Commit graph

401 commits

Author SHA1 Message Date
bunnie
9cf48fcaa5 add storage limits on FIDO cred count
perfomance regression on the PDDB with the latest fixes make
the device stop passing compliance when the database grows
to about 40-ish keys. For now, we set the cred limit to 32,
until the performance issues are addressed. (note that yubico
has a limit of 25 credentials, so, 32 is not insanely low).
2022-06-27 00:04:46 +08:00
bunnie
79c20eb498 fix in-place edit of keys
The behavior with a secret basis is as follows:

- The key stays in the basis in which it was created.
- So, if you created k1 in .System basis, and k2 in Secret basis,
  if you edit k1 while Secret basis is open, k1 stays in .System
  basis. it does *not* migrate into Secret basis.

I think this is the "least surprising" behavior. There is an
argument that "maybe I want to keep secret notes in an overlay
on a key" but I feel like creating a second shadow copy of a key
with a different set of notes is not the most desired behavior
for this application (NB: it could be desired for others applications,
and that behavior can be programmed with the PDDB API; which is
why the API is a bit explicit at the moment. This flexibility goes
against the goal of "foolproof deniability", unfortunately, but
I think it is too early to know which default to set for the system).
2022-06-26 17:37:57 +08:00
bunnie
42b84ad890 adjust some of the default basis references in OpenSK storage implementation
These adjustments ideally make the system less confusing to use in
the face of secret Bases, but the spec calls for some information that
could be a leakage channel (see below).

- items that are assumed to be consistent regardless of the basis
are now forced into the System basis. This reduces potentially confusing
behavior if the system has han unlocked secret basis and things like
PIN hashes are set which appear to be global.

- There is a "global signature counter" which could leak some information
about how many credentials are in the device. However, the FIDO2 spec
enforces some plausible deniability by requiring that the counter increments
by a random value instead of by 1. In the OpenSK implementation, this increments
by a random integer from 1-8. So, if you store dozens of credentials in secret
bases, it may become less deniable due to this global signature counter.

The global signature counter seems optional, so maybe that is a feature to
be stripped out...
2022-06-26 14:25:54 +08:00
bunnie
a4c6b65b11 Merge branch 'main' of github.com:betrusted-io/xous-core into main 2022-06-26 04:06:40 +08:00
bunnie
9f409fda2d fix basis search order issues
at least, it passes some simple tests. much more testing is needed.
2022-06-26 03:52:27 +08:00
bunnie
e62c9bf3d3 add basis management functions and debug some basis issues
there's some problems with the automatic basis unionization
feature. now debugging...
2022-06-26 02:11:13 +08:00
bunnie
475fadb255 make search strings persist between modes and menus 2022-06-26 02:10:50 +08:00
bunnie
819b84954a add autotype to TOTP. why not? 2022-06-25 23:02:12 +08:00
bunnie
6aadbf3cae fix error message on edit of FIDO2 entry 2022-06-25 23:02:01 +08:00
bunnie
4319ac4f45
now changes when time is set
fix logic
2022-06-25 09:01:02 +08:00
bunnie
68db7e8c7a reduce noise from the prereqs debug loop 2022-06-25 00:23:46 +08:00
bunnie
22702cc03b add a prereqs checkpoint
this checks to make sure that the system is in a usable state
for the vault to run before allowing it to be launched.
in particular, the PDDB must be initialized, and so must the
system time.
2022-06-25 00:22:09 +08:00
bunnie
7deb6fd4a1 add notes about how to clear test entries 2022-06-24 18:58:07 +08:00
bunnie
4771f04451 update progress while making test entries 2022-06-24 18:57:44 +08:00
bunnie
c6aecc08dc fix bug where entries don't redraw scrolling up/down to new page 2022-06-24 18:57:24 +08:00
bunnie
3c82898780 fix some minor bugs
- test entries for passwords at 100 entries. it's...pretty slow
at 300 passwords, about 8 seconds to scan/read-in the database.
once the database is read-in, sorting and viewing is fast. good
news is that heap usage is about 300k for that many entries,
no OOM crashes found. just slow.

- fix an unimplemented()! error

- strip out some debug
2022-06-24 17:23:00 +08:00
bunnie
847e64c46a add totp record add/edit rough draft UX flows 2022-06-24 15:53:57 +08:00
bunnie
024b9d9ce1 make it so that attempts to list empty dictionaries don't fail
this might point at a larger error of empty dictionaries not
being recorded, but this is a work-around for now...?
2022-06-24 15:53:23 +08:00
bunnie
9a5c04cc73 add a time progress bar in TOTP mode 2022-06-24 03:40:21 +08:00
bunnie
7765f3dec3 add other hash support; notch back the hmac dependencies
our engine is a version older in API rev, so notch back the
dependencies so we're all at the same rev
2022-06-24 01:46:09 +08:00
bunnie
5d8b85c0bf totp demo working 2022-06-24 01:18:37 +08:00
bunnie
57845b6fb6 clean up warnings in ctap 2022-06-24 01:18:32 +08:00
bunnie
729a6ad186 add the beginning of the TOTP functions 2022-06-24 00:31:25 +08:00
bunnie
ce461a06fe add routines to populate test entries, and use less synthetic data 2022-06-24 00:30:45 +08:00
bunnie
e998e96323 change visibility of some routines so we can access storage in the UX 2022-06-24 00:30:29 +08:00
bunnie
512f435b2a move the FIDO random secret into the FIDO_DICT
originally we stuck it in the CRED_DICT because it forces its
creation -- but there's no reason to *really* do that. We just
set the dictionary to be created if it's not existent in the later
API calls.

By moving this to FIDO_DICT instead of FIDO_CRED_DICT it makes
key access & sorting much cleaner in the UX, because you don't
have this odd key hanging out, and it also means that you can't
have an attack on the system where you try to name a key with
the name held in FIDO_CRED_DICT to try and read it out.
2022-06-24 00:28:58 +08:00
bunnie
5e08941950 change display order for passwords
sort by description not username
2022-06-23 04:22:11 +08:00
bunnie
ad22618c8d add some debug around deserialization errors 2022-06-23 04:17:37 +08:00
bunnie
8c986035b8 add trailing newline to serializations
the final entry can accrue cruft after it if there is no newline
2022-06-23 04:11:58 +08:00
bunnie
57f9e1a5f9 fix font setting 2022-06-23 03:01:44 +08:00
bunnie
304c865bb3 fix count on password access; make USB HID selected when vault is active 2022-06-23 02:27:05 +08:00
bunnie
09046854db add password edit/delete functions 2022-06-23 01:32:06 +08:00
bunnie
9671ef65f5 add password deserialization framework and UX routines 2022-06-22 21:20:19 +08:00
bunnie
f7f2868802 refactor to put blocking modals into its own thread
UX interaction via blocking modals are put into its own thread
so that the background UX can update while menus and modals are
popped up.

Still some bugs to work out to make this fully implemented,
but this is the major work fleshed out.
2022-06-22 04:52:40 +08:00
bunnie
ad92cbe1cb add customized menu contexts per sub-app 2022-06-21 02:41:03 +08:00
bunnie
8d4cdb3284 fix indexing bugs 2022-06-21 01:14:37 +08:00
bunnie
82e9c57ed4 add dirty rectangle flags to improve drawing performance 2022-06-20 22:20:28 +08:00
bunnie
b310967cf2 strip out excess redraws 2022-06-20 03:09:24 +08:00
bunnie
5a72076768 add dynamic search as text is typed 2022-06-19 19:04:25 +08:00
bunnie
72e6be8679 add some initial selectable list framework 2022-06-19 17:33:26 +08:00
bunnie
b945a085b0 more prototyping of the list browser UI 2022-06-19 03:17:12 +08:00
bunnie
b7453b627e add API tokens to prediction API
The prediction API now takes an API token. This is a 128-bit
number that gets passed to the prediction engine, and it's meant
to be unique for every context that requires a predictor.

It's up to the prediction engine to do anything with the token.
For the "null" engines (such as the icontray for `vault` and
`tts`) the token is just ignored. But for example, in `shellchat`,
the API token is used to store a history record in a HashMap on
a per-token basis. This means that every app that uses the
`shellchat` history function gets their own history that can only
be read out by that app.

The actual token itself is mostly passed through the API. The
token storage is in the Context layer (each Context has a record
for a token, that starts as `None`), and is passed down to
the predictor every time the context swaps, and the predictor
receives the token and passes it back if the token was `Some`,
or allocates a new random token if it was `None`.

Note that everything is a Context, including, for example, menus and
modal dialog boxes, but not all contexts have predictors. Only
Contexts that ask for a predictor have them. This mechanism
works smoothly in this model (as opposed to other hacks and work-arounds
previously tried getting to this solution).
2022-06-19 01:01:25 +08:00
bunnie
d874e50030 stash the latest changes 2022-06-18 13:59:26 +08:00
bunnie
c46579117b add the "trusted init" check to hosted mode
need to debug a problem with the IMEF connecting/disconnecting
causing trusted init to fail when there is more than one
plugin in the system.
2022-06-18 01:30:49 +08:00
bunnie
11932898ac oops, broke CI with something in *this* code
push a fix that will get CI running, but breaks this dev flow
2022-06-18 01:16:39 +08:00
bunnie
645fcf46fd refactor file locations so the structure is more scalable 2022-06-17 23:52:04 +08:00
bunnie
e5b4f4ad7f clean up a minor warning for a merge to main
let's put these GAM changes into production so others can use
them, even though the app is still coming together.
2022-06-17 21:38:45 +08:00
bunnie
f77324a4e5 add some test text 2022-06-17 21:37:39 +08:00
bunnie
37b2ee444d add basic submenu support to vault app 2022-06-17 18:51:13 +08:00
bunnie
3d4f689ef5 lay in the groundwork for the "icontray"
this allows the predictor area to serve as hot keys to
swap between modes of operation for the vault app.
2022-06-17 12:42:22 +08:00
bunnie
d2aff7da44 Fix #181
This is still a bogus AAGUID but at least it's less likely to
collide with other bogus AAGUIDs.
2022-06-13 17:49:14 +08:00
bunnie
cce617366f move to the release version of usbd-human-interface-device
thanks @dlkj!
2022-06-13 00:20:14 +08:00
bunnie
9ab3f71f23 modify naming of FidoInterface to RawFidoInterface
per maintainer request of usbd-human-interface-device for PR
2022-06-12 23:05:20 +08:00
bunnie
3f791f5979 sync pddb after any credential change state
help prevent data loss in case of power outage/crash
2022-06-11 03:08:23 +08:00
bunnie
cae1c3267e add sentinels for test automation 2022-06-10 21:16:59 +08:00
bunnie
42996a8177 fix U2F flow
accidentally broke it while debugging FIDO2. added notes about how
the flow is implemented.
2022-06-09 02:57:31 +08:00
bunnie
e68c976735 configure vault for production and add maintainer's notes 2022-06-09 00:20:36 +08:00
bunnie
3fc038cca9 add some missing translations, and add a listener to the unplug state 2022-06-08 23:48:09 +08:00
bunnie
52211f078e adjust autotest flow for UDP-based button pressing 2022-06-08 20:15:32 +08:00
bunnie
ec5d61e678 add comment on test 2022-06-08 04:35:04 +08:00
bunnie
7db4db2701 fix ECDH error
was missing the FIDO-required sha256 hash after ECDH key exchange
2022-06-08 04:33:09 +08:00
bunnie
4a12b458c9 seems like the DH key exchange is broken somehow? needs more digging.
i've extracted the key parameters from the test suite and confirmed that
that the hash value is computed correctly on their side.

i've also confirmed that the hash value is computed correctly from
the locally computed key.

so what's gone wrong is the devired secret key doesn't match what's
on the other side.

possible issues could be either a PK serialization issue, or something
wrong in the DH exchange itself.
2022-06-07 18:55:59 +08:00
bunnie
ae9c2d5b99 add authentication request UI; fix tests again.
trying to debug why PIN authentication doesn't work. looks like
Cosekey conversion is fine, the data is arriving OK...but the
dh key exchange is not generated the expected shared secret.
2022-06-07 17:36:28 +08:00
bunnie
51be4700ee improve UX for FIDO2 and try to ferret out FIDO2 test problems. 2022-06-07 03:56:52 +08:00
bunnie
d61732a3aa add note to self about not updating the hex dependency in the tests for crypto libraries 2022-06-06 22:55:19 +08:00
bunnie
6be6537194 fix dependency for tests to work 2022-06-06 22:54:48 +08:00
bunnie
b0c7895e81 add total use count to statistics stored on auth keys 2022-06-06 22:45:57 +08:00
bunnie
052c63aea3 why do these old files keep creeping in after a merge... 2022-06-06 19:05:53 +08:00
bunnie
86a1f9741e add UX flow for U2F transactions: register app name and display it
this allows users to know what exactly they are approving.
2022-06-06 19:00:43 +08:00
bunnie
73f3ef827c cleanup warnings and unused code 2022-06-06 04:39:43 +08:00
bunnie
25bbfc835e u2f is now "working" in that it passes some basic tests 2022-06-06 03:46:11 +08:00
bunnie
8fd94f061e fix default backend issue on configuration tool 2022-06-05 02:11:37 +08:00
bunnie
1024cebe55 add primitive UX layer to U2F application. fix numerous bugs
The OpenSK UX layer was completely gutted and re-written using
Xous-native idioms. There's a bit of ambiguity on what exactly
OpenSK meant to do with the one button and the LEDs, but for
Xous we want to show things like what you're signing for and
why.

For now, we just show the app hash during registration and signing,
but down the road that will need to be matched up against a
plaintext description of the application upon registration, so
users can have a client-side description of what they are approving.

At the moment, we're stuck at provisioning the attestation certificate
and key. The transaction goes through, but is rejected by the
test web site right now with not too much explanation of what went
wrong, but maybe the next step is to try and use the OpenSK
native tools to provision a set of keys onto the device to see
if that fixes it (instead of the current bodge of hard-coding
some manually extracted hex values into the database).
2022-06-05 01:49:11 +08:00
bunnie
272abfd157 cleanup pass on spew and warnings 2022-06-04 00:56:47 +08:00
bunnie
b2b4529c40 quiet down debug spew 2022-06-04 00:35:45 +08:00
bunnie
454ebd0b52 inital USB integration to stack 2022-06-03 01:08:08 +08:00
bunnie
5812cef8f3 add standard features to the config & fixup code paths that broke 2022-06-01 14:15:49 +08:00
bunnie
d995912a97 fix remaining crypto tests 2022-06-01 04:14:57 +08:00
bunnie
3647c38cb0 fix ecdh tests 2022-06-01 03:41:58 +08:00
bunnie
01a10ee709 fix tests so that they are runnable 2022-06-01 03:11:26 +08:00
bunnie
5a118972e4 adapt the PersistentStore object to use the PDDB 2022-06-01 00:19:04 +08:00
bunnie
9d73b4732d stash to look at #171 2022-05-31 14:16:09 +08:00
bunnie
6f84c76b9c fixup AES key references
the implementation uses an explicitly expanded key schedule for
every AES key, with separate enc/dec keys. This does save setting
up the key scheduler every time, but, it also breaks a lot of API
assumptions. Since we have the performance to afford the key
setup/teardown for every call, this commit basically turns
the key refs into actual slices of the key itself. This creates
a weird bit of code where you have an enc key being equated to
a dec key directly; those lines are left in to make it a bit easier
to diff against the original source.

But the high bit is that the original implementation, by expanding
the key scheduler, had to have a different enc and dec key schedule
(even tho they come from the same key) but this compacted form
makes them identical.

This commit also starts working in at converting the PersistentStore
to use the PDDB.

I think I basically understand the scope of the problem, and it's
mostly just some grunt recoding of the PersistentStore calls
to refer to PDDB entries by k/v names. The main thing I've done
so far is to recode the macro that creates the partitions in
OpenSK to just create string names for every item that needs
to be stored. These string names will be used to index into a dictionary
that is dedicated to the FIDO2 function.
2022-05-31 02:24:59 +08:00
bunnie
dd9aa25516 remove unused trait 2022-05-31 02:24:35 +08:00
bunnie
d4244b28af remove custom aes256 implementation 2022-05-31 00:41:13 +08:00
bunnie
a407f7e80d remove now-defunct custom elliptic curve implementation 2022-05-31 00:35:01 +08:00
bunnie
22181422b2 add ecdsa port first cut
this one is very iffy, but once we can run tests it should
chase out all the questions about which APIs go where.
2022-05-31 00:33:11 +08:00
bunnie
58ab2a059e add ecdh calls to "official" p256 crate 2022-05-30 18:52:12 +08:00
bunnie
11e8ce39e8 fix missing import 2022-05-30 16:05:17 +08:00
bunnie
f2ebfd88f4 patch sha2 to Xous implementation 2022-05-30 16:04:54 +08:00
bunnie
723b230a39 that was a long walk for a drink of water: cbc encryption
Woo? we have CBC encryption using Rust's most favored AES library
now.

I have a feeling I'm going to be re-doing all of this once they
decide to finally take GenericArray out back and shoot it....
2022-05-30 15:36:00 +08:00
bunnie
764a8ad195 partial progress on CBC porting.
man this GenericArray shit is so annoying.
2022-05-30 13:46:09 +08:00
bunnie
e84e378ef2 partial progress on aes integration
need to retrench and update AES crate to support some APIs
that allow for CBC modes
2022-05-29 20:35:11 +08:00
bunnie
0b1768115a retrench and try a new strategy for porting crypto libraries 2022-05-29 13:59:06 +08:00
bunnie
a4c72cc9bb fixup merge 2022-05-29 03:35:45 +08:00
bunnie
844543f72d beginning of ecdsa port...need to resolve rand_core dependency
*everoyne* is relying on rand_core, a later version, that uses
a "getrandom" thing that tries to make the call for random
number generators generic.

I mean, it's kind of cool, but, it's also sort of a libc
non-pure-rust nightmare because basically it always delegates
to a C-style API for a variety of platforms.

Next step is to see if we can't...patch...getrandom to point
to a much simplified, xous-only version that is 100% pure Rust.
2022-05-28 23:38:23 +08:00
bunnie
0a795ffaf7 add some shims for the CTAP module
the shims save some rewriting of heavily used Tock APIs to be
Xous-specific.
2022-05-28 18:20:16 +08:00
bunnie
2da0f0b812 some progess porting ctap to Xous 2022-05-28 18:19:52 +08:00
bunnie
9e1e80f089 organize the vault app, trim some things, get a dummy build going 2022-05-28 16:12:08 +08:00
bunnie
389ea55b00 add also the new files associated with the vault 2022-05-28 01:20:05 +08:00
bunnie
81f823ff34 migrate u2f to the "vault" app
The vault app is the beginning of a multi-purpose authentication vault.
It will feature:

1. FIDO2 authentication
2. TOTP authenticators
3. Password storage to PDDB with auto-typing via keyboard HID

This is the first step on a very very long journey.
2022-05-28 01:17:54 +08:00