From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stefan Roese Date: Tue, 29 Jul 2008 16:59:44 +0200 Subject: [U-Boot-Users] ppc4xx: [PATCH] ML507 Board Support In-Reply-To: <20080729142929.6D91C248B9@gemini.denx.de> References: <20080729142929.6D91C248B9@gemini.denx.de> Message-ID: <200807291659.44619.sr@denx.de> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de Hi Wolfgang, On Tuesday 29 July 2008, Wolfgang Denk wrote: > In message <200807171039.38227.sr@denx.de> you wrote: > > > > #define CFG_MBYTES_SDRAM 256 > > > > > > No. All boards, especially all PowerPC boards, should use dynamic > > > memory size detection. > > > > You are probably referring to calling get_ram_size() instead of just > > returning the configured fixed value. But what do we gain here. > > get_ram_size() also needs the ram-size as parameter. Not sure if this is > > an improvement on such fixed systems. > > It just needs a maximum possible size. OK, this would be the 256 MB from above again. > One major advantage is for example that get_ram_size() verifies that > RAM is actually working, and of expected size. It is a very efficient > test to make sure your hardware is actually working. But what exactly should happen if such a board with only one possible memory size configuration detects ram_size != configured ram_size? OK, this small check is an improvement over no check at all. Is it acceptable that I add this get_ram_size() with an additional patch? This would safe me from resetting my "next" branch or reverting patches. Best regards, Stefan ===================================================================== DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: +49-8142-66989-0 Fax: +49-8142-66989-80 Email: office at denx.de =====================================================================