sys/random: add option to use HWRNG as source of randomness#14453
Merged
miri64 merged 2 commits intoRIOT-OS:masterfrom Jul 8, 2020
Merged
sys/random: add option to use HWRNG as source of randomness#14453miri64 merged 2 commits intoRIOT-OS:masterfrom
miri64 merged 2 commits intoRIOT-OS:masterfrom
Conversation
benpicco
added
Type: enhancement
benpicco
added
the
CI: ready for build
Member
|
Isn't this what #14324 is already trying? |
miri64
reviewed
Jul 7, 2020
| USEMODULE += hashes | ||
| endif | ||
|
|
||
| ifneq (,$(filter prng_hwrng,$(USEMODULE))) |
Member
There was a problem hiding this comment.
Is it really a PRNG if it is generating random numbers from hardware? ;-)
Contributor
Author
There was a problem hiding this comment.
Yea that feels like a hack - we could rename them all prng_% -> rng_impl_%
Having a single API for randomness in random.h makes sense - the caller doesn't care if that randomness is provided by a PRNG or a HWRNG
Member
There was a problem hiding this comment.
They might, when it comes to timing, or when the deterministic behavior of a PRNG is actually required for some reason.
Contributor
Author
There was a problem hiding this comment.
They might, when it comes to timing
In most cases the HWRNG will be plenty fast. It could be faster (#11924 😉), but hardly noticeable.
or when the deterministic behavior of a PRNG is actually required
Then the application should use a private instance of a PRNG.
The system RNG will be called by 'everyone', thereby severely hampering determinism.
Also mind you, this is just an option. An application developer can chose to select the HWRNG for everything and will get
- cryptographically sound randomness
- reduced code size
if they accept the drawbacks you mentioned.
Member
There was a problem hiding this comment.
if they accept the drawbacks you mentioned.
Perfect call-in for a documentation upgrade ;-).
Contributor
Author
There was a problem hiding this comment.
I would add it to the documentation about the characteristics of the different PRNG implementations, but I couldn't find those either.
Member
There was a problem hiding this comment.
For most PRNG you can look it up in literature just searching for the name. This is not the case for the HWNRG.
Contributor
Author
|
I only skimmed #14324, but it looks like a replacement for when no HWRNG is available. |
Member
|
#14324 focuses entropy collection from alternative sources than a HWRNG, likely used for PRNG seeding. It is very unlikely however, that anyone would use them for more than seed generation due to the resource overhead. I like the idea of this PR as an option to connect the HWRNG directly to IMO we miss separation between entropy sources, seed values, standard PRNGs and crypto PRNGs in RIOT and I am planning on improving the situation step by step. Back to this specific PR, I don't have a concrete opinion by now. |
miri64
reviewed
Jul 8, 2020
| * - Simple Park-Miller PRNG | ||
| * - Musl C PRNG | ||
| * - Fortuna (CS)PRNG | ||
| * - Hardware Random Number Generator (non-seedable) |
Member
There was a problem hiding this comment.
I would add it to the documentation about the characteristics of the different PRNG implementations, but I couldn't find those either.
I would just add it here with all the implications @PeterKietzmann mentioned with regards to true randomness and pseudo-randomness.
Contributor
Author
There was a problem hiding this comment.
Hm, is there more to add than 'read the data sheet of your MCU' 😃 ?
Member
There was a problem hiding this comment.
At least that the HWNRGs vary in how they generate random numbers and that not all are pseudo-number random generators but provide true randomness should also be added as a note.
Member
There was a problem hiding this comment.
A bit off-topic but anyway: One could spend a note here but IMO the HWRNG implementation should as well indicate its hardware random technology.
benpicco
removed
the
CI: ready for build
miri64
requested changes
Jul 8, 2020
miri64
added
Reviewed: 1-fundamentals
miri64
requested changes
Jul 8, 2020
Member
miri64
left a comment
miri64
left a comment
There was a problem hiding this comment.
Getting an error when trying to compile:
$ USEMODULE=prng_hwnrg BOARD=iotlab-m3 IOTLAB_NODES=1 make -C tests/rng/ iotlab-exp
make: Entering directory '/home/mlenders/Repositories/RIOT-OS/RIOT/tests/rng'
Building application "tests_rng" for "iotlab-m3" with MCU "stm32".
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/boards/iotlab-m3
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/boards/common/iotlab
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/core
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/cpu/stm32
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/cpu/cortexm_common
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/cpu/cortexm_common/periph
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/cpu/stm32/bootloader
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/cpu/stm32/periph
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/cpu/stm32/stmclk
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/cpu/stm32/vectors
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/drivers
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/drivers/periph_common
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/auto_init
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/div
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/fmt
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/isrpipe
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/luid
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/newlib_syscalls_default
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/pm_layered
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/random
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/shell
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/stdio_uart
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/test_utils/interactive_sync
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/tsrb
"make" -C /home/mlenders/Repositories/RIOT-OS/RIOT/sys/xtimer
/usr/lib/gcc/arm-none-eabi/10.1.0/../../../../arm-none-eabi/bin/ld: /home/mlenders/Repositories/RIOT-OS/RIOT/tests/rng/bin/iotlab-m3/application_tests_rng/test.o: in function `test_init':
/home/mlenders/Repositories/RIOT-OS/RIOT/tests/rng/test.c:42: undefined reference to `random_init'
/usr/lib/gcc/arm-none-eabi/10.1.0/../../../../arm-none-eabi/bin/ld: /home/mlenders/Repositories/RIOT-OS/RIOT/tests/rng/bin/iotlab-m3/application_tests_rng/test.o: in function `test_get_uint32':
/home/mlenders/Repositories/RIOT-OS/RIOT/tests/rng/test.c:88: undefined reference to `random_uint32'
/usr/lib/gcc/arm-none-eabi/10.1.0/../../../../arm-none-eabi/bin/ld: /home/mlenders/Repositories/RIOT-OS/RIOT/tests/rng/bin/iotlab-m3/random/random.o: in function `auto_init_random':
/home/mlenders/Repositories/RIOT-OS/RIOT/sys/random/random.c:60: undefined reference to `random_init'
/usr/lib/gcc/arm-none-eabi/10.1.0/../../../../arm-none-eabi/bin/ld: /home/mlenders/Repositories/RIOT-OS/RIOT/tests/rng/bin/iotlab-m3/random/random.o: in function `random_uint32_range':
/home/mlenders/Repositories/RIOT-OS/RIOT/sys/random/random.c:99: undefined reference to `random_uint32'
collect2: error: ld returned 1 exit status
make: *** [/home/mlenders/Repositories/RIOT-OS/RIOT/tests/rng/../../Makefile.include:572: /home/mlenders/Repositories/RIOT-OS/RIOT/tests/rng/bin/iotlab-m3/tests_rng.elf] Error 1
make: Leaving directory '/home/mlenders/Repositories/RIOT-OS/RIOT/tests/rng'
Contributor
Author
|
When I add |
Member
I assumed the IoT-LAB has that feature. However, from how the module is structured, the linker should at least find the |
Member
|
|
Member
|
Ok, now I get the more expected result: |
Member
|
Then let's use |
Member
|
Please squash! |
Add the `prng_hwrng` module to enable the HWRNG as source of all randomness, not just for seeding a PRNG. saves ~260 bytes compared to using tinymt32.
benpicco
force-pushed
the
sys/random-hwrng
branch
from
1330b18 to
2f42347
Compare
Contributor
Author
|
I tried on |
benpicco
requested review from
aabadie,
fjmolinas and
leandrolanzieri
as code owners
benpicco
added
the
CI: ready for build
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Contribution description
Add the
prng_hwrngmodule to enable the HWRNG as source of all randomness, not just for seeding a PRNG.saves ~260 bytes compared to using tinymt32.
Testing procedure
add
USEMODULE += prng_hwrngto any application that uses therandommodule.Issues/PRs references