Hello Kim,
Kim Phillips wrote:
On Tue, 15 Sep 2009 23:51:31 -0500 Kumar Gala galak@kernel.crashing.org wrote:
On Aug 27, 2009, at 3:53 PM, Kim Phillips wrote:
Heiko - let's go with the SCCR approach of setting bits in the LCRR, and have board config files only specify values for fields they're modifying from the reset value for their processor (this can be extended to 85xx-world).
Did you guys ever come to resolution on this? Realizing we have same issue on 85xx & 86xx (we dont actually set LCRR at all in cpu/ code for 85xx or 86xx.)
this should probably be extended to SICRH, SICRL, etc., but how's this:?
From 15d01649e403ec7da20f5fdd25b8d2c1bccb6a8d Mon Sep 17 00:00:00 2001
From: Kim Phillips kim.phillips@freescale.com Date: Fri, 25 Sep 2009 18:07:29 -0500 Subject: [PATCH] mpc83xx: retain POR values of non-configured ACR, SPCR, SCCR, and LCRR bitfields
some LCRR bits are not documented throughout the 83xx family RMs. New board porters copying similar board configurations might omit setting e.g., DBYP since it was not documented in their SoC's RM.
Prevent them bricking their board by retaining power on reset values in bit fields that the board porter doesn't explicitly configure via CONFIG_SYS_<registername>_<bitfield> assignments in the board config file.
also start to use i/o accessors.
Signed-off-by: Kim Phillips kim.phillips@freescale.com
cpu/mpc83xx/cpu_init.c | 255 ++++++++++++++++++++++++----------------- include/configs/MPC8313ERDB.h | 3 +- include/configs/MPC8315ERDB.h | 3 +- include/configs/MPC8323ERDB.h | 3 +- include/configs/MPC832XEMDS.h | 3 +- include/configs/MPC8349EMDS.h | 3 +- include/configs/MPC8349ITX.h | 3 +- include/configs/MPC8360EMDS.h | 3 +- include/configs/MPC8360ERDK.h | 3 +- include/configs/MPC837XEMDS.h | 3 +- include/configs/MPC837XERDB.h | 3 +- include/configs/MVBLM7.h | 3 +- include/configs/SIMPC8313.h | 4 +- include/configs/TQM834x.h | 3 +- include/configs/kmeter1.h | 4 +- include/configs/sbc8349.h | 3 +- include/configs/vme8349.h | 3 +- 17 files changed, 183 insertions(+), 122 deletions(-)
diff --git a/cpu/mpc83xx/cpu_init.c b/cpu/mpc83xx/cpu_init.c index 5c930d3..95dbc76 100644 --- a/cpu/mpc83xx/cpu_init.c +++ b/cpu/mpc83xx/cpu_init.c @@ -23,8 +23,8 @@ #include <common.h> #include <mpc83xx.h> #include <ioports.h> -#ifdef CONFIG_USB_EHCI_FSL #include <asm/io.h> +#ifdef CONFIG_USB_EHCI_FSL #include <usb/ehci-fsl.h> #endif
@@ -63,149 +63,192 @@ static void config_qe_ioports(void) */ void cpu_init_f (volatile immap_t * im) {
- /* Pointer is writable since we allocated a register for it */
[...]
- /* LCRR - Clock Ratio Register (10.3.1.16) */
- im->lbus.lcrr = CONFIG_SYS_LCRR;
- /* LCRR - Clock Ratio Register (10.3.1.16)
* write, read, and isync per MPC8379ERM rev.1 CLKDEV field description*/- clrsetbits_be32(&im->lbus.lcrr, lcrr_mask, lcrr_val);
- __raw_readl(&im->lbus.lcrr);
- isync();
Hmm.. shouldn;t this be done when running from RAM, as DaveLiu suggested?
bye Heiko