So thumb2 can also use the instructions similar to thumb1, right? It
potentially has better performance and smaller code size.
thanks
Carrot
On Tue, Jan 5, 2010 at 7:06 PM, Richard Earnshaw <rearn...@arm.com> wrote:
>
> On Tue, 2010-01-05 at 15:42 +0800, Carrot Wei wrote:
>> Hi
>>
>> In function arm_load_pic_register in file arm.c there are following code:
>>
>> if (TARGET_ARM)
>> {
>> ...
>> }
>> else if (TARGET_THUMB2)
>> {
>> /* Thumb-2 only allows very limited access to the PC. Calculate the
>> address in a temporary register. */
>> if (arm_pic_register != INVALID_REGNUM)
>> {
>> pic_tmp = gen_rtx_REG (SImode,
>> thumb_find_work_register (saved_regs));
>> }
>> else
>> {
>> gcc_assert (can_create_pseudo_p ());
>> pic_tmp = gen_reg_rtx (Pmode);
>> }
>>
>> emit_insn (gen_pic_load_addr_thumb2 (pic_reg, pic_rtx));
>> emit_insn (gen_pic_load_dot_plus_four (pic_tmp, labelno));
>> emit_insn (gen_addsi3 (pic_reg, pic_reg, pic_tmp));
>> }
>> else /* TARGET_THUMB1 */
>> {
>> ...
>> }
>>
>> The comment said "Thumb-2 only allows very limited access to the PC.
>> Calculate the address in a temporary register.". So the generated code
>> is a little more complex than thumb1. Could anybody help to give more
>> explanation on the limitation thumb2 has compared to thumb1?
>>
>> The generated instructions by this function for thumb1 is listed
>> following, both instructions are available under thumb2.
>>
>> ldr r3, .L2
>> .LPIC0:
>> add r3, pc
>
> I didn't write the comment, but I think the limitation is relative to
> ARM not thumb1. Thumb2 code generator is a variation on the ARM code
> generator, so the author was most likely thinking of that rather than
> Thumb1.
>
> R.
>
>
