symtab/1313: bad line numbers on very tiny functions

[mec at shout dot net]

>Number:         1313
>Category:       symtab
>Synopsis:       bad line numbers on very tiny functions
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Aug 04 18:08:00 UTC 2003
>Closed-Date:
>Last-Modified:
>Originator:     mec@shout.net
>Release:        gdb 5.3
>Organization:
>Environment:
target=native
host=i686-pc-linux-gnu
osversion=red-hat-8.0
gdb=5.3
gcc=HEAD 2003-08-02 15:36:58 UTC
binutils=2.14
glibc=2.2.93-5-rh
gformat=dwarf-2
glevel=2
>Description:
gdb has bad line numbers on very tiny functions.  This causes regressions in unexpected places in the gdb test suite.

The symptoms are: when gdb hits a breakpoint, it prints:

  Breakpoint 2, 0x08048397 in handle_USR1 (sig=10) at ...

Note the hexadecimal address.  This is incorrect, because the breakpoint really is at the start of the line.  The correct output is:

  Breakpoint 2, handle_USR1 (sig=10) at ...

The bad behavior happened with gcc HEAD.
The good behavior happened with gcc 3.3.

In the attachment, there is a new test script which shows the problem.  After the gdb 6.0 release I'll come back and submit this script for the test suite.


>How-To-Repeat:
Run the empty.exp test script in the attachment.

Or, run gdb.base/attach1.exp and look carefully what happens with the breakpoint on handle_USR1.  Note that the problem is not related to 

Use gcc 3.3 to see good behavior, gcc HEAD to see bad behavior.  It's insensitive to debug format and binutils version.
>Fix:
I don't know whether gcc needs to be fixed to emit different line numbers, or whether gdb needs to be fixed to accept what gcc produces.
>Release-Note:
>Audit-Trail:
>Unformatted:
----gnatsweb-attachment----
Content-Type: application/x-gzip; name="empty.tar.gz"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="empty.tar.gz"
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