Error with Simple example running on Arduino Nano

[datedave]

Using an Arduino Nano, with the "simple" example provided, the decrkypt fails (output below) - any ideas?

because the decrypt looks almost there, I wonder if there is some buffer over write somewhere?

readBuffer length: 19 Calling encrypt (string)... Encrypted length = 44 Encrypted. Decrypting... 44 Calling decrypt...; Decrypted bytes: 32 Decrypted cleartext of length: 32 Decrypted cleartext: username:passwor⸮s⸮lڶ--⸮⸮n=⸮ Decryption test failed.

[suculent]

This seems more like the aes_iv changes on encryption and forgot being reset to same initial state on decryption...

On 17 Sep 2020, at 15:54, datedave notifications@github.com wrote:

Using an Arduino Nano, with the "simple" example provided, the decrkypt fails (output below) - any ideas?

because the decrypt looks almost there, I wonder if there is some buffer over write somewhere?

readBuffer length: 19 Calling encrypt (string)... Encrypted length = 44 Encrypted. Decrypting... 44 Calling decrypt...; Decrypted bytes: 32 Decrypted cleartext of length: 32 Decrypted cleartext: username:passwor⸮s⸮lڶ--�⸮�⸮n=⸮ Decryption test failed.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/suculent/thinx-aes-lib/issues/34, or unsubscribe https://github.com/notifications/unsubscribe-auth/AABWFRY3Z5AQ6UXELGFYKNTSGIIK7ANCNFSM4RQPI54A.

[datedave]

Ok, so I think I found the error with you example... but you will need to check as I am no expert!

The call to aesLib.encrypt() writes the cypher data back to the same buffer that ir read the clear data from. This causes a problem if the cypher data is longer than the clear data.

So looking at the error above, readbuffer was length 19, but Encrypted was length 44, hence a buffer over write.

Not being familiar with the maths behind AES encryption, but assuming there is always a risk that the encrypted data could be longer than the clear data, I suggest its better to return a new newly allocated buffer with the encrypted data, rather than ovwe write??

hope that helps

[datedave]

to check the above I increased the size of the readBuffer to longer that the input string..

i.e. unsigned char readBuffer[200] = "username:password";

and got the output readBuffer length: 200 Calling encrypt (string)... Encrypted length = 280 Encrypted. Decrypting... 280 Calling degth = 280 Encrypted. Decrypting... 280 lling decDecrypted bytes: 208 Decrypted cleartext of length: 208 Decrypted cleartext: username:password Decrypted correctly.

Although that could have been a fluke.

[datedave]

Again though it is still returning a length bigger than the input - so assuming you are re-allocating buffer internal to lib?

[datedave]

Digging a bit deeper.. I think my fix above was a fluke.. I can see now that you return the txt into a differnt global buffer

[suculent]

Well, I'm trying to solve it but probably will have to back to it later with better focus.

I'm just refactoring some another crazy project...

On 17 Sep 2020, at 16:17, datedave notifications@github.com wrote:

Digging a bit deeper.. I think my fix above was a fluke.. I can see now that you return the txt into a differnt global buffer

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/suculent/thinx-aes-lib/issues/34#issuecomment-694267319, or unsubscribe https://github.com/notifications/unsubscribe-auth/AABWFRY7EFAZ4CGU35XQC5TSGIK77ANCNFSM4RQPI54A.

[datedave]

Thanks - much appreciated. I'll have a look around myself, but this is probably way out of my league.

[suculent]

I don't think so. This is actually a beginner's termination error in C-string. I'm sometimes repeating those even after 20 years...

On 17 Sep 2020, at 16:25, datedave notifications@github.com wrote:

Thanks - much appreciated. I'll have a look around myself, but this is probably way out of my league.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/suculent/thinx-aes-lib/issues/34#issuecomment-694272106, or unsubscribe https://github.com/notifications/unsubscribe-auth/AABWFR6LF6G2JH5SSYNIAGDSGIL35ANCNFSM4RQPI54A.

[datedave]

If I understand you, you are saying that somewhere you are not adding the null '\0' to terminate the C str correctly - right?

Or are you not allowing enough space at the end of a buffer for a '\0' somewhere?

[irineu]

getting the same problem here

[suculent]

There has been issue with missing aesLib.set_paddingmode(0) in examples.