Testing Done: |
|
---|
8982 Support building with OpenSSL 1.1
Review Request #827 — Created Jan. 31, 2018 and submitted
Information | |
---|---|
citrus | |
illumos-gate | |
master | |
8982 | |
c6ccd50... | |
Reviewers | |
general | |
8982 Support building with OpenSSL 1.1
- libkrb5/pkinit - tested certificate-based preauth/pkinit between machine and local KDC. Confirmed modified code was being used via dtrace.
- libkmf - Test kmf_openssl plugin with
pktool
. - sendmail - tested basic TLS interop
Testing Done: |
|
---|
Change Summary:
re-based following wanboot removal
Testing Done: |
|
||||||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Commit: |
|
||||||||||||||||||||||||||||||||||||
Diff: |
Revision 2 (+1541 -692) |
-
-
usr/src/Makefile (Diff revision 2) This check is somewhat bothersome -- we don't really need the openssl utility to build the gate, only the openssl libs? I understand that I'm nitpicking here, but still..
Change Summary:
Do not abort if no openssl utility found (only libraries are required to build)
Commit: |
|
||||
---|---|---|---|---|---|
Diff: |
Revision 3 (+1540 -692) |
Change Summary:
Updated information following additional testing.
Testing Done: |
|
||||||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Bugs: |
|
-
Aside from the OPENSSL macro question, there was nothing obvious from a first glance. I'll plan to look over everything some more a bit later as well, so there might be more questions later.
-
usr/src/Makefile.master (Diff revision 3) Is this strictly for version detection/printing in nightly(1) (via usr/src/Makefile)?
Change Summary:
Remove changes to pkcs11_tpm since '9156 Remove openssl dependency from pkcs11_tpm'
Remove changes to libipsecutil since it is being dealt with under https://illumos.org/rb/r/966/
Testing Done: |
|
||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Commit: |
|
||||||||||||||||||||||||
Diff: |
Revision 4 (+1437 -679) |
-
-
usr/src/cmd/sendmail/src/tls.c (Diff revision 4) Can the DSA_new() call succeed while the DSA_generate_parameters_ex() call fail? If so, this could leak memory.
-
usr/src/lib/krb5/plugins/preauth/pkinit/pkinit_crypto_openssl.c (Diff revision 4) It seems like if 'q = BN_new()' fails, g will get leaked.
-
usr/src/lib/krb5/plugins/preauth/pkinit/pkinit_crypto_openssl.c (Diff revision 4) Similar concern about leaking g as above.
-
usr/src/lib/krb5/plugins/preauth/pkinit/pkinit_crypto_openssl.c (Diff revision 4) if
q = BN_new()
fails, it appears g will be leaked. -
usr/src/lib/libkmf/plugins/kmf_openssl/common/openssl_spi.c (Diff revision 4) Won't this leak
rsa
if it fails? -
usr/src/lib/libkmf/plugins/kmf_openssl/common/openssl_spi.c (Diff revision 4) Won't this leak
rsa
andn
if this fails? -
usr/src/lib/libkmf/plugins/kmf_openssl/common/openssl_spi.c (Diff revision 4) Similar questions as above.
Commit: |
|
||||
---|---|---|---|---|---|
Diff: |
Revision 5 (+1497 -715) |