From: Richard Sandiford <richard.sandiford@arm.com>
To: Hao Liu OS <hliu@os.amperecomputing.com>
Cc: "GCC-patches\@gcc.gnu.org" <gcc-patches@gcc.gnu.org>
Subject: Re: [PATCH] AArch64: Do not increase the vect reduction latency by multiplying count [PR110625]
Date: Tue, 25 Jul 2023 10:44:04 +0100 [thread overview]
Message-ID: <mptsf9cfhqz.fsf@arm.com> (raw)
In-Reply-To: <SJ2PR01MB8635D1095B5BDDAE1CA19877E103A@SJ2PR01MB8635.prod.exchangelabs.com> (Hao Liu's message of "Tue, 25 Jul 2023 09:10:31 +0000")
Hao Liu OS <hliu@os.amperecomputing.com> writes:
> Hi,
>
> Thanks for the suggestion. I tested it and found a gcc_assert failure:
> gcc.target/aarch64/sve/cost_model_13.c (internal compiler error: in info_for_reduction, at tree-vect-loop.cc:5473)
>
> It is caused by empty STMT_VINFO_REDUC_DEF.
When was STMT_VINFO_REDUC_DEF empty? I just want to make sure that
we're not papering over an issue elsewhere.
Thanks,
Richard
So, I added an extra check before checking single_defuse_cycle. The updated patch is below. Is it OK for trunk?
>
> ---
>
> The new costs should only count reduction latency by multiplying count for
> single_defuse_cycle. For other situations, this will increase the reduction
> latency a lot and miss vectorization opportunities.
>
> Tested on aarch64-linux-gnu.
>
> gcc/ChangeLog:
>
> PR target/110625
> * config/aarch64/aarch64.cc (count_ops): Only '* count' for
> single_defuse_cycle while counting reduction_latency.
>
> gcc/testsuite/ChangeLog:
>
> * gcc.target/aarch64/pr110625_1.c: New testcase.
> * gcc.target/aarch64/pr110625_2.c: New testcase.
> ---
> gcc/config/aarch64/aarch64.cc | 13 ++++--
> gcc/testsuite/gcc.target/aarch64/pr110625_1.c | 46 +++++++++++++++++++
> gcc/testsuite/gcc.target/aarch64/pr110625_2.c | 14 ++++++
> 3 files changed, 69 insertions(+), 4 deletions(-)
> create mode 100644 gcc/testsuite/gcc.target/aarch64/pr110625_1.c
> create mode 100644 gcc/testsuite/gcc.target/aarch64/pr110625_2.c
>
> diff --git a/gcc/config/aarch64/aarch64.cc b/gcc/config/aarch64/aarch64.cc
> index 560e5431636..478a4e00110 100644
> --- a/gcc/config/aarch64/aarch64.cc
> +++ b/gcc/config/aarch64/aarch64.cc
> @@ -16788,10 +16788,15 @@ aarch64_vector_costs::count_ops (unsigned int count, vect_cost_for_stmt kind,
> {
> unsigned int base
> = aarch64_in_loop_reduction_latency (m_vinfo, stmt_info, m_vec_flags);
> -
> - /* ??? Ideally we'd do COUNT reductions in parallel, but unfortunately
> - that's not yet the case. */
> - ops->reduction_latency = MAX (ops->reduction_latency, base * count);
> + if (STMT_VINFO_REDUC_DEF (stmt_info)
> + && STMT_VINFO_FORCE_SINGLE_CYCLE (
> + info_for_reduction (m_vinfo, stmt_info)))
> + /* ??? Ideally we'd use a tree to reduce the copies down to 1 vector,
> + and then accumulate that, but at the moment the loop-carried
> + dependency includes all copies. */
> + ops->reduction_latency = MAX (ops->reduction_latency, base * count);
> + else
> + ops->reduction_latency = MAX (ops->reduction_latency, base);
> }
>
> /* Assume that multiply-adds will become a single operation. */
> diff --git a/gcc/testsuite/gcc.target/aarch64/pr110625_1.c b/gcc/testsuite/gcc.target/aarch64/pr110625_1.c
> new file mode 100644
> index 00000000000..0965cac33a0
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/aarch64/pr110625_1.c
> @@ -0,0 +1,46 @@
> +/* { dg-do compile } */
> +/* { dg-options "-Ofast -mcpu=neoverse-n2 -fdump-tree-vect-details -fno-tree-slp-vectorize" } */
> +/* { dg-final { scan-tree-dump-not "reduction latency = 8" "vect" } } */
> +
> +/* Do not increase the vector body cost due to the incorrect reduction latency
> + Original vector body cost = 51
> + Scalar issue estimate:
> + ...
> + reduction latency = 2
> + estimated min cycles per iteration = 2.000000
> + estimated cycles per vector iteration (for VF 2) = 4.000000
> + Vector issue estimate:
> + ...
> + reduction latency = 8 <-- Too large
> + estimated min cycles per iteration = 8.000000
> + Increasing body cost to 102 because scalar code would issue more quickly
> + ...
> + missed: cost model: the vector iteration cost = 102 divided by the scalar iteration cost = 44 is greater or equal to the vectorization factor = 2.
> + missed: not vectorized: vectorization not profitable. */
> +
> +typedef struct
> +{
> + unsigned short m1, m2, m3, m4;
> +} the_struct_t;
> +typedef struct
> +{
> + double m1, m2, m3, m4, m5;
> +} the_struct2_t;
> +
> +double
> +bar (the_struct2_t *);
> +
> +double
> +foo (double *k, unsigned int n, the_struct_t *the_struct)
> +{
> + unsigned int u;
> + the_struct2_t result;
> + for (u = 0; u < n; u++, k--)
> + {
> + result.m1 += (*k) * the_struct[u].m1;
> + result.m2 += (*k) * the_struct[u].m2;
> + result.m3 += (*k) * the_struct[u].m3;
> + result.m4 += (*k) * the_struct[u].m4;
> + }
> + return bar (&result);
> +}
> diff --git a/gcc/testsuite/gcc.target/aarch64/pr110625_2.c b/gcc/testsuite/gcc.target/aarch64/pr110625_2.c
> new file mode 100644
> index 00000000000..7a84aa8355e
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/aarch64/pr110625_2.c
> @@ -0,0 +1,14 @@
> +/* { dg-do compile } */
> +/* { dg-options "-Ofast -mcpu=neoverse-n2 -fdump-tree-vect-details -fno-tree-slp-vectorize" } */
> +/* { dg-final { scan-tree-dump "reduction latency = 8" "vect" } } */
> +
> +/* The reduction latency should be multiplied by the count for
> + single_defuse_cycle. */
> +
> +long
> +f (long res, short *ptr1, short *ptr2, int n)
> +{
> + for (int i = 0; i < n; ++i)
> + res += (long) ptr1[i] << ptr2[i];
> + return res;
> +}
next prev parent reply other threads:[~2023-07-25 9:44 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-07-19 4:33 Hao Liu OS
2023-07-24 1:58 ` Hao Liu OS
2023-07-24 11:10 ` Richard Sandiford
2023-07-25 9:10 ` Hao Liu OS
2023-07-25 9:44 ` Richard Sandiford [this message]
2023-07-26 2:01 ` Hao Liu OS
2023-07-26 8:47 ` Richard Biener
2023-07-26 9:14 ` Richard Sandiford
2023-07-26 10:02 ` Richard Biener
2023-07-26 10:12 ` Richard Sandiford
2023-07-26 12:00 ` Richard Biener
2023-07-26 12:54 ` Hao Liu OS
2023-07-28 10:06 ` Hao Liu OS
2023-07-28 17:35 ` Richard Sandiford
2023-07-31 2:39 ` Hao Liu OS
2023-07-31 9:11 ` Richard Sandiford
2023-07-31 9:25 ` Hao Liu OS
2023-08-01 9:43 ` Hao Liu OS
2023-08-02 3:45 ` Hao Liu OS
2023-08-03 9:33 ` Hao Liu OS
2023-08-03 10:10 ` Richard Sandiford
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=mptsf9cfhqz.fsf@arm.com \
--to=richard.sandiford@arm.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=hliu@os.amperecomputing.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 read-only IMAP folder(s) and NNTP newsgroup(s).