extranotes

-

 b) Reserve 1 Mbyte of kernel memory for your driver (Similar to PSTORE) 

root@apalis-imx8:~$ cat /proc/iomem

....

8020_0000-83ffffff : System RAM

8028_0000-8161ffff : Kernel code

8162_0000-8188ffff : reserved

8189_0000-81a85fff : Kernel data

8640_0000-87ffffff : System RAM

9000_0000-901fffff : System RAM

9050_0000-91ffffff : System RAM

94c0_0000-94ffffff : System RAM

9540_0000-ffffffff : System RAM

9600_0000-d1ffffff : reserved

f960_3000-fd62afff : reserved

8_8000_0000-8ffffffff : System RAM

8_fae0_0000-8fb5fffff : reserved

8_fb7f_e000-8ff5fffff : reserved

8_ff6d_b000-8ff73afff : reserved

8_ff73_b000-8ff73bfff : reserved

8_ff73_c000-8ff7cbfff : reserved

8_ff7c_e000-8ff7cefff : reserved

8_ff7c_f000-8ff7d3fff : reserved

8_ff7d_4000-8ff7d4fff : reserved

8_ff7d_5000-8ffffffff : reserved

  

check available memory of 1 MB which is not reserved, e.g. 8_8000_0000


Now in device tree



  

pstore and ramoops

The ramoops driver allows us to store kernel oopses, the kernel console output and user messages in a reserved memory region inside RAM


configs:


CONFIG_PSTORE=y

CONFIG_PSTORE_CONSOLE=y

CONFIG_PSTORE_PMSG=y

CONFIG_PSTORE_RAM=y


device tree


reserved-memory {

#address-cells = <2>;

#size-cells = <2>;

ranges;


ramoops@0x880000000 {

compatible = "ramoops";

reg = <0 0x880000000 0 0x100000>;

record-size = <0x4000>;

console-size = <0x4000>;

ftrace-size = <0x4000>;

};

};


or from cmdline

ramoops.mem_address=0x30000000 ramoops.mem_size=0x100000 memmap=0x100000$0x30000000


record-size - kernel dmesg

console-size - console output from the kernel 

pmsg-size - user messages 


After crash mount pstore

mount -t pstore pstore /mnt/

then

cat /mnt/dmesg-ramoops-0


and it will dump store dmesg


---------------------------------------------

DMA and cache




----------------------------------------------

https://stackoverflow.com/questions/8265657/how-does-copy-from-user-from-the-linux-kernel-work-internally

copy_from_user: copy user data to kernel


The implementation of copy_from_user() system call is done using two buffers from different address spaces:


The user-space buffer in user virtual address space.

The kernel-space buffer in kernel virtual address space.

When the copy_from_user() system call is invoked, data is copied from user buffer to kernel buffer.


implementation is here: arch/arm64/lib/copy_from_user.S

SYM_FUNC_START(__arch_copy_from_user)

add end, x0, x2

mov srcin, x1

#include "copy_template.S"

mov x0, #0 // Nothing to copy

ret


// Exception fixups

9997: cmp dst, dstin

b.ne 9998f

// Before being absolutely sure we couldn't copy anything, try harder

USER(9998f, ldtrb tmp1w, [srcin])

strb tmp1w, [dst], #1

9998: sub x0, end, dst // bytes not copied

ret

SYM_FUNC_END(__arch_copy_from_user)

EXPORT_SYMBOL(__arch_copy_from_user)


copy_to_user: copy kernel data to user

copy_to_user() works in exactly the same way, except that the source and destination addresses are swapped - it reads from kernel addresses and writes to userspace addresses. 

The only check that is done in both is that the userspace address really is below the kernel/user split; there is no need to check for addresses belonging to other user processes, because copy_*_user() are called only in process context, 

which means that all addresses are either kernel addresses or belong to the current process


implementation is here: arch/arm64/lib/copy_to_user.S


SYM_FUNC_START(__arch_copy_to_user)

add end, x0, x2

mov srcin, x1

#include "copy_template.S"

mov x0, #0

ret


// Exception fixups

9997: cmp dst, dstin

b.ne 9998f

// Before being absolutely sure we couldn't copy anything, try harder

ldrb tmp1w, [srcin]

USER(9998f, sttrb tmp1w, [dst])

add dst, dst, #1

9998: sub x0, end, dst // bytes not copied

ret

SYM_FUNC_END(__arch_copy_to_user)

EXPORT_SYMBOL(__arch_copy_to_user)


-----------------------------------------------------------------------

linux scheduler

https://blog.csdn.net/weixin_44390164/article/details/132253387?spm=1001.2101.3001.6650.5&utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7ECTRLIST%7ERate-5-132253387-blog-105787081.235%5Ev43%5Epc_blog_bottom_relevance_base1&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2%7Edefault%7ECTRLIST%7ERate-5-132253387-blog-105787081.235%5Ev43%5Epc_blog_bottom_relevance_base1



https://blog.csdn.net/wellt606/article/details/132331378?spm=1001.2101.3001.6650.6&utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromBaidu%7ERate-6-132331378-blog-132253387.235%5Ev43%5Epc_blog_bottom_relevance_base1&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromBaidu%7ERate-6-132331378-blog-132253387.235%5Ev43%5Epc_blog_bottom_relevance_base1


-----------------------------------------------------------------------

system calls:




----------------------------------------------------------------------

Watchdog:


On Linux-based systems, the standard user space interface to the watchdog is the /dev/watchdog file, 

through which a daemon will notify the kernel watchdog driver that the

user space is still alive. The watchdog starts right after the file is opened, and gets pinged

by periodically writing into this file.

When the notification occurs, the underlying driver will notify the watchdog device,

which will result in resetting its timeout; the watchdog will then wait for yet another

timeout duration prior to resetting the system. However, if for any reason the user

space does not perform the notification before the timeout is elapsed, the watchdog will

reset the system (causing a reboot).


--------------------------------------------------------------------

Linux Device Driver code to load another module: request_module

http://embeddedguruji.blogspot.com/2019/05/linux-device-driver-code-to-load.html

------------------------------------------------------------------

memory barrier

https://developer.arm.com/documentation/den0042/a/Memory-Ordering/Memory-barriers

-----------------------------------------------------------------

What kind of driver you have written

accelerometer, fuel guage as i2c client: https://github.com/tejasatgithub/Linux-i2c-client-driver/blob/master/src/i2c_client_driver.c

pwm driver for brightness

pstore driver

device tree manipulator driver

touch driver

bootinfo driver

Comments

Post a Comment

Popular posts from this blog

dev_get_platdata understanding

-
So I see this type of code in platform driver inside kernel, struct gpio_led_platform_data *pdata = dev_get_platdata(&pdev->dev); or this, struct aat2870_platform_data *pdata = dev_get_platdata(&client->dev); So how dev_ get_platdata is helping us. In device.h the definition here is, static inline void *dev_get_platdata(const struct device *dev) { return dev->platform_data; } So if I see from kernel it is always done for to get platform specific data. Like Here is a definition of platform_device, in platform_device.h struct platform_device { const char *name; int id; bool id_auto; struct device dev;  <-------------------- u32 num_resources; struct resource *resource; const struct platform_device_id *id_entry; char *driver_override; /* Driver name to force a match */ /* MFD cell pointer */ struct mfd_cell *mfd_cell; /* arch specific additions */ struct pdev_archdata archdata; }; further struct devic...
Read more

Getting started with pinctrl subsystem linux

-
A pin controller is a piece of hardware, usually a set of registers, that can control PINs. It may be able to multiplex, bias, set load capacitance, set drive strength, etc. for individual pins or groups of pins.   PINMUX   PINCONF - electronic properties e.g. high impedance, tristate, input pin to VDD or GND using a certain resistor value - pull up and pull down so that the pin has a stable value when nothing is driving the rail it is connected to, or when it’s unconnected. Pin configuration can be programmed by adding configuration entries into the mapping table     say that we have a group of pins dealing with an SPI interface on { 0, 8, 16, 24 }, and a group of pins dealing with an I2C interface on pins on { 24, 25 }.   make a group static const unsigned int spi0_pins[] = { 0, 8, 16, 24 }; static const unsigned int i2c0_pins[] = { 24, 25 }; //combine them  static const struct foo_group foo_groups[] = { { .name = "spi0_grp", ...
Read more

How to take systrace in android

-
Image
Install python 2.7 . As systrace does not support python 3.6. Python has its own package manager 'pip'. Use it to install new packages.         download python2.7 install pywin32 and six module on windows. Open power shell in C:/Python27/Scripts path,          pip install pywin32          pip install six   Above packages are not required to be installed on Ubuntu. But if you want other packages of python, we have to install python-pip, however not required for systrace.           sudo apt install python-pip        Then do pip list to see what packages are installed in Ubuntu. Install android studio        Start android studio from the downloaded android studio. That will        have a script to start an...
Read more