some bug in rhash md5.c about rhash HOT 1 CLOSED

Hello!
the function rhash_md5_process_block() accepts array of unsigned values,
so ctx->message[] should not be specially filled for BigEndian CPUs.
This md5 algorithm was tested on PPC emulator (BigEndian) and works fine.

20.02.2013 10:27, sailzeng wrote:

In function
void rhash_md5_update(md5_ctx /ctx, const unsigned char/ msg, size_t size)
//----------------------------------------------------------------------------------------------------------
if(index) {
unsigned left = md5_block_size - index;
le32_copy((char*)ctx->message, index, msg, (size < left ? size : left));
if(size < left) return;

/* process partial block */
rhash_md5_process_block(ctx->hash, ctx->message);
msg += left;
size -= left;
}

//----------------------------------------------------------------------------------------------------------
this codes is unnecessary.

In function
void rhash_md5_final(md5_ctx /ctx, unsigned char/ result)
//------------------------------------------------------------------------------------------------------
ctx->message[14] = (unsigned)(ctx->length << 3);
ctx->message[15] = (unsigned)(ctx->length >> 29);
//------------------------------------------------------------------------------------------------------

this code must think about bytes endian ?

—
Reply to this email directly or view it on GitHub
#4.

from rhash.