qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Richard Henderson <richard.henderson@linaro.org>
To: Laurent Desnogues <laurent.desnogues@gmail.com>
Cc: rajav@quicinc.com, qemu-arm <qemu-arm@nongnu.org>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
	apazos@quicinc.com
Subject: Re: [PATCH 21/31] target/arm: Implement SVE2 integer absolute difference and accumulate long
Date: Mon, 13 Apr 2020 16:19:39 -0700	[thread overview]
Message-ID: <ae2cb680-2200-e9cb-afa1-e55593ef6d6d@linaro.org> (raw)
In-Reply-To: <CABoDooNE=R_jQjg-dZrdaK-atmuQ4XjMkGjKnmebKvuXcTpDQg@mail.gmail.com>

On 4/13/20 9:15 AM, Laurent Desnogues wrote:
> On Fri, Mar 27, 2020 at 12:18 AM Richard Henderson
> <richard.henderson@linaro.org> wrote:
> [...]
>> diff --git a/target/arm/sve_helper.c b/target/arm/sve_helper.c
>> index a3653007ac..a0995d95c7 100644
>> --- a/target/arm/sve_helper.c
>> +++ b/target/arm/sve_helper.c
>> @@ -1216,6 +1216,30 @@ DO_ZZZ_NTB(sve2_eoril_d, uint64_t,     , DO_EOR)
>>
>>  #undef DO_ZZZ_NTB
>>
>> +#define DO_ABAL(NAME, TYPE, TYPEN) \
>> +void HELPER(NAME)(void *vd, void *va, void *vn, void *vm, uint32_t desc) \
>> +{                                                              \
>> +    intptr_t i, opr_sz = simd_oprsz(desc);                     \
>> +    int sel1 = (simd_data(desc) & 1) * sizeof(TYPE);           \
>> +    int sel2 = (simd_data(desc) & 2) * (sizeof(TYPE) / 2);     \
>> +    for (i = 0; i < opr_sz; i += sizeof(TYPE)) {               \
>> +        TYPE nn = (TYPEN)(*(TYPE *)(vn + i) >> sel1);          \
>> +        TYPE mm = (TYPEN)(*(TYPE *)(vm + i) >> sel2);          \
>> +        TYPE aa = *(TYPE *)(va + i);                           \
>> +        *(TYPE *)(vd + i) = DO_ABD(nn, mm) + aa;               \
>> +    }                                                          \
>> +}
> 
> ABAL is either top or bottom not a mix of two.  So only sel1 is needed
> and its multiplicand should be the number of bits of TYPEN.

Yep.

> vd is both a source and a destination so a temporary should be used.

In what way am I not?  Both sources are read before the write.  The operands
are all in columns of the wide type (unlike the addp case you pointed out).


r~


  reply	other threads:[~2020-04-13 23:20 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-03-26 23:08 [PATCH for-5.1 00/31] target/arm: SVE2, part 1 Richard Henderson
2020-03-26 23:08 ` [PATCH 01/31] target/arm: Add ID_AA64ZFR0 fields and isar_feature_aa64_sve2 Richard Henderson
2020-04-11 18:45   ` Alex Bennée
2020-03-26 23:08 ` [PATCH 02/31] target/arm: Implement SVE2 Integer Multiply - Unpredicated Richard Henderson
2020-03-26 23:08 ` [PATCH 03/31] target/arm: Implement SVE2 integer pairwise add and accumulate long Richard Henderson
2020-03-26 23:08 ` [PATCH 04/31] target/arm: Remove fp_status from helper_{recpe, rsqrte}_u32 Richard Henderson
2020-03-26 23:08 ` [PATCH 05/31] target/arm: Implement SVE2 integer unary operations (predicated) Richard Henderson
2020-03-26 23:08 ` [PATCH 06/31] target/arm: Split out saturating/rounding shifts from neon Richard Henderson
2020-03-26 23:08 ` [PATCH 07/31] target/arm: Implement SVE2 saturating/rounding bitwise shift left (predicated) Richard Henderson
2020-03-26 23:08 ` [PATCH 08/31] target/arm: Implement SVE2 integer halving add/subtract (predicated) Richard Henderson
2020-03-26 23:08 ` [PATCH 09/31] target/arm: Implement SVE2 integer pairwise arithmetic Richard Henderson
2020-04-13 16:02   ` Laurent Desnogues
2020-03-26 23:08 ` [PATCH 10/31] target/arm: Implement SVE2 saturating add/subtract (predicated) Richard Henderson
2020-03-26 23:08 ` [PATCH 11/31] target/arm: Implement SVE2 integer add/subtract long Richard Henderson
2020-04-13 16:09   ` Laurent Desnogues
2020-03-26 23:08 ` [PATCH 12/31] target/arm: Implement SVE2 integer add/subtract interleaved long Richard Henderson
2020-03-26 23:08 ` [PATCH 13/31] target/arm: Implement SVE2 integer add/subtract wide Richard Henderson
2020-04-13 16:11   ` Laurent Desnogues
2020-03-26 23:08 ` [PATCH 14/31] target/arm: Implement SVE2 integer multiply long Richard Henderson
2020-03-26 23:08 ` [PATCH 15/31] target/arm: Implement PMULLB and PMULLT Richard Henderson
2020-03-26 23:08 ` [PATCH 16/31] target/arm: Tidy SVE tszimm shift formats Richard Henderson
2020-03-26 23:08 ` [PATCH 17/31] target/arm: Implement SVE2 bitwise shift left long Richard Henderson
2020-03-26 23:08 ` [PATCH 18/31] target/arm: Implement SVE2 bitwise exclusive-or interleaved Richard Henderson
2020-03-26 23:08 ` [PATCH 19/31] target/arm: Implement SVE2 bitwise permute Richard Henderson
2020-03-26 23:08 ` [PATCH 20/31] target/arm: Implement SVE2 complex integer add Richard Henderson
2020-04-13 16:20   ` Laurent Desnogues
2020-03-26 23:08 ` [PATCH 21/31] target/arm: Implement SVE2 integer absolute difference and accumulate long Richard Henderson
2020-04-13 16:15   ` Laurent Desnogues
2020-04-13 23:19     ` Richard Henderson [this message]
2020-04-14  7:04       ` Laurent Desnogues
2020-03-26 23:08 ` [PATCH 22/31] target/arm: Implement SVE2 integer add/subtract long with carry Richard Henderson
2020-04-13 16:18   ` Laurent Desnogues
2020-03-26 23:08 ` [PATCH 23/31] target/arm: Create arm_gen_gvec_[us]sra Richard Henderson
2020-03-26 23:08 ` [PATCH 24/31] target/arm: Create arm_gen_gvec_{u,s}{rshr,rsra} Richard Henderson
2020-03-26 23:08 ` [PATCH 25/31] target/arm: Implement SVE2 bitwise shift right and accumulate Richard Henderson
2020-03-26 23:08 ` [PATCH 26/31] target/arm: Create arm_gen_gvec_{sri,sli} Richard Henderson
2020-03-26 23:08 ` [PATCH 27/31] target/arm: Tidy handle_vec_simd_shri Richard Henderson
2020-03-26 23:08 ` [PATCH 28/31] target/arm: Implement SVE2 bitwise shift and insert Richard Henderson
2020-03-26 23:08 ` [PATCH 29/31] target/arm: Vectorize SABD/UABD Richard Henderson
2020-03-26 23:08 ` [PATCH 30/31] target/arm: Vectorize SABA/UABA Richard Henderson
2020-03-26 23:08 ` [PATCH 31/31] target/arm: Implement SVE2 integer absolute difference and accumulate Richard Henderson
2020-04-01 21:17 ` [PATCH for-5.1 00/31] target/arm: SVE2, part 1 Ana Pazos
2020-04-22  2:51 ` LIU Zhiwei
2020-04-22  2:55   ` Richard Henderson

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ae2cb680-2200-e9cb-afa1-e55593ef6d6d@linaro.org \
    --to=richard.henderson@linaro.org \
    --cc=apazos@quicinc.com \
    --cc=laurent.desnogues@gmail.com \
    --cc=qemu-arm@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=rajav@quicinc.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).