Oh. Ok Thanks Richard so much.
I will merge V6 after I finished regression.
Previously, I didn't understand whether you want V7 (I tried use google
translator to translate your words :)
Now I understand you are happy with V6.
Thanks.
juzhe.zh...@rivai.ai
From: Richard Sandiford
Date: 2023-06-23 16:03
To: Bernhard Reutner-Fischer
CC: juzhe.zhong; gcc-patches; rguenther
Subject: Re: [PATCH V6] VECT: Apply LEN_MASK_{LOAD,STORE} into vectorizer
Bernhard Reutner-Fischer <rep.dot....@gmail.com> writes:
> On 23 June 2023 01:51:12 CEST, juzhe.zh...@rivai.ai wrote:
>>From: Ju-Zhe Zhong <juzhe.zh...@rivai.ai>
>
> I am sorry but I somehow overlooked a trivial spot in V5.
> Nit which does not warrant an immediate next version, but please consider it
> before pushing iff approved:
>
>>+ if (final_len)
>>+ {
>>+ signed char biasval
>>+ = LOOP_VINFO_PARTIAL_LOAD_STORE_BIAS (loop_vinfo);
>>+
>>+ bias = build_int_cst (intQI_type_node, biasval);
>>+ }
>>+
>>+ /* Arguments are ready. Create the new vector stmt. */
>>+ if (final_len)
>>+ {
>
> Fuse the block below into the one above as the condition seems to be
> identical?
Yeah, true, but I think the idea is that the code above “Arguments are
ready” is calculating argument values, and the code after it is creating
code. These are two separate steps, and the fact that the two final_len
blocks end up being consecutive is something of a coincidence.
So personally I think we should keep the structure in the patch.
Thanks,
Richard
