Article View: gnu.gdb.bug
Article #10110coff .file symbol
From: Jay Norwood
Date: Wed, 16 May 1990 17:07
Date: Wed, 16 May 1990 17:07
15 lines
1585 bytes
1585 bytes
I've been looking at the coffread module and how it processes the
symbols of ".file". I'm working with version 3.5. In the read_coff_symtab
function there is a block of code which calls complete_symtab ("_globals_"...
It seems that the only entry into this block is based upon reading a
bogus ".file" symbol that doesn't have a C_FILE n_sclass. I haven't
found any documentation of such a ".file" symbol in the Understanding
and Using COFF book that I am using for reference. What I have found
(on page 107) is that the .file symbol value is an index which forms
a circular list, with the last .file value having an index to the first
.file symbol. If implemented this way, the variable "next_file_symnum"
will eventually get set back to a value smaller than the current
symbol number, and the section which calls complete_symtab ("_globals_"...
will never be entered. So - the code seems to be expecting this
undocumented termination '.file' block. Can you comment on this?
Jay Norwood
Technical Details
Message-ID:
<9005161707.AA11775@ray.sun.com>Path:
arc1-ca.pugleaf.net!gmdzi!unido!mcsun!uunet!tut.cis.ohio-state.edu!CS.UTEXAS.EDU!oakhill!ray!jayShow Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: f5053,5aae7926cc088b5b,start X-Google-Attributes: gidf5053,public X-Google-ArrivalTime: 1990-05-16 15:08:23 PST Path: arc1-ca.pugleaf.net!gmdzi!unido!mcsun!uunet!tut.cis.ohio-state.edu!CS.UTEXAS.EDU!oakhill!ray!jay From: oakhill!ray!jay@CS.UTEXAS.EDU (Jay Norwood) Newsgroups: gnu.gdb.bug Subject: coff .file symbol Message-ID: <9005161707.AA11775@ray.sun.com> Date: 16 May 90 17:07:49 GMT Sender: daemon@tut.cis.ohio-state.edu Distribution: gnu Organization: GNUs Not Usenet Lines: 15 Posted: Wed May 16 18:07:49 1990