From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from proofpoint-cluster.metrocast.net ([65.175.128.136]:59240 "EHLO proofpoint-cluster.metrocast.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752349Ab1AJMHK (ORCPT ); Mon, 10 Jan 2011 07:07:10 -0500 Subject: Re: Enable IR on hdpvr From: Andy Walls To: Jarod Wilson Cc: Jason Gauthier , Linux Media Mailing List , Janne Grunau In-Reply-To: <8AFBEFD7-69E3-4E71-B155-EA773C2FED43@wilsonet.com> References: <65DE7931C559BF4DBEE42C3F8246249A0B686EB0@V-EXMAILBOX.ctg.com> <8AFBEFD7-69E3-4E71-B155-EA773C2FED43@wilsonet.com> Content-Type: text/plain; charset="UTF-8" Date: Mon, 10 Jan 2011 07:07:52 -0500 Message-ID: <1294661272.2084.10.camel@morgan.silverblock.net> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit List-ID: Sender: On Mon, 2011-01-10 at 01:05 -0500, Jarod Wilson wrote: > On Jan 9, 2011, at 7:36 PM, Jason Gauthier wrote: > Janne, I've heard many success stories w/the hdpvr IR lately, and almost no reports > of lockups, so I'm thinking a firmware update may have helped out here, and thus, > maybe its time we just go ahead and push this patch along upstream? We still > require someone to load lirc_zilog manually, so it seems like a fairly low-risk > thing to do. FYI, the code I added to hdpvr-i2c.c will perform 2 accesses to the chip to check for existence, by virtue of a call to i2c_new_probed_device() (or whatever it is called). The I2C subsystem tries to talk the chip to see if it exists. If you are really concerned about corner cases that may hang, add a module option to hdpvr to disable I2C and/or IR in the hdpvr driver. With that users in the field can work-around the problem without rebuilding modules. Regards, Andy