linux-kernel - [PATCH 1/3] memory-failure: Introduce memory failure notifier

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20220520070648.1794132-2-pizhenwei@bytedance.com>
Date:   Fri, 20 May 2022 15:06:46 +0800
From:   zhenwei pi <pizhenwei@...edance.com>
To:     akpm@...ux-foundation.org, naoya.horiguchi@....com, mst@...hat.com,
        david@...hat.com
Cc:     linux-mm@...ck.org, linux-kernel@...r.kernel.org,
        jasowang@...hat.com, virtualization@...ts.linux-foundation.org,
        pbonzini@...hat.com, peterx@...hat.com, qemu-devel@...gnu.org,
        zhenwei pi <pizhenwei@...edance.com>
Subject: [PATCH 1/3] memory-failure: Introduce memory failure notifier

Introduce memory failure notifier, once hardware memory failure
occurs, after the kernel handles the corrupted page successfully,
someone who registered this chain gets noticed of the corrupted PFN.

Signed-off-by: zhenwei pi <pizhenwei@...edance.com>
---
 include/linux/mm.h  |  2 ++
 mm/memory-failure.c | 33 +++++++++++++++++++++++++++++++++
 2 files changed, 35 insertions(+)

diff --git a/include/linux/mm.h b/include/linux/mm.h
index 9f44254af8ce..665873c2788c 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -3197,6 +3197,8 @@ extern int sysctl_memory_failure_recovery;
 extern void shake_page(struct page *p);
 extern atomic_long_t num_poisoned_pages __read_mostly;
 extern int soft_offline_page(unsigned long pfn, int flags);
+extern int register_memory_failure_notifier(struct notifier_block *nb);
+extern int unregister_memory_failure_notifier(struct notifier_block *nb);
 #ifdef CONFIG_MEMORY_FAILURE
 extern int __get_huge_page_for_hwpoison(unsigned long pfn, int flags);
 #else
diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index 2d590cba412c..95c218bb0a37 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -68,6 +68,35 @@ int sysctl_memory_failure_recovery __read_mostly = 1;
 
 atomic_long_t num_poisoned_pages __read_mostly = ATOMIC_LONG_INIT(0);
 
+static BLOCKING_NOTIFIER_HEAD(mf_notifier_list);
+
+/**
+ * register_memory_failure_notifier - Register function to be called if a
+ *                                    corrupted page gets handled successfully
+ * @nb: Info about notifier function to be called
+ *
+ * Currently always returns zero, as blocking_notifier_chain_register()
+ * always returns zero.
+ */
+int register_memory_failure_notifier(struct notifier_block *nb)
+{
+	return blocking_notifier_chain_register(&mf_notifier_list, nb);
+}
+EXPORT_SYMBOL_GPL(register_memory_failure_notifier);
+
+/**
+ * unregister_memory_failure_notifier - Unregister previously registered
+ *                                      memory failure notifier
+ * @nb: Hook to be unregistered
+ *
+ * Returns zero on success, or %-ENOENT on failure.
+ */
+int unregister_memory_failure_notifier(struct notifier_block *nb)
+{
+	return blocking_notifier_chain_unregister(&mf_notifier_list, nb);
+}
+EXPORT_SYMBOL_GPL(unregister_memory_failure_notifier);
+
 static bool __page_handle_poison(struct page *page)
 {
 	int ret;
@@ -1136,6 +1165,10 @@ static void action_result(unsigned long pfn, enum mf_action_page_type type,
 	num_poisoned_pages_inc();
 	pr_err("Memory failure: %#lx: recovery action for %s: %s\n",
 		pfn, action_page_types[type], action_name[result]);
+
+	/* notify the chain if we handle successfully only */
+	if (result == MF_RECOVERED)
+		blocking_notifier_call_chain(&mf_notifier_list, pfn, NULL);
 }
 
 static int page_action(struct page_state *ps, struct page *p,
-- 
2.20.1