[prev in list] [next in list] [prev in thread] [next in thread] 

List:       gdb-patches
Subject:    Patch: Handle relative paths in .debug_line
From:       garrison () mail ! ru (Garrison)
Date:       2004-09-30 0:49:00
Message-ID: 415B5898.1020203 () mail ! ru
[Download RAW message or body]

Garrison wrote:
> Hi!
> 
> I've taken a look at dwarf standard, and I have a question (possibly
> already answered elsewhere): it is specified that relative paths in list
> are related to compilation directory, so why not make the substitution
> while parsing the list? If it's OK please take a look at attached patch
> (sorry if it wraps this time). It's against current CVS HEAD.
> 
> I'm not sure I got those make_cleanup() calls right, if so please correct
> me.
> 
Also I looked at regressions - this patch makes approximately the same number
of failures as original relative paths patch. Test case gdb.base/unload.exp
might show the reason - there seems to be a problem in handling relative paths
within regression testing scripts:

Running ../../../gdb-dejagnu/gdb/testsuite/gdb.base/unload.exp ...
ERROR: couldn't open \
"../../../gdb-dejagnu/gdb/testsuite/gdb.base/../../../gdb-dejagnu/gdb/testsuite/gdb.base/unloadshr.c":
 no such file or directory
FAIL: gdb.base/unload.exp: rerun to shared library breakpoint

--
Kind regards,
Igor V. Kovalenko


[prev in list] [next in list] [prev in thread] [next in thread]