On 03/02/2015 10:55 AM, Aldy Hernandez wrote:
>
>>> Do you actually need the reload completed? Couldn't this be legitimately
>>> split
>>> before reload (and then parts of it DCE'd as necessary)?
>>>
>>> Otherwise, the split condition does need "&&" as mentioned by Uros.
>>
>> Oh, I forgot -- ev
Do you actually need the reload completed? Couldn't this be legitimately split
before reload (and then parts of it DCE'd as necessary)?
Otherwise, the split condition does need "&&" as mentioned by Uros.
Oh, I forgot -- even if you can exclude reload_completed,
you'd then need to use "&& 1"
On 03/02/2015 09:55 AM, Richard Henderson wrote:
> On 02/26/2015 05:46 PM, Aldy Hernandez wrote:
>> +;; Optimize division or modulo by constant power of 2, if the constant
>> +;; materializes only after expansion.
>> +(define_insn_and_split "*udivmod4_pow2"
>> + [(set (match_operand:SWI48 0 "regis
On 03/02/2015 10:25 AM, Aldy Hernandez wrote:
> On 03/02/2015 09:55 AM, Richard Henderson wrote:
>> On 02/26/2015 05:46 PM, Aldy Hernandez wrote:
>>> +;; Optimize division or modulo by constant power of 2, if the constant
>>> +;; materializes only after expansion.
>>> +(define_insn_and_split "*udiv
On 03/02/2015 09:55 AM, Richard Henderson wrote:
On 02/26/2015 05:46 PM, Aldy Hernandez wrote:
+;; Optimize division or modulo by constant power of 2, if the constant
+;; materializes only after expansion.
+(define_insn_and_split "*udivmod4_pow2"
+ [(set (match_operand:SWI48 0 "register_operand
On 02/26/2015 05:46 PM, Aldy Hernandez wrote:
> +;; Optimize division or modulo by constant power of 2, if the constant
> +;; materializes only after expansion.
> +(define_insn_and_split "*udivmod4_pow2"
> + [(set (match_operand:SWI48 0 "register_operand" "=r")
> + (udiv:SWI48 (match_operand:S
Hello!
> This is actually Jakub's patch from the PR, with a few minor tweaks that were
> needed to bootstrap
> and pass the regression suite. The splitter was using operand 0 without
> setting it first. It should've
> been operand 2. Also, there was a division by zero that was causing an
> inva
This is actually Jakub's patch from the PR, with a few minor tweaks that
were needed to bootstrap and pass the regression suite.
The splitter was using operand 0 without setting it first. It should've
been operand 2. Also, there was a division by zero that was causing an
invalid insn; fixed
