Why the STM32_AES_CBC_Encrypt function doesn't return multiple of 64 bytes encrypted message

[emil28]

Hi,

I am using the function STM32_AES_CBC_Encrypt from the ST library Crypto 3.1.0. I don't understand why this function doesn't give me an encrypted message which is a multiple of 64 bytes.

For example, I would like to encrypt "thisisatest12345". The function returns a tab which is the same size of the original tab:

Thes encrypted message is "0x1E 0xDD 0x1B 0x2A 0x8B 0xA6 0xDC 0x5E 0xBE 0xD6 0x9D 0x78 0xA4 0xA9 0xCA 0x3B ". Following the website https://www.devglan.com/online-tools/aes-encryption-decryption, the encypted message should be "1EDD1B2A8BA6DC5EBED69D78A4A9CA3B813290DF65F7ABD59239929768C8603F", which is 64 bytes. But, when I transform the encrypted message into Base 64, I have the same result as the previsous website, but there is a missing part:

From website;

From my code:

The beginning is the same, but a part is missing, and I don't know why.

Can you help me, I don't know how to solve this problem.

Thanks.

[emil28]

I made some progress concerning this issue. Even the project for the Nucleo F401RE from the ST library doesn't work. I have this warning:

I am working on IAR 8.42.2. Does this problem come from my IAR version or from the ST library ?

[Jocelyn RICARD]

Hello Emil,

I don't understand your issue on first post. AES constraint is 16 bytes alignment only.

Regarding this wchar_t size, you need to get a patch of the crypto library that includes the binaries compiled with IAR8.

Best regards

Jocelyn

[emil28]

Hello Jocelyn,

Thanks for your answer. Yes, AES is 16 bytes alignment. I used the cyrptoLib patch V3.1.3 and I don't have the warning anymore.

But I still have a problem: the result of the function " STM32_AES_CBC_Encrypt" always gives me an array which is the same lenght of the entry array. For example, if I have an array of 20 bytes to encrypt, the encrypted result will be also 20 bytes... The result should a multiple of 16 bytes ?

I am working on the stm32l476ccu6.

 

 Does the buffer to be encrypted need to be a multiple of 16 bytes ?

[emil28]

Hello,

I really need an answer because I don't understand how this function works.

Thanks.

[Pavel A.]

@emil28​ If you encrypt multiples of 16 bytes, do you get correct results ?

-- pa

[emil28]

Hello,

Yes, if I encrypt multiples of 16 bytes, I get the correct results. The function doesn't work if I encrypt an array which is not a multiple of 16 bytes.

[emil28]

Hello,

Another question, is it possible to use PKCS5 padding with this library ? I don't find informations about this in the documentation.

I am a little bit disappointed by the lack of information from ST about this subject...

[Jocelyn RICARD]

Hello Emil,

I'm not sure to catch your point. If you need to encrypt a buffer that is not aligned with 16 bytes, just pad it to make it aligned and encrypt it.

Best regards

Jocelyn

[Jocelyn RICARD]

Hello Emil,

the documentation of the X-CUBE-CRYPTOLIB is UM1924

https://www.st.com/resource/en/user_manual/dm00215061-stm32-crypto-library-stmicroelectronics.pdf

The is no PKCS5 mentioned in this documentation meaning this is not supported.

What kind of information would you expect ?

Best regards

Jocelyn