If we have multiple configs, we need to flash DUT with different binaries. But if we don't close DUT before apply new config, the old DUT will be reused, so new config name will not be applied.
It gave us a better performance of RSA operations. (2~11 times)
The old modexp implementation (Z = X ^ Y mod M) loaded all the data into
the hw registers and was waiting for completion, but due to
the hardware RSA implementation, the calculations always started with 4096 bit,
which took a lot of time.
Measurement results (measurements were made for keys: 2048, 3072 and 4096 bits)
(Old) - Sliding-window exponentiation (HAC 14.85):
keysize = 2048 bits
RSA key operation (performance): public [93206 us], private [280189 us]
keysize = 3072 bits
RSA key operation (performance): public [293614 us], private [858157 us]
keysize = 4096 bits
RSA key operation (performance): public [653192 us], private [1912126 us]
Instead (Old) - Sliding-window exponentiation (HAC 14.85) was implemented
(New) - Montgomery exponentiation (HAC 14.94) which showed
better performance on private and public keys.
keysize = 2048 bits
RSA key operation (performance): public [14504 us], private [149456 us]
keysize = 3072 bits
RSA key operation (performance): public [35073 us], private [392743 us]
keysize = 4096 bits
RSA key operation (performance): public [58650 us], private [787186 us]
For this reason, the old implementation was removed
and the MBEDTLS_HARDWARE_MPI option was turned on by default.
Why the MPI_INTERRUPT option is removed:
the old implementation used calculations on the hardware and
it took a lot of time (10ms - 500ms). And in order not to stand idle
while waiting for completion, an interrupt option was added.
This made it possible to carry out other tasks during the calculation,
and this one to block. The new method is free from such a drawback and
the maximum duration of one RSA HW operation does not exceed 70us (usually 2-70 μs).
This option is no longer needed.
Closes: IDF-965
The workaround for PSRAM that will occupy an SPI bus is enabled only when:
1. used on 32MBit ver 0 PSRAM.
2. work at 80MHz.
The test used to only check 32MBit by the config option, but for PSRAM
on Wrover-B module seems to use a newer version of 32MBit PSRAM. So it
expects the workaround to be enabled, but actually not.
This commit split the unit test into two parts:
1. check all SPI buses are available, for all configs except psram_hspi
and psram_vspi, run on regular runners (including Wrover and Wrover-B).
a hidden option is enabled so that the compiler knows it's not building
psram_hspi or psram_vspi.
2. check the specified bus are acquired, for config psram_hspi and
psram_vspi. This only run on special runner (legacy Wrover module).
DEFAULT_CPU_FREQ and ULP_COPROC_ENABLED options have chip-specific
names, and should be set in a chip-specific sdkconfig.defaults file.
This commit also changes the default CPU frequency for ESP32S2 unit
tests to 240 MHz.
Bug was this sequence:
1. old sdkconfig file has some settings (maybe target=esp32s2beta)
2. idf.py builds a new sdkconfig.defaults file with full new settings
3. new settings includes something that conflicts with the old settings (for example,
CONFIG_IDF_TARGET=esp32 and BT_ENABLE=y)
4. confgen tries to apply the new "defaults" to the existing sdkconfig, settings end up a mix of both due to the conflicts
Fix is to generate the sdkconfig file directly.
Since pyyaml 5.1 yaml.load without specifing loader is deprecated
Details: https://msg.pyyaml.org/load
To keep code compatible with older versions of pyyaml
and keep best perfomance CLoader with fallback to Loader is used.
The build system automatically determines offsets of partitions from
the partition table, so no manual changes are needed. Instead, add a
note that partition offsets may need to be updated when increasing
the bootloader size.
We could split cases of same config into multiple binaries as we have limited rom space. So we should regard those configs like `default` and `default_2` as the same config.
