i.MX8QXP Reserved Memory Size

[Kazuma_Sasaki]

Hi,

My customer is making a product with i.MX8QuadXPlus and Linux BSP L4.14.98_2.0.0_ga.

They would like to reduce a memory size from cost point of view.

However, current linux BSP has consumed around 270MiB for reserved for system.

We seems it is difficult to apply low-specification platform such as DRAM density 512MiB.

How do we reduce these reserved memory? Can you provide us any advice to reduce reserved memory size? 

[    0.000000] Linux version 4.14.78-imx_4.14.78_1.0.0_ga+g66620c3 (oe-user@oe-host) (gcc version 7.3.0 (GCC)) #1 SMP PREEMPT Tue Dec 4 04:00:25 UTC 2018

[    0.000000] Memory: 1815668K/3068928K available (13500K kernel code, 1660K rwdata, 6012K rodata, 1280K init, 444K bss, 270220K reserved, 983040K cma-reserved)

root@imx8qxpmek:~# free -h -k

total used free shared buff/cache available
Mem: 2.7G 362M 2.2G 8.8M 83M 2.2G
Swap: 0B 0B 0B

Best Regards,

Kazuma Sasaki.

[Kazuma_Sasaki]

Hi Markus, igor,

I really appreciate your support. I will try reserved-memory adjustment.

Best Regards,

Kazuma Sasaki.

[igorpadykov]

Hi Kazuma

one can try to adjust "reserved-memory" in fsl-imx8dx.dtsi

fsl-imx8dx.dtsi\freescale\dts\boot\arm64\arch - linux-imx - i.MX Linux kernel 

For finding optimal settings for customer design and software configuration may be

recommended to proceed with help of extended support of NXP Professional Services | NXP 

Best regards
igor
-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------

[Niebel-TQ]

Hello,

just a few thoughts. If there is any IP with reserved memory that you do not use in your design, you could just remove the nodes (both for ip and the reserved memory) from the device tree. Depending on the dispaly in use you can also eyperiment with lowering the GPU reserved memory size.

Best Regards

Markus Niebel