WIP: Sha1 runtime unittest

[pemensik]

These changes complements PR #660, which added some support into unbound for runtime disabled SHA1 validation. Depending on setting in crypto policy and resulting codes in crypto library, it either considers signature indeterminate. That is roughly equivalent to insecure, but we have some signatures present and no proof about missing DS record.

This fixes unittest to pass on RHEL9, but rpl tests do not yet pass.

[pemensik]

Currently fails to me make test with:

Oct 19 18:16:42 unbound[23932:0] info: testbound: do STEP 11 CHECK_AUTOTRUST
/tmp/testbound_23932_auto_example.com.tmp: 1 ok : ; autotrust trust anchor file
/tmp/testbound_23932_auto_example.com.tmp: 2 ok : ;;id: example.com. 1
/tmp/testbound_23932_auto_example.com.tmp: 3 ok : ;;last_queried: 1258969600 ;;Mon Nov 23 09:46:40 2009
Oct 19 18:16:42 unbound[23932:0] error: mismatch in file /tmp/testbound_23932_auto_example.com.tmp, line 4
Oct 19 18:16:42 unbound[23932:0] error: file has : ;;last_success: 1258962400 ;;Mon Nov 23 07:46:40 2009
Oct 19 18:16:42 unbound[23932:0] error: should be: ;;last_success: 1258969600 ;;Mon Nov 23 09:46:40 2009
/tmp/testbound_23932_auto_example.com.tmp: 5 ok : ;;next_probe_time: 1258972867 ;;Mon Nov 23 10:41:07 2009
Oct 19 18:16:42 unbound[23932:0] error: mismatch in file /tmp/testbound_23932_auto_example.com.tmp, line 6
Oct 19 18:16:42 unbound[23932:0] error: file has : ;;query_failed: 2
Oct 19 18:16:42 unbound[23932:0] error: should be: ;;query_failed: 0
Oct 19 18:16:42 unbound[23932:0] error: mismatch in file /tmp/testbound_23932_auto_example.com.tmp, line 7
Oct 19 18:16:42 unbound[23932:0] error: file has : ;;query_interval: 5400
Oct 19 18:16:42 unbound[23932:0] error: should be: ;;query_interval: 3600
/tmp/testbound_23932_auto_example.com.tmp: 8 ok : ;;retry_time: 3600
Oct 19 18:16:42 unbound[23932:0] error: mismatch in file /tmp/testbound_23932_auto_example.com.tmp, line 9
Oct 19 18:16:42 unbound[23932:0] error: file has : example.com.	10800	IN	DNSKEY	257 3 5 AwEAAc3Z5DQDJpH4oPdNtC4BUQHk50XMD+dHr4r8psHmivIa83hxR5CRgCtd9sENCW9Ae8OIO19xw9t/RPaEAqQa+OE= ;{id = 55582 (ksk), size = 512b} ;;state=2 [  VALID  ] ;;count=0 ;;lastchange=1258962400 ;;Mon Nov 23 07:46:40 2009
Oct 19 18:16:42 unbound[23932:0] error: should be: example.com.	3600	IN	DNSKEY	257 3 5 AwEAAdz+Xe5qS3BRnw1hBy2wL2wi0o3Nh94lDxtDtfOsmyJ0WD/25Ova9Pb27Yzh5XW/baENkI+xGJTFsljbi9bdSd0= ;{id = 63067 (ksk), size = 512b} ;;state=1 [ ADDPEND ] ;;count=1 ;;lastchange=1258969600 ;;Mon Nov 23 09:46:40 2009
Oct 19 18:16:42 unbound[23932:0] error: mismatch in file /tmp/testbound_23932_auto_example.com.tmp, line 10
Oct 19 18:16:42 unbound[23932:0] error: file has : example.com.	10800	IN	DNSKEY	257 3 5 AwEAAas/cAhCFXvBUgTSNZCvQp0pLx1dY+7rXR0hH4/3EUgWmsmbYUpI1qD0xhwKD/oYGEwAm291fyWJ9c0oVxXDEK8= ;{id = 16486 (ksk), size = 512b} ;;state=2 [  VALID  ] ;;count=0 ;;lastchange=1258962400 ;;Mon Nov 23 07:46:40 2009
Oct 19 18:16:42 unbound[23932:0] error: should be: example.com.	3600	IN	DNSKEY	257 3 5 AwEAAd9vx7tR9cd9MMDh0gL/qHNTG4ykehjT3UzIIEtAi3Z4DI3/FFw9U/GjpYcqVC6hx2Yo1lbc4tVIa/uA0mbU7uE= ;{id = 58687 (ksk), size = 512b} ;;state=1 [ ADDPEND ] ;;count=1 ;;lastchange=1258969600 ;;Mon Nov 23 09:46:40 2009
Oct 19 18:16:42 unbound[23932:0] error: autotrust check failed, could not read line
Oct 19 18:16:42 unbound[23932:0] error: file /tmp/testbound_23932_auto_example.com.tmp, line 11
Oct 19 18:16:42 unbound[23932:0] error: should be: example.com.	3600	IN	DNSKEY	257 3 5 AwEAAeu99txoU5i2Z4BVatCVi9PTj93oOPft8ZB9ovcjfzRZLpfc/woges07k5Ru+H44qSRxjtDKDqtf4QSo3RkkGLk= ;{id = 56782 (ksk), size = 512b} ;;state=1 [ ADDPEND ] ;;count=1 ;;lastchange=${$t0} ;;${ctime $t0}
Oct 19 18:16:42 unbound[23932:0] fatal error: autotrust_check failed 
../testdata/autotrust_10key.rpl  failed

[pemensik]

It seems many tests should be recreated with non-SHA1 algorithms if that is not required. Many of those tests would be just ignored and not checked on RHEL9-like systems. There is quite a lot of results when using command grep 'RRSIG\s\+\w\+\s\+[57]\s' testdata/*.rpl. Is there documented any original data, how these test recipes were created? Were they created by hand or by some tool?