linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Ankur Arora <ankur.a.arora@oracle.com>
To: linux-kernel@vger.kernel.org, linux-mm@kvack.org, x86@kernel.org
Cc: akpm@linux-foundation.org, david@redhat.com, bp@alien8.de,
	dave.hansen@linux.intel.com, hpa@zytor.com, mingo@redhat.com,
	mjguzik@gmail.com, luto@kernel.org, peterz@infradead.org,
	acme@kernel.org, namhyung@kernel.org, tglx@linutronix.de,
	willy@infradead.org, raghavendra.kt@amd.com,
	boris.ostrovsky@oracle.com, konrad.wilk@oracle.com,
	ankur.a.arora@oracle.com
Subject: [PATCH v5 11/14] x86/clear_page: Introduce clear_pages()
Date: Wed,  9 Jul 2025 17:59:23 -0700	[thread overview]
Message-ID: <20250710005926.1159009-12-ankur.a.arora@oracle.com> (raw)
In-Reply-To: <20250710005926.1159009-1-ankur.a.arora@oracle.com>

Performance when clearing with string instructions (x86-64-stosq and
similar) can vary significantly based on the chunk-size used.

  $ perf bench mem memset -k 4KB -s 4GB -f x86-64-stosq
  # Running 'mem/memset' benchmark:
  # function 'x86-64-stosq' (movsq-based memset() in arch/x86/lib/memset_64.S)
  # Copying 4GB bytes ...

      13.748208 GB/sec

  $ perf bench mem memset -k 2MB -s 4GB -f x86-64-stosq
  # Running 'mem/memset' benchmark:
  # function 'x86-64-stosq' (movsq-based memset() in
  # arch/x86/lib/memset_64.S)
  # Copying 4GB bytes ...

      15.067900 GB/sec

  $ perf bench mem memset -k 1GB -s 4GB -f x86-64-stosq
  # Running 'mem/memset' benchmark:
  # function 'x86-64-stosq' (movsq-based memset() in arch/x86/lib/memset_64.S)
  # Copying 4GB bytes ...

      38.104311 GB/sec

(Both on AMD Milan.)

With a change in chunk-size of 4KB to 1GB, we see the performance go
from 13.7 GB/sec to 38.1 GB/sec. For a chunk-size of 2MB the change isn't
quite as drastic but it is worth adding a clear_page() variant that can
handle contiguous page-extents.

Signed-off-by: Ankur Arora <ankur.a.arora@oracle.com>
---
 arch/x86/include/asm/page_32.h | 15 +++++++++++----
 arch/x86/include/asm/page_64.h | 24 ++++++++++++++++--------
 2 files changed, 27 insertions(+), 12 deletions(-)

diff --git a/arch/x86/include/asm/page_32.h b/arch/x86/include/asm/page_32.h
index a8ff43bb9652..561f416b61e5 100644
--- a/arch/x86/include/asm/page_32.h
+++ b/arch/x86/include/asm/page_32.h
@@ -18,16 +18,23 @@ extern unsigned long __phys_addr(unsigned long);
 #include <linux/string.h>
 
 /*
- * clear_page() - clear kernel page.
- * @page: address of kernel page
+ * clear_pages() - clear kernel page range.
+ * @addr: start address of page range
+ * @npages: number of pages
  *
+ * Assumes that (@addr, +@npages) references a kernel region.
  * Does absolutely no exception handling.
  */
-static inline void clear_page(void *page)
+static inline void clear_pages(void *addr, u64 npages)
 {
-	memset(page, 0, PAGE_SIZE);
+	for (u64 i = 0; i < npages; i++)
+		memset(addr + i * PAGE_SIZE, 0, PAGE_SIZE);
 }
 
+static inline void clear_page(void *addr)
+{
+	clear_pages(addr, 1);
+}
 static inline void copy_page(void *to, void *from)
 {
 	memcpy(to, from, PAGE_SIZE);
diff --git a/arch/x86/include/asm/page_64.h b/arch/x86/include/asm/page_64.h
index 28b9adbc5f00..5625d616bd00 100644
--- a/arch/x86/include/asm/page_64.h
+++ b/arch/x86/include/asm/page_64.h
@@ -43,8 +43,11 @@ extern unsigned long __phys_addr_symbol(unsigned long);
 void memzero_page_aligned_unrolled(void *addr, u64 len);
 
 /*
- * clear_page() - clear kernel page.
- * @page: address of kernel page
+ * clear_pages() - clear kernel page range.
+ * @addr: start address of page range
+ * @npages: number of pages
+ *
+ * Assumes that (@addr, +@npages) references a kernel region.
  *
  * Switch between three implementations of page clearing based on CPU
  * capabilities:
@@ -65,22 +68,27 @@ void memzero_page_aligned_unrolled(void *addr, u64 len);
  *
  * Does absolutely no exception handling.
  */
-static inline void clear_page(void *page)
+static inline void clear_pages(void *addr, u64 npages)
 {
-	u64 len = PAGE_SIZE;
+	u64 len = npages * PAGE_SIZE;
 	/*
-	 * Clean up KMSAN metadata for the page being cleared. The assembly call
-	 * below clobbers @page, so we perform unpoisoning before it.
+	 * Clean up KMSAN metadata for the pages being cleared. The assembly call
+	 * below clobbers @addr, so we perform unpoisoning before it.
 	 */
-	kmsan_unpoison_memory(page, len);
+	kmsan_unpoison_memory(addr, len);
 	asm volatile(ALTERNATIVE_2("call memzero_page_aligned_unrolled",
 				   "shrq $3, %%rcx; rep stosq", X86_FEATURE_REP_GOOD,
 				   "rep stosb", X86_FEATURE_ERMS)
-			: "+c" (len), "+D" (page), ASM_CALL_CONSTRAINT
+			: "+c" (len), "+D" (addr), ASM_CALL_CONSTRAINT
 			: "a" (0)
 			: "cc", "memory");
 }
 
+static inline void clear_page(void *addr)
+{
+	clear_pages(addr, 1);
+}
+
 void copy_page(void *to, void *from);
 KCFI_REFERENCE(copy_page);
 
-- 
2.43.5



  parent reply	other threads:[~2025-07-10  1:00 UTC|newest]

Thread overview: 52+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-07-10  0:59 [PATCH v5 00/14] mm: folio_zero_user: clearing of page-extents Ankur Arora
2025-07-10  0:59 ` [PATCH v5 01/14] perf bench mem: Remove repetition around time measurement Ankur Arora
2025-07-15 20:04   ` Namhyung Kim
2025-07-10  0:59 ` [PATCH v5 02/14] perf bench mem: Defer type munging of size to float Ankur Arora
2025-07-15 20:05   ` Namhyung Kim
2025-07-16  2:17     ` Ankur Arora
2025-07-10  0:59 ` [PATCH v5 03/14] perf bench mem: Move mem op parameters into a structure Ankur Arora
2025-07-15 20:06   ` Namhyung Kim
2025-07-10  0:59 ` [PATCH v5 04/14] perf bench mem: Pull out init/fini logic Ankur Arora
2025-07-15 20:09   ` Namhyung Kim
2025-07-10  0:59 ` [PATCH v5 05/14] perf bench mem: Switch from zalloc() to mmap() Ankur Arora
2025-07-15 20:09   ` Namhyung Kim
2025-07-10  0:59 ` [PATCH v5 06/14] perf bench mem: Allow mapping of hugepages Ankur Arora
2025-07-15 20:12   ` Namhyung Kim
2025-07-16  2:32     ` Ankur Arora
2025-07-10  0:59 ` [PATCH v5 07/14] perf bench mem: Allow chunking on a memory region Ankur Arora
2025-07-15 20:17   ` Namhyung Kim
2025-07-16  2:34     ` Ankur Arora
2025-07-10  0:59 ` [PATCH v5 08/14] perf bench mem: Refactor mem_options Ankur Arora
2025-07-15 20:18   ` Namhyung Kim
2025-07-10  0:59 ` [PATCH v5 09/14] perf bench mem: Add mmap() workloads Ankur Arora
2025-07-15 20:20   ` Namhyung Kim
2025-07-16  2:40     ` Ankur Arora
2025-07-10  0:59 ` [PATCH v5 10/14] x86/mm: Simplify clear_page_* Ankur Arora
2025-07-11 11:47   ` David Hildenbrand
2025-07-11 17:26     ` Ankur Arora
2025-07-11 19:03       ` David Hildenbrand
2025-07-11 19:24         ` Ankur Arora
2025-07-11 19:27           ` David Hildenbrand
2025-07-10  0:59 ` Ankur Arora [this message]
2025-07-10  0:59 ` [PATCH v5 12/14] mm: add config option for clearing page-extents Ankur Arora
2025-07-10  7:58   ` Andrew Morton
2025-07-10 16:31     ` Ankur Arora
2025-07-11 11:39   ` David Hildenbrand
2025-07-11 17:25     ` Ankur Arora
2025-07-11 19:14       ` David Hildenbrand
2025-07-11 19:35         ` Ankur Arora
2025-07-11 11:40   ` David Hildenbrand
2025-07-11 17:32     ` Ankur Arora
2025-07-11 19:26       ` David Hildenbrand
2025-07-11 19:42         ` Ankur Arora
2025-07-14 20:35         ` Ankur Arora
2025-07-15 20:59           ` David Hildenbrand
2025-07-10  0:59 ` [PATCH v5 13/14] mm: memory: support " Ankur Arora
2025-07-11 11:44   ` David Hildenbrand
2025-07-11 13:27     ` Raghavendra K T
2025-07-11 17:39     ` Ankur Arora
2025-07-15 22:08   ` David Hildenbrand
2025-07-16  3:19     ` Ankur Arora
2025-07-16  8:03       ` David Hildenbrand
2025-07-16 17:54         ` Ankur Arora
2025-07-10  0:59 ` [PATCH v5 14/14] x86/clear_pages: Support clearing of page-extents Ankur Arora

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=20250710005926.1159009-12-ankur.a.arora@oracle.com \
    --to=ankur.a.arora@oracle.com \
    --cc=acme@kernel.org \
    --cc=akpm@linux-foundation.org \
    --cc=boris.ostrovsky@oracle.com \
    --cc=bp@alien8.de \
    --cc=dave.hansen@linux.intel.com \
    --cc=david@redhat.com \
    --cc=hpa@zytor.com \
    --cc=konrad.wilk@oracle.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=luto@kernel.org \
    --cc=mingo@redhat.com \
    --cc=mjguzik@gmail.com \
    --cc=namhyung@kernel.org \
    --cc=peterz@infradead.org \
    --cc=raghavendra.kt@amd.com \
    --cc=tglx@linutronix.de \
    --cc=willy@infradead.org \
    --cc=x86@kernel.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 NNTP newsgroup(s).