Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Dinghao Liu <dinghao.liu@zju.edu.cn>
To:     dinghao.liu@zju.edu.cn, kjlu@umn.edu
Cc:     Oded Gabbay <ogabbay@kernel.org>, Arnd Bergmann <arnd@arndb.de>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Tomer Tayar <ttayar@habana.ai>,
        Omer Shpigelman <oshpigelman@habana.ai>,
        Ofir Bitton <obitton@habana.ai>,
        Moti Haimovski <mhaimovski@habana.ai>,
        linux-kernel@vger.kernel.org
Subject: [PATCH] habanalabs: Fix memleak in hl_device_reset
Date:   Sat, 26 Dec 2020 15:27:14 +0800
Message-Id: <20201226072717.32001-1-dinghao.liu@zju.edu.cn>
Precedence: bulk

When kzalloc() fails, we should execute hl_mmu_fini()
to release the MMU module. It's the same when
hl_ctx_init() fails.

Signed-off-by: Dinghao Liu <dinghao.liu@zju.edu.cn>
---
 drivers/misc/habanalabs/common/device.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/misc/habanalabs/common/device.c b/drivers/misc/habanalabs/common/device.c
index 5871162a8442..e5028890ead1 100644
--- a/drivers/misc/habanalabs/common/device.c
+++ b/drivers/misc/habanalabs/common/device.c
@@ -1092,6 +1092,7 @@ int hl_device_reset(struct hl_device *hdev, bool hard_reset,
 						GFP_KERNEL);
 		if (!hdev->kernel_ctx) {
 			rc = -ENOMEM;
+			hl_mmu_fini(hdev);
 			goto out_err;
 		}
 
@@ -1103,6 +1104,7 @@ int hl_device_reset(struct hl_device *hdev, bool hard_reset,
 				"failed to init kernel ctx in hard reset\n");
 			kfree(hdev->kernel_ctx);
 			hdev->kernel_ctx = NULL;
+			hl_mmu_fini(hdev);
 			goto out_err;
 		}
 	}
-- 
2.17.1