linux-kernel - Re: [PATCH v2] mtd: rawnand: cafe: fix drivers probe/remove methods

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 for Android: free password hash cracker in your pocket

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

Message-ID: <2c788d15-e771-377a-2dd2-796c86cf6f8a@wanadoo.fr>
Date:   Fri, 20 May 2022 19:09:12 +0200
From:   Marion & Christophe JAILLET <christophe.jaillet@...adoo.fr>
To:     Peng Wu <wupeng58@...wei.com>, miquel.raynal@...tlin.com,
        richard@....at, vigneshr@...com
Cc:     linux-mtd@...ts.infradead.org, linux-kernel@...r.kernel.org,
        stable@...r.kernel.org, liwei391@...wei.com,
        Kernel Janitors <kernel-janitors@...r.kernel.org>
Subject: Re: [PATCH v2] mtd: rawnand: cafe: fix drivers probe/remove methods



Le 20/05/2022 à 18:53, Christophe JAILLET a écrit :
> Le 20/05/2022 à 10:44, Peng Wu a écrit :
>> Driver should call pci_disable_device() if it returns from
>> cafe_nand_probe() with error.
>>
>> Meanwhile, the driver calls pci_enable_device() in
>> cafe_nand_probe(), but never calls pci_disable_device()
>> during removal.
>>
>> Signed-off-by: Peng Wu <wupeng58@...wei.com>
>> ---
>> v2:
>> - fix the subject prefix with "mtd: ranwnand: cafe:"
>> ---
>>   drivers/mtd/nand/raw/cafe_nand.c | 9 +++++++--
>>   1 file changed, 7 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/mtd/nand/raw/cafe_nand.c 
>> b/drivers/mtd/nand/raw/cafe_nand.c
>> index 9dbf031716a6..af119e376352 100644
>> --- a/drivers/mtd/nand/raw/cafe_nand.c
>> +++ b/drivers/mtd/nand/raw/cafe_nand.c
>> @@ -679,8 +679,10 @@ static int cafe_nand_probe(struct pci_dev *pdev,
>>       pci_set_master(pdev);
>>       cafe = kzalloc(sizeof(*cafe), GFP_KERNEL);
>> -    if (!cafe)
>> -        return  -ENOMEM;
>> +    if (!cafe) {
>> +        err = -ENOMEM;
>> +        goto out_disable_device;
>> +    }
>>       mtd = nand_to_mtd(&cafe->nand);
>>       mtd->dev.parent = &pdev->dev;
>> @@ -801,6 +803,8 @@ static int cafe_nand_probe(struct pci_dev *pdev,
>>       pci_iounmap(pdev, cafe->mmio);
>>    out_free_mtd:
>>       kfree(cafe);
>> + out_disable_device:
>> +    pci_disable_device(pdev);
>>    out:
>>       return err;
>>   }
>> @@ -822,6 +826,7 @@ static void cafe_nand_remove(struct pci_dev *pdev)
>>       pci_iounmap(pdev, cafe->mmio);
>>       dma_free_coherent(&cafe->pdev->dev, 2112, cafe->dmabuf, 
>> cafe->dmaaddr);
> 
> Hi,
> 
> Not related to this patch , but I wonder if this dma_free_coherent() is 
> needed.
> It is already part of cafe_nand_detach_chip() which is a .detach_chip() 
> function.
> 
> Is this .detach_chip() function already called (by nand_cleanup()) 
> somewhere in the removal process?
> 

In fact the "somewhere" is just 3 lines above the dma_free_coherent() 
call. So I definitively think that this line should be axed.

CJ


> CJ
> 
> 
>>       kfree(cafe);
>> +    pci_disable_device(pdev);
>>   }
>>   static const struct pci_device_id cafe_nand_tbl[] = {
> 
>