dwz vs xul

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

dwz vs xul

Marc Espie-2
Haven't had time to investigate, but it seems libxul might have
issues.

Errors show up in all 3 ports (www/firefox-esr, www/mozilla-firefox
and mail/thunderbird)

Looks like:
> Extracting debug info from lib/thunderbird/libxul.so.34.0
dwz: lib/thunderbird/.debug/libxul.so.34.0.dbg: Couldn't find DIE referenced by DW_AT_abstract_origin


I've found a small tool called dwarf-check on github (in go... well)

hub$ dwarf-check libxul.so.34.0                                                
error initializing dwarf state examiner: malformed dwarf at offset 450264449: nstack underflow

so both info seem to point at a bug in clang/lld

Quick&dirty port of dwarf-check attached

dwarf-check.tgz (888 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: dwz vs xul

Jeremie Courreges-Anglas-2
On Wed, Mar 25 2020, Marc Espie <[hidden email]> wrote:

> Haven't had time to investigate, but it seems libxul might have
> issues.
>
> Errors show up in all 3 ports (www/firefox-esr, www/mozilla-firefox
> and mail/thunderbird)
>
> Looks like:
>> Extracting debug info from lib/thunderbird/libxul.so.34.0
> dwz: lib/thunderbird/.debug/libxul.so.34.0.dbg: Couldn't find DIE referenced by DW_AT_abstract_origin
>
>
> I've found a small tool called dwarf-check on github (in go... well)
>
> hub$ dwarf-check libxul.so.34.0                                                
> error initializing dwarf state examiner: malformed dwarf at offset 450264449: nstack underflow
>
> so both info seem to point at a bug in clang/lld
>
> Quick&dirty port of dwarf-check attached
Could be useful and appears quite low maintenance.

Please see the attached tarball (standard Makefile ordering, fold
DESCR), ok jca@ to import



--
jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF  DDCC 0DFA 74AE 1524 E7EE

dwarf-check.2.tgz (890 bytes) Download Attachment