On Thu, Feb 04, 2021 at 10:41:20PM +0100, Heinrich Schuchardt wrote:
On 2/4/21 9:52 PM, Sean Anderson wrote:
On 2/4/21 3:41 PM, Tom Rini wrote:
On Thu, Feb 04, 2021 at 03:29:48PM -0500, Sean Anderson wrote:
This is almost identical to the imx7 version, except that the register names are different.
Signed-off-by: Sean Anderson sean.anderson@seco.com
arch/arm/mach-imx/mx6/soc.c | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+)
diff --git a/arch/arm/mach-imx/mx6/soc.c b/arch/arm/mach-imx/mx6/soc.c index bf6dddfdc9..cb729be46f 100644 --- a/arch/arm/mach-imx/mx6/soc.c +++ b/arch/arm/mach-imx/mx6/soc.c @@ -11,6 +11,7 @@ #include <linux/delay.h> #include <linux/errno.h> #include <asm/io.h> +#include <asm/setup.h> #include <asm/arch/imx-regs.h> #include <asm/arch/clock.h> #include <asm/arch/sys_proto.h> @@ -705,6 +706,38 @@ int arch_misc_init(void) } #endif
+#ifdef CONFIG_SERIAL_TAG +/*
- UNIQUE_ID describes a unique ID based on silicon wafer
- and die X/Y position
- UNIQUE_ID offset 0x410
- 31:0 fuse 0
- FSL-wide unique, encoded LOT ID STD II/SJC CHALLENGE/ Unique ID
- UNIQUE_ID offset 0x420
- 31:24 fuse 1
- The X-coordinate of the die location on the wafer/SJC CHALLENGE/
Unique ID
- 23:16 fuse 1
- The Y-coordinate of the die location on the wafer/SJC CHALLENGE/
Unique ID
- 15:11 fuse 1
- The wafer number of the wafer on which the device was
fabricated/SJC
- CHALLENGE/ Unique ID
- 10:0 fuse 1
- FSL-wide unique, encoded LOT ID STD II/SJC CHALLENGE/ Unique ID
- */
+void get_board_serial(struct tag_serialnr *serialnr) +{ + struct ocotp_regs *ocotp = (struct ocotp_regs *)OCOTP_BASE_ADDR; + struct fuse_bank *bank = &ocotp->bank[0]; + struct fuse_bank0_regs *fuse = + (struct fuse_bank0_regs *)bank->fuse_regs;
+ serialnr->low = fuse->uid_low; + serialnr->high = fuse->uid_high; +} +#endif
/* * gpr_init() function is common for boards using MX6S, MX6DL, MX6D, * MX6Q and MX6QP processors
NAK. This needs to use the normal mechanism to populate "serial#" directly and not abuse the old ATAG infrastructure.
What is the "normal" mechanism? Define something like omap_die_id_serial which sets serial# directly? From what I can tell, get_board_serial is the "generic" function for separating the getting of the serial from the setting of serial#.
--Sean
get_board_serial() is defined as empty function in bootm.h if CONFIG_SERIAL_TAG=n. But CONFIG_SERIAL_TAG is not documented anywhere.
Correct, it's ATAGs related stuff, so quite old and at best may have been documented in the README at some point.
board/gateworks/gw_ventana/gw_ventana.c has this comment:
#ifdef CONFIG_SERIAL_TAG /*
- called when setting up ATAGS before booting kernel
- populate serialnum from the following (in order of priority):
- serial# env var
- eeprom
*/ void get_board_serial(struct tag_serialnr *serialnr)
Right.
serial# seems to be set generally in misc_init_r(). But doing this on a per board level seems to be an inadequate design for i.mx6. The best place for i.mx6 seems to be arch_misc_init().
Yes, there's a number of reasonable examples of where to populate serial# from, and a few examples from when we had to support the kernel with it.
@Tom: Linux Kconfig has:
CONFIG_ATAGS: This is the traditional way of passing data to the kernel at boot time. If you are solely relying on the flattened device tree (or the ARM_ATAG_DTB_COMPAT option) then you may unselect this option to remove ATAGS support from your kernel binary. If unsure, leave this to y.
CONFIG_DEPRECATED_PARAM_STRUCT depends on ATAGS This was deprecated in 2001 and announced to live on for 5 years. Some old boot loaders still use this way.
As this has been deprecated 20 years ago, why don't we remove it from U-Boot?
The history is a lot more complex than that. The earliest starting date for "ATAGs is dead, use DT" would be 2010 or so when Linaro started real work on migrating ARM to DT. I also _think_ that deprecated param struct note is about what predated ATAGs actually.
But, we can't universally drop ATAGs support from U-Boot as we support hardware that runs on pre-device tree kernels only (Nokia N900) or older VxWorks that also used ATAGs I _think__. But we certainly have some cases that should be cleaned up and disabled, hence my series yesterday.