From: "Uladzislau Rezki (Sony)" <urezki@gmail.com>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mm@kvack.org, LKML <linux-kernel@vger.kernel.org>,
Christoph Hellwig <hch@infradead.org>,
Matthew Wilcox <willy@infradead.org>,
Nicholas Piggin <npiggin@gmail.com>,
Uladzislau Rezki <urezki@gmail.com>,
Oleksiy Avramchenko <oleksiy.avramchenko@sony.com>
Subject: [PATCH 5/5] lib/test_vmalloc: Switch to prandom_u32()
Date: Tue, 7 Jun 2022 11:34:49 +0200 [thread overview]
Message-ID: <20220607093449.3100-6-urezki@gmail.com> (raw)
In-Reply-To: <20220607093449.3100-1-urezki@gmail.com>
A get_random_bytes() function can cause a high contention if it is
called across CPUs simultaneously. Because it shares one lock per
all CPUs:
<snip>
class name con-bounces contentions waittime-min waittime-max waittime-total waittime-avg acq-bounces acquisitions holdtime-min holdtime-max holdtime-total holdtime-avg
&crng->lock: 663145 665886 0.05 8.85 261966.66 0.39 7188152 13731279 0.04 11.89 2181582.30 0.16
-----------
&crng->lock 307835 [<00000000acba59cd>] _extract_crng+0x48/0x90
&crng->lock 358051 [<00000000f0075abc>] _crng_backtrack_protect+0x32/0x90
-----------
&crng->lock 234241 [<00000000f0075abc>] _crng_backtrack_protect+0x32/0x90
&crng->lock 431645 [<00000000acba59cd>] _extract_crng+0x48/0x90
<snip>
Switch from the get_random_bytes() to prandom_u32() that does not
have any internal contention when a random value is needed for the
tests.
The reason is to minimize CPU cycles introduced by the test-suite
itself from the vmalloc performance metrics.
Signed-off-by: Uladzislau Rezki (Sony) <urezki@gmail.com>
---
lib/test_vmalloc.c | 15 ++++++++-------
1 file changed, 8 insertions(+), 7 deletions(-)
diff --git a/lib/test_vmalloc.c b/lib/test_vmalloc.c
index cf41fd6df42a..4f2f2d1bac56 100644
--- a/lib/test_vmalloc.c
+++ b/lib/test_vmalloc.c
@@ -74,12 +74,13 @@ test_report_one_done(void)
static int random_size_align_alloc_test(void)
{
- unsigned long size, align, rnd;
+ unsigned long size, align;
+ unsigned int rnd;
void *ptr;
int i;
for (i = 0; i < test_loop_count; i++) {
- get_random_bytes(&rnd, sizeof(rnd));
+ rnd = prandom_u32();
/*
* Maximum 1024 pages, if PAGE_SIZE is 4096.
@@ -150,7 +151,7 @@ static int random_size_alloc_test(void)
int i;
for (i = 0; i < test_loop_count; i++) {
- get_random_bytes(&n, sizeof(i));
+ n = prandom_u32();
n = (n % 100) + 1;
p = vmalloc(n * PAGE_SIZE);
@@ -294,14 +295,14 @@ pcpu_alloc_test(void)
for (i = 0; i < 35000; i++) {
unsigned int r;
- get_random_bytes(&r, sizeof(i));
+ r = prandom_u32();
size = (r % (PAGE_SIZE / 4)) + 1;
/*
* Maximum PAGE_SIZE
*/
- get_random_bytes(&r, sizeof(i));
- align = 1 << ((i % 11) + 1);
+ r = prandom_u32();
+ align = 1 << ((r % 11) + 1);
pcpu[i] = __alloc_percpu(size, align);
if (!pcpu[i])
@@ -396,7 +397,7 @@ static void shuffle_array(int *arr, int n)
int i, j;
for (i = n - 1; i > 0; i--) {
- get_random_bytes(&rnd, sizeof(rnd));
+ rnd = prandom_u32();
/* Cut the range. */
j = rnd % i;
--
2.30.2
next prev parent reply other threads:[~2022-06-07 9:35 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-06-07 9:34 [PATCH 0/5] Reduce a vmalloc internal lock contention preparation work Uladzislau Rezki (Sony)
2022-06-07 9:34 ` [PATCH 1/5] mm/vmalloc: Make link_va()/unlink_va() common to different rb_root Uladzislau Rezki (Sony)
2022-06-08 3:45 ` Baoquan He
2022-06-07 9:34 ` [PATCH 2/5] mm/vmalloc: Extend __alloc_vmap_area() with extra arguments Uladzislau Rezki (Sony)
2022-06-07 9:49 ` Christoph Hellwig
2022-06-07 13:02 ` Uladzislau Rezki
2022-06-10 8:18 ` Christoph Hellwig
2022-06-08 3:46 ` Baoquan He
2022-06-07 9:34 ` [PATCH 3/5] mm/vmalloc: Initialize VA's list node after unlink Uladzislau Rezki (Sony)
2022-06-08 3:19 ` Baoquan He
2022-06-09 12:36 ` Uladzislau Rezki
2022-06-09 13:30 ` Baoquan He
2022-06-09 13:53 ` Uladzislau Rezki
2022-06-07 9:34 ` [PATCH 4/5] mm/vmalloc: Extend __find_vmap_area() with one more argument Uladzislau Rezki (Sony)
2022-06-07 9:47 ` Christoph Hellwig
2022-06-07 13:03 ` Uladzislau Rezki
2022-06-08 3:47 ` Baoquan He
2022-06-07 9:34 ` Uladzislau Rezki (Sony) [this message]
2022-06-07 22:35 ` [PATCH 0/5] Reduce a vmalloc internal lock contention preparation work Andrew Morton
2022-06-08 10:05 ` Uladzislau Rezki
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=20220607093449.3100-6-urezki@gmail.com \
--to=urezki@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=hch@infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=npiggin@gmail.com \
--cc=oleksiy.avramchenko@sony.com \
--cc=willy@infradead.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.