From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <4C1A4E1B.8090905@domain.hid> Date: Thu, 17 Jun 2010 18:32:27 +0200 From: Gilles Chanteperdrix MIME-Version: 1.0 References: <181804936ABC2349BE503168465576460EBD5C0D@exchserver.basler.com> <1274047063.2972.1.camel@domain.hid> <2319761F7FA0D1479BA77EC2E0A8E7BCE8A603@domain.hid> <1275033915.27918.31.camel@domain.hid> <2319761F7FA0D1479BA77EC2E0A8E7BCEDCE09@domain.hid> <4C191D07.4050203@domain.hid> <4C191D91.6060209@domain.hid> <2319761F7FA0D1479BA77EC2E0A8E7BCEDCE78@domain.hid> <4C19E1D8.6010705@domain.hid> <1276792000.4890.600.camel@domain.hid> In-Reply-To: <1276792000.4890.600.camel@domain.hid> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Subject: Re: [Xenomai-help] rt_task_sleep causing crashes on PowerPC List-Id: Help regarding installation and common use of Xenomai List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Travis Stratman Cc: xenomai@xenomai.org, Steve Deiters Travis Stratman wrote: > Gilles, > > On Thu, 2010-06-17 at 10:50 +0200, Gilles Chanteperdrix wrote: >> Abhijit Majumdar wrote: >>> We use adeos-ipipe-2.6.28.10-arm-1.12-07.patch. >> Ok. That is the latest I-pipe patch for 2.6.28. However, why are you >> using a 2.6.28 kernel? If you are starting a new project, there is no >> reason not to pick the latest kernel. > > The 2.6.28.10 kernel that they are using is the latest release that we > provide that has support patches for the board that they are using. We > pick a stable version and develop with that for awhile before updating > to a newer version, as we simply don't have enough developers to > continually update and re-test newer kernel versions for all of our > products. We are working on a newer version for a new board but that > hasn't been tested yet. > > Abhijit: it would be possible to compile a new kernel (i.e. 2.6.33) with > Xenomai 2.5.3 for the AT91SAM9G20 eval board that would work with your > board (after some configuration changes). It wouldn't support all of the > custom I/O but you would be able to test the failure that you are > seeing. The AT91 patch set for 2.6.33 would also need to be applied. The > at91 experimental patch set is not ported to 2.6.33 yet, but this > shouldn't be a problem for initial testing. Ok, forget it, 2.6.28 is fine. -- Gilles.