2
0
mirror of https://github.com/mozilla/cipherscan.git synced 2024-11-22 14:23:41 +01:00
Commit Graph

61 Commits

Author SHA1 Message Date
Hubert Kario
b673fb976a separate AES-CBC from AES-GCM 2015-05-30 00:01:32 +02:00
Hubert Kario
d773b73e45 don't divide by zero on empty results folder 2015-05-30 00:01:32 +02:00
Hubert Kario
b9b3a221ce add Firefox 35 cipher settings 2015-05-30 00:01:32 +02:00
Hubert Kario
82f643244e don't count export grade ciphers towards PFS 2015-05-30 00:01:32 +02:00
Hubert Kario
1b360153a0 sum servers that support SSL3 or TLS1 as the highest protocol 2015-05-30 00:01:32 +02:00
Hubert Kario
341f657e83 better detection for EXP and low grade ciphers in stats
EXP is self explanatory - export grade
DES-CBC3-MD5 is available only in SSLv2 - not secure
RC4-64-MD5 is also a weakened version (though not marked as export grade)
2015-05-30 00:01:32 +02:00
Julien Vehent
4a6ff56b81 Add back support for old curve json format in parse results 2015-04-02 04:39:59 -04:00
Julien Vehent
b2a399617f Use new JSON format in parse_results 2015-04-01 14:50:49 -04:00
Hubert Kario
c52e008347 add support for testing supported curves
since early versions of 1.0.2 openssl supports -curves command line
option, it allows us to set the curves advertised as supported

use the same approach to testing: advertise all, check what server
accepts, remove the accepted from list, repeat. When server aborts
connection or selects non ECC cipher, we know that we've tested all.
2015-03-27 10:04:15 -04:00
Hubert Kario
1eae0cc71b use CApath for certificates and store certificates (v2)
CApath is about 20% faster than CAfile so use it, also
save the received certificates from the servers for later analysis
(proper hostname checking, looking for certificates sharing private key,
etc.)

Use the mechanism from cipherscan to find location of ca cert bundle
2014-11-05 18:13:39 +01:00
Julien Vehent
dca3457d5a Merge pull request #28 from tomato42/certificate-stats
Certificate stats
2014-11-03 22:15:44 -05:00
Hubert Kario
5a6eaaac41 parse_CAs.c - implement error checking, remove magic numbers, compile fix 2014-10-30 23:37:43 +01:00
Hubert Kario
aac3e9a9db parse_CAs.py - add few comments 2014-10-30 01:41:46 +01:00
Hubert Kario
edab545f3e add Makefile for the C utility 2014-10-30 01:33:58 +01:00
Julien Vehent
ebc6939299 Merge pull request #29 from tomato42/client-handshake-simulation
Client handshake simulation
2014-10-29 19:22:52 -04:00
Hubert Kario
11ce6187de small fixes for delay
firstly, test_cipher_on_target() will try at least 4 connections before
incurring the sleep, for aggressive rate limiter on server side it may be
too much, so sleep before every connection

secondly, because running external commands like sleep incurs a fork
penalty, we first check if it is necessary
2014-10-28 16:44:43 +01:00
Hubert Kario
29c739faa9 count EDH-DES as PFS too in general stats 2014-10-25 16:23:41 +02:00
Hubert Kario
af2e25ec89 fix EDH checking
old ciphers have names that use EDH instead of DHE so we need check
for both names
2014-10-25 16:11:18 +02:00
Hubert Kario
76d791fcbe make cipher selection simulation generic
it's relatively easy to make the cipher selection generic,
so that adding different clients is as easy as converting their
client hello cipher ordering to openssl cipher names
2014-10-12 20:39:39 +02:00
Hubert Kario
c82bc44558 report cipher ordering in scanning stats, use it to simulate handshakes
since now we know if server honours client order or not, we can use it
to properly simulate handshakes for a given client, also report
the general stats of this server configuration variable
2014-10-12 20:39:39 +02:00
Hubert Kario
42fa7d9ecb report what ciphers Firefox would select while connecting to server 2014-10-12 20:39:39 +02:00
Hubert Kario
1b4dcc4393 report ciphers causing incompatibility for Firefox
It turns out that the situation is even more bleak for Firefox
with regards to RC4, add it to report
2014-10-12 20:39:39 +02:00
Hubert Kario
142726c4fd count ECDH-RSA ciphers as ECDSA
the ECDH parameters come from server certificate - the point
on elliptic curve. The RSA comes from the signature on the certificate
which comes from CA
2014-10-12 20:39:39 +02:00
Hubert Kario
ac18195b21 process-certificate-statistics.sh - the script HOWTO to turn results to CA stats 2014-10-12 20:38:25 +02:00
Hubert Kario
3cfd7b76cc collect statistics about found certificates 2014-10-12 20:38:25 +02:00
Hubert Kario
3699acfc2d helper application for finding cert chains
because neither M2crypto nor OpenSSL packages provide extensive
enough API to do certificate chain building, verification
and outputting of details, we have to pre-parse the data
with a C app that can access the full OpenSSL API.

I've also tried monkey patching the packages, but unfortunately
the result wasn't working reliably

The actual statistic collection (both about the chains and
specific certificates) will be done in a python script
2014-10-12 20:34:53 +02:00
Hubert Kario
ca0ef2fc5c fixes for the pull request #18
there were few small issues with the pull #18 even though jvehent merged
it, this fixes them
2014-10-06 13:26:53 -04:00
Hubert Kario
29109f1e64 update SEED and IDEA classification, do a total of broken ciphers
SEED and IDEA are not good ciphers, but not broken, so count them
separately, do a total count of servers that support broken and insecure
ciphers
2014-10-06 13:25:04 -04:00
Hubert Kario
8a0c9190a9 sort reported TLS session ticket hint using natural sort 2014-10-06 13:20:37 -04:00
Julien Vehent
ecd77f94fc Merge pull request #18 from tomato42/wip
Hodgepodge of fixes
2014-08-28 16:02:19 -04:00
Hubert Kario
7591062bbc parse_results.py: compatibility with old results files 2014-06-04 18:52:39 +02:00
Hubert Kario
be0439ef99 provide statistics for all key exchange methods, not DHE and ECDHE only 2014-06-04 18:17:41 +02:00
Hubert Kario
3667b04ad7 correctly count broken cipher suites with "no reporting of untrusted" 2014-06-04 18:17:02 +02:00
Hubert Kario
86ff1122cc parse_results.py: don't count anonymous cipher suites toward correct config stats 2014-06-04 15:15:32 +02:00
Julien Vehent
50f4959e79 updated license on parse_results.py 2014-05-20 08:23:57 -04:00
Hubert Kario
2f56f0515e don't scan the same host twice 2014-05-16 18:16:45 +02:00
Hubert Kario
f1d3b51749 update the top 1M list to the version from 2014-05-16 2014-05-16 17:34:22 +02:00
Hubert Kario
4e94d95bd8 ask for OCSP stapling by default
for now, no option to disable
2014-05-16 17:31:44 +02:00
Hubert Kario
0777682aa6 collect TLS ticket lifetime hints 2014-05-16 16:55:19 +02:00
Hubert Kario
1a78172936 scan just one host per hostname 2014-05-16 16:11:01 +02:00
Hubert Kario
cdbf596466 properly handle pure IP adressess
(it's illegal to use IP in SNI)
2014-05-16 15:42:47 +02:00
Hubert Kario
5ef53dda9c increase paralelism of jobs
because sometimes tcping takes a long time to timeout for a lot
of hosts in batch use also load average to keep the cpu busy
2014-05-13 13:41:16 +02:00
Hubert Kario
a213fc45d0 remove the folder/file part from url
some hostnames in the top-1m.csv file have folder or site specified
in them, cut it off before using
2014-05-13 13:41:16 +02:00
Hubert Kario
00b20a20ed perform SNI enabled scan
for example, youtube requires SNI extension to be present to return
ECDSA certificates, use it for scanning
2014-05-13 13:41:16 +02:00
Hubert Kario
8817a7b1c8 testtop1m.sh: correct counting of background jobs
`jobs` command returns multiple lines for a jobs with `if` so counting
number of background jobs was off
2014-05-13 13:41:16 +02:00
Hubert Kario
686d7c958b extend reporting of RC4-related stats
While preferring RC4 in TLS1.0 or SSL3 was recommended before,
it was always known that TLS1.1 and TLS1.2 were not vulnerable against
BEAST, so forcing RC4 there is a mistake. Report number of such servers.
2014-04-19 23:14:57 +02:00
Hubert Kario
21bba67df0 extend SSL stats
Two interesting server configurations are ones that support
only SSL3 or TLS1 only (old, but otherwise correctly configured servers)
and ones that support only TLS1.1 or up (brave admins that support
only new clients)
2014-04-19 23:14:57 +02:00
Hubert Kario
349d4ebc3c more detailed PFS report
Just because server supports some bad DH params, doesn't mean
it will force them on users. Report number of servers
that prefer specific DH params.
2014-04-19 23:14:57 +02:00
Hubert Kario
d3b6f9b507 fix reporting of the TLS1.2 but not TLS1.1
Some servers may be configured to support only TLS1.2, it would
count them towards the number of servers affected by the OpenSSL bug
2014-04-19 23:14:57 +02:00
Hubert Kario
c8abfb53e8 add support for Chacha20 based ciphers
Basically all Google servers support Chacha20 now and it is
not a bad choice, so report it as a regular cipher
2014-04-19 23:14:57 +02:00