Commit graph

11673 commits

Author SHA1 Message Date
Gilles Peskine e39b903de5 entropy_nv_seed: clean up properly
Call mbedtls_entropy_free on test failure.

Restore the previous NV seed functions which the call to
mbedtls_platform_set_nv_seed() changed. This didn't break anything,
but only because the NV seed functions used for these tests happened
to work for the tests that got executed later in the .data file.
2019-07-19 17:08:48 +02:00
Gilles Peskine 9bb1f64706 Don't call memset after calloc
memset has undefined behavior when either pointer can be NULL, which
is the case when it's the result of malloc/calloc with a size of 0.
The memset calls here are useless anyway since they come immediately
after calloc.
2019-07-19 17:08:48 +02:00
Gilles Peskine 614faa26ac Test PSA functions against PSA_SUCCESS, not 0
Writing 0 instead of PSA_SUCCESS is correct, but bad form.
2019-07-19 17:08:35 +02:00
Gilles Peskine 69c0ea26c7 Test suites: cope with psa_crypto_init failure
psa_crypto_init() can fail. Do check its return code. Don't call it
before initializing local objects that are going to be cleaned up.
2019-07-19 17:08:35 +02:00
Hanno Becker 80bb77e16d ECP restart: Don't calculate address of sub ctx if ctx is NULL
All modules using restartable ECC operations support passing `NULL`
as the restart context as a means to not use the feature.

The restart contexts for ECDSA and ECP are nested, and when calling
restartable ECP operations from restartable ECDSA operations, the
address of the ECP restart context to use is calculated by adding
the to the address of the ECDSA restart context the offset the of
the ECP restart context.

If the ECP restart context happens to not reside at offset `0`, this
leads to a non-`NULL` pointer being passed to restartable ECP
operations from restartable ECDSA-operations; those ECP operations
will hence assume that the pointer points to a valid ECP restart
address and likely run into a segmentation fault when trying to
dereference the non-NULL but close-to-NULL address.

The problem doesn't arise currently because luckily the ECP restart
context has offset 0 within the ECDSA restart context, but we should
not rely on it.

This commit fixes the passage from restartable ECDSA to restartable ECP
operations by propagating NULL as the restart context pointer.

Apart from being fragile, the previous version could also lead to
NULL pointer dereference failures in ASanDbg builds which dereferenced
the ECDSA restart context even though it's not needed to calculate the
address of the offset'ed ECP restart context.
2019-07-19 14:44:36 +01:00
Hanno Becker bf84d503b3 Adapt ChangeLog 2019-07-19 13:03:10 +01:00
Hanno Becker 59c92ed89b ECP restart: Don't calculate address of sub ctx if ctx is NULL
All modules using restartable ECC operations support passing `NULL`
as the restart context as a means to not use the feature.

The restart contexts for ECDSA and ECP are nested, and when calling
restartable ECP operations from restartable ECDSA operations, the
address of the ECP restart context to use is calculated by adding
the to the address of the ECDSA restart context the offset the of
the ECP restart context.

If the ECP restart context happens to not reside at offset `0`, this
leads to a non-`NULL` pointer being passed to restartable ECP
operations from restartable ECDSA-operations; those ECP operations
will hence assume that the pointer points to a valid ECP restart
address and likely run into a segmentation fault when trying to
dereference the non-NULL but close-to-NULL address.

The problem doesn't arise currently because luckily the ECP restart
context has offset 0 within the ECDSA restart context, but we should
not rely on it.

This commit fixes the passage from restartable ECDSA to restartable ECP
operations by propagating NULL as the restart context pointer.

Apart from being fragile, the previous version could also lead to
NULL pointer dereference failures in ASanDbg builds which dereferenced
the ECDSA restart context even though it's not needed to calculate the
address of the offset'ed ECP restart context.

dummy
2019-07-19 13:03:10 +01:00
Gilles Peskine 3b3b34f608 Replace some macros by functions
Replace some frequently-used macros by inline functions: instead of
calling MOD_{ADD,SUB,MUL} after the mbedtls_mpi_{add,sub,mul}_mpi,
call a function mbedtls_mpi_xxx_mod that does the same.

In the baremetal config, with "gcc -Os -mthumb -mcpu=cortex-m0plus",
ecp.o goes down from 13878 bytes to 12234.

No noticeable performance change for benchmarks on x86_64 with either
"gcc -O2" or "gcc -Os".
2019-07-18 21:08:27 +02:00
Manuel Pégourié-Gonnard 49d65ba929 Re-roll main loop with SHA512_SMALLER
Saves 1924 bytes (same measurement as before).
2019-07-17 13:16:54 +02:00
Manuel Pégourié-Gonnard 0270ed99bb Use tables and roll up some loops
Saves 108 bytes (measured as in previous commit).
2019-07-17 13:08:02 +02:00
Manuel Pégourié-Gonnard 7f0719598f Make SHA512_SMALLER turn a macro into a function
Saves 356 bytes on sha512.o compiling for Cortex-M0+ with ARM-GCC

Size measured with:
arm-none-eabi-gcc -Wall -Wextra -Iinclude -Os -mcpu=cortex-m0plus -mthumb -c library/sha512.c
arm-none-eabi-size sha512.o

GCC version:
arm-none-eabi-gcc (GNU Tools for Arm Embedded Processors 7-2018-q2-update) 7.3.1 20180622 (release) [ARM/embedded-7-branch revision 261907]
2019-07-17 13:06:55 +02:00
Manuel Pégourié-Gonnard 2306d15344 Declare new config.h option MBEDTLS_SHA512_SMALLER 2019-07-17 13:05:41 +02:00
Jaeden Amero f7dca86522 Allow building with the -Wunused flag
Make some functions non-static, to avoid Wunused function warnings. Make
a function scoped variable block scoped instead, to avoid Wunused
variable warnings in some configurations.
2019-07-15 15:52:25 +01:00
Jaeden Amero 3a0f08d911 Revert "cpp_dummy_build: Remove dependency on compat-1.3.h"
There is now a test that ensures all headers are included in the
cpp_dummy_build test, so we can't remove compat-1.3.h from the
cpp_dummy_build test until we remove compat-1.3.h.

This reverts commit 2b725ef727.
2019-07-15 15:52:25 +01:00
Jaeden Amero b8e4ae18cf Remove certs.h
certs.h is not needed in Mbed Crypto. No programs or other library code
use it.
2019-07-15 15:52:25 +01:00
Jaeden Amero 8045cfbaa8 Enable ALT implementations of ripemd160
In configurations wanting an alternative ripemd160 implementation, We
were including the ordinary Mbed Crypto ripemd160.h instead of the
user-provided ripemd160_alt.h. Use the user-provided header instead.
2019-07-15 15:52:25 +01:00
Jaeden Amero c49fbbf3eb Use mbedtls-based path for includes
To help the build system find the correct include files, paths starting
with "mbedtls/" or "psa/" must be used. Otherwise, you can run into
build failures like the following when building Mbed Crypto as a
submodule.

    In file included from chachapoly.c:31:0:
    ../../include/mbedtls/chachapoly.h:43:10: fatal error: poly1305.h: No such file or directory
     #include "poly1305.h"
              ^~~~~~~~~~~~
    compilation terminated.

Includes for ALT implementations are not modified, as the alt headers
are provided by system integrators and not Mbed TLS or Mbed Crypto.
2019-07-15 15:52:25 +01:00
Gilles Peskine f61bf9c13b
Merge pull request #170 from gilles-peskine-arm/at-misra-2012
Use stdint.h types
2019-07-15 14:42:15 +02:00
Gilles Peskine c11c4dcf95 Favor stdint.h types in internal types
Use uint8_t for PSA buffers. Keep unsigned char for generic libc
buffers and for mbedtls buffers.
2019-07-15 11:17:53 +02:00
Gilles Peskine 7228da25f9 Favor stdint.h types in implementation-specific API 2019-07-15 11:16:18 +02:00
Gilles Peskine f82088a5f4 Favor stdint.h types in example code 2019-07-15 11:16:18 +02:00
Andrew Thoelke d16bdac9b5 Use stdint.h types in multipart AEAD functions
No implementation yet.
2019-07-15 11:14:56 +02:00
Andrew Thoelke 163639b830 Apply same changes to implementation source code 2019-07-15 11:14:56 +02:00
Andrew Thoelke 47629d076e Use stdint.h types
Follow MISRA C 2012 rules by using exact width types from stdint.h.
2019-07-15 11:14:56 +02:00
Ron Eldor 991a05b411 Add support for all SHA modes in cert_write
Add support for `MBEDTLS_SHA_224` and `MBEDTLS_SHA_384` in
`cert_write`, to support generating such certificates in
`tests/data_files/Makefile`.
2019-07-14 09:17:57 +03:00
Gilles Peskine 5dc742c36a SE keys: smoke test import, export, destroy 2019-07-12 23:47:47 +02:00
Gilles Peskine 5d309672af SE keys: support import and export 2019-07-12 23:47:28 +02:00
Gilles Peskine 354f7671f4 SE keys: support destroy
When destroying a key in a secure element, call the driver's destroy
method and update the driver's persistent data in storage.
2019-07-12 23:46:38 +02:00
Gilles Peskine cbaff467ef SE keys: allocate a slot before creating the key 2019-07-12 23:46:04 +02:00
Gilles Peskine 73167e128f SE keys: store the slot number in the memory slot 2019-07-12 23:44:37 +02:00
Gilles Peskine 8abe6a2d5c Driver table entries are now mutable
Since driver table entries contain the driver context, which is
mutable, they can't be const anymore.
2019-07-12 23:42:20 +02:00
Gilles Peskine 5243a202c3 Driver context manipulation functions
Create the driver context when registering the driver.

Implement some helper functions to access driver information.
2019-07-12 23:42:20 +02:00
Gilles Peskine 94cc42c28f Pass a writable pointer to the persistent data when needed
Most driver methods are not allowed to modify the persistent data, so
the driver context structure contains a const pointer to it. Pass a
non-const pointer to the persstent data to the driver methods that
need it: init, allocate, destroy.
2019-07-12 23:34:20 +02:00
Gilles Peskine f2223c868d New driver method: allocate
Add a driver method to allocate a key slot for a key that is about to
be created.
2019-07-12 23:33:02 +02:00
Gilles Peskine 8597bc13e7 Pass the driver context to most driver methods
Pass the driver context to all driver methods except the ones that
operate on an already-setup operation context.

Rename `p_context` arguments to `op_context` to avoid confusion
between contexts.
2019-07-12 23:32:27 +02:00
Gilles Peskine 7a86da1d42 Define a driver context structure type
Define a structure that is to be instantiated once per driver
instance.

Define a driver initialization method and pass it the driver context.
2019-07-12 23:25:59 +02:00
Gilles Peskine f03143a4d1 Change driver key slot numbers to 64 bits
This slightly increases storage requirements, but works in more use
cases. In particular, it allows drivers to treat choose slot numbers
with a monotonic counter that is incremented each time a key is
created, without worrying about overflow in practice.
2019-07-12 23:18:29 +02:00
k-stachowiak 2cd9051d5f Add decoding empty buffer test calls for cipher modes that benefit from this 2019-07-12 14:54:04 +02:00
Gilles Peskine 011e4284a1 Look up the SE driver when creating a key
When creating a key with a lifetime that places it in a secure
element, retrieve the appropriate driver table entry.

This commit doesn't yet achieve behavior: so far the code only
retrieves the driver, it doesn't call the driver.
2019-07-12 11:47:50 +02:00
Gilles Peskine 6e59c42d1d Split the secure element driver method table memory layout
Instead of having one giant table containing all possible methods,
represent a driver's method table as a structure containing pointers
to substructures. This way a driver that doesn't implement a certain
class of operations can use NULL for this class as a whole instead of
storing NULL for each method.
2019-07-12 11:47:50 +02:00
Gilles Peskine f989dbe6d8 SE driver lookup functions
Expose the type of an entry in the SE driver table as an opaque type
to other library modules. Soon, driver table entries will have state,
and callers will need to be able to access this state through
functions using this opaque type.

Provide functions to look up a driver by its lifetime and to retrieve
the method table from an entry.
2019-07-12 11:47:50 +02:00
Gilles Peskine e62b74e68f Add public-key export method 2019-07-12 11:47:50 +02:00
Gilles Peskine 1168ef947b
Merge pull request #167 from adrianlshaw/document_algorithms2
Documented some algorithms
2019-07-12 11:03:40 +02:00
Jaeden Amero ff645d9838 Merge remote-tracking branch 'origin/pr/2727' into development
* origin/pr/2727:
  tests: Limit each log to 10 GiB
2019-07-11 16:19:02 +01:00
Jaeden Amero 8306508250 Merge remote-tracking branch 'origin/pr/2660' into development
* origin/pr/2660:
  Fix parsing issue when int parameter is in base 16
  Refactor receive_uint32()
  Refactor get_byte function
  Make the script portable to both pythons
  Update the test encoding to support python3
  update the test script
2019-07-11 16:17:38 +01:00
Jaeden Amero 072959f5c2 Merge remote-tracking branch 'origin/pr/1622' into development
* origin/pr/1622: (29 commits)
  Do not build fuzz on windows
  No booleans and import config
  Removing space before opening parenthesis
  Style corrections
  Syntax fix
  Fixes warnings from MSVC
  Add a linker flag to enable gcov in basic-build-test.sh
  checks MBEDTLS_PEM_PARSE_C
  Restore programs/fuzz/Makefile after in-tree cmake
  Move fuzz directory to programs
  Documentation for corpus generation
  Restore tests/fuzz/Makefile after in-tree cmake
  Adding ifdefs to avoid warnings for unused globals
  Adds LDFLAGS fsanitize=address
  Ignore compiled object files and executables
  Also clean the fuzz subdirectory
  copyediting README.md
  Protecting client/server fuzz targts with ifdefs
  Makefile support 1
  Fuzz README and direct compilation
  ...
2019-07-11 16:17:18 +01:00
Adrian L. Shaw 2282cfa660 Remove GMAC algorithm (for now)
It can't be implemented with the current version of the API
2019-07-11 15:51:45 +01:00
Adrian L. Shaw fd2aed4d76 Document cipher modes 2019-07-11 15:47:40 +01:00
Philippe Antoine a864db0dd6 Do not build fuzz on windows 2019-07-10 20:37:57 +02:00
Gilles Peskine 2c8f909782 Correct version number for 1.1.0 format (formerly 1.0.1)
Update the 1.1.0 format description now that its version number has
been decided. This release turned out to be 1.1.0, not 1.0.1.
2019-07-10 17:19:46 +02:00