From: "H.J. Lu" <hjl.tools@gmail.com>
To: Luis Machado <luis.machado@linaro.org>
Cc: Jeff Law <law@redhat.com>, GCC Patches <gcc-patches@gcc.gnu.org>,
James Greenhalgh <james.greenhalgh@arm.com>,
Richard Earnshaw <Richard.Earnshaw@arm.com>
Subject: Re: [PATCH 1/2] Introduce prefetch-minimum stride option
Date: Mon, 07 May 2018 15:15:00 -0000 [thread overview]
Message-ID: <CAMe9rOou84WnBTk0RniHgJacishueWGQzr4fgrD76HgMQQHChA@mail.gmail.com> (raw)
In-Reply-To: <702726ab-1287-6042-3d7c-ac2119d23fb9@linaro.org>
On Mon, May 7, 2018 at 7:09 AM, Luis Machado <luis.machado@linaro.org> wrote:
>
>
> On 05/01/2018 03:30 PM, Jeff Law wrote:
>>
>> On 01/22/2018 06:46 AM, Luis Machado wrote:
>>>
>>> This patch adds a new option to control the minimum stride, for a memory
>>> reference, after which the loop prefetch pass may issue software prefetch
>>> hints for. There are two motivations:
>>>
>>> * Make the pass less aggressive, only issuing prefetch hints for bigger
>>> strides
>>> that are more likely to benefit from prefetching. I've noticed a case in
>>> cpu2017
>>> where we were issuing thousands of hints, for example.
>>>
>>> * For processors that have a hardware prefetcher, like Falkor, it allows
>>> the
>>> loop prefetch pass to defer prefetching of smaller (less than the
>>> threshold)
>>> strides to the hardware prefetcher instead. This prevents conflicts
>>> between
>>> the software prefetcher and the hardware prefetcher.
>>>
>>> I've noticed considerable reduction in the number of prefetch hints and
>>> slightly positive performance numbers. This aligns GCC and LLVM in terms
>>> of
>>> prefetch behavior for Falkor.
>>>
>>> The default settings should guarantee no changes for existing targets.
>>> Those
>>> are free to tweak the settings as necessary.
>>>
>>> No regressions in the testsuite and bootstrapped ok on aarch64-linux.
>>>
>>> Ok?
>>>
>>> 2018-01-22 Luis Machado <luis.machado@linaro.org>
>>>
>>> Introduce option to limit software prefetching to known constant
>>> strides above a specific threshold with the goal of preventing
>>> conflicts with a hardware prefetcher.
>>>
>>> gcc/
>>> * config/aarch64/aarch64-protos.h (cpu_prefetch_tune)
>>> <minimum_stride>: New const int field.
>>> * config/aarch64/aarch64.c (generic_prefetch_tune): Update to
>>> include
>>> minimum_stride field.
>>> (exynosm1_prefetch_tune): Likewise.
>>> (thunderxt88_prefetch_tune): Likewise.
>>> (thunderx_prefetch_tune): Likewise.
>>> (thunderx2t99_prefetch_tune): Likewise.
>>> (qdf24xx_prefetch_tune): Likewise. Set minimum_stride to 2048.
>>> (aarch64_override_options_internal): Update to set
>>> PARAM_PREFETCH_MINIMUM_STRIDE.
>>> * doc/invoke.texi (prefetch-minimum-stride): Document new option.
>>> * params.def (PARAM_PREFETCH_MINIMUM_STRIDE): New.
>>> * params.h (PARAM_PREFETCH_MINIMUM_STRIDE): Define.
>>> * tree-ssa-loop-prefetch.c (should_issue_prefetch_p): Return
>>> false if
>>> stride is constant and is below the minimum stride threshold.
>>
>> OK for the trunk.
>> jeff
>>
>
> Thanks. Committed as revision 259995 now.
This breaks bootstrap on x86:
../../src-trunk/gcc/tree-ssa-loop-prefetch.c: In function ‘bool
should_issue_prefetch_p(mem_ref*)’:
../../src-trunk/gcc/tree-ssa-loop-prefetch.c:1010:54: error:
comparison of integer expressions of different signedness: ‘long long
unsigned int’ and ‘int’ [-Werror=sign-compare]
&& absu_hwi (int_cst_value (ref->group->step)) < PREFETCH_MINIMUM_STRIDE)
../../src-trunk/gcc/tree-ssa-loop-prefetch.c:1014:4: error: format
‘%d’ expects argument of type ‘int’, but argument 5 has type ‘long
long int’ [-Werror=format=]
"Step for reference %u:%u (%d) is less than the mininum "
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"required stride of %d\n",
~~~~~~~~~~~~~~~~~~~~~~~~~
ref->group->uid, ref->uid, int_cst_value (ref->group->step),
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
--
H.J.
next prev parent reply other threads:[~2018-05-07 15:15 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-01-22 13:46 [PATCH 0/2] Add a couple new options to control loop prefetch pass Luis Machado
2018-01-22 14:01 ` [PATCH 1/2] Introduce prefetch-minimum stride option Luis Machado
2018-01-23 9:46 ` Kyrill Tkachov
2018-01-23 13:23 ` Luis Machado
2018-05-01 18:30 ` Jeff Law
2018-05-07 14:10 ` Luis Machado
2018-05-07 15:15 ` H.J. Lu [this message]
2018-05-07 15:51 ` Luis Machado
2018-05-14 21:21 ` Luis Machado
2018-05-15 9:59 ` Kyrill Tkachov
2018-05-15 11:21 ` Luis Machado
2018-05-16 9:22 ` Kyrill Tkachov
2018-05-16 11:53 ` Luis Machado
2018-05-22 18:56 ` Luis Machado
2018-05-22 21:21 ` Jeff Law
2018-05-23 20:27 ` H.J. Lu
2018-05-23 22:34 ` Luis Machado
2018-05-23 22:41 ` H.J. Lu
2018-05-23 22:42 ` H.J. Lu
2018-05-23 22:45 ` H.J. Lu
2018-05-23 23:29 ` Luis Machado
2018-05-24 2:51 ` Jeff Law
2018-05-24 12:21 ` Luis Machado
2018-01-22 14:10 ` [PATCH 2/2] Introduce prefetch-dynamic-strides option Luis Machado
2018-01-23 9:53 ` Kyrill Tkachov
2018-01-23 13:32 ` Luis Machado
2018-05-01 18:31 ` Jeff Law
2018-05-07 14:13 ` Luis Machado
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=CAMe9rOou84WnBTk0RniHgJacishueWGQzr4fgrD76HgMQQHChA@mail.gmail.com \
--to=hjl.tools@gmail.com \
--cc=Richard.Earnshaw@arm.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=james.greenhalgh@arm.com \
--cc=law@redhat.com \
--cc=luis.machado@linaro.org \
/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).