AM625: How to set up AM625 4GB RAM on SDK 9.2?

Andrew Tan

Prodigy 240 points

Part Number: AM625

Tool/software:

CPU: am6254ATCGGAALW

RAM: MT40A2G16

Dear Team,

The following is a patch that works successfully on 9.0.

edit:

board-support/ti-linux-kernel/arch/arm64/boot/dts/ti/k3-am625-sk.dts

board-support/ti-linux-kernel/arch/arm64/boot/dts/ti/k3-am62x-sk-common.dtsi

board-support/ti-u-boot/arch/arm/dts/k3-am625-sk.dts

memory@80000000 {
    device_type = "memory";
-   /* 2G RAM */
-   reg = <0x00000000 0x80000000 0x00000000 0x80000000>;
+   /* 4G RAM */
+   reg = <0x00000000 0x80000000 0x00000000 0x80000000>,
+   <0x00000008 0x80000000 0x00000000 0x80000000>;

};

board-support/ti-u-boot/arch/arm/dts/k3-am625-r5-sk.dts

-#include "k3-am62x-sk-ddr4-1600MTs.dtsi"
+#include "k3-am62x-ddr-config.dtsi"

/ {
	memory@80000000 {
		device_type = "memory";
-		/* 2G RAM */
-		reg = <0x00000000 0x80000000 0x00000000 0x80000000>;
+		/* 4G RAM */
+		reg = <0x00000000 0x80000000 0x00000000 0x80000000>,
+		<0x00000008 0x80000000 0x00000000 0x80000000>;
		bootph-pre-ram;
	};
};

add board-support/ti-u-boot/arch/arm/dts/k3-am62x-ddr-config.dtsi from sysconfig

board-support\ti-u-boot\drivers\ram\k3-ddrss\k3-ddrss.c

/* AM64x supports only up to 2 GB SDRAM */

-writel(0x000001EF, ddrss->ddrss_ss_cfg + DDRSS_V2A_CTL_REG);

+writel(0x00000210, ddrss->ddrss_ss_cfg + DDRSS_V2A_CTL_REG);

writel(0x0, ddrss->ddrss_ss_cfg + DDRSS_ECC_CTRL_REG);

The following is bdinfo and free -h use same patch on SDK9.0 and SDK9.2.

SDK9.0:

U-Boot SPL 2023.04-00001-gf5b119738d-dirty (Jun 19 2024 - 19:58:43 +0800)
SYSFW ABI: 3.1 (firmware rev 0x0009 '9.0.5--v09.00.05 (Kool Koala)')
SPL initial stack usage: 13376 bytes
Trying to boot from MMC2
Starting ATF on ARM64 core...

NOTICE:  BL31: v2.8(release):v2.8-226-g2fcd408bb3-dirty
NOTICE:  BL31: Built : 00:42:57, Jan 13 2023

U-Boot SPL 2023.04-00001-gf5b119738d-dirty (Jun 19 2024 - 19:58:52 +0800)
SYSFW ABI: 3.1 (firmware rev 0x0009 '9.0.5--v09.00.05 (Kool Koala)')
SPL initial stack usage: 1856 bytes
i2c_write: error waiting for data ACK (status=0x116)
pca953x gpio@22: Error reading output register
Error (-2): cannot determine file size
Trying to boot from MMC2


U-Boot 2023.04-00001-gf5b119738d-dirty (Jun 19 2024 - 19:58:52 +0800)

SoC:   AM62X SR1.0 GP
Model: Texas Instruments AM625 SK
EEPROM not available at 80, trying to read at 81
Reading on-board EEPROM at 0x51 failed -1
DRAM:  2 GiB (effective 4 GiB)
Core:  70 devices, 31 uclasses, devicetree: separate
MMC:   mmc@fa10000: 0, mmc@fa00000: 1
Loading Environment from nowhere... OK
In:    serial
Out:   serial
Err:   serial
EEPROM not available at 80, trying to read at 81
Reading on-board EEPROM at 0x51 failed -1
Net:   eth0: ethernet@8000000port@1
Autoboot in 3 seconds
i2c_write: error waiting for data ACK (status=0x116)
pca953x gpio@22: Error reading output register
switch to partitions #0, OK
mmc1 is current device
i2c_write: error waiting for data ACK (status=0x116)
pca953x gpio@22: Error reading output register
SD/MMC found on device 1
Failed to load 'boot.scr'
677 bytes read in 14 ms (46.9 KiB/s)
Loaded env from uEnv.txt
Importing environment from mmc1 ...
=> bdinfo
boot_params = 0x0000000000000000
DRAM bank   = 0x0000000000000000
-> start    = 0x0000000080000000
-> size     = 0x0000000080000000
DRAM bank   = 0x0000000000000001
-> start    = 0x0000000880000000
-> size     = 0x0000000080000000
flashstart  = 0x0000000000000000
flashsize   = 0x0000000000000000
flashoffset = 0x0000000000000000
baudrate    = 115200 bps
relocaddr   = 0x00000000ff61b000
reloc off   = 0x000000007ee1b000
Build       = 64-bit
current eth = ethernet@8000000port@1
ethaddr     = 1c:63:49:1b:f9:cf
IP addr     = <NULL>
fdt_blob    = 0x00000000fd5ec020
new_fdt     = 0x00000000fd5ec020
fdt_size    = 0x000000000000fd60
Video       = dss@30200000 inactive
multi_dtb_fit= 0x0000000000000000
lmb_dump_all:
 memory.cnt  = 0x2
 memory[0]      [0x80000000-0xffffffff], 0x80000000 bytes flags: 0
 memory[1]      [0x880000000-0x8ffffffff], 0x80000000 bytes flags: 0
 reserved.cnt  = 0x6
 reserved[0]    [0x9ca00000-0x9cafffff], 0x00100000 bytes flags: 0
 reserved[1]    [0x9cb00000-0x9e6fffff], 0x01c00000 bytes flags: 4
 reserved[2]    [0x9e780000-0x9fffffff], 0x01880000 bytes flags: 4
 reserved[3]    [0xfc5e0000-0xffffffff], 0x03a20000 bytes flags: 0
 reserved[4]    [0xfd5e69d0-0xffffffff], 0x02a19630 bytes flags: 0
 reserved[5]    [0x880000000-0x8ffffffff], 0x80000000 bytes flags: 0
devicetree  = separate
arch_number = 0x0000000000000000
TLB addr    = 0x00000000ffff0000
irq_sp      = 0x00000000fd5eaff0
sp start    = 0x00000000fd5eaff0
Early malloc usage: 3c28 / 8000
=>

root@AM62:~# free -h
               total        used        free      shared  buff/cache   available
Mem:           3.8Gi       390Mi       3.3Gi        11Mi       284Mi       3.4Gi
Swap:             0B          0B          0B
root@AM62:~#

SDK9.2:

U-Boot SPL 2023.04-dirty (Jul 16 2024 - 16:59:51 +0800)
SYSFW ABI: 3.1 (firmware rev 0x0009 '9.2.7--v09.02.07 (Kool Koala)')
SPL initial stack usage: 13408 bytes
Trying to boot from MMC2
Starting ATF on ARM64 core...

NOTICE:  BL31: v2.10.0(release):v2.10.0-367-g00f1ec6b87-dirty
NOTICE:  BL31: Built : 16:09:05, Feb  9 2024

U-Boot SPL 2023.04-dirty (Jul 16 2024 - 16:59:59 +0800)
SYSFW ABI: 3.1 (firmware rev 0x0009 '9.2.7--v09.02.07 (Kool Koala)')
SPL initial stack usage: 1856 bytes
i2c_write: error waiting for data ACK (status=0x116)
pca953x gpio@22: Error reading output register
Trying to boot from MMC2


U-Boot 2023.04-dirty (Jul 16 2024 - 16:59:59 +0800)

SoC:   AM62X SR1.0 GP
Model: Texas Instruments AM625 SK
EEPROM not available at 80, trying to read at 81
Reading on-board EEPROM at 0x51 failed -1
DRAM:  2 GiB (effective 4 GiB)
Core:  72 devices, 32 uclasses, devicetree: separate
MMC:   mmc@fa10000: 0, mmc@fa00000: 1
Loading Environment from nowhere... OK
In:    serial
Out:   serial
Err:   serial
EEPROM not available at 80, trying to read at 81
Reading on-board EEPROM at 0x51 failed -1
Net:   eth0: ethernet@8000000port@1
Hit any key to stop autoboot:  0
=> bdinfo
boot_params = 0x0000000000000000
DRAM bank   = 0x0000000000000000
-> start    = 0x0000000080000000
-> size     = 0x0000000080000000
DRAM bank   = 0x0000000000000001
-> start    = 0x0000000080000000
-> size     = 0x0000000080000000
flashstart  = 0x0000000000000000
flashsize   = 0x0000000000000000
flashoffset = 0x0000000000000000
baudrate    = 115200 bps
relocaddr   = 0x00000000ff616000
reloc off   = 0x000000007ee16000
Build       = 64-bit
current eth = ethernet@8000000port@1
ethaddr     = 1c:63:49:1b:f9:cf
IP addr     = <NULL>
fdt_blob    = 0x00000000fd5e7820
new_fdt     = 0x00000000fd5e7820
fdt_size    = 0x000000000000f560
Video       = dss@30200000 inactive
multi_dtb_fit= 0x0000000000000000
lmb_dump_all:
 memory.cnt  = 0x1
 memory[0]      [0x80000000-0xffffffff], 0x80000000 bytes flags: 0
 reserved.cnt  = 0x4
 reserved[0]    [0x9ca00000-0x9cafffff], 0x00100000 bytes flags: 0
 reserved[1]    [0x9cb00000-0x9e6fffff], 0x01c00000 bytes flags: 4
 reserved[2]    [0x9e780000-0x9fffffff], 0x01880000 bytes flags: 4
 reserved[3]    [0xfc5e2000-0xffffffff], 0x03a1e000 bytes flags: 0
devicetree  = separate
arch_number = 0x0000000000000000
TLB addr    = 0x00000000ffff0000
irq_sp      = 0x00000000fd5e6ff0
sp start    = 0x00000000fd5e6ff0
Early malloc usage: 3c28 / 8000
=>

root@am62xx-evm:~# free -h
               total        used        free      shared  buff/cache   available
Mem:           1.9Gi       151Mi       1.7Gi       900Ki        99Mi       1.7Gi
Swap:             0B          0B          0B
root@am62xx-evm:~#

Thanks,

Andrew

5 months ago

0 Nick Saulnier 5 months ago

TI__Guru* 89295 points

Hello Andrew,

It is going to take me a couple of days to get back to you. Please feel free to ping the thread if I have not responded by the end of the week.

Regards,

Nick

0 Andrew Tan 5 months ago in reply to Nick Saulnier

Prodigy 240 points

Dear Nick,

Do you have any new info?

Thanks,

Andrew

0 Nick Saulnier 5 months ago in reply to Andrew Tan

TI__Guru* 89295 points

Hello Andrew,

I'm still discussing with our hardware & SW teams (and Rich) to try to understand the updated DDR driver code. It might be another couple days before I have anything to report, but please do keep us honest and ping the thread again if I haven't replied by Wednesday of next week.

Regards,

Nick

0 Andrew Tan 5 months ago in reply to Nick Saulnier

Prodigy 240 points

Dear Nick,

Do you have any new info?

Thanks,

Andrew

0 Nick Saulnier 5 months ago in reply to Andrew Tan

TI__Guru* 89295 points

Hello Andrew,

Apologies for the delay here, and thank you for the ping. I haven't heard back from the SW team since they confirmed they were looking into the updated code. I will try to chat with them at the beginning of this week and get back to you.

Regards,

Nick

0 Nick Saulnier 5 months ago in reply to Nick Saulnier

TI__Guru* 89295 points

Hello Andrew,

The team is still discussing exactly what kinds of fixes are appropriate, so this is NOT a tested solution. However, I wanted to get information to you now instead of making you wait another week.

It seems like the issue might be caused here:
https://git.ti.com/cgit/ti-u-boot/ti-u-boot/tree/drivers/ram/k3-ddrss/k3-ddrss.c?h=09.02.00.010#n786

where this command

writel((((ilog2(ddrss->ddr_ram_size) - 16) << 5) | 0xF),
                  ddrss->ddrss_ss_cfg + DDRSS_V2A_CTL_REG);

Should be ORed with 0x1F instead of 0xF, since the region_idx parameter is 5 bits.

Could I get you to try changing 0xF to 0x1F, rebuilding uboot, and seeing if your full 4GB DDR allocation becomes available?

Regards,

Nick

0 Rich Chen 4 months ago in reply to Nick Saulnier

TI__Expert 3305 points

Nick,

Andrew has not yet tested this change on their own board.

He is busy on other task and asked whether this fixed had been merged into the new release SDK 10 or not?

BR, Rich

0 Nick Saulnier 4 months ago in reply to Rich Chen

TI__Guru* 89295 points

Hello Rich,

Thanks for the ping. The developer said it tentatively looked like AM62x should be working with 4GB DDR on SDK 10.0, since that code I pointed to above is no longer there for AM62x:
https://git.ti.com/cgit/ti-u-boot/ti-u-boot/tree/drivers/ram/k3-ddrss/k3-ddrss.c?h=10.00.08#n736

However, AM64x customers would run into the same behavior as on SDK 9.2.

I'm saying "tentatively" in the response above because I do not have an EVM with 4GB of DDR that I could use to test.

Regards,

Nick

Processors

Processors forum

AM625: How to set up AM625 4GB RAM on SDK 9.2?