From: Jean-Christophe PLAGNIOL-VILLARD Date: Thu, 30 Apr 2009 09:18:02 +0200 Subject: [U-Boot] [patch u-boot git arm/next] davinci: display correct clock info In-Reply-To: <200904292335.33059.david-b@pacbell.net> References: <200904291538.09721.david-b@pacbell.net> <200904291713.39025.david-b@pacbell.net> <20090430061017.GC31304@game.jcrosoft.org> <200904292335.33059.david-b@pacbell.net> Message-ID: <20090430071802.GF31304@game.jcrosoft.org> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de On 23:35 Wed 29 Apr , David Brownell wrote: > On Wednesday 29 April 2009, Jean-Christophe PLAGNIOL-VILLARD wrote: > > my idea is more this > > the lowlovel will init the pll (lowlevel_init.S or other stage bootloader) > > Right ... > > > > so instead of hardcoding the PPLDIV read it in the register > > and then calculate the clock rate > > That's all this code does: read the PLL registers > (and divider taps), and display settings in use by > the current board. yes but it's does not provice help full api to use it other where in the code > It happens that some SoCs have > slightly different PLL configuration (like pre/post > dividers) and feed important components (ARM, DSP, > DDR) from different PLL dividers. the PLL dividers are hardcoded why? > > > > and it need it set the other non used clock > > depending on the IP used by u-boot (at runtime) > > I don't know what you mean here. PLL/tap configuration > is distinct from clock distribution on the chip. > > The cpu/arm926ejs/davinci/psc.c code handles clock > distribution (and resets). So for example the dm355 > chips leave most modules clocked but in reset, while > the dm6446 leaves them unclocked and in reset ... so > modules needed by u-boot get uniformly taken out of > reset and clocked. I known that's why I've in mind to set only the clock if we use the drivers/IP at run time othewise let it off >But that's *unrelated* to this > PLL display code. I agree this not for this patch but for an improvement Best Regards, J.