Hello,

I was running the or1ksim tool check regression, and I have the following results which do not match the expected results, please resolve this.


=== libsim tests ===

Schedule of variations:
unix

Running target unix
Using ../../or1ksim/testsuite/config/unix.exp as board description file for target.
Running ../../or1ksim/testsuite/libsim.tests/int-edge.exp ...
FAIL: int-edge simple 2: hit EOF seeking match line 7
Running ../../or1ksim/testsuite/libsim.tests/int-level.exp ...
Running ../../or1ksim/testsuite/libsim.tests/jtag-basic.exp ...
Running ../../or1ksim/testsuite/libsim.tests/jtag-go-command-read.exp ...
Running ../../or1ksim/testsuite/libsim.tests/jtag-go-command-write.exp ...
Running ../../or1ksim/testsuite/libsim.tests/jtag-read-command.exp ...
Running ../../or1ksim/testsuite/libsim.tests/jtag-read-control.exp ...
Running ../../or1ksim/testsuite/libsim.tests/jtag-select-module.exp ...
Running ../../or1ksim/testsuite/libsim.tests/jtag-write-command.exp ...
Running ../../or1ksim/testsuite/libsim.tests/jtag-write-control.exp ...
Running ../../or1ksim/testsuite/libsim.tests/lib-iftest.exp ...
Running ../../or1ksim/testsuite/libsim.tests/upcalls.exp ...

=== libsim Summary ===

# of expected passes 263
# of unexpected failures 1
Test Run By lrajani on Wed Jun 20 15:15:28 2012
Target is or32-unknown-elf
Host is x86_64-unknown-linux-gnu

=== or1ksim tests ===

Schedule of variations:
unix

Running target unix
Using ../../or1ksim/testsuite/config/unix.exp as board description file for target.
Running ../../or1ksim/testsuite/or1ksim.tests/basic.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/cache.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/cbasic.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/cfg.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/dhry.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/dmatest.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/eth.exp ...
FAIL: eth: hit EOF seeking match line 2
Running ../../or1ksim/testsuite/or1ksim.tests/except-test.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/exit.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/ext.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/fbtest.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/flag.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/fp.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/functest.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/inst-set-test.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/int-test.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/kbdtest.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/local-global.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/mem-test.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/mmu.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/mul.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/mycompress.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/testfloat.exp ...
Running ../../or1ksim/testsuite/or1ksim.tests/tick.exp ...

=== or1ksim Summary ===

# of expected passes 2263
# of unexpected failures 1
make[3]: *** [check-DEJAGNU] Error 1
make[3]: Leaving directory `/home/lrajani/ortool/build_or1ksim/testsuite'
make[2]: *** [check-am] Error 2
make[2]: Leaving directory `/home/lrajani/ortool/build_or1ksim/testsuite'
make[1]: *** [check-recursive] Error 1
make[1]: Leaving directory `/home/lrajani/ortool/build_or1ksim/testsuite'
make: *** [check-recursive] Error 1


-Rajani

Hi Rajani,

Are you using the latest version of Or1ksim from OpenCores SVN?

If so, the tests should all pass. So your failure may be due to a platform/OS we haven't used in testing. What hardware and operating system are you using?

Jeremy

--
Tel: +44 (1590) 610184
Cell: +44 (7970) 676050
SkypeID: jeremybennett
Email: jeremy.bennett@embecosm.com
Web: www.embecosm.com

I am using the 0.5.0rc2 version of or1ksim and these are the platform details:

UBUNTU 10.10 server

I am attaching the config.log file generated during configuration, please go through it.

-Rajani

Hi Rajani,

Is this 32-bit or 64-bit. I've just rerun all the tests using the latest version of SVN on a 64-bit Fedora 16 machine and it all worked fine. Could you try using the latest version from SVN and see if this cures your problem.

HTH

Jeremy

Hello Jeremy,

I checked out the latest version of OR1kSIM and find the same issue, 1 test from libsim and 1 test from or1ksim are failing.

BTW, I had to reconfigure the OR1KSIM after building the GNU toolchain, could this be an issue??

I am using a 64-bit machine.

Also, in the summary report it says:

"Target is 0r32-unknown-elf
Host is x86_64-unknown-linux-gnu".

Please clarify why does it say "unknown"


-Rajani

Hi Rajani,

I don't understand what this problem is then. There are detailed logs of the tests in bd/testsuite/or1ksim.log and bd/testsuite/libsim.log. Could you post them here, so I can look at them.

This is normal. You have to allow the build system to find the new compiler.

I've run the tests on 64-bit Fedora 16 here, and it works find for me.

This is the fully expanded triplet describing the machines used as target and host. This has the general form cpu-vendor-os. In the first case, the CPU is or32 and it is a bare metal target taking ELF binaries, so we show the OS as "elf". The manufacturer is not specified, so it is just shown as "unknown". It would be quite possible to fully specify the target as or32-opencores-elf, to indicate that "OpenCores" was the vendor.

In the second case, the CPU ix x86_64, the operating system is GNU/Linux (the obsession with reminding everyone that Linux is just the kernel, the rest of the operating system comes from the GNU project). The vendor might be Intel or AMD, but is not specified.

HTH

Jeremy

attaching or1ksim.log

attaching libsim.log

Hi Rajani,

I can't see what is causing the failure and I cannot reproduce it.

However looking at the code, this Ethernet test is ancient. It is one of the handful of tests that existed before I got involved with Or1ksim. It is not representative (it uses files to model the Ethernet interface, rather than the TUN/TAP interface) and I suspect continues to work for me by happy accident. It needs to be completely rewritten.

Would you file this as a bug in Bugzilla, along with the log files, so it can be addressed in the future. If you would like to contribute a rewritten test, that would be much appreciated. However for now I suggest you just ignore this failure.

Best wishes,

Jeremy

Thanks Jeremy for looking into this.

I have been trying to create a login for myself in Bugzilla, but it is not accepting my email IDs, I have sent a separate email to you regarding this. Once I get the login, I will add a bug for this test.
Regarding the test, I am still at the initial stage of getting to know about OR1KSim. I can re-write this test as an experiment once I am comfortable with the simulator.

Thanks,
Rajani

Hi Rajani,

You should be able to self-register with your opencores email address (vjraj@opencores.org) using the "Open a New Account" button on the main bugzilla page.

If that doesn't work, please email me.

HTH

Jeremy