>From: [email protected] On Behalf Of Taraniteja Vishwanatha
>Sent: Thursday, 25 April, 2013 16:43
>I was using the low level aes APIs and now have switched to EVP ones.
>My string encryption and decryption always work fine. But when it comes
>to files, I am getting malloc errors: malloc: *** error for object
>: incorrect checksum for freed object - object was probably modified
>after being freed.
Actually your encrypt is wrong; you just didn't notice the error.
>unsigned char* encryptBlockAES(unsigned char *plainText, int dataLength,
>int *outLength,const unsigned char* keyData, int pageNo)
>{
> unsigned char key[AES_BLOCK_SIZE*2], iv[AES_BLOCK_SIZE*2];
The fact that an AES-256 key is twice the size of an AES data block (128)
-- and a -128 or -192 key is the same or half more -- is just coincidence.
Use 256/8, or create your own name (#define or enum).
OTOH the IV for (any) CBC is ONE data block. Allocating and setting
a second block is just wasted -- and if you use values that differ
only in the second block they won't actually protect against some
attacks as intended.
> enLength = dataLength + (AES_BLOCK_SIZE);
> encryptedString = (unsigned char*)calloc(enLength, sizeof(unsigned
char));
Nit: you don't need this buffer zeroed; malloc() would be fine.
> EVP_CIPHER_CTX_init(&enCtx);
> EVP_EncryptInit_ex(&enCtx, EVP_aes_256_cbc(), NULL, key, iv);
> EVP_EncryptUpdate(&enCtx, encryptedString, &outLen, plainText,
enLength);
The length of plainText is dataLength not enLength. Using enLength creates
a ciphertext that is longer than the buffer you allocated, usually
corrupting your "heap" (malloc arena) with symptoms like the above,
and is not the correct ciphertext for your data anyway.
> EVP_EncryptFinal_ex(&enCtx, encryptedString + outLen, &tempLen);
> *outLength = outLen + tempLen;
> EVP_CIPHER_CTX_cleanup(&enCtx);
> return encryptedString;
______________________________________________________________________
OpenSSL Project http://www.openssl.org
User Support Mailing List [email protected]
Automated List Manager [email protected]