All of lore.kernel.org
 help / color / mirror / Atom feed
From: Catalin Marinas <catalin.marinas@arm.com>
To: linux-arm-kernel@lists.infradead.org
Cc: linux-mm@kvack.org, linux-arch@vger.kernel.org,
	Will Deacon <will@kernel.org>,
	Dave P Martin <Dave.Martin@arm.com>,
	Vincenzo Frascino <vincenzo.frascino@arm.com>,
	Szabolcs Nagy <szabolcs.nagy@arm.com>,
	Kevin Brodsky <kevin.brodsky@arm.com>,
	Andrey Konovalov <andreyknvl@google.com>,
	Peter Collingbourne <pcc@google.com>
Subject: [PATCH v4 09/26] arm64: mte: Tags-aware aware memcmp_pages() implementation
Date: Fri, 15 May 2020 18:15:55 +0100	[thread overview]
Message-ID: <20200515171612.1020-10-catalin.marinas@arm.com> (raw)
In-Reply-To: <20200515171612.1020-1-catalin.marinas@arm.com>

When the Memory Tagging Extension is enabled, two pages are identical
only if both their data and tags are identical.

Make the generic memcmp_pages() a __weak function and add an
arm64-specific implementation which returns non-zero if any of the two
pages contain valid MTE tags (PG_mte_tagged set). There isn't much
benefit in comparing the tags of two pages since these are normally used
for heap allocations and likely to differ anyway.

Co-developed-by: Vincenzo Frascino <vincenzo.frascino@arm.com>
Signed-off-by: Vincenzo Frascino <vincenzo.frascino@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
---

Notes:
    v4:
    - Remove page tag comparison. This is not very useful to detect
      identical pages as long as set_pte_at() can zero the tags on a page
      without copy-on-write if mapped with PROT_MTE. This can be improved
      if a real case appears but it's unlikely for heap pages to be
      identical across multiple processes.
    - Move the memcmp_pages() function to mte.c.

 arch/arm64/kernel/mte.c | 26 ++++++++++++++++++++++++++
 mm/util.c               |  2 +-
 2 files changed, 27 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/kernel/mte.c b/arch/arm64/kernel/mte.c
index 65a2f8490d18..da2d70178a4b 100644
--- a/arch/arm64/kernel/mte.c
+++ b/arch/arm64/kernel/mte.c
@@ -5,6 +5,7 @@
 
 #include <linux/bitops.h>
 #include <linux/mm.h>
+#include <linux/string.h>
 #include <linux/thread_info.h>
 
 #include <asm/cpufeature.h>
@@ -23,6 +24,31 @@ void mte_sync_tags(pte_t *ptep, pte_t pte)
 	mte_clear_page_tags(page_address(page), page_size(page));
 }
 
+int memcmp_pages(struct page *page1, struct page *page2)
+{
+	char *addr1, *addr2;
+	int ret;
+
+	addr1 = page_address(page1);
+	addr2 = page_address(page2);
+	ret = memcmp(addr1, addr2, PAGE_SIZE);
+
+	if (!system_supports_mte() || ret)
+		return ret;
+
+	/*
+	 * If the page content is identical but at least one of the pages is
+	 * tagged, return non-zero to avoid KSM merging. If only one of the
+	 * pages is tagged, set_pte_at() may zero or change the tags of the
+	 * other page via mte_sync_tags().
+	 */
+	if (test_bit(PG_mte_tagged, &page1->flags) ||
+	    test_bit(PG_mte_tagged, &page2->flags))
+		return addr1 != addr2;
+
+	return ret;
+}
+
 void flush_mte_state(void)
 {
 	if (!system_supports_mte())
diff --git a/mm/util.c b/mm/util.c
index 988d11e6c17c..662fb3da6d01 100644
--- a/mm/util.c
+++ b/mm/util.c
@@ -899,7 +899,7 @@ int get_cmdline(struct task_struct *task, char *buffer, int buflen)
 	return res;
 }
 
-int memcmp_pages(struct page *page1, struct page *page2)
+int __weak memcmp_pages(struct page *page1, struct page *page2)
 {
 	char *addr1, *addr2;
 	int ret;

WARNING: multiple messages have this Message-ID (diff)
From: Catalin Marinas <catalin.marinas@arm.com>
To: linux-arm-kernel@lists.infradead.org
Cc: linux-arch@vger.kernel.org, Szabolcs Nagy <szabolcs.nagy@arm.com>,
	Andrey Konovalov <andreyknvl@google.com>,
	Kevin Brodsky <kevin.brodsky@arm.com>,
	Peter Collingbourne <pcc@google.com>,
	linux-mm@kvack.org, Vincenzo Frascino <vincenzo.frascino@arm.com>,
	Will Deacon <will@kernel.org>,
	Dave P Martin <Dave.Martin@arm.com>
Subject: [PATCH v4 09/26] arm64: mte: Tags-aware aware memcmp_pages() implementation
Date: Fri, 15 May 2020 18:15:55 +0100	[thread overview]
Message-ID: <20200515171612.1020-10-catalin.marinas@arm.com> (raw)
In-Reply-To: <20200515171612.1020-1-catalin.marinas@arm.com>

When the Memory Tagging Extension is enabled, two pages are identical
only if both their data and tags are identical.

Make the generic memcmp_pages() a __weak function and add an
arm64-specific implementation which returns non-zero if any of the two
pages contain valid MTE tags (PG_mte_tagged set). There isn't much
benefit in comparing the tags of two pages since these are normally used
for heap allocations and likely to differ anyway.

Co-developed-by: Vincenzo Frascino <vincenzo.frascino@arm.com>
Signed-off-by: Vincenzo Frascino <vincenzo.frascino@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
---

Notes:
    v4:
    - Remove page tag comparison. This is not very useful to detect
      identical pages as long as set_pte_at() can zero the tags on a page
      without copy-on-write if mapped with PROT_MTE. This can be improved
      if a real case appears but it's unlikely for heap pages to be
      identical across multiple processes.
    - Move the memcmp_pages() function to mte.c.

 arch/arm64/kernel/mte.c | 26 ++++++++++++++++++++++++++
 mm/util.c               |  2 +-
 2 files changed, 27 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/kernel/mte.c b/arch/arm64/kernel/mte.c
index 65a2f8490d18..da2d70178a4b 100644
--- a/arch/arm64/kernel/mte.c
+++ b/arch/arm64/kernel/mte.c
@@ -5,6 +5,7 @@
 
 #include <linux/bitops.h>
 #include <linux/mm.h>
+#include <linux/string.h>
 #include <linux/thread_info.h>
 
 #include <asm/cpufeature.h>
@@ -23,6 +24,31 @@ void mte_sync_tags(pte_t *ptep, pte_t pte)
 	mte_clear_page_tags(page_address(page), page_size(page));
 }
 
+int memcmp_pages(struct page *page1, struct page *page2)
+{
+	char *addr1, *addr2;
+	int ret;
+
+	addr1 = page_address(page1);
+	addr2 = page_address(page2);
+	ret = memcmp(addr1, addr2, PAGE_SIZE);
+
+	if (!system_supports_mte() || ret)
+		return ret;
+
+	/*
+	 * If the page content is identical but at least one of the pages is
+	 * tagged, return non-zero to avoid KSM merging. If only one of the
+	 * pages is tagged, set_pte_at() may zero or change the tags of the
+	 * other page via mte_sync_tags().
+	 */
+	if (test_bit(PG_mte_tagged, &page1->flags) ||
+	    test_bit(PG_mte_tagged, &page2->flags))
+		return addr1 != addr2;
+
+	return ret;
+}
+
 void flush_mte_state(void)
 {
 	if (!system_supports_mte())
diff --git a/mm/util.c b/mm/util.c
index 988d11e6c17c..662fb3da6d01 100644
--- a/mm/util.c
+++ b/mm/util.c
@@ -899,7 +899,7 @@ int get_cmdline(struct task_struct *task, char *buffer, int buflen)
 	return res;
 }
 
-int memcmp_pages(struct page *page1, struct page *page2)
+int __weak memcmp_pages(struct page *page1, struct page *page2)
 {
 	char *addr1, *addr2;
 	int ret;

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

  parent reply	other threads:[~2020-05-15 17:16 UTC|newest]

Thread overview: 123+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-15 17:15 [PATCH v4 00/26] arm64: Memory Tagging Extension user-space support Catalin Marinas
2020-05-15 17:15 ` Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 01/26] arm64: mte: system register definitions Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 02/26] arm64: mte: CPU feature detection and initial sysreg configuration Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 03/26] arm64: mte: Use Normal Tagged attributes for the linear map Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 04/26] arm64: mte: Add specific SIGSEGV codes Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 05/26] arm64: mte: Handle synchronous and asynchronous tag check faults Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 06/26] mm: Add PG_ARCH_2 page flag Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 07/26] arm64: mte: Clear the tags when a page is mapped in user-space with PROT_MTE Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 08/26] arm64: mte: Tags-aware copy_page() implementation Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:15 ` Catalin Marinas [this message]
2020-05-15 17:15   ` [PATCH v4 09/26] arm64: mte: Tags-aware aware memcmp_pages() implementation Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 10/26] mm: Introduce arch_calc_vm_flag_bits() Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 11/26] arm64: mte: Add PROT_MTE support to mmap() and mprotect() Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-27 18:57   ` Peter Collingbourne
2020-05-27 18:57     ` Peter Collingbourne
2020-05-27 18:57     ` Peter Collingbourne
2020-05-27 18:57     ` Peter Collingbourne
2020-05-28  9:14     ` Catalin Marinas
2020-05-28  9:14       ` Catalin Marinas
2020-05-28 11:05       ` Szabolcs Nagy
2020-05-28 11:05         ` Szabolcs Nagy
2020-05-28 11:05         ` Szabolcs Nagy
2020-05-28 16:34         ` Catalin Marinas
2020-05-28 16:34           ` Catalin Marinas
2020-05-28 18:35           ` Evgenii Stepanov
2020-05-28 18:35             ` Evgenii Stepanov
2020-05-29 11:19             ` Catalin Marinas
2020-05-29 11:19               ` Catalin Marinas
2020-06-01  8:55           ` Dave Martin
2020-06-01  8:55             ` Dave Martin
2020-06-01 14:45             ` Catalin Marinas
2020-06-01 14:45               ` Catalin Marinas
2020-06-01 15:04               ` Dave Martin
2020-06-01 15:04                 ` Dave Martin
2020-05-15 17:15 ` [PATCH v4 12/26] mm: Introduce arch_validate_flags() Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:15 ` [PATCH v4 13/26] arm64: mte: Validate the PROT_MTE request via arch_validate_flags() Catalin Marinas
2020-05-15 17:15   ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 14/26] mm: Allow arm64 mmap(PROT_MTE) on RAM-based files Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 15/26] arm64: mte: Allow user control of the tag check mode via prctl() Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-27  7:46   ` Will Deacon
2020-05-27  7:46     ` Will Deacon
2020-05-27  8:32     ` Dave Martin
2020-05-27  8:32       ` Dave Martin
2020-05-27  8:48       ` Will Deacon
2020-05-27  8:48         ` Will Deacon
2020-05-27 11:16       ` Catalin Marinas
2020-05-27 11:16         ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 16/26] arm64: mte: Allow user control of the generated random tags " Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 17/26] arm64: mte: Restore the GCR_EL1 register after a suspend Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 18/26] arm64: mte: Add PTRACE_{PEEK,POKE}MTETAGS support Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-29 21:25   ` Luis Machado
2020-05-29 21:25     ` Luis Machado
2020-06-01 12:07     ` Catalin Marinas
2020-06-01 12:07       ` Catalin Marinas
2020-06-01 15:17       ` Luis Machado
2020-06-01 15:17         ` Luis Machado
2020-06-01 16:33         ` Catalin Marinas
2020-06-01 16:33           ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 19/26] fs: Handle intra-page faults in copy_mount_options() Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 20/26] mm: Add arch hooks for saving/restoring tags Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 21/26] arm64: mte: Enable swap of tagged pages Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 22/26] arm64: mte: Save tags when hibernating Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 23/26] arm64: mte: Check the DT memory nodes for MTE support Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 24/26] arm64: mte: Introduce early param to disable " Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-18 11:26   ` Vladimir Murzin
2020-05-18 11:26     ` Vladimir Murzin
2020-05-18 11:31     ` Will Deacon
2020-05-18 11:31       ` Will Deacon
2020-05-18 17:20       ` Catalin Marinas
2020-05-18 17:20         ` Catalin Marinas
2020-05-22  5:57         ` Patrick Daly
2020-05-22  5:57           ` Patrick Daly
2020-05-22 10:37           ` Catalin Marinas
2020-05-22 10:37             ` Catalin Marinas
2020-05-27  2:11             ` Patrick Daly
2020-05-27  2:11               ` Patrick Daly
2020-05-27  2:11               ` Patrick Daly
2020-05-27  9:55               ` Will Deacon
2020-05-27  9:55                 ` Will Deacon
2020-05-27 10:37                 ` Szabolcs Nagy
2020-05-27 10:37                   ` Szabolcs Nagy
2020-05-27 11:12                 ` Catalin Marinas
2020-05-27 11:12                   ` Catalin Marinas
2020-05-19 16:14     ` Catalin Marinas
2020-05-19 16:14       ` Catalin Marinas
2021-01-21 19:37   ` Andrey Konovalov
2021-01-21 19:37     ` Andrey Konovalov
2021-01-21 19:37     ` Andrey Konovalov
2021-01-22  2:03     ` Andrey Konovalov
2021-01-22  2:03       ` Andrey Konovalov
2021-01-22  2:03       ` Andrey Konovalov
2021-01-22 14:41     ` Catalin Marinas
2021-01-22 14:41       ` Catalin Marinas
2021-01-22 17:28       ` Andrey Konovalov
2021-01-22 17:28         ` Andrey Konovalov
2021-01-22 17:28         ` Andrey Konovalov
2020-05-15 17:16 ` [PATCH v4 25/26] arm64: mte: Kconfig entry Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas
2020-05-15 17:16 ` [PATCH v4 26/26] arm64: mte: Add Memory Tagging Extension documentation Catalin Marinas
2020-05-15 17:16   ` Catalin Marinas

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=20200515171612.1020-10-catalin.marinas@arm.com \
    --to=catalin.marinas@arm.com \
    --cc=Dave.Martin@arm.com \
    --cc=andreyknvl@google.com \
    --cc=kevin.brodsky@arm.com \
    --cc=linux-arch@vger.kernel.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-mm@kvack.org \
    --cc=pcc@google.com \
    --cc=szabolcs.nagy@arm.com \
    --cc=vincenzo.frascino@arm.com \
    --cc=will@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 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.