From mboxrd@z Thu Jan 1 00:00:00 1970 From: Scott Wood Date: Wed, 25 Apr 2012 13:59:22 -0500 Subject: [U-Boot] [PATCH 2/4][v2] powerpc/85xx:Fix MSR[DE] bit in MSR to support debugger In-Reply-To: References: <1335342343-28501-1-git-send-email-prabhakar@freescale.com> <4F9845CA.2090405@freescale.com> Message-ID: <4F98498A.3090903@freescale.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de On 04/25/2012 01:55 PM, Joakim Tjernlund wrote: > > Scott Wood wrote on 2012/04/25 20:43:22: >> >> On 04/25/2012 05:57 AM, Joakim Tjernlund wrote: >>>> >>>> Debugging of e500 and e500v1 processer requires MSR[DE] bit to be set always. >>>> Where MSR = Machine State register >>>> >>>> Make sure of MSR[DE] bit is set uniformaly across the different execution >>>> address space i.e. AS0 and AS1. >>> >>> Hi >>> >>> We are trying to bring up our custom P2010 RDB based board. boot is NOR >>> based and we cannot get past the rfi below. >>> lis r6,MSR_IS|MSR_DS at h >>> ori r6,r6,MSR_IS|MSR_DS at l >>> lis r7,switch_as at h >>> ori r7,r7,switch_as at l >>> >>> mtspr SPRN_SRR0,r7 >>> mtspr SPRN_SRR1,r6 >>> rfi >>> >>> switch_as: >>> >>> We end up with a TLB exception no matter what we do, even after applying this patch. >> >> Did you apply the entire patchset, and define CONFIG_SYS_PPC_E500_DEBUG_TLB? > > No, but this code is executed before any of the other parts of the patch. Anyhow, I just > found the problem(really obvious once I found it). > During bring up we had to load uboot in the middle of the flash instead of > the end because we have a flash burn problem in the end of the flash that we do not > understand yet. We think it may be related to DDR3 being misconfigured by the emulator(BDI3000). > I do not understand why this emulator can not use the L2SRAM instead? Is there something > magic behind the L2SRAM so it is impossible to use it as a work area for > flash burning? I don't know of any reason L2SRAM couldn't be used for this. My guess is they just don't want to have to support more than one way of creating RAM. -Scott