All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Ilpo Järvinen" <ilpo.jarvinen@linux.intel.com>
To: linux-kselftest@vger.kernel.org, linux-kernel@vger.kernel.org,
	Reinette Chatre <reinette.chatre@intel.com>,
	Fenghua Yu <fenghua.yu@intel.com>, Shuah Khan <shuah@kernel.org>
Cc: "Shaopeng Tan" <tan.shaopeng@jp.fujitsu.com>,
	"Ilpo Järvinen" <ilpo.jarvinen@linux.intel.com>
Subject: [PATCH 19/22] selftests/resctrl: Read in less obvious order to defeat prefetch optimizations
Date: Wed, 12 Apr 2023 16:21:19 +0300	[thread overview]
Message-ID: <20230412132122.29452-20-ilpo.jarvinen@linux.intel.com> (raw)
In-Reply-To: <20230412132122.29452-1-ilpo.jarvinen@linux.intel.com>

When reading memory in order, HW prefetching optimizations will
interfere with measuring how caches and memory are being accessed. This
adds noise into the results.

Change the fill_buf reading loop to not use an obvious in-order
access.

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
---
 tools/testing/selftests/resctrl/fill_buf.c | 17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)

diff --git a/tools/testing/selftests/resctrl/fill_buf.c b/tools/testing/selftests/resctrl/fill_buf.c
index 316800715386..742b3878ac2f 100644
--- a/tools/testing/selftests/resctrl/fill_buf.c
+++ b/tools/testing/selftests/resctrl/fill_buf.c
@@ -96,14 +96,17 @@ static void *malloc_and_init_memory(size_t s)
 
 static int fill_one_span_read(unsigned char *start_ptr, unsigned char *end_ptr)
 {
-	unsigned char sum, *p;
-
+	unsigned int size = (end_ptr - start_ptr) / (CL_SIZE / 2);
+	unsigned int count = size;
+	unsigned char sum;
+
+	/*
+	 * Read the buffer in an order that is unexpected by HW prefetching
+	 * optimizations to prevent them interfering with the caching pattern.
+	 */
 	sum = 0;
-	p = start_ptr;
-	while (p < end_ptr) {
-		sum += *p;
-		p += (CL_SIZE / 2);
-	}
+	while (count--)
+		sum += start_ptr[((count * 59) % size) * CL_SIZE / 2];
 
 	return sum;
 }
-- 
2.30.2


  parent reply	other threads:[~2023-04-12 13:25 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-04-12 13:21 [PATCH 00/22] selftests/resctrl: Fixes, cleanups, and rewritten CAT test Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 01/22] selftests/resctrl: Add resctrl.h into build deps Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 02/22] selftests/resctrl: Check also too low values for CBM bits Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 03/22] selftests/resctrl: Make span unsigned long everywhere Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 04/22] selftests/resctrl: Express span in bytes Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 05/22] selftests/resctrl: Remove duplicated preparation for span arg Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 06/22] selftests/resctrl: Don't use variable argument list for ->setup() Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 07/22] selftests/resctrl: Remove "malloc_and_init_memory" param from run_fill_buf() Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 08/22] selftests/resctrl: Split run_fill_buf() to alloc, work, and dealloc helpers Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 09/22] selftests/resctrl: Remove start_buf local variable from buffer alloc func Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 10/22] selftests/resctrl: Don't pass test name to fill_buf Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 11/22] selftests/resctrl: Add flush_buffer() " Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 12/22] selftests/resctrl: Remove test type checks from cat_val() Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 13/22] selftests/resctrl: Refactor get_cbm_mask() Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 14/22] selftests/resctrl: Create cache_alloc_size() helper Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 15/22] selftests/resctrl: Replace count_bits with count_consecutive_bits() Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 16/22] selftests/resctrl: Exclude shareable bits from schemata in CAT test Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 17/22] selftests/resctrl: Pass the real number of tests to show_cache_info() Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 18/22] selftests/resctrl: Move CAT/CMT test global vars to func they are used Ilpo Järvinen
2023-04-12 13:21 ` Ilpo Järvinen [this message]
2023-04-12 13:21 ` [PATCH 20/22] selftests/resctrl: Split measure_cache_vals() function Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 21/22] selftests/resctrl: Split show_cache_info() to test specific and generic parts Ilpo Järvinen
2023-04-12 13:21 ` [PATCH 22/22] selftests/resctrl: Rewrite Cache Allocation Technology (CAT) test Ilpo Järvinen
2023-04-13 10:12   ` Ilpo Järvinen

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=20230412132122.29452-20-ilpo.jarvinen@linux.intel.com \
    --to=ilpo.jarvinen@linux.intel.com \
    --cc=fenghua.yu@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-kselftest@vger.kernel.org \
    --cc=reinette.chatre@intel.com \
    --cc=shuah@kernel.org \
    --cc=tan.shaopeng@jp.fujitsu.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.