home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky gnu.gcc.bug:3106 gnu.gdb.bug:1343
- From: Eberhard Mattes <mattes@azu.informatik.uni-stuttgart.de>
- Date: Sun, 3 Jan 93 17:56:04 +0100
- Message-ID: <9301031656.AA03447@azu.informatik.uni-stuttgart.de>
- To: bug-gcc@prep.ai.mit.edu
- Cc: bug-gdb@prep.ai.mit.edu
- Subject: More MKTEMP_EACH_FILE trouble
- X-Created-Offline: 02-Jan-1993 14:00:10
- Distribution: world
- Approved: info-gnu@prep.ai.mit.edu
- Path: sparky!uunet!usc!cs.utexas.edu!qt.cs.utexas.edu!yale.edu!yale!mintaka.lcs.mit.edu!ai-lab!prep.ai.mit.edu!gnulists
- Newsgroups: gnu.gcc.bug,gnu.gdb.bug
- Followup-To: gnu.gcc.bug,gnu.gdb.bug
- Lines: 19
-
- When using MKTEMP_EACH_FILE for GCC 2.3.3 with DBX debugging info, GDB
- no longer works correctly as it explicitely checks for ".o" at the end
- of symbols. For instance, the backtrace command is affected.
-
- Quotation from partial-stab.h of GDB 4.7:
-
- if ((namestring[0] == '-' && namestring[1] == 'l')
- || (namestring [(nsl = strlen (namestring)) - 1] == 'o'
- && namestring [nsl - 2] == '.')
-
- (BTW this is a bad idea if the length of namestring is less than 2.)
-
- I solved the problem by letting ld append `.o' to the symbol
- indicating the start of the text segment unless that symbol starts
- with `-l' or already ends with `.o'. (I'm still using the pre-BFD
- version of GNU ld.)
-
- Eberhard Mattes (mattes@azu.informatik.uni-stuttgart.de)
-
-