Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH] scsi/mvsas/mv_init.c: Use dma_zalloc_coherent
To:     Souptick Joarder <jrdr.linux@gmail.com>
References: <5bf56248.1c69fb81.21a6e.5a72@mx.google.com>
 <CAJr6mf11gp9iV=NiFEFyo7kSZfCfMsEySdijem_PhMFEAwqucA@mail.gmail.com>
 <CAJr6mf2kr4BmCWZU0=EwXURWUZWRHHLifi_2aiUrJSCjnWusZg@mail.gmail.com>
 <CAJr6mf29hJft=rurUK8EJRQ1MS_3_MXb7f-zX0vyXUrQNLcWYA@mail.gmail.com>
 <ff090818-2f5f-1fa0-2eff-6988e1660345@huawei.com>
 <CAJr6mf1ebw16j0AZSDDxnK2t_=8wdzZBm=X_3DW=zP+XRbRFsw@mail.gmail.com>
 <8af8ef0a-add1-1010-6ef3-7e5c6ff51aa9@huawei.com>
 <CAFqt6zbZaOA9+SX8CwYobBN_3hhhWOWkcBJ_qdgEuELoOBU5=w@mail.gmail.com>
CC:     Sabyasachi Gupta <sabyasachi.linux@gmail.com>,
        James Bottomley <jejb@linux.vnet.ibm.com>,
        "Martin K. Petersen" <martin.petersen@oracle.com>,
        Johannes Thumshirn <jthumshirn@suse.de>,
        linux-scsi <linux-scsi@vger.kernel.org>,
        <linux-kernel@vger.kernel.org>,
        Brajeswar Ghosh <brajeswar.linux@gmail.com>
From:   John Garry <john.garry@huawei.com>
Message-ID: <1e3d2653-67b5-07a2-37c3-d3e39532cbcb@huawei.com>
Date:   Mon, 7 Jan 2019 13:50:29 +0000
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:45.0) Gecko/20100101
 Thunderbird/45.3.0
MIME-Version: 1.0
In-Reply-To: <CAFqt6zbZaOA9+SX8CwYobBN_3hhhWOWkcBJ_qdgEuELoOBU5=w@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"; format=flowed
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 04/01/2019 19:49, Souptick Joarder wrote:
> On Fri, Jan 4, 2019 at 9:34 PM John Garry <john.garry@huawei.com> wrote:
>>
>> On 04/01/2019 15:11, Sabyasachi Gupta wrote:
>>> On Fri, Jan 4, 2019 at 6:43 PM John Garry <john.garry@huawei.com> wrote:
>>>>
>>>> On 04/01/2019 12:48, Sabyasachi Gupta wrote:
>>>>> On Wed, Dec 19, 2018 at 6:49 PM Sabyasachi Gupta
>>>>> <sabyasachi.linux@gmail.com> wrote:
>>>>>>
>>>>>> On Sat, Dec 1, 2018 at 6:40 PM Sabyasachi Gupta
>>>>>> <sabyasachi.linux@gmail.com> wrote:
>>>>>>>
>>>>>>> On Wed, Nov 21, 2018 at 7:18 PM Sabyasachi Gupta
>>>>>>> <sabyasachi.linux@gmail.com> wrote:
>>>>>>>>
>>>>>>>> Replace dma_alloc_coherent + memset with dma_zalloc_coherent
>>>>>>>>
>>>>
>>>> If you're going to make this change, then how about change these to the
>>>> managed version, so that we can avoid the explicit free'ing at driver
>>>> removal?
>>>
>>> I can't get it
>>
>> Please see dmam_alloc_coherent(). You can set __GFP_ZERO in the gfp
>> argument to memset(0).
>
> and dma_zalloc_coherent() did the same. So both are same right ? No ?
>

Firstly I would to say that I am not so keen on changes like this since 
(I assume) it will not be tested at all.

However, having said that, the reason to use dmam_alloc_coherent() is 
that you can also drop the dma_free_coherent() in mvs_free(). But 
changes like that would actually require some code analysis effort...

John

>>
>> I would say that this is a more useful change.
>>
>>>
>>>>
>>>>>>>> Signed-off-by: Sabyasachi Gupta <sabyasachi.linux@gmail.com>
>>>>>>>
>>>>>>> Any comment on this patch?
>>>>>>
>>>>>> Any comment on this patch?
>>>>>
>>>>> Any comment on this patch?
>>>>>
>>>>>>
>>>>>>>
>>>>>>>> ---
>>>>>>>>  drivers/scsi/mvsas/mv_init.c | 12 ++++--------
>>>>>>>>  1 file changed, 4 insertions(+), 8 deletions(-)
>>>>>>>>
>>>>>>>> diff --git a/drivers/scsi/mvsas/mv_init.c b/drivers/scsi/mvsas/mv_init.c
>>>>>>>> index 3ac3437..495bddb 100644
>>>>>>>> --- a/drivers/scsi/mvsas/mv_init.c
>>>>>>>> +++ b/drivers/scsi/mvsas/mv_init.c
>>>>>>>> @@ -253,33 +253,29 @@ static int mvs_alloc(struct mvs_info *mvi, struct Scsi_Host *shost)
>>>>>>>>         /*
>>>>>>>>          * alloc and init our DMA areas
>>>>>>>>          */
>>>>>>>> -       mvi->tx = dma_alloc_coherent(mvi->dev,
>>>>>>>> +       mvi->tx = dma_zalloc_coherent(mvi->dev,
>>>>>>>>                                      sizeof(*mvi->tx) * MVS_CHIP_SLOT_SZ,
>>>>>>>>                                      &mvi->tx_dma, GFP_KERNEL);
>>>>
>>>> I'm guessing that this does not pass checkpatch with --strict option.
>>>>
>>>> Thanks,
>>>> John
>>>
>>> I have not not checked with --strict option
>>
>> It may warn that you're not maintaining alignment.
>>
>>>
>>>>
>>>>>>>>         if (!mvi->tx)
>>>>>>>>                 goto err_out;
>>>>>>>> -       memset(mvi->tx, 0, sizeof(*mvi->tx) * MVS_CHIP_SLOT_SZ);
>>>>>>>> -       mvi->rx_fis = dma_alloc_coherent(mvi->dev, MVS_RX_FISL_SZ,
>>>>>>>> +       mvi->rx_fis = dma_zalloc_coherent(mvi->dev, MVS_RX_FISL_SZ,
>>>>>>>>                                          &mvi->rx_fis_dma, GFP_KERNEL);
>>>>>>>>         if (!mvi->rx_fis)
>>>>>>>>                 goto err_out;
>>>>>>>> -       memset(mvi->rx_fis, 0, MVS_RX_FISL_SZ);
>>>>>>>>
>>>>>>>> -       mvi->rx = dma_alloc_coherent(mvi->dev,
>>>>>>>> +       mvi->rx = dma_zalloc_coherent(mvi->dev,
>>>>>>>>                                      sizeof(*mvi->rx) * (MVS_RX_RING_SZ + 1),
>>>>>>>>                                      &mvi->rx_dma, GFP_KERNEL);
>>>>>>>>         if (!mvi->rx)
>>>>>>>>                 goto err_out;
>>>>>>>> -       memset(mvi->rx, 0, sizeof(*mvi->rx) * (MVS_RX_RING_SZ + 1));
>>>>>>>>         mvi->rx[0] = cpu_to_le32(0xfff);
>>>>>>>>         mvi->rx_cons = 0xfff;
>>>>>>>>
>>>>>>>> -       mvi->slot = dma_alloc_coherent(mvi->dev,
>>>>>>>> +       mvi->slot = dma_zalloc_coherent(mvi->dev,
>>>>>>>>                                        sizeof(*mvi->slot) * slot_nr,
>>>>>>>>                                        &mvi->slot_dma, GFP_KERNEL);
>>>>>>>>         if (!mvi->slot)
>>>>>>>>                 goto err_out;
>>>>>>>> -       memset(mvi->slot, 0, sizeof(*mvi->slot) * slot_nr);
>>>>>>>>
>>>>>>>>         mvi->bulk_buffer = dma_alloc_coherent(mvi->dev,
>>>>>>>>                                        TRASH_BUCKET_SIZE,
>>>>>>>> --
>>>>>>>> 2.7.4
>>>>>>>>
>>>>>
>>>>> .
>>>>>
>>>>
>>>>
>>>
>>> .
>>>
>>
>>
>
> .
>