Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <7f7aab67-69a3-b2ca-0bed-30cb6a5bca16@quicinc.com>
Date:   Wed, 2 Aug 2023 19:19:33 +0530
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:102.0) Gecko/20100101
 Thunderbird/102.9.0
Subject: Re: [PATCH v3] misc: fastrpc: Fix incorrect DMA mapping unmap request
Content-Language: en-US
To:     Srinivas Kandagatla <srinivas.kandagatla@linaro.org>,
        <linux-arm-msm@vger.kernel.org>
CC:     <ekangupt@qti.qualcomm.com>, <gregkh@linuxfoundation.org>,
        <linux-kernel@vger.kernel.org>,
        <fastrpc.upstream@qti.qualcomm.com>, stable <stable@kernel.org>
References: <1690953032-17070-1-git-send-email-quic_ekangupt@quicinc.com>
 <4059684f-2e44-ccd6-4c65-5137cc76492c@linaro.org>
From:   Ekansh Gupta <quic_ekangupt@quicinc.com>
In-Reply-To: <4059684f-2e44-ccd6-4c65-5137cc76492c@linaro.org>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk


On 8/2/2023 7:13 PM, Srinivas Kandagatla wrote:
> 
> 
> On 02/08/2023 06:10, Ekansh Gupta wrote:
>> Scatterlist table is obtained during map create request and the same
>> table is used for DMA mapping unmap. In case there is any failure
>> while getting the sg_table, ERR_PTR is returned instead of sg_table.
>>
>> When the map is getting freed, there is only a non-NULL check of
>> sg_table which will also be true in case failure was returned instead
>> of sg_table. This would result in improper unmap request. Add proper
>> check before setting map table to avoid bad unmap request.
>>
>> Fixes: c68cfb718c8f ("misc: fastrpc: Add support for context Invoke 
>> method")
>> Cc: stable <stable@kernel.org>
>> Signed-off-by: Ekansh Gupta <quic_ekangupt@quicinc.com>
>> ---
>> Changes in v2:
>>    - Added fixes information to commit text
>> Changes in v3:
>>    - Set map->table only if attachment for successful
>>
>>   drivers/misc/fastrpc.c | 8 +++++---
>>   1 file changed, 5 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
>> index 9666d28..de7c812 100644
>> --- a/drivers/misc/fastrpc.c
>> +++ b/drivers/misc/fastrpc.c
>> @@ -756,6 +756,7 @@ static int fastrpc_map_create(struct fastrpc_user 
>> *fl, int fd,
>>   {
>>       struct fastrpc_session_ctx *sess = fl->sctx;
>>       struct fastrpc_map *map = NULL;
>> +    struct sg_table *table;
>>       int err = 0;
>>       if (!fastrpc_map_lookup(fl, fd, ppmap, true))
>> @@ -783,11 +784,12 @@ static int fastrpc_map_create(struct 
>> fastrpc_user *fl, int fd,
>>           goto attach_err;
>>       }
>> -    map->table = dma_buf_map_attachment_unlocked(map->attach, 
>> DMA_BIDIRECTIONAL);
>> -    if (IS_ERR(map->table)) {
>> -        err = PTR_ERR(map->table);
>> +    table = dma_buf_map_attachment(map->attach, DMA_BIDIRECTIONAL);
> 
> Any reason why dma_buf_map_attachment_unlocked changed to 
> dma_buf_map_attachment?
This is a mistake from my end. My local workspace had older version due 
to which the function also got reverted. I will fix this in new patch. 
Apologies for the confusion.
> 
> --srini
>> +    if (IS_ERR(table)) {
>> +        err = PTR_ERR(table);
>>           goto map_err;
>>       }
>> +    map->table = table;
>>       if (attr & FASTRPC_ATTR_SECUREMAP) {
>>           map->phys = sg_phys(map->table->sgl);