Non-dwarf blocks detected by valgrind

Non-dwarf blocks detected by valgrind

[Paul Richard Thomas]

Hi All,

Has anybody else seen this popping up in executables produced from master?

[pault@pc30 pr64290]$ valgrind ./a.out -s
==1627334== Memcheck, a memory error detector
==1627334== Copyright (C) 2002-2017, and GNU GPL'd, by Julian Seward et al.
==1627334== Using Valgrind-3.16.1 and LibVEX; rerun with -h for copyright
info
==1627334== Command: ./a.out -s
==1627334==
--1627334-- WARNING: Serious error when reading debug info
--1627334-- When reading debug info from /home/pault/prs/pr64290/a.out:
--1627334-- Ignoring non-Dwarf2/3/4 block in .debug_info
,,,,,repeats many times
--1627334-- WARNING: Serious error when reading debug info
--1627334-- When reading debug info from /home/pault/prs/pr64290/a.out:
--1627334-- parse_CU_Header: is neither DWARF2 nor DWARF3 nor DWARF4
==1627334==
==1627334== HEAP SUMMARY:
==1627334==     in use at exit: 0 bytes in 0 blocks
==1627334==   total heap usage: 42 allocs, 42 frees, 13,691 bytes allocated
==1627334==
==1627334== All heap blocks were freed -- no leaks are possible
==1627334==
==1627334== For lists of detected and suppressed errors, rerun with: -s
==1627334== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)

Paul

Re: Non-dwarf blocks detected by valgrind

[Tobias Burnus]

Hi Paul,

On 23.01.21 09:52, Paul Richard Thomas via Fortran wrote:
> Has anybody else seen this popping up in executables produced from master?
>
> --1627334-- Ignoring non-Dwarf2/3/4 block in .debug_info

This error looks as if Valgrind does not support DWARF5. Try to compile 
with -gdwarf-4.

GCC switched to defaulting to DWARF 5 in GCC 11, cf. 
https://gcc.gnu.org/g:3804e937b0e252a7e42632fe6d9f898f1851a49c (Mark 
Wielaard; committed 17 Jan 2021).

(Not yet documented in the release notes.)

Cheers,

Tobias

Re: Non-dwarf blocks detected by valgrind

[Paul Richard Thomas]

Hi Tobias,

Thanks for the quick reply. However, -gdwarf-4 does not quieten valgrind
down.

Does this constitute a regression?

Cheers

Paul


On Sat, 23 Jan 2021 at 10:51, Tobias Burnus <burnus@net-b.de> wrote:

> Hi Paul,
>
> On 23.01.21 09:52, Paul Richard Thomas via Fortran wrote:
> > Has anybody else seen this popping up in executables produced from
> master?
> >
> > --1627334-- Ignoring non-Dwarf2/3/4 block in .debug_info
>
> This error looks as if Valgrind does not support DWARF5. Try to compile
> with -gdwarf-4.
>
> GCC switched to defaulting to DWARF 5 in GCC 11, cf.
> https://gcc.gnu.org/g:3804e937b0e252a7e42632fe6d9f898f1851a49c (Mark
> Wielaard; committed 17 Jan 2021).
>
> (Not yet documented in the release notes.)
>
> Cheers,
>
> Tobias
>
>

-- 
"If you can't explain it simply, you don't understand it well enough" -
Albert Einstein

Re: Non-dwarf blocks detected by valgrind

[Thomas Koenig]

Hi Paul,

> Thanks for the quick reply. However, -gdwarf-4 does not quieten valgrind
> down.
> 
> Does this constitute a regression?

Very much so, IMHO.

Do you want to submit the bug report?

Best regards

	Thomas

Re: Non-dwarf blocks detected by valgrind

[Paul Richard Thomas]

OK - I'll get on to it tomorrow.

Thanks

Paul


On Sun, 24 Jan 2021 at 20:37, Thomas Koenig <tkoenig@netcologne.de> wrote:

> Hi Paul,
>
> > Thanks for the quick reply. However, -gdwarf-4 does not quieten valgrind
> > down.
> >
> > Does this constitute a regression?
>
> Very much so, IMHO.
>
> Do you want to submit the bug report?
>
> Best regards
>
>         Thomas
>


-- 
"If you can't explain it simply, you don't understand it well enough" -
Albert Einstein

Re: Non-dwarf blocks detected by valgrind

[Tobias Burnus]

Hi Paul,

On 23.01.21 14:32, Paul Richard Thomas wrote:

> Thanks for the quick reply. However, -gdwarf-4 does not quieten valgrind down.

Works here, trying with the third attachment to PR64290,
current GCC on x86_64-gnu-linux, and Valgrind-3.15.0 of
  Ubuntu focal (20.04.1 LTS).

With "gfortran -g" (or "gfortran -gdwarf-5), I get:

--721834-- WARNING: Serious error when reading debug info
--721834-- When reading debug info from /dev/shm/a.out:
--721834-- Ignoring non-Dwarf2/3/4 block in .debug_info

But when I use "gfortran -gdwarf-4" (or "gfortran -g -gdwarf-4"), no error is shown.

> Does this constitute a regression?

 From your description, yes. Can you give more details how to reproduce it?

Tobias

-----------------
Mentor Graphics (Deutschland) GmbH, Arnulfstraße 201, 80634 München / Germany
Registergericht München HRB 106955, Geschäftsführer: Thomas Heurung, Alexander Walter

Re: Non-dwarf blocks detected by valgrind

[Thomas Koenig]

Hi Tobias,

>> Does this constitute a regression?
> 
>  From your description, yes. Can you give more details how to reproduce it?

$ cat hello.f90
   print *,"Hello, world!"
end
$ valgrind --version
valgrind-3.15.0
$ gfortran -g hello.f90
$ valgrind ./a.out
[...]
--4184-- WARNING: Serious error when reading debug info
--4184-- When reading debug info from /home/ig25/lib64/libgfortran.so.5.0.0:
--4184-- Ignoring non-Dwarf2/3/4 block in .debug_info
--4184-- WARNING: Serious error when reading debug info
--4184-- When reading debug info from /home/ig25/lib64/libgfortran.so.5.0.0:

[...]

--4184-- WARNING: Serious error when reading debug info
--4184-- When reading debug info from /home/ig25/lib64/libgcc_s.so.1:
--4184-- Ignoring non-Dwarf2/3/4 block in .debug_info


[...]

--4184-- WARNING: Serious error when reading debug info
--4184-- When reading debug info from /home/ig25/lib64/libquadmath.so.0.0.0:
--4184-- Ignoring non-Dwarf2/3/4 block in .debug_info

So, the problem (at least for me) is that libgfortran and
other libraries are built in a format that valgrind can not
understand, so it becomes much harder to debug libgfortran :-(

Is there a way to configure the build so that the old debug
format is used for the libraries?

Best regards

	Thomas

Re: Non-dwarf blocks detected by valgrind

[Mark Wielaard]

Hi,

On Mon, 2021-01-25 at 15:05 +0100, Thomas Koenig via Gcc wrote:
> $ valgrind --version
> valgrind-3.15.0
> $ gfortran -g hello.f90
> $ valgrind ./a.out
> [...]
> --4184-- WARNING: Serious error when reading debug info
> --4184-- When reading debug info from
> /home/ig25/lib64/libgfortran.so.5.0.0:
> --4184-- Ignoring non-Dwarf2/3/4 block in .debug_info
> --4184-- WARNING: Serious error when reading debug info
> --4184-- When reading debug info from
> /home/ig25/lib64/libgfortran.so.5.0.0:
> 
> [...]
> 
> --4184-- WARNING: Serious error when reading debug info
> --4184-- When reading debug info from /home/ig25/lib64/libgcc_s.so.1:
> --4184-- Ignoring non-Dwarf2/3/4 block in .debug_info
> 
> 
> [...]
> 
> --4184-- WARNING: Serious error when reading debug info
> --4184-- When reading debug info from
> /home/ig25/lib64/libquadmath.so.0.0.0:
> --4184-- Ignoring non-Dwarf2/3/4 block in .debug_info
> 
> So, the problem (at least for me) is that libgfortran and
> other libraries are built in a format that valgrind can not
> understand, so it becomes much harder to debug libgfortran :-(

I have a patch for that:
https://code.wildebeest.org/git/user/mjw/valgrind/commit/?h=dwarf5

It needs a little bit more cleanup, but should already work as
expected.

Cheers,

Mark

Re: Non-dwarf blocks detected by valgrind

[Tobias Burnus]

Hi Thomas,

which binutil version do you have? I am asking because Jakub just
submitted:
https://gcc.gnu.org/pipermail/gcc-patches/2021-January/564206.html

If that does not help, it might be that valgrind simply does not like
the DWARF which GCC now produces by default.

Defaulting to 4: Well, one way would be to use '4' in "init(5)" of
"gdwarf-" in gcc/common.opt.

Tobias

PS: When checking myself, I did not set LD_LIBRARY_PATH, using the *.so
of the Linux distro; hence, I might have missed the issue you are seeing.

On 25.01.21 15:05, Thomas Koenig via Fortran wrote:
>
> Hi Tobias,
>
>>> Does this constitute a regression?
>>
>>  From your description, yes. Can you give more details how to
>> reproduce it?
>
> $ cat hello.f90
>   print *,"Hello, world!"
> end
> $ valgrind --version
> valgrind-3.15.0
> $ gfortran -g hello.f90
> $ valgrind ./a.out
> [...]
> --4184-- WARNING: Serious error when reading debug info
> --4184-- When reading debug info from
> /home/ig25/lib64/libgfortran.so.5.0.0:
> --4184-- Ignoring non-Dwarf2/3/4 block in .debug_info
> --4184-- WARNING: Serious error when reading debug info
> --4184-- When reading debug info from
> /home/ig25/lib64/libgfortran.so.5.0.0:
>
> [...]
>
> --4184-- WARNING: Serious error when reading debug info
> --4184-- When reading debug info from /home/ig25/lib64/libgcc_s.so.1:
> --4184-- Ignoring non-Dwarf2/3/4 block in .debug_info
>
>
> [...]
>
> --4184-- WARNING: Serious error when reading debug info
> --4184-- When reading debug info from
> /home/ig25/lib64/libquadmath.so.0.0.0:
> --4184-- Ignoring non-Dwarf2/3/4 block in .debug_info
>
> So, the problem (at least for me) is that libgfortran and
> other libraries are built in a format that valgrind can not
> understand, so it becomes much harder to debug libgfortran :-(
>
> Is there a way to configure the build so that the old debug
> format is used for the libraries?
>
> Best regards
>
>     Thomas
-----------------
Mentor Graphics (Deutschland) GmbH, Arnulfstraße 201, 80634 München / Germany
Registergericht München HRB 106955, Geschäftsführer: Thomas Heurung, Alexander Walter

Re: Non-dwarf blocks detected by valgrind

[Thomas Koenig]

Hi Tobias,

> which binutil version do you have?

$ as -v
GNU assembler version 2.35.1 (x86_64-suse-linux) using BFD version (GNU 
Binutils; openSUSE Leap 15.1) 2.35.1.20201123-lp151.3.12


> I am asking because Jakub just
> submitted:
> https://gcc.gnu.org/pipermail/gcc-patches/2021-January/564206.html

Looks like this could be the issue.  I'll re-check once that patch
has hit the trunk.

Best regards

	Thomas

Re: Non-dwarf blocks detected by valgrind

[Mark Wielaard]

On Mon, 2021-01-25 at 15:38 +0100, Mark Wielaard wrote:
> On Mon, 2021-01-25 at 15:05 +0100, Thomas Koenig via Gcc wrote:
> > --4184-- WARNING: Serious error when reading debug info
> > --4184-- When reading debug info from
> > /home/ig25/lib64/libquadmath.so.0.0.0:
> > --4184-- Ignoring non-Dwarf2/3/4 block in .debug_info
> > 
> > So, the problem (at least for me) is that libgfortran and
> > other libraries are built in a format that valgrind can not
> > understand, so it becomes much harder to debug libgfortran :-(
> 
> I have a patch for that:
> https://code.wildebeest.org/git/user/mjw/valgrind/commit/?h=dwarf5
> 
> It needs a little bit more cleanup, but should already work as
> expected.

Cleaned up and submitted upstream:
https://bugs.kde.org/show_bug.cgi?id=432102
Also backported to Fedora 34 valgrind package:
https://bodhi.fedoraproject.org/updates/FEDORA-2021-6335fad95d

Cheers,

Mark