All of lore.kernel.org
 help / color / mirror / Atom feed
From: Raghavendra K T <raghavendra.kt@amd.com>
To: <linux-kernel@vger.kernel.org>, <linux-mm@kvack.org>
Cc: Ingo Molnar <mingo@redhat.com>,
	Peter Zijlstra <peterz@infradead.org>,
	"Mel Gorman" <mgorman@suse.de>,
	Andrew Morton <akpm@linux-foundation.org>,
	"David Hildenbrand" <david@redhat.com>, <rppt@kernel.org>,
	Juri Lelli <juri.lelli@redhat.com>,
	Vincent Guittot <vincent.guittot@linaro.org>,
	Bharata B Rao <bharata@amd.com>,
	Aithal Srikanth <sraithal@amd.com>,
	"kernel test robot" <oliver.sang@intel.com>,
	Raghavendra K T <raghavendra.kt@amd.com>,
	Sapkal Swapnil <Swapnil.Sapkal@amd.com>,
	K Prateek Nayak <kprateek.nayak@amd.com>
Subject: [RFC PATCH V1 6/6] sched/numa: Allow scanning of shared VMAs
Date: Tue, 29 Aug 2023 11:36:14 +0530	[thread overview]
Message-ID: <0ff0d56b88a64e2017b427f37b15ab41aa5d4019.1693287931.git.raghavendra.kt@amd.com> (raw)
In-Reply-To: <cover.1693287931.git.raghavendra.kt@amd.com>

VMAs being accessed by more than two tasks are critical. Scan them
unconditionally. Note that current patch considers full history of
VMA access.

Signed-off-by: Raghavendra K T <raghavendra.kt@amd.com>
---
 kernel/sched/fair.c | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 6529da7f370a..70c8e62c1a89 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -2981,6 +2981,19 @@ static inline bool vma_accessed_recent(struct vm_area_struct *vma)
 	return (bitmap_weight(pids, BITS_PER_LONG) >= 1);
 }
 
+#define SHARED_VMA_THRESH	3
+
+static inline bool vma_shared_access(struct vm_area_struct *vma)
+{
+	int i;
+	unsigned long pids = 0;
+
+	for (i = 0; i < NR_ACCESS_PID_HIST; i++)
+		pids  |= vma->numab_state->access_pids[i];
+
+	return (bitmap_weight(&pids, BITS_PER_LONG) >= SHARED_VMA_THRESH);
+}
+
 static bool vma_is_accessed(struct vm_area_struct *vma)
 {
 	/* Check at least one task had accessed VMA recently. */
@@ -2991,7 +3004,8 @@ static bool vma_is_accessed(struct vm_area_struct *vma)
 	if (vma_test_access_pid_history(vma))
 		return true;
 
-	return false;
+	/* Check if VMA is shared by many tasks. */
+	return vma_shared_access(vma);
 }
 
 #define VMA_PID_RESET_PERIOD (4 * sysctl_numa_balancing_scan_delay)
-- 
2.34.1


  parent reply	other threads:[~2023-08-29  6:08 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-29  6:06 [RFC PATCH V1 0/6] sched/numa: Enhance disjoint VMA scanning Raghavendra K T
2023-08-29  6:06 ` [RFC PATCH V1 1/6] sched/numa: Move up the access pid reset logic Raghavendra K T
2023-08-29  6:06 ` [RFC PATCH V1 2/6] sched/numa: Add disjoint vma unconditional scan logic Raghavendra K T
2023-09-12  7:50   ` kernelt test robot
2023-09-13  6:21     ` Raghavendra K T
2023-08-29  6:06 ` [RFC PATCH V1 3/6] sched/numa: Remove unconditional scan logic using mm numa_scan_seq Raghavendra K T
2023-08-29  6:06 ` [RFC PATCH V1 4/6] sched/numa: Increase tasks' access history Raghavendra K T
2023-09-12 14:24   ` kernel test robot
2023-09-13  6:15     ` Raghavendra K T
2023-09-13  7:34       ` Oliver Sang
2023-08-29  6:06 ` [RFC PATCH V1 5/6] sched/numa: Allow recently accessed VMAs to be scanned Raghavendra K T
2023-09-10 15:29   ` kernel test robot
2023-09-11 11:25     ` Raghavendra K T
2023-09-12  2:22       ` Oliver Sang
2023-09-12  6:43         ` Raghavendra K T
2023-08-29  6:06 ` Raghavendra K T [this message]
2023-09-13  5:28 ` [RFC PATCH V1 0/6] sched/numa: Enhance disjoint VMA scanning Swapnil Sapkal
2023-09-13  6:24   ` Raghavendra K T
2023-09-19  6:30 ` Raghavendra K T
2023-09-19  7:15   ` Ingo Molnar
2023-09-19  8:06     ` Raghavendra K T
2023-09-19  9:28 ` Peter Zijlstra
2023-09-19 16:22   ` Mel Gorman
2023-09-19 19:11     ` Peter Zijlstra
2023-09-20 10:42     ` Raghavendra K T

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=0ff0d56b88a64e2017b427f37b15ab41aa5d4019.1693287931.git.raghavendra.kt@amd.com \
    --to=raghavendra.kt@amd.com \
    --cc=Swapnil.Sapkal@amd.com \
    --cc=akpm@linux-foundation.org \
    --cc=bharata@amd.com \
    --cc=david@redhat.com \
    --cc=juri.lelli@redhat.com \
    --cc=kprateek.nayak@amd.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mgorman@suse.de \
    --cc=mingo@redhat.com \
    --cc=oliver.sang@intel.com \
    --cc=peterz@infradead.org \
    --cc=rppt@kernel.org \
    --cc=sraithal@amd.com \
    --cc=vincent.guittot@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 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.