gdb tests with gcc 3.0.3 versus gcc 3.0.4-200202-15

[Michael Elizabeth Chastain <mec at shout dot net>]

I did some test runs with several versions of gdb versus gcc 3.0.3,
gcc 3.0.4-20020215, and gcc gcc-3_0-branch.

The tables are here:

  http://www.shout.net/~mec/sunday_project/by-gcc-2002-02-16/index.html

The test matrix has 18 configurations:

  target=native
  host=i686-pc-linux-gnulibc2.2:rh-7.2)
  gdb={5.1.1, gdb_5_1-2001-07-29-branch:20020216, HEAD}
  gcc={3.0.3, 3.0.4-20020215, gcc-3_0-branch:20020216}
  goption={-gdwarf2, -gstabs+}

gdb test regressions from gcc 3.0.3 to gcc 3.0.4-20020215:

  *none*

gdb test improvements from gcc 3.0.3 to gcc 3.0.4-20020215:

  gdb.base/break.exp: breakpoint at start of multi line if conditional
  gdb.base/break.exp: breakpoint at start of multi line while conditional
  gdb.base/break.exp: breakpoint info

    FAIL -> PASS with gdb-HEAD-g-dwarf-2
    FAIL -> PASS with gdb-HEAD-g-stabs+

  gdb.base/step-line.exp: next over dummy 1
  gdb.base/step-line.exp: next over dummy 10
  gdb.base/step-line.exp: next over dummy 2
  gdb.base/step-line.exp: next over dummy 4
  gdb.base/step-line.exp: next over dummy 6
  gdb.base/step-line.exp: next over dummy 8

    FAIL -> PASS with gdb-5.1.1-g-dwarf-2
    FAIL -> PASS with gdb-5.1.1-g-stabs+
    FAIL -> PASS with gdb-gdb_5_1-2001-07-29-branch-g-dwarf-2
    FAIL -> PASS with gdb-gdb_5_1-2001-07-29-branch-g-stabs+
    FAIL -> PASS with gdb-HEAD-g-dwarf-2
    FAIL -> PASS with gdb-HEAD-g-stabs+

gdb test regressions from gcc 3.0.3 to gcc gcc-3_0-branch:

  gdb.base/condbreak.exp, gdb.base/funcargs.exp, gdb.c++/ovldbreak.exp:
  regressions in dwarf-2 configurations.
  I don't have time to drill down.

gdb test improvements from gcc 3.0.3 to gcc gcc-3_0-branch:

  same as gcc 3.0.3 to gcc 3.0.4-20020215

Notes:

  I am cross-posting these results to lists: gcc-testresults and
  gdb-testresults.  Someone kindly correct me if I should do something
  different.

  gdb 5.1.1 is the latest released version of gdb.

  gdb gdb_5_1-2001-07-29-branch is a nearly dead branch.

  I am not investigating the regressions on gcc-3_0-branch because
  it's not a release candidate at this time and I have limited time.
  If another tarball gets made then I will test the tarballs.

Michael Elizabeth Chastain
<mailto:mec@shout.net>
"love without fear"