Date: Tue, 3 Jul 2012 21:27:54 +0200 (CEST) From: Thierry BULTEL Message-ID: <1012208988.15650.1341343674156.JavaMail.www@wwinf1m30> In-Reply-To: <4FEDF65A.50401@grandegger.com> References: <4FED4948.8080906@basystemes.fr> <4FED4B42.6000808@wanadoo.fr> <4FEDF65A.50401@grandegger.com> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Subject: Re: [Xenomai] RTDM driver for the Advantech 1680U CANbus Adapter Reply-To: Thierry BULTEL List-Id: Discussions about the Xenomai project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Wolfgang Grandegger Cc: Linux-CAN , xenomai@xenomai.org Hi Wolfgang, =C2=A0 Thanks for your feedback. =C2=A0 The issue is that I did not realize that the patch would not have been acce= pted as such, so I will have to invest more time on than expected ... so I will not be able to do so immediately. =C2=A0 Several comments, to your comments - I agree for the multiple devices, theoretically supported by the original= Linux driver. I intentionnally removed them, since I do not have the appro= priate hardware to test. - Linux code style. I really thought I had respected it, furthermore I took= another (ixxat) RTDM driver as a model. Ok, I can doublecheck ... - patch high-check vs plx. I do no see what you mean. If I made something w= rong it is not intentional. Can you be more specific please ? - Avantech specific ... not much but the classical vendor ID, and also, PCI= banks mapping. Again, it is just like the ixxat one, all the driver intell= igence is in the generic sja1000 code - patches inline ... Really ??? In the mail content ? But that makes huge m= ails. I can do so of course, but I do not see the point. =C2=A0 Cheers Thierry > Message du 29/06/12 20:39 > De : "Wolfgang Grandegger"=20 > A : "Thierry Bultel"=20 > Copie =C3=A0 : xenomai@xenomai.org, "Linux-CAN"=20 > Objet : Re: [Xenomai] RTDM driver for the Advantech 1680U CANbus Adapter >=20 > Hi Thierry, >=20 > On 06/29/2012 08:29 AM, Thierry Bultel wrote: > > Hi, > >=20 > > I have ported the adv_pci driver from > > http://git.toiit.sgu.ru/people/samarkinpa/public/?p=3Dadv_pci.git;a=3Dt= ree;h=3Da57c5d22415f8dbf92eeae0e05ff482839d8c2f3;hb=3Da57c5d22415f8dbf92eea= e0e05ff482839d8c2f3 >=20 > Thanks for your contribution. This driver is not yet mainline and it's > the first time that I see it. I added the Linux-CAN ML to the CC as it > might be of general interest. >=20 > > I has been quite intensively been tested and cycled for several days in > > real industrial conditions. > >=20 > > The patch for 2.6.38-8 is attached to this mail. > >=20 > > Is it of interest and does is have a chance to be integrated in the > > xenomai mainline ? >=20 > There is interest, of course. But we need a proper patch against the git > head of xenomai-2.6. At a first glance I realized the following issues: >=20 > - The patch does high-check and remove the plx_pci driver, please fix. >=20 > - What is Advantech specific in that driver? It looks pretty generic to > me and it could easily support other similar devices. >=20 > - The Linux driver mentioned above does support much more Advantech > devices. Any chance to support them in your driver as well. >=20 > - Please send patches *inline*. >=20 > - Please respect the Linux coding style (check CodingStyle and use > checkpatch.pl). >=20 > Thanks in advance. >=20 > Wolfgang. >