Wez Furlong
4cf2e65c94
dkim: add comparitive benchmark against mail-auth
...
```
signing/sign Simple time: [1.2990 ms 1.3002 ms 1.3015 ms]
thrpt: [2.3581 MiB/s 2.3603 MiB/s 2.3626 MiB/s]
Found 1 outliers among 100 measurements (1.00%)
1 (1.00%) high mild
signing/sign Relaxed time: [1.2942 ms 1.2955 ms 1.2968 ms]
thrpt: [2.3665 MiB/s 2.3689 MiB/s 2.3712 MiB/s]
Found 4 outliers among 100 measurements (4.00%)
1 (1.00%) low mild
1 (1.00%) high mild
2 (2.00%) high severe
mail-auth signing/sign Simple
time: [487.97 µs 488.22 µs 488.48 µs]
thrpt: [6.2826 MiB/s 6.2860 MiB/s 6.2892 MiB/s]
Found 1 outliers among 100 measurements (1.00%)
1 (1.00%) high severe
mail-auth signing/sign Relaxed
time: [494.68 µs 495.68 µs 496.68 µs]
thrpt: [6.1789 MiB/s 6.1913 MiB/s 6.2039 MiB/s]
```
2023-06-15 15:41:18 -07:00
Wez Furlong
30b8522873
dkim: precompile list of headers to sign during signing
...
We can exploit eg: knowing that the list of headers is a unique
list (no repeated headers) and save some overhead when using the
same signer multiple times.
2023-06-15 15:18:17 -07:00
Wez Furlong
4805f60837
dkim: avoid allocating a vec when signing headers
2023-06-15 14:45:40 -07:00
Wez Furlong
02d1378495
dkim: avoid re-parsing header list during signing
2023-06-15 14:28:11 -07:00
Wez Furlong
e5d03052be
dkim: tidy up handling of l= body length tag
...
We should parse at the edge and pass down the parsed number
2023-06-15 13:23:44 -07:00
Wez Furlong
cf11540f84
dkim: remove another extraneous allocation, tidy up canon interface
2023-06-15 13:13:28 -07:00
Wez Furlong
28f0786e42
dkim: update crate deps
2023-06-15 11:37:59 -07:00
Wez Furlong
19c9201e44
dkim: remove some redundant clone calls
2023-06-15 11:30:44 -07:00
Wez Furlong
2e40c00121
dkim: introduce ParsedEmail type
...
This help transitions to a cheaper email parser, that doesn't need to
parse the entire MIME structure
Note that as part of this, I found that several tests were running
against email messages using unix line endings instead of canonical
CRLF.
This was silently being masked and an empty or otherwise incorrect byte
slice was being hashed; I've updated the affected tests and their test
expectations accordingly, as well as made this error bubble up at the
time of parsing the email, rather than silently squashing it in the
depths of the hasher.
2023-06-15 11:27:03 -07:00
Wez Furlong
3af3f52f66
dkim: pick some lint various bits
2023-06-15 10:27:33 -07:00
Wez Furlong
19aca29007
dkim: don't require moving data into verify_signature
2023-06-15 10:07:33 -07:00
Wez Furlong
424192eb5f
dkim: tidy up DKIMHeader construction
2023-06-15 10:07:26 -07:00
Wez Furlong
aecd187c21
dkim: remove some string copies
2023-06-15 09:53:53 -07:00
Wez Furlong
ed8ccbc810
dkim: remove slog dependency
2023-06-15 09:27:28 -07:00
Wez Furlong
af3236a72e
dkim: remove Cargo.lock
2023-06-15 09:17:43 -07:00
Wez Furlong
eb44e0780e
Use vendored in dkim fork
2023-06-15 09:11:34 -07:00
Wez Furlong
49c1533435
Merge in my DKIM crate fork
2023-06-15 09:10:39 -07:00
Wez Furlong