[PATCH] Fixed issues with processing of DW_FORM_loclistx / DW_FORM_rnglistx
Simon Marchi
simon.marchi@polymtl.ca
Thu Jan 21 05:53:37 GMT 2021
On 2021-01-21 12:49 a.m., Sharma, Alok Kumar via Gdb-patches wrote:
> Hi all,
>
> I request you all to please review the patch. Below are the details.
>
> Problem Description:
> GDB produces internal error while processing DW_FORM_loclistx / DW_FORM_rnglistx,
> $ make check RUNTESTFLAGS="CXX_FOR_TARGET=/path/to/clang++ " TEST=gdb.dwarf2/dw5-rnglist-test.exp
> Below is the snippet from gdb.log file.
> ---------
> (gdb) file /dir/gdb/testsuite/outputs/gdb.dwarf2/dw5-rnglist-test/dw5-rnglist-test
> Reading symbols from /dir/gdb/testsuite/outputs/gdb.dwarf2/dw5-rnglist-test/dw5-rnglist-test...
> gdb/dwarf2/attribute.h:232: internal-error: void attribute::set_unsigned_reprocess(ULONGEST): Assertion `form_requires_reprocessing ()' failed.
> A problem internal to GDB has been detected,
> further debugging may prove unreliable.
> Quit this debugging session? (y or n) ERROR: Couldn't load dw5-rnglist-test into GDB (GDB internal error).
> Resyncing due to internal error.
> ---------
> This is due to incomplete handling of DW_FORM_loclistx / DW_FORM_rnglistx,
> which is available with -gdwarf-5 option for clang compiler.
> Resolution:
> Forms DW_FORM_loclistx / DW_FORM_rnglistx need two phases of processing.
> The required modifications in various routines are added to fix the issue.
>
> gdb/ChangeLog:
>
> 2021-01-21 Alok Kumar Sharma <AlokKumar.Sharma@amd.com>
>
> * dwarf2/attribute.c (form_is_unsigned): Handle DW_FORM_loclistx.
> (form_requires_reprocessing): Handle DW_FORM_loclistx and
> DW_FORM_rnglistx.
> * dwarf2/attribute.h (set_address): Handle DW_FORM_loclistx and
> DW_FORM_rnglistx.
> * dwarf2/read.c (read_attribute_reprocess): Replaced as_unsigned with
> as_unsigned_reprocess for DW_FORM_loclistx and DW_FORM_rnglistx.
>
> Regards,
> Alok
>
Hi Alok,
Could you please take a look at this series I posted yesterday?
https://sourceware.org/pipermail/gdb-patches/2021-January/175221.html
I think it contains pretty much the same changes (and more).
Simon
More information about the Gdb-patches
mailing list