From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mga11.intel.com ([192.55.52.93]) by linuxtogo.org with esmtp (Exim 4.72) (envelope-from ) id 1RTv53-0005T7-Rt for openembedded-devel@lists.openembedded.org; Fri, 25 Nov 2011 13:40:38 +0100 Received: from fmsmga002.fm.intel.com ([10.253.24.26]) by fmsmga102.fm.intel.com with ESMTP; 25 Nov 2011 04:34:02 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.69,571,1315206000"; d="scan'208";a="95307031" Received: from unknown (HELO helios.localnet) ([10.252.121.214]) by fmsmga002.fm.intel.com with ESMTP; 25 Nov 2011 04:34:01 -0800 From: Paul Eggleton To: openembedded-devel@lists.openembedded.org Date: Fri, 25 Nov 2011 12:34 +0000 Message-ID: <1734371.1v8iFp6Vj1@helios> Organization: Intel Corporation User-Agent: KMail/4.7.3 (Linux/3.0.0-13-generic-pae; KDE/4.7.3; i686; ; ) In-Reply-To: <20111121115213.GF3565@jama.jama.net> References: <1321783600-23772-1-git-send-email-Martin.Jansa@gmail.com> <20111121115213.GF3565@jama.jama.net> MIME-Version: 1.0 Cc: Martin Jansa Subject: Re: [meta-handheld v2] linux-kexecboot: use kernel_conf_variable like linux.inc does and respect CMDLINE_DEBUG X-BeenThere: openembedded-devel@lists.openembedded.org X-Mailman-Version: 2.1.11 Precedence: list Reply-To: openembedded-devel@lists.openembedded.org List-Id: Using the OpenEmbedded metadata to build Distributions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 25 Nov 2011 12:40:38 -0000 Content-Transfer-Encoding: 7Bit Content-Type: text/plain; charset="us-ascii" On Monday 21 November 2011 12:52:13 Martin Jansa wrote: > On Mon, Nov 21, 2011 at 11:43:46AM +0100, Andrea Adami wrote: > > On Sun, Nov 20, 2011 at 12:26 PM, Martin Jansa > > wrote: Martin, > > > > I agree for the use kernel_conf_variable but I'm unsure about > > CMDLINE_DEBUG. In fact, I hardcoded it purposedly with commit 8929166 > > with the > > following reasons: > > > > * Being a second-stage bootloader we don't expect any output on screen. > > * Adding 'debug' as commandline arg is pointless because > > * linux-kexecboot is compiled without debug and printk in order > > * to minimize the size. > > * Full debug implies not only editing the commandline (e.g. earlyprintk) > > * but re-enabling full debug options in the defconfigs > > I had some issues with 3.2.0-rc2 when used as linux-kexecboot, that's > why I've added this, but later found out that it's really a bit > incomplete, that's why I have followup patch for this: > https://gitorious.org/shr/meta-handheld/commit/6bc710615b587fc0bfd8fc4372cad > 0de5599dd82?format=patch > > But with latest SRCREV it works again, so in theory I don't need it, but > still it would be usefull to have easy way to debug even this stage > (maybe you still remember older kernel version where it worked fine > as 3rd stage and failed to boot as 2nd stage..). > > BTW: here is upgrade to 3.2.0-rc2 with those changes to > linux-kexecboot.inc applied > https://gitorious.org/shr/meta-handheld/commit/27ca3d24353a5473f4f1aade00ba7 > a52b667e28c?format=patch > > (be aware that I do rebase that branch sometimes so if those commits are > not valid anymore just look at last few commits in that branch). As discussed on IRC I've merged the latest version of this branch to meta- handheld master now, thanks. Cheers, Paul -- Paul Eggleton Intel Open Source Technology Centre