backtrace?
> > > > > > >
> > > > > > > Markus M
> > > > > >
> > > > > > Thank you Markus,
> > > > > >
> > > > > > My system is debian Stretch, and the following lines
> describe
> > > my
> > > > > > attempt at generating a gdb backtrace. Please let me know
> if I
> > > > > > should
> > > > > > do otherwise.
> > > > >
> > > > > So far so good. The module crashes in a library function that
> > > comes
> > > > > from libavl without modifications, but I can still not see
> why it
> > > > > could cause a segmentation fault. Can you recompile with the
> > > compiler
> > > > > debugging option -g? gdb will then be able to provide more
> > > > > information:
> > > > >
> > > > > make distclean
> > > > > CFLAGS="-g" ./configure 
> > > > > make
> > > > >
> > > > > optionally make install
> > > > >
> > > > > Then start GRASS and use gdb for v.net.iso as before.
> > > > >
> > > > > Markus M
> > > > >
> > > >
> > > > Dear Markus,
> > > >
> > > > Following those instructions does not seem to have changed the
> > > > resulting backtrace info (copied below). Maybe I missed
> something
> > > along
> > > > the way?
> > > >
> > > > -
> > >
> > >
> > > Is this something Debian specific that debugging symbols must be
> > > explicitly activated? Are you using GCC or another compiler?
> > >
> > > Alternatively, you can also try valgrind:
> > >
> > > CMD="v.net.iso input=myroads_net output=myroads_net_iso
> > > center_cats=1-10 costs=1000,2000,5000 --o"
> > > valgrind --tool=memcheck --log-file=v.net.iso.vg.log $CMD
> > >
> > > Markus M
> > >
> >
> > You were right: it was my fault not having grass-core-dbgsym and
> grass-
> > dev-dbgsym packages installed.
> 
> You do not need these packages when you compile from source.
> 
> > I copied below the results of the
> > debugging commands:
> >
> > 
> > GRASS 7.0.5 (spearfish60_grass7):~ > gdb `which v.net.iso`
> > GNU gdb (Debian 7.11.1-2+b1) 7.11.1
> > Copyright (C) 2016 Free Software Foundation, Inc.
> > License GPLv3+: GNU GPL version 3 or later  /gpl
> > .html>
> > This is free software: you are free to change and redistribute it.
> > There is NO WARRANTY, to the extent permitted by law.  Type "show
> > copying"
> > and "show warranty" for details.
> > This GDB was configured as "x86_64-linux-gnu".
> > Type "show configuration" for configuration details.
> > For bug reporting instructions, please see:
> > .
> > Find the GDB manual and other documentation resources online at:
> > .
> > For help, type "help".
> > Type "apropos word" to search for commands related to "word"...
> > Reading symbols from /usr/lib/grass70/bin/v.net.iso...Reading
> symbols
> > from /usr/lib/debug/.build-
> > id/5a/060fc6c3da41153d82601e3fa37ba6d88e9c6c.debug...done.
> > done.
> >
> >
> > (gdb) run input=myroads_net output=myroads_net_iso center_cats=1-
> 10
> > costs=1000,2000,5000
> 
> This seems to be the GRASS version installed with your package
> manager because make install installs by default to /usr/local/:
> > Starting program: 

Re: [GRASS-user] v.net.iso - segmentation fault error

[Markus Metz]

On Wed, Nov 16, 2016 at 7:14 PM, Eduardo Corbelle Rico <
eduardo.corbe...@gmx.net> wrote:
>
> O Mér, 16-11-2016 ás 15:22 +0100, Markus Metz escribiu:
> >
> > On Wed, Nov 16, 2016 at 11:44 AM, Eduardo Corbelle Rico  > e...@gmx.net> wrote:
> > >
> > > O Xov, 27-10-2016 ás 14:56 +0200, Markus Metz escribiu:
> > > > On Thu, Oct 27, 2016 at 11:16 AM, Eduardo Corbelle Rico
> > > >  wrote:
> > > > >
> > > > > O Mér, 26-10-2016 ás 23:18 +0200, Markus Metz escribiu:
> > > > > > On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle Rico
> > > > > >  wrote:
> > > > > > >
> > > > > > > Dear all,
> > > > > > >
> > > > > > > I am unable to use v.net.iso because of a "segmentation
> > fault"
> > > > > > > error. I
> > > > > > >  have a script that used to work correctly in GRASS 6.4 (32
> > > > > > > bit)
> > > > > > > but
> > > > > > > produces this error after I switched to GRASS 7.0.4 (64
> > bit).
> > > > > > >
> > > > > > > If I try to follow both examples shown in https://grass.osg
> > eo.o
> > > > > > > rg/g
> > > > > > > rass
> > > > > > > 70/manuals/v.net.iso.html (using the Spearfish dataset for
> > > > > > > GRASS 7)
> > > > > > > the
> > > > > > > same error appears:
> > > > > > >
> > > > > > > > Building graph...
> > > > > > > > Registering arcs...
> > > > > > > > Segmentation fault
> > > > > >
> > > > > > The segmentation fault must happen in the vector libraries. I
> > > > > > tested
> > > > > > on Linux and can not reproduce the segmentation fault.
> > Valgrind
> > > > > > also
> > > > > > does not show anything that could cause a segmentation fault.
> > Can
> > > > > > you
> > > > > > provide a gdb backtrace?
> > > > > >
> > > > > > Markus M
> > > > >
> > > > > Thank you Markus,
> > > > >
> > > > > My system is debian Stretch, and the following lines describe
> > my
> > > > > attempt at generating a gdb backtrace. Please let me know if I
> > > > > should
> > > > > do otherwise.
> > > >
> > > > So far so good. The module crashes in a library function that
> > comes
> > > > from libavl without modifications, but I can still not see why it
> > > > could cause a segmentation fault. Can you recompile with the
> > compiler
> > > > debugging option -g? gdb will then be able to provide more
> > > > information:
> > > >
> > > > make distclean
> > > > CFLAGS="-g" ./configure 
> > > > make
> > > >
> > > > optionally make install
> > > >
> > > > Then start GRASS and use gdb for v.net.iso as before.
> > > >
> > > > Markus M
> > > >
> > >
> > > Dear Markus,
> > >
> > > Following those instructions does not seem to have changed the
> > > resulting backtrace info (copied below). Maybe I missed something
> > along
> > > the way?
> > >
> > > -
> >
> >
> > Is this something Debian specific that debugging symbols must be
> > explicitly activated? Are you using GCC or another compiler?
> >
> > Alternatively, you can also try valgrind:
> >
> > CMD="v.net.iso input=myroads_net output=myroads_net_iso
> > center_cats=1-10 costs=1000,2000,5000 --o"
> > valgrind --tool=memcheck --log-file=v.net.iso.vg.log $CMD
> >
> > Markus M
> >
>
> You were right: it was my fault not having grass-core-dbgsym and grass-
> dev-dbgsym packages installed.

You do not need these packages when you compile from source.

> I copied below the results of the
> debugging commands:
>
> 
> GRASS 7.0.5 (spearfish60_grass7):~ > gdb `which v.net.iso`
> GNU gdb (Debian 7.11.1-2+b1) 7.11.1
> Copyright (C) 2016 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later  .html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.  Type "show
> copying"
> and "show warranty" for details.
> This GDB was configured as "x86_64-linux-gnu".
> Type "show configuration" for configuration details.
> For bug reporting instructions, please see:
> .
> Find the GDB manual and other documentation resources online at:
> .
> For help, type "help".
> Type "apropos word" to search for commands related to "word"...
> Reading symbols from /usr/lib/grass70/bin/v.net.iso...Reading symbols
> from /usr/lib/debug/.build-
> id/5a/060fc6c3da41153d82601e3fa37ba6d88e9c6c.debug...done.
> done.
>
>
> (gdb) run input=myroads_net output=myroads_net_iso center_cats=1-10
> costs=1000,2000,5000

This seems to be the GRASS version installed with your package manager
because make install installs by default to /usr/local/:
> Starting program: /usr/lib/grass70/bin/v.net.iso input=myroads_net
> output=myroads_net_iso center_cats=1-10 costs=1000,2000,5000
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib/x86_64-linux-
> gnu/libthread_db.so.1".
> Building graph...
> Registering arcs...
>
> Program received signal SIGSEGV, Segmentation fault.
> tavl_probe 

Re: [GRASS-user] v.net.iso - segmentation fault error

[Eduardo Corbelle Rico]

O Mér, 16-11-2016 ás 15:22 +0100, Markus Metz escribiu:
> 
> On Wed, Nov 16, 2016 at 11:44 AM, Eduardo Corbelle Rico  e...@gmx.net> wrote:
> >
> > O Xov, 27-10-2016 ás 14:56 +0200, Markus Metz escribiu:
> > > On Thu, Oct 27, 2016 at 11:16 AM, Eduardo Corbelle Rico
> > >  wrote:
> > > >
> > > > O Mér, 26-10-2016 ás 23:18 +0200, Markus Metz escribiu:
> > > > > On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle Rico
> > > > >  wrote:
> > > > > >
> > > > > > Dear all,
> > > > > >
> > > > > > I am unable to use v.net.iso because of a "segmentation
> fault"
> > > > > > error. I
> > > > > >  have a script that used to work correctly in GRASS 6.4 (32
> > > > > > bit)
> > > > > > but
> > > > > > produces this error after I switched to GRASS 7.0.4 (64
> bit).
> > > > > >
> > > > > > If I try to follow both examples shown in https://grass.osg
> eo.o
> > > > > > rg/g
> > > > > > rass
> > > > > > 70/manuals/v.net.iso.html (using the Spearfish dataset for
> > > > > > GRASS 7)
> > > > > > the
> > > > > > same error appears:
> > > > > >
> > > > > > > Building graph...
> > > > > > > Registering arcs...
> > > > > > > Segmentation fault
> > > > >
> > > > > The segmentation fault must happen in the vector libraries. I
> > > > > tested
> > > > > on Linux and can not reproduce the segmentation fault.
> Valgrind
> > > > > also
> > > > > does not show anything that could cause a segmentation fault.
> Can
> > > > > you
> > > > > provide a gdb backtrace?
> > > > >
> > > > > Markus M
> > > >
> > > > Thank you Markus,
> > > >
> > > > My system is debian Stretch, and the following lines describe
> my
> > > > attempt at generating a gdb backtrace. Please let me know if I
> > > > should
> > > > do otherwise.
> > >
> > > So far so good. The module crashes in a library function that
> comes
> > > from libavl without modifications, but I can still not see why it
> > > could cause a segmentation fault. Can you recompile with the
> compiler
> > > debugging option -g? gdb will then be able to provide more
> > > information:
> > >
> > > make distclean
> > > CFLAGS="-g" ./configure 
> > > make
> > >
> > > optionally make install
> > >
> > > Then start GRASS and use gdb for v.net.iso as before.
> > >
> > > Markus M
> > >
> >
> > Dear Markus,
> >
> > Following those instructions does not seem to have changed the
> > resulting backtrace info (copied below). Maybe I missed something
> along
> > the way?
> >
> > -
> 
> 
> Is this something Debian specific that debugging symbols must be
> explicitly activated? Are you using GCC or another compiler?
> 
> Alternatively, you can also try valgrind:
> 
> CMD="v.net.iso input=myroads_net output=myroads_net_iso
> center_cats=1-10 costs=1000,2000,5000 --o"
> valgrind --tool=memcheck --log-file=v.net.iso.vg.log $CMD
> 
> Markus M
> 

You were right: it was my fault not having grass-core-dbgsym and grass-
dev-dbgsym packages installed. I copied below the results of the
debugging commands:


GRASS 7.0.5 (spearfish60_grass7):~ > gdb `which v.net.iso`
GNU gdb (Debian 7.11.1-2+b1) 7.11.1
Copyright (C) 2016 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later 
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show
copying"
and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
.
Find the GDB manual and other documentation resources online at:
.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/lib/grass70/bin/v.net.iso...Reading symbols
from /usr/lib/debug/.build-
id/5a/060fc6c3da41153d82601e3fa37ba6d88e9c6c.debug...done.
done.


(gdb) run input=myroads_net output=myroads_net_iso center_cats=1-10 
costs=1000,2000,5000
Starting program: /usr/lib/grass70/bin/v.net.iso input=myroads_net
output=myroads_net_iso center_cats=1-10 costs=1000,2000,5000
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-
gnu/libthread_db.so.1".
Building graph...
Registering arcs...

Program received signal SIGSEGV, Segmentation fault.
tavl_probe (tree=tree@entry=0x5599b6c0, item=item@entry=0x5599f
7b0)
at tavl.c:146
146 tavl.c: Non hai tal ficheiro ou directorio.
(gdb) bt full
#0  tavl_probe (tree=tree@entry=0x5599b6c0, item=item@entry=0x5
599f7b0)
at tavl.c:146
y = 0x0
z = 0x5599b6c0
p = 0x0
q = 
n = 0x5599f7d0
w = 
dir = 
da =
"\023\000\000\000\000\000\000\000\000\vb\367\377\177\000\000\300\266\23
1UUU\000\000\030\000\000\000\000\000\000"
k = 0
__PRETTY_FUNCTION__ = "tavl_probe"
#1  

Re: [GRASS-user] v.net.iso - segmentation fault error

[Markus Metz]

On Wed, Nov 16, 2016 at 11:44 AM, Eduardo Corbelle Rico <
eduardo.corbe...@gmx.net> wrote:
>
> O Xov, 27-10-2016 ás 14:56 +0200, Markus Metz escribiu:
> > On Thu, Oct 27, 2016 at 11:16 AM, Eduardo Corbelle Rico
> >  wrote:
> > >
> > > O Mér, 26-10-2016 ás 23:18 +0200, Markus Metz escribiu:
> > > > On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle Rico
> > > >  wrote:
> > > > >
> > > > > Dear all,
> > > > >
> > > > > I am unable to use v.net.iso because of a "segmentation fault"
> > > > > error. I
> > > > >  have a script that used to work correctly in GRASS 6.4 (32
> > > > > bit)
> > > > > but
> > > > > produces this error after I switched to GRASS 7.0.4 (64 bit).
> > > > >
> > > > > If I try to follow both examples shown in https://grass.osgeo.o
> > > > > rg/g
> > > > > rass
> > > > > 70/manuals/v.net.iso.html (using the Spearfish dataset for
> > > > > GRASS 7)
> > > > > the
> > > > > same error appears:
> > > > >
> > > > > > Building graph...
> > > > > > Registering arcs...
> > > > > > Segmentation fault
> > > >
> > > > The segmentation fault must happen in the vector libraries. I
> > > > tested
> > > > on Linux and can not reproduce the segmentation fault. Valgrind
> > > > also
> > > > does not show anything that could cause a segmentation fault. Can
> > > > you
> > > > provide a gdb backtrace?
> > > >
> > > > Markus M
> > >
> > > Thank you Markus,
> > >
> > > My system is debian Stretch, and the following lines describe my
> > > attempt at generating a gdb backtrace. Please let me know if I
> > > should
> > > do otherwise.
> >
> > So far so good. The module crashes in a library function that comes
> > from libavl without modifications, but I can still not see why it
> > could cause a segmentation fault. Can you recompile with the compiler
> > debugging option -g? gdb will then be able to provide more
> > information:
> >
> > make distclean
> > CFLAGS="-g" ./configure 
> > make
> >
> > optionally make install
> >
> > Then start GRASS and use gdb for v.net.iso as before.
> >
> > Markus M
> >
>
> Dear Markus,
>
> Following those instructions does not seem to have changed the
> resulting backtrace info (copied below). Maybe I missed something along
> the way?
>
> -
> make distclean
> CFLAGS="-g" ./configure --with-freetype-includes=/usr/include/freetype2
> make
> sudo make install
>
>
> GRASS 7.0.5 (spearfish60_grass7):~ > gdb `which v.net.iso`
> GNU gdb (Debian 7.11.1-2+b1) 7.11.1
> Copyright (C) 2016 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later  .html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.  Type "show
> copying"
> and "show warranty" for details.
> This GDB was configured as "x86_64-linux-gnu".
> Type "show configuration" for configuration details.
> For bug reporting instructions, please see:
> .
> Find the GDB manual and other documentation resources online at:
> .
> For help, type "help".
> Type "apropos word" to search for commands related to "word"...

This is the problem:
> Reading symbols from /usr/lib/grass70/bin/v.net.iso...(no debugging
> symbols found)...done.

It should be
Reading symbols from /usr/lib/grass70/bin/v.net.iso...done.

For testing purposes I introduced a bug in my local copy to trigger a
segmentation fault. The backtrace shows:

(gdb) bt
#0  tavl_probe (tree=tree@entry=0x827d30, item=item@entry=0x82fb20) at
tavl.c:131
#1  0x76c3d3e7 in dglTreeNodeAdd (pavl=0x827d30, nKey=nKey@entry=1)
at tree.c:73
#2  0x76c32d17 in dgl_add_edge_V1 (pgraph=0x7fffccf8,
nHead=, nTail=, nCost=1231014,
nEdge=1, pvHeadAttr=pvHeadAttr@entry=0x0, pvTailAttr=0x0,
pvEdgeAttr=0x0, nFlags=0) at edgemgmt-template.c:59
#3  0x76c30ea2 in dglAddEdge (pGraph=pGraph@entry=0x7fffccf8,
nHead=, nTail=,
nCost=, nEdge=nEdge@entry=1) at graph.c:624
#4  0x77bb5d09 in Vect_net_build_graph (Map=Map@entry=0x7fffc920,
ltype=ltype@entry=6, afield=afield@entry=1,
nfield=nfield@entry=2, afcol=0x0, abcol=0x0, ncol=0x0, geo=0,
version=1) at net_build.c:918
#5  0x004029e1 in main (argc=, argv=)
at main.c:224

Note that file and line are given where the segfault occured.

Is this something Debian specific that debugging symbols must be explicitly
activated? Are you using GCC or another compiler?

Alternatively, you can also try valgrind:

CMD="v.net.iso input=myroads_net output=myroads_net_iso
center_cats=1-10 costs=1000,2000,5000 --o"
valgrind --tool=memcheck --log-file=v.net.iso.vg.log $CMD

Markus M

>
>
> (gdb) run input=myroads_net output=myroads_net_iso center_cats=1-10
> costs=1000,2000,5000
> Starting program: /usr/lib/grass70/bin/v.net.iso input=myroads_net
> output=myroads_net_iso center_cats=1-10 costs=1000,2000,5000
> [Thread debugging using 

Re: [GRASS-user] v.net.iso - segmentation fault error

[Eduardo Corbelle Rico]

O Xov, 27-10-2016 ás 14:56 +0200, Markus Metz escribiu:
> On Thu, Oct 27, 2016 at 11:16 AM, Eduardo Corbelle Rico
>  wrote:
> > 
> > O Mér, 26-10-2016 ás 23:18 +0200, Markus Metz escribiu:
> > > On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle Rico
> > >  wrote:
> > > > 
> > > > Dear all,
> > > > 
> > > > I am unable to use v.net.iso because of a "segmentation fault"
> > > > error. I
> > > >  have a script that used to work correctly in GRASS 6.4 (32
> > > > bit)
> > > > but
> > > > produces this error after I switched to GRASS 7.0.4 (64 bit).
> > > > 
> > > > If I try to follow both examples shown in https://grass.osgeo.o
> > > > rg/g
> > > > rass
> > > > 70/manuals/v.net.iso.html (using the Spearfish dataset for
> > > > GRASS 7)
> > > > the
> > > > same error appears:
> > > > 
> > > > > Building graph...
> > > > > Registering arcs...
> > > > > Segmentation fault
> > > 
> > > The segmentation fault must happen in the vector libraries. I
> > > tested
> > > on Linux and can not reproduce the segmentation fault. Valgrind
> > > also
> > > does not show anything that could cause a segmentation fault. Can
> > > you
> > > provide a gdb backtrace?
> > > 
> > > Markus M
> > 
> > Thank you Markus,
> > 
> > My system is debian Stretch, and the following lines describe my
> > attempt at generating a gdb backtrace. Please let me know if I
> > should
> > do otherwise.
> 
> So far so good. The module crashes in a library function that comes
> from libavl without modifications, but I can still not see why it
> could cause a segmentation fault. Can you recompile with the compiler
> debugging option -g? gdb will then be able to provide more
> information:
> 
> make distclean
> CFLAGS="-g" ./configure 
> make
> 
> optionally make install
> 
> Then start GRASS and use gdb for v.net.iso as before.
> 
> Markus M
> 

Dear Markus,

Following those instructions does not seem to have changed the
resulting backtrace info (copied below). Maybe I missed something along
the way?

-
make distclean
CFLAGS="-g" ./configure --with-freetype-includes=/usr/include/freetype2
make
sudo make install


GRASS 7.0.5 (spearfish60_grass7):~ > gdb `which v.net.iso`
GNU gdb (Debian 7.11.1-2+b1) 7.11.1
Copyright (C) 2016 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later 
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show
copying"
and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
.
Find the GDB manual and other documentation resources online at:
.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/lib/grass70/bin/v.net.iso...(no debugging
symbols found)...done.


(gdb) run input=myroads_net output=myroads_net_iso center_cats=1-10 
costs=1000,2000,5000
Starting program: /usr/lib/grass70/bin/v.net.iso input=myroads_net
output=myroads_net_iso center_cats=1-10 costs=1000,2000,5000
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-
gnu/libthread_db.so.1".
Building graph...
Registering arcs...

Program received signal SIGSEGV, Segmentation fault.
0x76c64708 in tavl_probe ()
   from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so


(gdb) bt full
#0  0x76c64708 in tavl_probe ()
   from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
No symbol table info available.
#1  0x76c65717 in dglTreeNodeAdd ()
   from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
No symbol table info available.
#2  0x76c5aa0c in dgl_add_edge_V1 ()
   from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
No symbol table info available.
#3  0x76c58a62 in dglAddEdge ()
   from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
No symbol table info available.
#4  0x77ba1d69 in Vect_net_build_graph ()
   from /usr/lib/grass70/lib/libgrass_vector.7.0.5.so
No symbol table info available.
#5  0x5b08 in main ()
No symbol table info available.


(gdb) l
1   ../sysdeps/x86_64/dl-procinfo.c: Non hai tal ficheiro ou
directorio.






> > 
> > > 
> > > > 
> > > > What I am missing?
> > > > 
> > > > Thanks in advance.
> > > > 
> > > > Greetings.
> > > > 
> > > > Eduardo Corbelle
> > > > 
> > > > 
> > > > 
> > > > --
> > > > Dr. Eduardo Corbelle Rico
> > > > 
> > > > Land laboratory (LaboraTe)
> > > > Department of Agricultural and Forest Engineering
> > > > Universidade de Santiago de Compostela
> > > > 
> > > > e-mail: eduardo.corbe...@usc.es
> > > > Tel:  +34 982 823 324
> > > > Fax:  +34 982 285 926
> > > > Web: http://laborate.usc.es
> > > >  http://masterterra.usc.es
> > > >  http://github.com/eduardocorbelle
> > > > 

Re: [GRASS-user] v.net.iso - segmentation fault error

[Markus Neteler]

On Fri, Oct 28, 2016 at 10:11 AM, Eduardo Corbelle Rico
 wrote:
> Thank you, I'll try. Grass was installed in my system using pre-
> compiled packages (via apt-get)... Forgive me if this is a silly
> question: from which folder am I supposed to execute the commands? (or
> should I download the source files and start all over?)

Please check the guide here:
https://grass.osgeo.org/development/compiling-grass-gis/

If it is not sufficient, let us know and we can improve it!

cheers
markusN
___
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user

Re: [GRASS-user] v.net.iso - segmentation fault error

[Eduardo Corbelle Rico]

O Xov, 27-10-2016 ás 14:56 +0200, Markus Metz escribiu:
> On Thu, Oct 27, 2016 at 11:16 AM, Eduardo Corbelle Rico
>  wrote:
> > 
> > O Mér, 26-10-2016 ás 23:18 +0200, Markus Metz escribiu:
> > > On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle Rico
> > >  wrote:
> > > > 
> > > > Dear all,
> > > > 
> > > > I am unable to use v.net.iso because of a "segmentation fault"
> > > > error. I
> > > >  have a script that used to work correctly in GRASS 6.4 (32
> > > > bit)
> > > > but
> > > > produces this error after I switched to GRASS 7.0.4 (64 bit).
> > > > 
> > > > If I try to follow both examples shown in https://grass.osgeo.o
> > > > rg/g
> > > > rass
> > > > 70/manuals/v.net.iso.html (using the Spearfish dataset for
> > > > GRASS 7)
> > > > the
> > > > same error appears:
> > > > 
> > > > > Building graph...
> > > > > Registering arcs...
> > > > > Segmentation fault
> > > 
> > > The segmentation fault must happen in the vector libraries. I
> > > tested
> > > on Linux and can not reproduce the segmentation fault. Valgrind
> > > also
> > > does not show anything that could cause a segmentation fault. Can
> > > you
> > > provide a gdb backtrace?
> > > 
> > > Markus M
> > 
> > Thank you Markus,
> > 
> > My system is debian Stretch, and the following lines describe my
> > attempt at generating a gdb backtrace. Please let me know if I
> > should
> > do otherwise.
> 
> So far so good. The module crashes in a library function that comes
> from libavl without modifications, but I can still not see why it
> could cause a segmentation fault. Can you recompile with the compiler
> debugging option -g? gdb will then be able to provide more
> information:
> 
> make distclean
> CFLAGS="-g" ./configure 
> make
> 
> optionally make install
> 
> Then start GRASS and use gdb for v.net.iso as before.
> 
> Markus M

Thank you, I'll try. Grass was installed in my system using pre-
compiled packages (via apt-get)... Forgive me if this is a silly
question: from which folder am I supposed to execute the commands? (or
should I download the source files and start all over?)



> 
> > 
> > 
> > GRASS 7.0.5 (spearfish60_grass7):~ > gdb `which v.net.iso`
> > GNU gdb (Debian 7.11.1-2) 7.11.1
> > Copyright (C) 2016 Free Software Foundation, Inc.
> > License GPLv3+: GNU GPL version 3 or later  > /gpl
> > .html>
> > This is free software: you are free to change and redistribute it.
> > There is NO WARRANTY, to the extent permitted by law.  Type "show
> > copying"
> > and "show warranty" for details.
> > This GDB was configured as "x86_64-linux-gnu".
> > Type "show configuration" for configuration details.
> > For bug reporting instructions, please see:
> > .
> > Find the GDB manual and other documentation resources online at:
> > .
> > For help, type "help".
> > Type "apropos word" to search for commands related to "word"...
> > Reading symbols from /usr/lib/grass70/bin/v.net.iso...(no debugging
> > symbols found)...done.
> > 
> > 
> > (gdb) run input=myroads_net output=myroads_net_iso center_cats=1-
> > 10
> > costs=1000,2000,5000
> > Starting program: /usr/lib/grass70/bin/v.net.iso input=myroads_net
> > output=myroads_net_iso center_cats=1-10 costs=1000,2000,5000
> > [Thread debugging using libthread_db enabled]
> > Using host libthread_db library "/lib/x86_64-linux-
> > gnu/libthread_db.so.1".
> > Building graph...
> > Registering arcs...
> > 
> > Program received signal SIGSEGV, Segmentation fault.
> > 0x76c64708 in tavl_probe ()
> >    from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
> > 
> > 
> > (gdb) bt full
> > #0  0x76c64708 in tavl_probe ()
> >    from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
> > No symbol table info available.
> > #1  0x76c65717 in dglTreeNodeAdd ()
> >    from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
> > No symbol table info available.
> > #2  0x76c5aa0c in dgl_add_edge_V1 ()
> >    from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
> > No symbol table info available.
> > #3  0x76c58a62 in dglAddEdge ()
> >    from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
> > No symbol table info available.
> > #4  0x77ba1d69 in Vect_net_build_graph ()
> >    from /usr/lib/grass70/lib/libgrass_vector.7.0.5.so
> > No symbol table info available.
> > #5  0x00401aa9 in main ()
> > No symbol table info available.
> > 
> > 
> > (gdb) l
> > 1   ../sysdeps/x86_64/dl-procinfo.c: Non hai tal ficheiro ou
> > directorio.
> > 
> > 
> > 
> > > 
> > > > 
> > > > What I am missing?
> > > > 
> > > > Thanks in advance.
> > > > 
> > > > Greetings.
> > > > 
> > > > Eduardo Corbelle
> > > > 
> > > > 
> > > > 
> > > > --
> > > > Dr. Eduardo Corbelle Rico
> > > > 
> > > > Land laboratory (LaboraTe)
> > > > Department of Agricultural and Forest Engineering
> > > > Universidade de Santiago de Compostela
> > > > 
> > > > 

Re: [GRASS-user] v.net.iso - segmentation fault error

[Markus Metz]

On Thu, Oct 27, 2016 at 11:16 AM, Eduardo Corbelle Rico
 wrote:
>
> O Mér, 26-10-2016 ás 23:18 +0200, Markus Metz escribiu:
>> On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle Rico
>>  wrote:
>> >
>> > Dear all,
>> >
>> > I am unable to use v.net.iso because of a "segmentation fault"
>> > error. I
>> >  have a script that used to work correctly in GRASS 6.4 (32 bit)
>> > but
>> > produces this error after I switched to GRASS 7.0.4 (64 bit).
>> >
>> > If I try to follow both examples shown in https://grass.osgeo.org/g
>> > rass
>> > 70/manuals/v.net.iso.html (using the Spearfish dataset for GRASS 7)
>> > the
>> > same error appears:
>> >
>> > > Building graph...
>> > > Registering arcs...
>> > > Segmentation fault
>>
>> The segmentation fault must happen in the vector libraries. I tested
>> on Linux and can not reproduce the segmentation fault. Valgrind also
>> does not show anything that could cause a segmentation fault. Can you
>> provide a gdb backtrace?
>>
>> Markus M
>
> Thank you Markus,
>
> My system is debian Stretch, and the following lines describe my
> attempt at generating a gdb backtrace. Please let me know if I should
> do otherwise.

So far so good. The module crashes in a library function that comes
from libavl without modifications, but I can still not see why it
could cause a segmentation fault. Can you recompile with the compiler
debugging option -g? gdb will then be able to provide more
information:

make distclean
CFLAGS="-g" ./configure 
make

optionally make install

Then start GRASS and use gdb for v.net.iso as before.

Markus M

>
>
> GRASS 7.0.5 (spearfish60_grass7):~ > gdb `which v.net.iso`
> GNU gdb (Debian 7.11.1-2) 7.11.1
> Copyright (C) 2016 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later  .html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.  Type "show
> copying"
> and "show warranty" for details.
> This GDB was configured as "x86_64-linux-gnu".
> Type "show configuration" for configuration details.
> For bug reporting instructions, please see:
> .
> Find the GDB manual and other documentation resources online at:
> .
> For help, type "help".
> Type "apropos word" to search for commands related to "word"...
> Reading symbols from /usr/lib/grass70/bin/v.net.iso...(no debugging
> symbols found)...done.
>
>
> (gdb) run input=myroads_net output=myroads_net_iso center_cats=1-10
> costs=1000,2000,5000
> Starting program: /usr/lib/grass70/bin/v.net.iso input=myroads_net
> output=myroads_net_iso center_cats=1-10 costs=1000,2000,5000
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib/x86_64-linux-
> gnu/libthread_db.so.1".
> Building graph...
> Registering arcs...
>
> Program received signal SIGSEGV, Segmentation fault.
> 0x76c64708 in tavl_probe ()
>from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
>
>
> (gdb) bt full
> #0  0x76c64708 in tavl_probe ()
>from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
> No symbol table info available.
> #1  0x76c65717 in dglTreeNodeAdd ()
>from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
> No symbol table info available.
> #2  0x76c5aa0c in dgl_add_edge_V1 ()
>from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
> No symbol table info available.
> #3  0x76c58a62 in dglAddEdge ()
>from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
> No symbol table info available.
> #4  0x77ba1d69 in Vect_net_build_graph ()
>from /usr/lib/grass70/lib/libgrass_vector.7.0.5.so
> No symbol table info available.
> #5  0x00401aa9 in main ()
> No symbol table info available.
>
>
> (gdb) l
> 1   ../sysdeps/x86_64/dl-procinfo.c: Non hai tal ficheiro ou
> directorio.
>
>
>
>>
>> >
>> > What I am missing?
>> >
>> > Thanks in advance.
>> >
>> > Greetings.
>> >
>> > Eduardo Corbelle
>> >
>> >
>> >
>> > --
>> > Dr. Eduardo Corbelle Rico
>> >
>> > Land laboratory (LaboraTe)
>> > Department of Agricultural and Forest Engineering
>> > Universidade de Santiago de Compostela
>> >
>> > e-mail: eduardo.corbe...@usc.es
>> > Tel:  +34 982 823 324
>> > Fax:  +34 982 285 926
>> > Web: http://laborate.usc.es
>> >  http://masterterra.usc.es
>> >  http://github.com/eduardocorbelle
>> > ___
>> > grass-user mailing list
>> > grass-user@lists.osgeo.org
>> > http://lists.osgeo.org/mailman/listinfo/grass-user
___
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user

Re: [GRASS-user] v.net.iso - segmentation fault error

[Eduardo Corbelle Rico]

O Mér, 26-10-2016 ás 23:18 +0200, Markus Metz escribiu:
> On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle Rico
>  wrote:
> > 
> > Dear all,
> > 
> > I am unable to use v.net.iso because of a "segmentation fault"
> > error. I
> >  have a script that used to work correctly in GRASS 6.4 (32 bit)
> > but
> > produces this error after I switched to GRASS 7.0.4 (64 bit).
> > 
> > If I try to follow both examples shown in https://grass.osgeo.org/g
> > rass
> > 70/manuals/v.net.iso.html (using the Spearfish dataset for GRASS 7)
> > the
> > same error appears:
> > 
> > > Building graph...
> > > Registering arcs...
> > > Segmentation fault
> 
> The segmentation fault must happen in the vector libraries. I tested
> on Linux and can not reproduce the segmentation fault. Valgrind also
> does not show anything that could cause a segmentation fault. Can you
> provide a gdb backtrace?
> 
> Markus M

Thank you Markus,

My system is debian Stretch, and the following lines describe my
attempt at generating a gdb backtrace. Please let me know if I should
do otherwise.


GRASS 7.0.5 (spearfish60_grass7):~ > gdb `which v.net.iso`
GNU gdb (Debian 7.11.1-2) 7.11.1
Copyright (C) 2016 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later 
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show
copying"
and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
.
Find the GDB manual and other documentation resources online at:
.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/lib/grass70/bin/v.net.iso...(no debugging
symbols found)...done.


(gdb) run input=myroads_net output=myroads_net_iso center_cats=1-10 
costs=1000,2000,5000
Starting program: /usr/lib/grass70/bin/v.net.iso input=myroads_net
output=myroads_net_iso center_cats=1-10 costs=1000,2000,5000
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-
gnu/libthread_db.so.1".
Building graph...
Registering arcs...

Program received signal SIGSEGV, Segmentation fault.
0x76c64708 in tavl_probe ()
   from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so


(gdb) bt full
#0  0x76c64708 in tavl_probe ()
   from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
No symbol table info available.
#1  0x76c65717 in dglTreeNodeAdd ()
   from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
No symbol table info available.
#2  0x76c5aa0c in dgl_add_edge_V1 ()
   from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
No symbol table info available.
#3  0x76c58a62 in dglAddEdge ()
   from /usr/lib/grass70/lib/libgrass_dgl.7.0.5.so
No symbol table info available.
#4  0x77ba1d69 in Vect_net_build_graph ()
   from /usr/lib/grass70/lib/libgrass_vector.7.0.5.so
No symbol table info available.
#5  0x00401aa9 in main ()
No symbol table info available.


(gdb) l
1   ../sysdeps/x86_64/dl-procinfo.c: Non hai tal ficheiro ou
directorio.



> 
> > 
> > What I am missing?
> > 
> > Thanks in advance.
> > 
> > Greetings.
> > 
> > Eduardo Corbelle
> > 
> > 
> > 
> > --
> > Dr. Eduardo Corbelle Rico
> > 
> > Land laboratory (LaboraTe)
> > Department of Agricultural and Forest Engineering
> > Universidade de Santiago de Compostela
> > 
> > e-mail: eduardo.corbe...@usc.es
> > Tel:  +34 982 823 324
> > Fax:  +34 982 285 926
> > Web: http://laborate.usc.es
> >  http://masterterra.usc.es
> >  http://github.com/eduardocorbelle
> > ___
> > grass-user mailing list
> > grass-user@lists.osgeo.org
> > http://lists.osgeo.org/mailman/listinfo/grass-user
___
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user

Re: [GRASS-user] v.net.iso - segmentation fault error

[Markus Metz]

On Thu, Oct 13, 2016 at 9:55 AM, Eduardo Corbelle Rico
 wrote:
>
> Dear all,
>
> I am unable to use v.net.iso because of a "segmentation fault" error. I
>  have a script that used to work correctly in GRASS 6.4 (32 bit) but
> produces this error after I switched to GRASS 7.0.4 (64 bit).
>
> If I try to follow both examples shown in https://grass.osgeo.org/grass
> 70/manuals/v.net.iso.html (using the Spearfish dataset for GRASS 7) the
> same error appears:
>
>> Building graph...
>> Registering arcs...
>> Segmentation fault

The segmentation fault must happen in the vector libraries. I tested
on Linux and can not reproduce the segmentation fault. Valgrind also
does not show anything that could cause a segmentation fault. Can you
provide a gdb backtrace?

Markus M

>
> What I am missing?
>
> Thanks in advance.
>
> Greetings.
>
> Eduardo Corbelle
>
>
>
> --
> Dr. Eduardo Corbelle Rico
>
> Land laboratory (LaboraTe)
> Department of Agricultural and Forest Engineering
> Universidade de Santiago de Compostela
>
> e-mail: eduardo.corbe...@usc.es
> Tel:  +34 982 823 324
> Fax:  +34 982 285 926
> Web: http://laborate.usc.es
>  http://masterterra.usc.es
>  http://github.com/eduardocorbelle
> ___
> grass-user mailing list
> grass-user@lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-user
___
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user

[GRASS-user] v.net.iso - segmentation fault error

[Eduardo Corbelle Rico]

Dear all,

I am unable to use v.net.iso because of a "segmentation fault" error. I
 have a script that used to work correctly in GRASS 6.4 (32 bit) but
produces this error after I switched to GRASS 7.0.4 (64 bit).

If I try to follow both examples shown in https://grass.osgeo.org/grass
70/manuals/v.net.iso.html (using the Spearfish dataset for GRASS 7) the
same error appears:

> Building graph...
> Registering arcs...
> Segmentation fault

What I am missing?

Thanks in advance. 

Greetings.

Eduardo Corbelle



-- 
Dr. Eduardo Corbelle Rico

Land laboratory (LaboraTe)
Department of Agricultural and Forest Engineering
Universidade de Santiago de Compostela

e-mail: eduardo.corbe...@usc.es
Tel:  +34 982 823 324
Fax:  +34 982 285 926
Web: http://laborate.usc.es
 http://masterterra.usc.es
 http://github.com/eduardocorbelle
___
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user