From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <51488665.3010505@wanadoo.fr> Date: Tue, 19 Mar 2013 16:38:13 +0100 From: Thierry Bultel MIME-Version: 1.0 References: <51482868.5020206@wanadoo.fr> <51483887.4020504@xenomai.org> <514854E3.2070707@xenomai.org> In-Reply-To: <514854E3.2070707@xenomai.org> Content-Type: text/plain; charset="utf-8"; format="flowed" Content-Transfer-Encoding: 8bit Subject: Re: [Xenomai] Limited kernel size for booting ? List-Id: Discussions about the Xenomai project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Gilles Chanteperdrix Cc: xenomai@xenomai.org Really many thanks. This works like a charm. I also add to hack into arch/arm/kernel/sleep.S, because the "cpu_resume" is located in .data, and it had a 24bits jump to cpu_do_resume that could not be relocated and led to a linker error. I have simply removed the line, because I am not insterested in cpu suspend for my application anyway. FYI, the freescale rel_imx_3.0.35_12.09.01 tag version of that file is closed enough to the committed patch for being taken as is. Also, this is not related, but I have been a little confused by failing attempts, because I had forgotten the "rootwait" bootarg, which is mandatory with that kernel. Cheers Thierry Le 19/03/2013 13:06, Gilles Chanteperdrix a écrit : > On 03/19/2013 11:05 AM, Gilles Chanteperdrix wrote: > >> On 03/19/2013 09:57 AM, Thierry Bultel wrote: >> >>> Hi Gilles, >>> >>> First, I apologize for posting such a question that is sincerely not >>> really related to xenomai. >>> But I am using the kernel from ipipe-gch.git, branch ipipe-3.0-imx6q >>> >>> The issue I have is that it is not booting when using a really big >>> built-in initramfs(about 60 Mo) >>> It does not seem to reach "start_kernel", there is nomore log after >>> Uboot says "Starting kernel" >>> >>> I have payed attention to address zones, and even tried to use an >>> uncompressed kernel >>> in the uImage container, to skip and discriminate the gzip phase. Same >>> result. >>> >>> I do not have an operational Jtag probe yet, thus I only would like to >>> get your opinion on >>> this. To your mind, is this assumed to work ? >> >> No, the maximum size of initramfs is limited, you need to change the >> linker script to put the initramfs at the end of the memory used by the >> kernel, a patch exists to do that on the linux arm kernel mailing list >> and has been merged in recent kernels. >> >> > > https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=3835d69a6c7048a28d0aea3cb8403d5e83a0f867 >