From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <5494411D.5030705@siemens.com>
Date: Fri, 19 Dec 2014 16:15:41 +0100
From: Jan Kiszka <jan.kiszka@siemens.com>
MIME-Version: 1.0
References: <CALH4WdbO8pfsUCvUUHVtv_Rzh22-CLxAuRAZ=HzuPWOcvY7KdQ@mail.gmail.com>
 <20141218174447.GV2012@hermes.click-hack.org>
 <CALH4WdZvLBWyUavA7=UV=RoJEg+Cd3nEELjnQac3uZSkfj8B+A@mail.gmail.com>
 <20141219145755.GA7589@hermes.click-hack.org>
In-Reply-To: <20141219145755.GA7589@hermes.click-hack.org>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 7bit
Subject: Re: [Xenomai] PATCH : msi patch rtcan_peak_pci
List-Id: Discussions about the Xenomai project <xenomai.xenomai.org>
List-Unsubscribe: <http://www.xenomai.org/mailman/options/xenomai>,
 <mailto:xenomai-request@xenomai.org?subject=unsubscribe>
List-Archive: <http://www.xenomai.org/pipermail/xenomai/>
List-Post: <mailto:xenomai@xenomai.org>
List-Help: <mailto:xenomai-request@xenomai.org?subject=help>
List-Subscribe: <http://www.xenomai.org/mailman/listinfo/xenomai>,
 <mailto:xenomai-request@xenomai.org?subject=subscribe>
To: Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org>, Frederik Bayart <frederik.bayart@triphase.com>
Cc: xenomai@xenomai.org

On 2014-12-19 15:57, Gilles Chanteperdrix wrote:
> On Fri, Dec 19, 2014 at 12:20:13PM +0100, Frederik Bayart wrote:
>> On 18 December 2014 at 18:44, Gilles Chanteperdrix
>> <gilles.chanteperdrix@xenomai.org> wrote:
>>> On Thu, Dec 18, 2014 at 05:53:33PM +0100, Frederik Bayart wrote:
>>>> Hallo,
>>>>
>>>> We have added msi to the rtcan_peak_pci driver
>>>> In attachment a zip file containing the old and the new rtcan_peak_pci.c
>>>> and the patch.
>>>> Can this patch be applied the the source tree ?
>>>
>>> Please send the patch alone, preferably inline, or as an attachment.
>>> This makes review and comment easier.
>>>
>>> --
>>>                                             Gilles.
>>
>> Gilles,
>>
>> Below the patch. I used  xenomai-3-3.0-rc2.tar because I haven't used
>> git yet but I verified that the
>> driver source didn't changed in git
>> (http://git.xenomai.org/xenomai-3.git/plain/kernel/drivers/can/sja1000/rtcan_peak_pci.c)
>>
>> Frederik
>>
>> --- a/kernel/drivers/can/sja1000/rtcan_peak_pci.c    2014-10-31
>> 15:18:44.000000000 +0100
>> +++ b/kernel/drivers/can/sja1000/rtcan_peak_pci.c    2014-12-18
>> 17:44:21.799350602 +0100
>> @@ -301,6 +301,8 @@
>>      if ((ret = pci_write_config_word(pdev, 0x44, 0)))
>>      goto failure_cleanup;
>>
>> +    pci_enable_msi(pdev);
>> +
> 
> You should get pci_enable_msi return status, and...
> 
>>      if (sub_sys_id > 3) {
>>      if ((ret = rtcan_peak_pci_add_chan(pdev, CHANNEL_MASTER,
>>                         &master_dev)))
>> @@ -337,6 +339,8 @@
>>      rtcan_peak_pci_del_chan(board->slave_dev, 0);
>>      rtcan_peak_pci_del_chan(dev, 0);
>>
>> +    pci_disable_msi(pdev);
>> +
> 
> if it failed, do not use pci_disable_msi.
> 

And I suspect that - if no other magic path does this - you also need
pci_set_master. Without master mode enabled, the device will not send
any MSIs (unless it violates the PCI spec).

Jan

-- 
Siemens AG, Corporate Technology, CT RTC ITP SES-DE
Corporate Competence Center Embedded Linux