This is the mail archive of the
gdb-prs@sources.redhat.com
mailing list for the GDB project.
Re: gdb/511: c++ GNU gdb Linux 7.x (5.0rh-15) - i386-redhat-linux, gdb print double
- From: drow at sources dot redhat dot com
- To: drow at sources dot redhat dot com, gdb-prs at sources dot redhat dot com, imack at ix dot netcom dot com, nobody at sources dot redhat dot com
- Date: 30 Apr 2002 17:41:59 -0000
- Subject: Re: gdb/511: c++ GNU gdb Linux 7.x (5.0rh-15) - i386-redhat-linux, gdb print double
- Reply-to: drow at sources dot redhat dot com, drow at sources dot redhat dot com, gdb-prs at sources dot redhat dot com, imack at ix dot netcom dot com, nobody at sources dot redhat dot com, gdb-gnats at sources dot redhat dot com
Synopsis: c++ GNU gdb Linux 7.x (5.0rh-15) - i386-redhat-linux, gdb print double
Responsible-Changed-From-To: unassigned->drow
Responsible-Changed-By: drow
Responsible-Changed-When: Tue Apr 30 10:41:59 2002
Responsible-Changed-Why:
C++
State-Changed-From-To: open->analyzed
State-Changed-By: drow
State-Changed-When: Tue Apr 30 10:41:59 2002
State-Changed-Why:
This appears to be a bug in GCC's DWARF-2 output, present in 2.96 and 3.0:
<2><19f49>: Abbrev Number: 3 (DW_TAG_member)
DW_AT_name : d
DW_AT_decl_file : 1
DW_AT_decl_line : 11
DW_AT_type : <188de>
DW_AT_data_member_location: 2 byte block: 23 8 (DW_OP_plus_uconst: 8; )
However, d is stored at +4, not +8.
I suspect this is fixed in GCC 3.1, but I need to confirm that. I can't think of a workaround for debug info quite
this wrong.
(Also not closing PR until I add a good testcase)
http://sources.redhat.com/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gdb&pr=511