Commit graph

16243 commits

Author SHA1 Message Date
Gilles Peskine
6e3187b212 RSA: Use hashlen as the hash input size as documented
Where hashlen was previously ignored when the hash length could be
inferred from an md_alg parameter, the two must now match.

Adapt the existing tests accordingly. Adapt the sample programs accordingly.

This commit does not add any negative testing.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-22 18:39:53 +02:00
Gilles Peskine
9dbbc297a3 PK signature function: require exact hash length
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-22 18:39:41 +02:00
Manuel Pégourié-Gonnard
e7885e5441 RSA: Require hashlen to match md_alg when applicable
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
2021-06-22 12:29:27 +02:00
Janos Follath
ab97e003f3 Improve psa_rsa_decode_md_type()
Remove a case that cannot be triggered as PSA_ALG_SIGN_GET_HASH always
returns 0 for raw algorithms.

Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-06-21 10:56:29 +01:00
Janos Follath
3dd3ff0096 Add ChangeLog entry
Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-06-21 10:39:36 +01:00
Janos Follath
4c0b60e5fa Fix exercise key test
Hash and sign algorithms require the alignment of the input length with
the hash length at verification as well not just when signing.

Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-06-21 10:39:36 +01:00
Janos Follath
ae50d28e4c Fix verify_hash tests for PSA RSA PSS
The psa_verify_hash() is the pre-hashed version of the API and supposed
to work on hashes generated by the user. There were tests passing that
were getting "hashes" of sizes different from the expected.

Transform these into properly failing tests.

Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-06-21 10:39:36 +01:00
Janos Follath
0af093b6c8 PSA RSA PSS: pass pre-hash algorithm to Mbed TLS
PSA Crypto always passed MBEDTLS_MD_NONE to Mbed TLS, which worked well
as Mbed TLS does not use this parameter for anything beyond determining
the input lengths.

Some alternative implementations however check the consistency of the
algorithm used for pre-hash and for other uses in verification (verify
operation and mask generation) and fail if they don't match. This makes
all such verifications fail.

Furthermore, the PSA Crypto API mandates that the pre-hash and internal
uses are aligned as well.

Fixes #3990.

Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-06-21 10:39:36 +01:00
Janos Follath
b795332401 mbedtls_rsa_rsassa_pss_*: improve documentation
Hashes used in RSA-PSS encoding (EMSA-PSS-ENCODE, see §9.1.1 in RFC
8017):

- H1: Hashing the message (step 2)
- H2: Hashing in the salt (step 6)
- H3: Mask generation function (step 9)

According to the standard:

- H1 and H2 MUST be done by the same hash function
- H3 is RECOMMENDED to be the same as the hash used for H1 and H2.

According to the implementation:

- H1 happens outside of the function call. It might or might not happen
and the implementation might or might not be aware of the hash used.
- H2 happens inside the function call, consistency with H1 is not
enforced and might not even be possible to detect.
- H3 is done with the same hash as H2 (with the exception of
mbedtls_rsassa_pss_verify_ext(), which takes a dedicated parameter for
the hash used in the MGF).

Issues with the documentation:

- The comments weren't always clear about the three hashes involved and
often only mentioned two of them (which two varied from function to
function).
- The documentation was giving the impression that the standard
recommends aligning H2 and H1 (which is not a recommendation but a
must).

Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-06-21 10:39:04 +01:00
Manuel Pégourié-Gonnard
8cad2e22fc
Merge pull request #4595 from gilles-peskine-arm/alt-dummy-headers-3.0
Lighten and test constraints on context types in alternative implementations
2021-06-15 12:12:46 +02:00
Gilles Peskine
e5702489f5 mbedtls_dhm_get_value parameter order: context first, output last
mbedtls_dhm_get_value can be seen as either a copy function or a getter
function. Given the name and the semantics, it's more of a getter, even if
it "gets" by doing a copy. Therefore, put the context first, and the
selector next, leaving the output for last.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:14:28 +02:00
Gilles Peskine
cadd3d860e Give examples of PLATFORM_XXX_ALT
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:14:28 +02:00
Gilles Peskine
bf26bef157 Discuss the durability of PSA drivers vs ALT when introducing them
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:14:28 +02:00
Gilles Peskine
a71db94c66 Document that contexts must be movable
Fix #4451.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:14:28 +02:00
Gilles Peskine
f35c42bdb9 Document the remaining constraints on ALT context types
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:14:28 +02:00
Gilles Peskine
1628a9c140 MBEDTLS_DEBUG_C is compatible with every whole-module ALT except DHM
It would be possible to make SSL debugging compatible with MBEDTLS_DHM_ALT,
but too much low-priority work right now, so don't require it.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:14:28 +02:00
Gilles Peskine
5c3f18d37c MBEDTLS_PK_PARSE_EC_EXTENDED is incompatible with MBEDTLS_ECP_ALT
... unless the alt implementation defines a group structure that's mostly
compatible with the built-in one and supports partially filled group
structures in the same way.

It would be possible to rewrite the SpecifiedECDomain parsing code to avoid
requiring support for partially filled group structures, but that's too
complicated to do now.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:13:02 +02:00
Gilles Peskine
ad7f53cdb3 Define public fields of mbedtls_ecp_group in alt test header
And don't define configuration macros that only apply to the built-in
implementation.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:12:37 +02:00
Gilles Peskine
bf69ea5b8f Make some fields of mbedtls_ecp_group public
The Mbed TLS code relies heavily on reading certain fields of
mbedtls_ecp_group directly. Make these fields public. Require
that MBEDTLS_ECP_ALT alternative implementations have them.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:12:37 +02:00
Gilles Peskine
cd07e22048 New function mbedtls_ecjpake_set_point_format
Use this instead of accessing the field directly.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:12:37 +02:00
Gilles Peskine
71acc6e8d9 New function mbedtls_dhm_get_value to copy a field of a DHM context
Reduce the need to break the DHM abstraction by accessing the context directly.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:12:37 +02:00
Gilles Peskine
487bbf6805 DHM: new functions to query the length of the modulus
Add two functions mbedtls_dhm_get_len() and mbedtls_dhm_get_bitlen() to
query the length of the modulus in bytes or bits.

Remove the len field: the cost of calling mbedtls_dhm_get_len() each time
it's needed is negligible, and this improves the abstraction of the DHM
module.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:12:37 +02:00
Gilles Peskine
85b1bc65a0 pk_debug: build with RSA_ALT implementations
When MBEDTLS_RSA_ALT is defined, mbedtls_rsa_context does not necessarily
have fields called N and E of type mbedtls_mpi. Don't emit pk_debug
information with MBEDTLS_RSA_ALT. This is not ideal but at least the
library compiles.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:12:37 +02:00
Gilles Peskine
d41d59e900 Timing: support ALT implementations in self-test
Don't assume that the context has a timer field if MBEDTLS_TIMING_ALT
is defined.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:12:37 +02:00
Gilles Peskine
cc73cc55e8 Test the build with whole-module alternative implementations
Use headers defining dummy context types.

The test does not pass yet. I plan to fix this in subsequent commits.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:12:18 +02:00
Gilles Peskine
8e90cf49ca config.py: new command line commands set-all and unset-all
The new method `Config.change_matching` and the new command-line
commands `set-all` and `unset-all` change a batch of existing boolean
settings to the desired state (active or inactive).

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
a0ebaefef9 Create threading_alt.h header for testing
Follow-up to "Create xxx_alt.h headers for testing". The inclusion of
threading_alt.h in include/mbedtls/threading.h does not follow the
same pattern as the others so it was missed by the script.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
c5048db24c Create xxx_alt.h headers for testing
These headers define the context types that alternative implementations must
provide. The context types are dummy types, suitable for building but not
meant to be usable by an implementation.

This is the output of the following script:

```
perl -0777 -ne '
m@^#if !defined\((MBEDTLS_\w+_ALT)\).*\n((?:.*\n)*?)#else.*\n#include "(.*_alt\.h)"\n#endif@m or next;
$symbol = $1; $content = $2; $header = $3;
$header_symbol = $header; $header_symbol =~ y/a-z./A-Z_/;
m@/\*[ *\n]*Copyright .*?\*/@s or die; $copyright = $&;
open OUT, ">tests/include/alt-dummy/$header" or die;
$content =~ s@//.*@@mg;
$content =~ s@/\*.*?\*/@@sg;
$content =~ s@\{.*?\}@{\n    int dummy;\n}@sg;
$content =~ s@ +$@@mg;
$content =~ s@\n{3,}@\n\n@g; $content =~ s@\A\n+@@; $content =~ s@\n*\Z@\n@;
print OUT "/* $header with dummy types for $symbol */\n$copyright\n\n#ifndef $header_symbol\n#define $header_symbol\n\n$content\n\n#endif /* $header */\n" or die;
close OUT or die;
' include/mbedtls/*.h
```

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
a1b44dd808 Changelog entry for MBEDTLS_ECP_MAX_BITS automatic determination
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
814d0129ec Test calculated MBEDTLS_ECP_MAX_xxx constants in unit tests
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
601d34396c Update TLS presets when adding a curve
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
6e1387abed Update MBEDTLS_ECP_MAX_BITS when adding a curve
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
d0b16303bb Determine MBEDTLS_ECP_MAX_BITS automatically
MBEDTLS_ECP_MAX_BITS doesn't make sense as a configuration option: it
must not be smaller than the largest supported curve, and it's useless
to set it to a larger value. So unconditionally set it to the size of
the largest supported curve. Remove it from the build configuration.

Alternative implementations must no longer need define this macro.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
9ef1ea7444 Move the definition of derived values out of the settings section
Alternative implementations must no longer define these macros.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
6a2fb61896 Rename library/ecp_alt.h to ecp_internal_alt.h
library/ecp_alt.h (declaring individual functions of the ECP module that can
be substituted, included when building the library with
MBEDTLS_ECP_INTERNAL_ALT enabled) clashes with ecp_alt.h (not provided,
declaring types of the ECP module when the whole implementation is
substituted, included when building the library with MBEDTLS_ECP_ALT enabled).
Depending on the search path during build, this can make MBEDTLS_ECP_ALT
unusable.

Rename library/ecp_alt.h to follow the naming convention of other alt headers:
MBEDTLS_XXX_ALT corresponds to xxx_alt.h.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
b9ccb25f33 Starter-class documentation of alternative implementations
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-06-15 00:10:37 +02:00
Gilles Peskine
cee21d76f1
Merge pull request #4606 from TRodziewicz/turn__SSL_SRV_RESPECT_CLIENT_PREFERENCE_config_opt_to_runtime_opt
Turn _SSL_SRV_RESPECT_CLIENT_PREFERENCE config opt to a runtime opt
2021-06-14 19:03:28 +02:00
Ronald Cron
823f594db5
Merge pull request #4511 from mstarzyk-mobica/mbedtls_private_with_python
Introduce scripts automating wrapping structs members with MBEDTLS_PRIVATE macro
2021-06-14 17:36:39 +02:00
Ronald Cron
c4c761e35e Merge remote-tracking branch 'mbedtls/development' into mbedtls_private_with_python
Conflicts:
    include/mbedtls/ssl.h
    include/psa/crypto_struct.h

Conflicts fixed by using the code from development branch
and manually re-applying the MBEDTLS_PRIVATE wrapping.
2021-06-14 16:17:32 +02:00
Ronald Cron
a90e0907c5 Remove scripts automating wrapping structs members with MBEDTLS_PRIVATE.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
2021-06-14 16:15:26 +02:00
TRodziewicz
3946f79cab Correction according to code review (function and param. names change
and docs rewording)

Signed-off-by: TRodziewicz <tomasz.rodziewicz@mobica.com>
2021-06-14 13:46:21 +02:00
TRodziewicz
8b223b6509 Addition of the migration guide entry file.
Signed-off-by: TRodziewicz <tomasz.rodziewicz@mobica.com>
2021-06-14 11:56:33 +02:00
TRodziewicz
8476f2f30a Turn _SSL_SRV_RESPECT_CLIENT_PREFERENCE config option to a runtime option
Signed-off-by: TRodziewicz <tomasz.rodziewicz@mobica.com>
2021-06-14 11:56:20 +02:00
Janos Follath
21f84643f8
Merge pull request #4317 from yanesca/psa-jpake-interface
J-PAKE interface definition for PSA Crypto
2021-06-11 13:47:25 +01:00
Ronald Cron
57864faf84
Merge pull request #4634 from Patater/mbed-can-do-timing
config: Allow Mbed to implement TIMING_C
2021-06-11 09:14:13 +02:00
Dave Rodgman
88c88eddf4
Merge pull request #4647 from daverodgman/travis-disable-osx-development
Disable OS X builds on Travis
2021-06-10 17:48:16 +01:00
Gilles Peskine
02b76b7d18
Merge pull request #4619 from TRodziewicz/remove_MBEDTLS_X509_CHECK_x_KEY_USAGE_options
Remove MBEDTLS_X509_CHECK_*_KEY_USAGE options but enable the code
2021-06-10 17:43:36 +02:00
Dave Rodgman
d0581e119b Disable OS X builds on Travis
Signed-off-by: Dave Rodgman <dave.rodgman@arm.com>
2021-06-10 15:47:18 +01:00
Manuel Pégourié-Gonnard
44eea8f067
Merge pull request #4477 from TRodziewicz/Remove__X509_ALLOW_UNSUPPORTED_CRITICAL_EXTENSION
Remove MBEDTLS_X509_ALLOW_UNSUPPORTED_CRITICAL_EXTENSION
2021-06-10 09:13:14 +02:00
TRodziewicz
2a5e5a2759 Correction to the migration guide entry wording
Signed-off-by: TRodziewicz <tomasz.rodziewicz@mobica.com>
2021-06-09 16:54:20 +02:00