[PATCH] imx8m: fix reading of DDR4 MR registers

I was trying to employ lpddr4_mr_read() to something similar to what the imx8mm-cl-iot-gate board is doing for auto-detecting the RAM type. However, the version in drivers/ddr/imx/imx8m/ddrphy_utils.c differs from the private one used by that board in how it extracts the byte value, and I was only getting zeroes. Adding a bit of debug printf'ing gives me
tmp = 0x00ffff00 tmp = 0x00070700 tmp = 0x00000000 tmp = 0x00101000
and indeed I was expecting a (combined) value of 0xff070010 (0xff being Manufacturer ID for Micron). I can't find any documentation that says how the values are supposed to be read, but clearly the iot-gate definition is the right one, both for its use case as well as my imx8mp-based board.
So lift the private definition of lpddr4_mr_read() from the imx8mm-cl-iot-gate board code to ddrphy_utils.c, and add a declaration in the ddr.h header where e.g. get_trained_CDD() is already declared.
This has only been compile-tested for the imx8mm-cl-iot-gate board (since I don't have the hardware), but since I've merely moved its definition of lpddr4_mr_read(), I'd be surprised if it changed anything for that board.
Signed-off-by: Rasmus Villemoes rasmus.villemoes@prevas.dk --- arch/arm/include/asm/arch-imx8m/ddr.h | 1 + board/compulab/imx8mm-cl-iot-gate/ddr/ddr.c | 27 --------------------- drivers/ddr/imx/imx8m/ddrphy_utils.c | 9 +++++-- 3 files changed, 8 insertions(+), 29 deletions(-)
diff --git a/arch/arm/include/asm/arch-imx8m/ddr.h b/arch/arm/include/asm/arch-imx8m/ddr.h index 0f1e832c03..2ce8a8f2d4 100644 --- a/arch/arm/include/asm/arch-imx8m/ddr.h +++ b/arch/arm/include/asm/arch-imx8m/ddr.h @@ -723,6 +723,7 @@ void ddrphy_init_read_msg_block(enum fw_type type);
void update_umctl2_rank_space_setting(unsigned int pstat_num); void get_trained_CDD(unsigned int fsp); +unsigned int lpddr4_mr_read(unsigned int mr_rank, unsigned int mr_addr);
static inline void reg32_write(unsigned long addr, u32 val) { diff --git a/board/compulab/imx8mm-cl-iot-gate/ddr/ddr.c b/board/compulab/imx8mm-cl-iot-gate/ddr/ddr.c index 5b93491923..b230478b61 100644 --- a/board/compulab/imx8mm-cl-iot-gate/ddr/ddr.c +++ b/board/compulab/imx8mm-cl-iot-gate/ddr/ddr.c @@ -24,33 +24,6 @@
#include <linux/delay.h>
-static unsigned int lpddr4_mr_read(unsigned int mr_rank, unsigned int mr_addr) -{ - unsigned int tmp; - - reg32_write(DRC_PERF_MON_MRR0_DAT(0), 0x1); - do { - tmp = reg32_read(DDRC_MRSTAT(0)); - } while (tmp & 0x1); - - reg32_write(DDRC_MRCTRL0(0), (mr_rank << 4) | 0x1); - reg32_write(DDRC_MRCTRL1(0), (mr_addr << 8)); - reg32setbit(DDRC_MRCTRL0(0), 31); - do { - tmp = reg32_read(DRC_PERF_MON_MRR0_DAT(0)); - } while ((tmp & 0x8) == 0); - tmp = reg32_read(DRC_PERF_MON_MRR1_DAT(0)); - reg32_write(DRC_PERF_MON_MRR0_DAT(0), 0x4); - while (tmp) { //try to find a significant byte in the word - if (tmp & 0xff) { - tmp &= 0xff; - break; - } - tmp >>= 8; - } - return tmp; -} - struct lpddr4_desc { char name[16]; unsigned int id; diff --git a/drivers/ddr/imx/imx8m/ddrphy_utils.c b/drivers/ddr/imx/imx8m/ddrphy_utils.c index a54449e5f1..975d553674 100644 --- a/drivers/ddr/imx/imx8m/ddrphy_utils.c +++ b/drivers/ddr/imx/imx8m/ddrphy_utils.c @@ -198,9 +198,14 @@ unsigned int lpddr4_mr_read(unsigned int mr_rank, unsigned int mr_addr) tmp = reg32_read(DRC_PERF_MON_MRR0_DAT(0)); } while ((tmp & 0x8) == 0); tmp = reg32_read(DRC_PERF_MON_MRR1_DAT(0)); - tmp = tmp & 0xff; reg32_write(DRC_PERF_MON_MRR0_DAT(0), 0x4); - + while (tmp) { //try to find a significant byte in the word + if (tmp & 0xff) { + tmp &= 0xff; + break; + } + tmp >>= 8; + } return tmp; }

Hi Paul,
On Mon, Apr 25, 2022 at 11:23 AM Rasmus Villemoes rasmus.villemoes@prevas.dk wrote:
I was trying to employ lpddr4_mr_read() to something similar to what the imx8mm-cl-iot-gate board is doing for auto-detecting the RAM type. However, the version in drivers/ddr/imx/imx8m/ddrphy_utils.c differs from the private one used by that board in how it extracts the byte value, and I was only getting zeroes. Adding a bit of debug printf'ing gives me
tmp = 0x00ffff00 tmp = 0x00070700 tmp = 0x00000000 tmp = 0x00101000
and indeed I was expecting a (combined) value of 0xff070010 (0xff being Manufacturer ID for Micron). I can't find any documentation that says how the values are supposed to be read, but clearly the iot-gate definition is the right one, both for its use case as well as my imx8mp-based board.
So lift the private definition of lpddr4_mr_read() from the imx8mm-cl-iot-gate board code to ddrphy_utils.c, and add a declaration in the ddr.h header where e.g. get_trained_CDD() is already declared.
This has only been compile-tested for the imx8mm-cl-iot-gate board (since I don't have the hardware), but since I've merely moved its definition of lpddr4_mr_read(), I'd be surprised if it changed anything for that board.
Signed-off-by: Rasmus Villemoes rasmus.villemoes@prevas.dk
Could you please test this patch?
I only have remote access to this board.

Hi Fabio,
I tested. It works. Tested-by: Ying-Chun Liu (PaulLiu) paul.liu@linaro.org
On Tue, 26 Apr 2022 at 20:59, Fabio Estevam festevam@gmail.com wrote:
Hi Paul,
On Mon, Apr 25, 2022 at 11:23 AM Rasmus Villemoes rasmus.villemoes@prevas.dk wrote:
I was trying to employ lpddr4_mr_read() to something similar to what the imx8mm-cl-iot-gate board is doing for auto-detecting the RAM type. However, the version in drivers/ddr/imx/imx8m/ddrphy_utils.c differs from the private one used by that board in how it extracts the byte value, and I was only getting zeroes. Adding a bit of debug printf'ing gives me
tmp = 0x00ffff00 tmp = 0x00070700 tmp = 0x00000000 tmp = 0x00101000
and indeed I was expecting a (combined) value of 0xff070010 (0xff being Manufacturer ID for Micron). I can't find any documentation that says how the values are supposed to be read, but clearly the iot-gate definition is the right one, both for its use case as well as my imx8mp-based board.
So lift the private definition of lpddr4_mr_read() from the imx8mm-cl-iot-gate board code to ddrphy_utils.c, and add a declaration in the ddr.h header where e.g. get_trained_CDD() is already declared.
This has only been compile-tested for the imx8mm-cl-iot-gate board (since I don't have the hardware), but since I've merely moved its definition of lpddr4_mr_read(), I'd be surprised if it changed anything for that board.
Signed-off-by: Rasmus Villemoes rasmus.villemoes@prevas.dk
Could you please test this patch?
I only have remote access to this board.

Hi Paul,
On 27/04/2022 03:48, Paul Liu wrote:
Hi Fabio,
I tested. It works. Tested-by: Ying-Chun Liu (PaulLiu) paul.liu@linaro.org
Thanks for testing.
Reviewed-by: Fabio Estevam festevam@denx.de

I was trying to employ lpddr4_mr_read() to something similar to what the imx8mm-cl-iot-gate board is doing for auto-detecting the RAM type. However, the version in drivers/ddr/imx/imx8m/ddrphy_utils.c differs from the private one used by that board in how it extracts the byte value, and I was only getting zeroes. Adding a bit of debug printf'ing gives me tmp = 0x00ffff00 tmp = 0x00070700 tmp = 0x00000000 tmp = 0x00101000 and indeed I was expecting a (combined) value of 0xff070010 (0xff being Manufacturer ID for Micron). I can't find any documentation that says how the values are supposed to be read, but clearly the iot-gate definition is the right one, both for its use case as well as my imx8mp-based board. So lift the private definition of lpddr4_mr_read() from the imx8mm-cl-iot-gate board code to ddrphy_utils.c, and add a declaration in the ddr.h header where e.g. get_trained_CDD() is already declared. This has only been compile-tested for the imx8mm-cl-iot-gate board (since I don't have the hardware), but since I've merely moved its definition of lpddr4_mr_read(), I'd be surprised if it changed anything for that board. Signed-off-by: Rasmus Villemoes rasmus.villemoes@prevas.dk Tested-by: Ying-Chun Liu (PaulLiu) paul.liu@linaro.org Reviewed-by: Fabio Estevam festevam@denx.de
Applied to u-boot-imx, master, thanks !
Best regards, Stefano Babic
participants (5)
-
Fabio Estevam
-
Fabio Estevam
-
Paul Liu
-
Rasmus Villemoes
-
sbabic@denx.de