Why isn't movl from memory to memory allowed?

Question

I was wondering if this is allowed in assembly,

 movl (%edx) (%eax) 

I would have guessed that it access the memory in the first operand a

Answer 1

It is not valid. You may not perform memory to memory moves directly on any architecture that I am familiar with except with a limited set of operands. The exception are string move and the like through the SI and DI registers on Intel compatible processors, for instance, though these should be avoided (see below). Most architectures do have something that assists in these limited memory to memory moves.

This makes a great deal of sense if you think about the hardware. There are address lines and data lines. The processor signals which memory address to access on the address lines and the data is then read or written via the data lines. Because of this data must pass through the cache or the processor to get to other memory. In fact, if you have a look at this reference on page 145, you'll see the strong statement that MOVS and its friends must never be used:

Note that while the REP MOVS instruction writes a word to the destination, it reads the next word from the source in the same clock cycle. You can have a cache bank conflict if bit 2-4 are the same in these two addresses on P2 and P3. In other words, you will get a penalty of one clock extra per iteration if ESI+WORDSIZE-EDI is divisible by 32. The easiest way to avoid cache bank conflicts is to align both source and destination by 8. Never use MOVSB or MOVSW in optimized code, not even in 16-bit mode.

On many processors, REP MOVS and REP STOS can perform fast by moving 16 bytes or an entire cache line at a time. This happens only when certain conditions are met. Depending on the processor, the conditions for fast string instructions are, typically, that the count must be high, both source and destination must be aligned, the direction must be forward, the distance between source and destination must be at least the cache line size, and the memory type for both source and destination must be either write-back or write-combining (you can normally assume the latter condition is met).

Under these conditions, the speed is as high as you can obtain with vector register moves or even faster on some processors. While the string instructions can be quite convenient, it must be emphasized that other solutions are faster in many cases. If the above conditions for fast move are not met then there is a lot to gain by using other methods.

This also, in a sense, explains is why register to register moves are ok (though there are other reasons). Perhaps I should say, it explains why they wouldn't require very special hardware on the board... The registers are all in the processor; there's no need to access the bus to read and write via addresses.