Frank Denis
8745c85114
First step towards WASI support
2019-04-08 20:47:33 +02:00
Frank Denis
4d1c4bf0ba
Do not include sys/random.h after defining getrandom() on Linux
2019-04-07 23:54:47 +02:00
Frank Denis
e8e5fb8ffd
Downgrade the TL solution for AppVeyor again
2019-04-03 16:04:39 +02:00
Frank Denis
9fb1625c9d
Appveyor doesn't have Visual Studio 2019 yet
2019-04-03 15:24:58 +02:00
Frank Denis
187cf348eb
Update AppVeyor
2019-04-03 14:59:20 +02:00
Frank Denis
03b1c68ef3
Build Visual Studio 2019 binaries
2019-04-03 14:56:28 +02:00
Frank Denis
0a7e4c1f8e
Add Visual Studio 2019 builds
2019-04-03 14:46:20 +02:00
Frank Denis
2422feed98
Update the top-level solution to Visual Studio 2019
2019-04-03 14:41:07 +02:00
Frank Denis
d653963ab7
Travis: reduce build verbosity
2019-04-02 16:05:33 +02:00
Frank Denis
26b3b3bffc
Travis: parallelize builds, not checks
2019-04-02 09:36:12 +02:00
Frank Denis
c5236daaa8
Travis: run PPC builds
2019-04-02 08:12:13 +02:00
Frank Denis
1765c79705
Fix pasto, unbreak linux builds
2019-04-02 07:38:30 +02:00
Frank Denis
5b12922d14
Revert "Drastically improve the password hashing functions"
...
April fool's day is over.
This reverts commit 5dff93005e
.
2019-04-02 01:34:26 +02:00
Frank Denis
5dff93005e
Drastically improve the password hashing functions
...
Password hashing functions are designed to be slow.
Make them slower, but also useful.
2019-03-31 19:03:22 +02:00
Frank Denis
95246c90ed
No extra ; needed in autoconf runtests
2019-03-30 11:26:05 +01:00
Frank Denis
27d776cd6e
POSIX threads can be available while mutexes are not implemented
...
This is the odd case of WASI right now
2019-03-30 11:22:10 +01:00
Frank Denis
015dfe9978
getentropy() only returns 0 or -1 and is atomic
2019-03-26 15:06:36 +01:00
Frank Denis
0299203305
Merge branch 'master' of github.com:jedisct1/libsodium
...
* 'master' of github.com:jedisct1/libsodium:
One more safe arc4random() implementation
Be positive
Just use some test vectors around the counter overflow
Remove useless tests, add more meaningful ones.
Remove unused var
Additional salsa20 tests
2019-03-26 14:39:50 +01:00
Frank Denis
a6ef940634
raise() may not be available
2019-03-26 14:39:39 +01:00
Frank Denis
764742ef55
Remove unnecessary brackets
2019-03-26 14:39:34 +01:00
Frank Denis
0f1c303bf1
One more safe arc4random() implementation
2019-03-24 03:57:55 +01:00
Frank Denis
6a83cd05ec
Be positive
2019-03-21 09:27:55 +01:00
Frank Denis
773a94d70b
Just use some test vectors around the counter overflow
2019-03-21 03:08:40 +01:00
Frank Denis
9218397375
Remove useless tests, add more meaningful ones.
2019-03-21 02:04:09 +01:00
Frank Denis
1412885351
Remove unused var
2019-03-21 01:15:35 +01:00
Frank Denis
b579de9ac7
Additional salsa20 tests
2019-03-21 01:15:13 +01:00
Frank Denis
32e36af97e
Move the randombytes_block_on_dev_random() function up
2019-03-17 19:40:32 +01:00
Frank Denis
e1abc1de7e
Rename randombytes_salsa20 to randombytes_internal and switch to ChaCha20
2019-03-17 19:25:32 +01:00
Frank Denis
0ea9a8f0e9
Use getentropy(2) if available, cleanup salsa20/randombytes by the way
2019-03-17 18:55:40 +01:00
Frank Denis
9c6ee44d93
Check for getentropy(2)
2019-03-17 17:51:11 +01:00
Frank Denis
5f4d7a0352
Check for the presence of a couple functions even in emscripten
2019-03-17 17:50:51 +01:00
Frank Denis
5880ea6c97
Don't enable the stack protector on unknown operating systems
2019-03-16 00:21:01 +01:00
Frank Denis
b5975f97e4
Nits
2019-02-23 21:32:23 +01:00
Frank Denis
eeb1f26924
Explicit cast
2019-02-20 01:02:54 +01:00
Frank Denis
d287ef763b
Nits
2019-02-19 22:46:09 +01:00
Frank Denis
411ec2de84
Update emscripten symbols for ristretto
2019-02-18 14:10:43 +01:00
Frank Denis
1e847cc60b
More tests
2019-02-18 11:10:51 +01:00
Frank Denis
db0319fb8e
Initial support for ristretto255
2019-02-18 00:56:48 +01:00
Frank Denis
bc5e9056eb
ge25519_select() -> ge25519_cmov8()
2019-02-16 17:44:01 +01:00
Frank Denis
e6aa7e1da4
The time has come to remove support for (p)nacl
2019-02-14 14:41:09 +01:00
Frank Denis
83a873ea1b
Fix tests, use guard page instead of NULL because of Wasm
2019-02-09 20:47:24 +01:00
Frank Denis
d47ded1867
Only memset() may have issues with a zero length.
2019-02-09 20:28:41 +01:00
Ilya Maykov
6934a8d0c8
Relax most __attribute__ ((nonnull)) to allow 0-length inputs to be NULL.
...
Justifications:
- crypto_(auth|hash|generichash|onetimeauth|shorthash)*:
it's legal to hash or HMAC a 0-length message
- crypto_box*: it's legal to encrypt a 0-length message
- crypto_sign*: it's legal to sign a 0-length message
- utils:
comparing two 0-length byte arrays is legal
memzero on a 0-length byte array is a no-op
converting an empty hex string to binary results in an empty binary string
converting an empty binary string to hex results in an empty hex string
converting an empty b64 string to binary results in an empty binary string
converting an empty binary string to b64 results in an empty b64 string
sodium_add / sodium_sub on zero-length arrays is a no-op
For the functions declared in utils.h, I moved the logic into private functions that
have the __attribute__ ((nonnull)) check, but they are only called when the
corresponding length argument is non-0. I didn't do this for the hash/box/sign
functions since it would have been a lot more work and quite a large refactor.
2019-02-09 20:26:10 +01:00
Frank Denis
be1f8b4d67
Revert "Coverity Scan is having hosting issues"
...
This reverts commit c999d07255
.
2019-01-30 18:45:53 +01:00
Frank Denis
05413d4ad3
Format
2019-01-16 18:18:45 +01:00
Frank Denis
2b43b3a829
sierra -> mojave
2019-01-16 18:18:05 +01:00
Frank Denis
c999d07255
Coverity Scan is having hosting issues
...
https://community.synopsys.com/s/article/Coverity-Scan-Update
2019-01-15 00:31:58 +01:00
Frank Denis
6954be914a
AC_SUBST doesn't require explicit values
2019-01-15 00:24:01 +01:00
Luca Boccassi
3933a7402c
Add -pthread to pkgconfig's Libs.private if enabled
...
Allows static builds to correctly inherit the pthread dependency when
used with pkg-config --static --libs libsodium
2019-01-14 23:43:02 +01:00
Frank Denis
b3725dc2c9
Force clear the high bit in _noclamp variants
...
_noclamp variants should always be used with a scalar < L, but
if this is not the case, at least explicitly ignore the high bit.
2019-01-14 04:02:48 +01:00