On 03/29/2011 07:55 PM, ext Nicolas Pitre wrote:

Sorry, I meant r15-indexed with a write back.

OK, I guess it was kind of implicit in your message but I missed it.

quoted

Now, I admit that it's possible that somewhere beyond the horizon of my
understanding there is some good reason to do two LDRs into adjacent registers
from adjacent memory addresses, instead of merging them into one LDRD.

In this case I suspect that the loaded values were pushed to the literal
pool, and it is hard for the compiler to ensure the placement is always
64-bit aligned.

I guess you are right, I missed that LDRD/STRD needs to be double word 
aligned for the older alignment models. In the ARMv7 model, word 
alignment is enough.

quoted

BTW, in my kernel, LDR indexed by r15 is a really common instruction at the
very beginning of functions. I am not sure why; it could have something to do
with the fact that the kernel is compiled without frame pointers.

No, it's all about literal pool usage.

Yes, of course it's the literals. I was silly to think otherwise :)

But nowhere will you find pc-indexed addressing with a writeback.
That's one of the cases I think should be rejected upfront instead of
evaluating this possibility which is likely to never happen in practice
each time the instruction is emulated.

Currently, we are not checking for that case at all, so the only missing 
part would be to modify the decoding logic. I just sent a patch for that.

best regards,

Viktor