IRIX Network Forums
OpenSSL 1.1.1g test failures - Printable Version

+- IRIX Network Forums (//forums.irixnet.org)
+-- Forum: SGI/MIPS (//forums.irixnet.org/forum-3.html)
+--- Forum: Development/Porting (//forums.irixnet.org/forum-9.html)
+--- Thread: OpenSSL 1.1.1g test failures (/thread-2225.html)

Pages: 1 2

OpenSSL 1.1.1g test failures - Raion - 06-18-2020

Code:
Test Summary Report
-------------------
../test/recipes/01-test_symbol_presence.t        (Wstat: 512 Tests: 4 Failed: 2)
  Failed tests:  2, 4
  Non-zero exit status: 2
../test/recipes/01-test_test.t                   (Wstat: 256 Tests: 1 Failed: 1)
  Failed test:  1
  Non-zero exit status: 1
../test/recipes/30-test_evp.t                    (Wstat: 256 Tests: 10 Failed: 1)
  Failed test:  7
  Non-zero exit status: 1
../test/recipes/70-test_asyncio.t                (Wstat: 256 Tests: 1 Failed: 1)
  Failed test:  1
  Non-zero exit status: 1
../test/recipes/70-test_recordlen.t              (Wstat: 256 Tests: 1 Failed: 1)
  Failed test:  1
  Non-zero exit status: 1
../test/recipes/70-test_servername.t             (Wstat: 256 Tests: 1 Failed: 1)
  Failed test:  1
  Non-zero exit status: 1
../test/recipes/80-test_ct.t                     (Wstat: 256 Tests: 1 Failed: 1)
  Failed test:  1
  Non-zero exit status: 1
../test/recipes/80-test_dtls.t                   (Wstat: 256 Tests: 1 Failed: 1)
  Failed test:  1
  Non-zero exit status: 1
../test/recipes/80-test_ssl_new.t                (Wstat: 6400 Tests: 29 Failed: 25)
  Failed tests:  1-18, 20-21, 24-28
  Non-zero exit status: 25
../test/recipes/80-test_ssl_old.t                (Wstat: 1280 Tests: 6 Failed: 5)
  Failed tests:  2-6
  Non-zero exit status: 5
../test/recipes/80-test_sslcorrupt.t             (Wstat: 256 Tests: 1 Failed: 1)
  Failed test:  1
  Non-zero exit status: 1
../test/recipes/90-test_sslapi.t                 (Wstat: 256 Tests: 1 Failed: 1)
  Failed test:  1
  Non-zero exit status: 1
../test/recipes/90-test_sslbuffers.t             (Wstat: 256 Tests: 1 Failed: 1)
  Failed test:  1
  Non-zero exit status: 1
../test/recipes/90-test_tls13ccs.t               (Wstat: 256 Tests: 1 Failed: 1)
  Failed test:  1
  Non-zero exit status: 1
Files=155, Tests=1461, 1584 wallclock secs (20.50 usr  1.17 sys + 1284.76 cusr 279.96 csys = 1586.39 CPU)
Result: FAIL
make[1]: *** [_tests] Error 1
make[1]: Leaving directory `/admin/openssl-1.1.1g'
make: *** [tests] Error 2

Tried this a few times with MIPSPro at different settings and opt levels, and it's still getting a fair number of test failures. Additionally, I've linked cURL, both 7.60 and 7.70 against it, and it's broken for SSL connections. This points to the OpenSSL library not working properly.

I'm unsure of what exactly all of these tests mean, but I do understand a few:

> ../test/recipes/01-test_symbol_presence.t

I'm gonna guess this refers to basic text or data symbols in the library not being detected?

> ../test/recipes/90-test_tls13ccs.t

This is obviously a TLS 1.3 test. Not crucial, but I'm guessing certain ciphers are not working or something.

The rest I'm not 100% on, but obviously there's issues here. I don't have a github, and refuse to sign back up with them, so ideally we can resolve this here if anyone knows some places to begin looking.

RE: OpenSSL 1.1.1g test failures - mgtremaine - 06-18-2020

You not alone on this one I found a thread from last year related to openssh builds that shows the same issue in the underlying openssl 1.1.1 branch

https://bugzilla.mindrot.org/show_bug.cgi?id=3032

IN that thread Darren Tucker says

"
> Failed test: 1
> Non-zero exit status: 1
> ../test/recipes/30-test_evp.t (Wstat: 256 Tests: 10 Failed: 1)
> Failed test: 7

This one makes me very suspicious. The tests in that file are:

my @files = ( "evpciph.txt", "evpdigest.txt", "evpencod.txt", "evpkdf.txt",
"evpmac.txt", "evppbe.txt", "evppkey.txt", "evppkey_ecc.txt",
"evpcase.txt", "evpccmcavs.txt" );

and #7 is "evppkey.txt". If we look at that file (test/recipes/30-test_evp_data/evppkey.txt) it's "Public key algorithm tests" including parsing private and public keys.
"

Which renders OpenSSH unable to read in any keys .. I do not know how to fix the 1.1.1 series of openssl but I'll keep my eyes open on this.

-Mike

RE: OpenSSL 1.1.1g test failures - Raion - 06-19-2020

Yeah, I've gotten the 1.0.2 branch working fine, but that's not going to continue to work properly in the future.

It's annoying, and GCC doesn't seem to fix it either.

RE: OpenSSL 1.1.1g test failures - mgtremaine - 06-19-2020

Did you catch this in the change log

Changes between 1.0.2h and 1.1.0 [25 Aug 2016]
.
.
*) Remove support for MIPS o32 ABI on IRIX (and IRIX only).
[Andy Polyakov]


I assume n32 still exists but o32 is gone? (At least I hope n32 is a valid target)

-Mike

RE: OpenSSL 1.1.1g test failures - Raion - 06-19-2020

The o32 ABI is not being targeted by me, I'm certainly passing the -n32 flag.

But it could be they broke something

RE: OpenSSL 1.1.1g test failures - jpstewart - 06-20-2020

I notice the OpenSSL test scripts are all written in Perl.  So is it possible that the tests are failing due to problems with the IRIX Perl installation?  They use '/usr/bin/env perl' in the shebang line, so if the ancient IRIX Perl is first in your path and the scripts are assuming a newer Perl that could cause all sorts of problems.  Even if you've got a more up-to-date Perl installation, might the scripts be trying to use Perl modules you don't have installed?

I'm not a Perl guy at all, so I may be way off base here.  But the number of failures makes me wonder if the test mechanism itself is to blame.  (I've certainly seen tests fail in the past for silly reasons like the difference between 0.0 and 0.000000 due to the test scripts making unfounded assumptions about how awk and other utilities truncate their output.)

RE: OpenSSL 1.1.1g test failures - Raion - 06-20-2020

I am sure I changed all the shebangs to where my perl actually is.

I'll go back to double check and report back if that fixes anything...

RE: OpenSSL 1.1.1g test failures - Raion - 07-02-2020

Same failures after I patched the pl files to use my perl. Clearly, there's still work to be done. It appears there's actual issues with the code.

RE: OpenSSL 1.1.1g test failures - hamei - 07-02-2020

Anybody good with C (jp, don't you go running off now ! I see you hiding behind the refrigerator !) to create a pass-through SSL ? The flocking thing is worthless anyhow, so just trick their entire useless "sekurite" apparatus right out the door. We're not doing banking on an O2 ....

RE: OpenSSL 1.1.1g test failures - Raion - 07-02-2020

SSL proxies exist, but openssl is also important for crypto and other things.