linux-kernel - [PATCH 2/4] crypto: hisilicon/qm - add concurrency protection for variable err

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: <20251025101258.2793179-3-huangchenghai2@huawei.com>
Date: Sat, 25 Oct 2025 18:12:56 +0800
From: Chenghai Huang <huangchenghai2@...wei.com>
To: <herbert@...dor.apana.org.au>, <davem@...emloft.net>
CC: <linux-kernel@...r.kernel.org>, <linux-crypto@...r.kernel.org>,
	<linuxarm@...neuler.org>, <liulongfang@...wei.com>, <qianweili@...wei.com>,
	<wangzhou1@...ilicon.com>, <fanghao11@...wei.com>, <nieweiqiang@...wei.com>
Subject: [PATCH 2/4] crypto: hisilicon/qm - add concurrency protection for variable err_threshold

From: nieweiqiang <nieweiqiang@...wei.com>

The isolate_strategy_store function is not protected
by a lock. If sysfs operations and functions that depend
on the err_threshold variable,such as qm_hw_err_isolate(),
execute concurrently, the outcome will be unpredictable.
Therefore, concurrency protection should be added for
the err_threshold variable.

Signed-off-by: nieweiqiang <nieweiqiang@...wei.com>
Signed-off-by: Chenghai Huang <huangchenghai2@...wei.com>
---
 drivers/crypto/hisilicon/qm.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/crypto/hisilicon/qm.c b/drivers/crypto/hisilicon/qm.c
index 8274da8b37ca..c58f67567c12 100644
--- a/drivers/crypto/hisilicon/qm.c
+++ b/drivers/crypto/hisilicon/qm.c
@@ -2655,10 +2655,10 @@ static int qm_hw_err_isolate(struct hisi_qm *qm)
 		}
 	}
 	list_add(&hw_err->list, &isolate->qm_hw_errs);
-	mutex_unlock(&isolate->isolate_lock);
 
 	if (count >= isolate->err_threshold)
 		isolate->is_isolate = true;
+	mutex_unlock(&isolate->isolate_lock);
 
 	return 0;
 }
@@ -2667,12 +2667,10 @@ static void qm_hw_err_destroy(struct hisi_qm *qm)
 {
 	struct qm_hw_err *err, *tmp;
 
-	mutex_lock(&qm->isolate_data.isolate_lock);
 	list_for_each_entry_safe(err, tmp, &qm->isolate_data.qm_hw_errs, list) {
 		list_del(&err->list);
 		kfree(err);
 	}
-	mutex_unlock(&qm->isolate_data.isolate_lock);
 }
 
 static enum uacce_dev_state hisi_qm_get_isolate_state(struct uacce_device *uacce)
@@ -2700,10 +2698,12 @@ static int hisi_qm_isolate_threshold_write(struct uacce_device *uacce, u32 num)
 	if (qm->isolate_data.is_isolate)
 		return -EPERM;
 
+	mutex_lock(&qm->isolate_data.isolate_lock);
 	qm->isolate_data.err_threshold = num;
 
 	/* After the policy is updated, need to reset the hardware err list */
 	qm_hw_err_destroy(qm);
+	mutex_unlock(&qm->isolate_data.isolate_lock);
 
 	return 0;
 }
@@ -2740,7 +2740,10 @@ static void qm_remove_uacce(struct hisi_qm *qm)
 	struct uacce_device *uacce = qm->uacce;
 
 	if (qm->use_sva) {
+		mutex_lock(&qm->isolate_data.isolate_lock);
 		qm_hw_err_destroy(qm);
+		mutex_unlock(&qm->isolate_data.isolate_lock);
+
 		uacce_remove(uacce);
 		qm->uacce = NULL;
 	}
-- 
2.33.0