setenv bootargs ‘console=none root=/dev/mmcblk0pX rootwait init=/sbin/init’

But this parameter broke the output of my start code (a shell script) located on /etc/rc.d/…

Checking for solutions I go back to getty and added two parameters “-l” and “-n” to specify your program (or script) and prompt for nologin.

$ cat /etc/inittab

::sysinit:/etc/rc.d/rcS
::respawn:/sbin/getty -L ttymxc0 115200 vt100 -l /path/to/your/program -n
::ctrlaltdel:/sbin/reboot
::shutdown:/etc/rc.d/rcS stop
::restart:/sbin/init

Now I can see the serial output without kernel boot messages.

> setenv mtdparts mtdparts=nand0:0x80000@0x0(uboot),0x400000@0x80000(kernel),-@0x480000(root)

If you type mtd you I’ll see:

device nand0 <nand0>, # parts = 3
 #: name        size        offset        mask_flags
 0: uboot     0x00080000    0x00000000    0
 1: kernel    0x00400000    0x00080000    0
 2: root      0x1fb80000    0x00480000    0

UBI deal with volume and not partitions. Let’s create one.

> ubi part kernel

If you got some -22 error, like:

UBI error: ubi_read_volume_table: the layout volume was not found
UBI error: ubi_init: cannot attach mtd1
UBI error: ubi_init: UBI error: cannot initialize UBI, error -22
UBI init error -22

you need erase the NAND region [ nand erase 0x00080000 0x400000 ] and ubi part command again.

Next step is create the volume:

> ubi create kernel_vol
Creating dynamic volume kernel_vol of size 3354624

The value in bold is the max size of that volume in bytes (~3MB). Note that is less than the 4MB (0×400000) defined in mtdparts. This happens because UBI works with logical blocks instead (LEB) of physical ones (PEB).

In order to write the kernel you need transfer the image to u-boot. Since Ethernet  isn’t working in my board I choose between serial or mmc.  As serial is too slow to large files I opted to write the image on FAT partition on SD card and load through:

> mmcinfo
> fatload mmc 0 ${loadaddr} uImage

The output will be something like

reading uImage
2845120 bytes read

Finally write it:

> ubi write ${loadaddr} kernel_vol 0x2b69c0

You can check if everything went fine comparing

> ubi read 0x90AC0000 kernel_vol
> cmp.b ${loadaddr} 0x90ac0000  0x2b69c0
Total of 2845120 bytes were the same

0×90AC0000 is some place on RAM different from ${loadaddr} (check using echo ${loadaddr}).

Now that I know that somebody at Academia reads my posts I’ll put code with a _little_ documentation information together. Today we will do a quick review about kernel threads or to be more precisely, a way to perform some operations in the background. Kernel threads are standard process that:

1. Exist solely in kernel-space.
2. Don’t have an address space.
3. Don’t context switch to user-space.
4. Are schedulable and preemptable as normal process.

And let’s see an example:

#include <linux/init.h>
#include <linux/module.h>
#include <linux/sched.h>
#include <linux/delay.h>

static int thread3(void *unused) {

	int count3 = 0;
	while (count3 < 1000) {
		msleep(100);
		printk("Thread 3: %d\n", count3++);
	}
	return 0;
} 

static int thread2(void *unused) {

	int count2 = 0;
	while (count2 < 1000) {
		msleep(10);
		printk("Thread 2: %d\n", count2++);
	}
	return 0;
}

static int __init threads_init(void)
{
	int count1;
	int ret1, ret2;

	count1 = 0;
	ret1 = kernel_thread(thread2, NULL, CLONE_FS | CLONE_FILES | CLONE_SIGHAND | SIGCHLD );
	ret2 = kernel_thread(thread3, NULL, CLONE_FS | CLONE_FILES | CLONE_SIGHAND | SIGCHLD );

	printk("-- kernel thread: module init\n");
	printk("-- kernel thread: spawning thread 1 ret=%d\n", ret1);
	printk("-- kernel thread: spawning thread 2 ret=%d\n", ret2);	

	/*
	   Don't use "long" loops on init().
	   If you uncomment the snippet below the insmod will lock until while finishes
	*/ 

	/*
	while (count1 < 1000) {
		msleep(10);
		printk("Thread 1: %d\n", count1++);
     	}
	*/

    return 0;
}

static void __exit threads_exit(void)
{
    printk("-- kernel thread: module removed\n");
}

module_init(threads_init);
module_exit(threads_exit);

MODULE_AUTHOR("Tiago Maluta <maluta@unifei.edu.br");
MODULE_DESCRIPTION("kthreads examples");
MODULE_LICENSE("GPL");

Things from code that are found in docs…

• CLONE_FS means that parent and child share filesystem information.
• CLONE_FILES means that parent and child open files.
• CLONE_SIGHAND means that parent and child share signal handlers.

I use arbitrarily msleep() just to allow user “see” what is happening when dumping kernel ring buffer.

Um dos pontos que a estatística não mostra é que – mesmo sendo empregado de grandes empresas – boa parte do trabalho é feita fora do horário “comercial”, se alguém observar as datas de todos os commits e organizá-los pela frequência em horas, você tem o seguinte resultado para diferentes releases.

v2.6.33-rc7		v2.6.32			  v2.6.24	

Hora	Commits		Hora	Commits		Hora	Commits
0	8069		0	7808		0	4745
1	7052		1	6800		1	4155
2	4715		2	4556		2	2835
3	2717		3	2601		3	1395
4	2236		4	2096		4	898
5	1426		5	1193		5	559
6	1299		6	1134		6	381
7	1933		7	1723		7	580
8	4108		8	3798		8	1445
9	6429		9	6003		9	2261
10	8550		10	7969		10	3027
11	10284		11	9640		11	3946
12	9191		12	8403		12	3313
13	11728		13	11024		13	4522
14	13127		14	12340		14	5052
15	14281		15	13295		15	5813
16	13685		16	12721		16	5212
17	11486		17	10793		17	5050
18	7938		18	7334		18	3335
19	7354		19	6933		19	2850
20	7460		20	6953		20	3161
21	8561		21	8138		21	3436
22	8953		22	8498		22	3872
23	7741		23	7335		23	3843

Comando utilizado para gerar esses dados:

git log v2.6.33 | grep ^Date: | perl -pe ’s/^(?:\S+\s+){4}(\d+).*/$1/’ | sort -g | uniq -c

A conclusão é que muitos desenvolvedores fazem o código durante a noite e acordam tarde.

Um exemplo com o ext4:

./scripts/get_maintainer.pl -f  fs/ext4/

“Theodore Ts’o” <tytso@…>

Aneesh Kumar K.V <aneesh.kumar@…>

Eric Sandeen <sandeen@…>

linux-kernel@vger.kernel.org (lembre-se de se cadastrar e sempre de copiar a lista de email junto)

Um exemplo com o driver de video da Intel:

./scripts/get_maintainer.pl -f drivers/video/intelfb/

Sylvain Meyer <sylvain.meyer@…>

Andrew Morton <akpm@…>

Thomas Hilber <sparkie@…>

Stefan Husemann <shusemann@…>

Hannes Eder <hannes@…>

Paul Menzel <paulepanter@…>

linux-fbdev-devel@lists.sourceforge.net

linux-kernel@vger.kernel.org

Dar o retorno do uso de algum programa é ótimo para o desenvolvedor e garante a qualidade do sotfware.

# Set the ARCH and CROSS_COMPILE default values
ARCH            ?= arm
CROSS_COMPILE   ?= arm-unknown-linux-gnu-

Latest Linux (~2.6.31) turns it deprecated (commit 575543347b5baed0ca927cb90ba8807396fe9cc9). Now the settings are saved in two files named:

include/generated/kernel.arch
include/generated/kernel.cross

What changes?

Using the new way you don’t edit Makefile anymore, just need set once your definitions for cross compiling:

$ make ARCH=arm CROSS_COMPILE=arm-unknown-linux-gnu-

And for next builds just run make

$ make

This works both for plain builds and for O=… builds.