From mboxrd@z Thu Jan 1 00:00:00 1970 From: Maxime Ripard Date: Mon, 24 Aug 2015 16:54:45 +0200 Subject: [U-Boot] [PATCH 07/15] sunxi_nand_spl: Turn off clocks when we're done with the nand In-Reply-To: <55D57528.2030603@redhat.com> References: <1439668968-3882-1-git-send-email-hdegoede@redhat.com> <1439668968-3882-8-git-send-email-hdegoede@redhat.com> <20150819134807.GF30520@lukather> <55D57528.2030603@redhat.com> Message-ID: <20150824145445.GX2413@lukather> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de On Thu, Aug 20, 2015 at 08:35:20AM +0200, Hans de Goede wrote: > On 08/19/2015 03:48 PM, Maxime Ripard wrote: > >On Sat, Aug 15, 2015 at 10:02:40PM +0200, Hans de Goede wrote: > >>Turn off the nand and dma clocks when we're done with the nand, this > >>puts the nand and dma controllers back into a clean state for when the > >>kernel boots. > >> > >>Without this the kernel will not boot properly when it is build with > >>dma-controller support. > > > >The kernel shouldn't behave like that. What issue are you seeing? > > The kernel does not output anything, not even on the serial console, > so it seems to crash quite early on. I guess that the dma controller > has an irq asserted and that that fires as soon as the dma driver > unmasks it. Hmmm, fishy. > I did not investigate this further since u-boot really should put > the hardware in a clean state before booting the os, so we want this > change even if the kernel were fixed to deal with the unclean state. I guess we should aim at "fixing" both. The kernel should not really care about the state the hardware has been left in, and U-boot should not really leave the dma controller running either. Maxime -- Maxime Ripard, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: Digital signature URL: