This is the mail archive of the
gdb-testers@sources.redhat.com
mailing list for the GDB project.
Sunday Project test results, gdb, 2002-03-22
- From: Michael Elizabeth Chastain <mec at shout dot net>
- To: gdb-testers at sources dot redhat dot com
- Date: Mon, 25 Mar 2002 12:05:49 -0600
- Subject: Sunday Project test results, gdb, 2002-03-22
This week, the configuration {gdb HEAD gcc 2.95.3, -gstabs+} is much
healthier. Indeed it is the healthiest gdb test run that I've ever seen:
0 ERRORs, 0 WARNINGs, and 22 FAILs.
These PR's have been fixed:
http://sources.redhat.com/cgi-bin/gnatsweb.pl?database=gdb&cmd=view&pr=379
Regression: The step command does not stop in the right place.
http://sources.redhat.com/cgi-bin/gnatsweb.pl?database=gdb&cmd=view&pr=380
Regression: "No line number known for main." and many other functions.
This PR has not been fixed yet:
http://sources.redhat.com/cgi-bin/gnatsweb.pl?database=gdb&cmd=view&pr=381
Regression: src_and_asm_line line attribute "line" for file gdb.mi/basics.c changed from 31 to 0 in several places.
There was one regression in this configuration: gdb lands on a worse
place when setting a breakpoint on a function in an optimized file.
The test suite considers this acceptable and I do too.
Aside from that, the bug fixes have no regression impact. It would be
great if someone could port them to the 5.2 branch.
Michael C
===
. Summary
. Test Matrix
. Matrix
target: native
host: i686-pc-linux-gnulibc2.2%rh-7.2
gdb: 5.1.1, gdb_5_2-branch%20020322, HEAD%20020322
gcc: 2.95.3, vendor, 3.0.4, gcc-3_1-branch%20020322, HEAD%20020322
glibc: vendor
goption: -gdwarf-2, -gstabs+
count: 30=1*1*3*5*1*2
. Notes
target and host names are gnu triples with extra information after
a '%' delimiter.
versions which start with a number are official releases or snapshots.
versions named "vendor" are the vendor-supplied facility.
versions named "HEAD" are the cvs HEAD (mainline).
versions with any other name are cvs branches.
cvs versions show the pull date after a '%' delimiter.
gcc releases and snapshots are built with binutils 2.11.2
gcc cvs versions are built with binutils HEAD with the same pull date.
. libiberty
All tests passed.
. gdb
. Tables
http://www.shout.net/~mec/sunday/2002-03-22/index.html
. Overview
gdb 5.1.1: 0 build aborts, 0 test aborts, 402 attention lines
gdb gdb_5_2-branch: 0 build aborts, 0 test aborts, 337 attention lines
gdb HEAD: 0 build aborts, 0 test aborts, 274 attention lines
an attention line is:
error warning fail kfail unresolved untested unsupported
an attention line is not:
pass xpass xfail
. Old bugs fixed
http://sources.redhat.com/cgi-bin/gnatsweb.pl?database=gdb&cmd=view&pr=379
Regression: The step command does not stop in the right place.
http://sources.redhat.com/cgi-bin/gnatsweb.pl?database=gdb&cmd=view&pr=380
Regression: "No line number known for main." and many other functions.
. New bugs detected
With gcc 2.95.3 -gstabs+, gdb lands on a worse place when setting
a breakpoint on a function in an optimized file. I don't consider
this important enough to file a bug report.
. Test protocol changes since last report
I added a line entry for unknown result types, just in case any weirdness
comes into the test suite. So far no cases of this have happened.
. Baseline software
. host i686-pc-linux-gnu%rh-7.2
make 3.79.1
binutils 2.11.2
gcc 3.0.4
flex 2.5.4
bison 1.33
tcl 8.3.4
expect 5.33.0
dejagnu 1.4.2
I am aware of binutils 2.12 but I have not deployed it yet.
. Analysis
. libiberty
. results
. target native
. host i686-pc-linux-gnu%rh-7.2
binutils HEAD All 648 tests passed
gcc gcc-3_1-branch All 648 tests passed
gcc HEAD All 648 tests passed
gdb gdb_5_2-branch All 648 tests passed
gdb HEAD All 648 tests passed
. gdb
The last report was 2002-03-17.
http://www.shout.net/~mec/sunday/2002-03-17/Analysis.txt
Configuration numbers for the difference grids are in:
http://www.shout.net/~mec/sunday/2002-03-22/Differences-compare-by-date-two.html
Difference grid key:
+ gdb improved
- gdb regressed
* new test with good result
_ new test with bad result
, noise: result changed but gdb behavior did not
. 5.1.1
. Difference Grid
gdb.base/interrupt.exp ,9
gdb.c++/annota2.exp ,4
. gdb_5_2-branch
. Difference Grid
gdb.c++/annota2.exp ,12 ,14 ,15 ,16 ,17 ,19 ,20
. HEAD
. Difference Grid
gdb.base/break.exp -22
gdb.base/call-ar-st.exp +22
gdb.base/callfuncs.exp +22
gdb.base/default.exp *21 *22 *23 *24 *25 *26 *27 *28 *29 *30
gdb.base/display.exp +22
gdb.base/ending-run.exp +22
gdb.base/funcargs.exp +22
gdb.base/help.exp *21 *22 *23 *24 *25 *26 *27 *28 *29 *30
gdb.base/list.exp +22
gdb.base/ptype.exp +22
gdb.base/shlib-call.exp +22
gdb.base/step-line.exp +22
gdb.base/step-test.exp +22
gdb.base/structs2.exp +22
gdb.c++/annota.exp ,23 ,25
gdb.c++/method.exp +22
gdb.c++/namespace.exp +22
gdb.c++/overload.exp +22
gdb.c++/ovldbreak.exp +22
gdb.mi/mi-disassemble.exp +22
gdb.mi/mi-simplerun.exp +22
gdb.mi/mi-var-display.exp +22
gdb.mi/mi0-disassemble.exp +22
gdb.mi/mi0-simplerun.exp +22
gdb.mi/mi0-var-display.exp +22
gdb.trace/tracecmd.exp +22
. Regression
. gdb.base/break.exp: run until function breakpoint, optimized file
. gcc 2.95.3, goption g-stabs+: PASS -> PASS
gdb lands on a worse place when setting a breakpoint on a function
in an optimized file. The test still considers this a PASS.
. Improvement
Many test scripts improved on gcc 2.95.3, goption g-stabs+.
PR gdb/379 and PR gdb/380 were fixed, as well as many other FAILs
in this configuration.