Device Drivers, Part 1: Linux Device Drivers for Your Girl Friend

78
43236
Driven...

Driven...

This series on Linux device drivers aims to present the usually technical topic in a way that is more interesting to a wider cross-section of readers.

“After a week of hard work, we finally got our driver working,” were Pugs’ first words when he met his girlfriend, Shweta.

“Why? What was your driver up to? Was he sick? And what hard work did you do?” asked Shweta. Confused, Pugs responded, “What are you talking about?”

Now it was Shweta’s turn to look puzzled, as she replied, “Why ask me? You tell me — which of your drivers are you talking about?”

When understanding dawned on him, Pugs groaned, “Ah c’mon! Not my car drivers — I am talking about a device driver on my computer.”

“I know about car and bus drivers, pilots, and even screwdrivers; but what is this ‘device driver’?” queried Shweta, puzzled.

That was all it took to launch Pugs into a passionate explanation of device drivers for the newbie — in particular, Linux device drivers, which he had been working on for many years.

Of drivers and buses

A driver drives, manages, controls, directs and monitors the entity under its command. What a bus driver does with a bus, a device driver does with a computer device (any piece of hardware connected to a computer) like a mouse, keyboard, monitor, hard disk, Web-camera, clock, and more.

Further, a “pilot” could be a person or even an automatic system monitored by a person (an auto-pilot system in airliners, for example). Similarly, a specific piece of hardware could be controlled by a piece of software (a device driver), or could be controlled by another hardware device, which in turn could be managed by a software device driver. In the latter case, such a controlling device is commonly called a device controller. This, being a device itself, often also needs a driver, which is commonly referred to as a bus driver.

General examples of device controllers include hard disk controllers, display controllers, and audio controllers that in turn manage devices connected to them. More technical examples would be an IDE controller, PCI controller, USB controller, SPI controller, I2C controller, etc. Pictorially, this whole concept can be depicted as in Figure 1.

Device and driver interaction
Figure 1: Device and driver interaction

Device controllers are typically connected to the CPU through their respectively named buses (collection of physical lines) — for example, the PCI bus, the IDE bus, etc. In today’s embedded world, we encounter more micro-controllers than CPUs; these are the CPU plus various device controllers built onto a single chip. This effective embedding of device controllers primarily reduces cost and space, making it suitable for embedded systems. In such cases, the buses are integrated into the chip itself. Does this change anything for the drivers, or more generically, on the software front?

The answer is, not much — except that the bus drivers corresponding to the embedded device controllers are now developed under the architecture-specific umbrella.

Drivers have two parts

Bus drivers provide hardware-specific interfaces for the corresponding hardware protocols, and are the bottom-most horizontal software layers of an operating system (OS). Over these sit the actual device drivers. These operate on the underlying devices using the horizontal layer interfaces, and hence are device-specific. However, the whole idea of writing these drivers is to provide an abstraction to the user, and so, at the other “end”, these do provide an interface (which varies from OS to OS). In short, a device driver has two parts, which are: a) device-specific, and b) OS-specific. Refer to Figure 2.

Linux device driver partition
Figure 2: Linux device driver partition

The device-specific portion of a device driver remains the same across all operating systems, and is more about understanding and decoding the device data sheets than software programming. A data sheet for a device is a document with technical details of the device, including its operation, performance, programming, etc. — in short a device user manual.

Later, I shall show some examples of decoding data sheets as well. However, the OS-specific portion is the one that is tightly coupled with the OS mechanisms of user interfaces, and thus differentiates a Linux device driver from a Windows device driver and from a MacOS device driver.

Verticals

In Linux, a device driver provides a “system call” interface to the user; this is the boundary line between the so-called kernel space and user-space of Linux, as shown in Figure 2. Figure 3 provides further classification.

Linux kernel overview
Figure 3: Linux kernel overview

Based on the OS-specific interface of a driver, in Linux, a driver is broadly classified into three verticals:

  • Packet-oriented or the network vertical
  • Block-oriented or the storage vertical
  • Byte-oriented or the character vertical

The CPU vertical and memory vertical, taken together with the other three verticals, give the complete overview of the Linux kernel, like any textbook definition of an OS: “An OS performs 5 management functions: CPU/process, memory, network, storage, device I/O.” Though these two verticals could be classified as device drivers, where CPU and memory are the respective devices, they are treated differently, for many reasons.

These are the core functionalities of any OS, be it a micro-kernel or a monolithic kernel. More often than not, adding code in these areas is mainly a Linux porting effort, which is typically done for a new CPU or architecture. Moreover, the code in these two verticals cannot be loaded or unloaded on the fly, unlike the other three verticals. Henceforth, when we talk about Linux device drivers, we mean to talk only about the latter three verticals in Figure 3.

Let’s get a little deeper into these three verticals. The network vertical consists of two parts: a) the network protocol stack, and b)the network interface card (NIC) device drivers, or simply network device drivers, which could be for Ethernet, Wi-Fi, or any other network horizontals. Storage, again, consists of two parts: a) File-system drivers, to decode the various formats on different partitions, and b) Block device drivers for various storage (hardware) protocols, i.e., horizontals like IDE, SCSI, MTD, etc.

With this, you may wonder if that is the only set of devices for which you need drivers (or for which Linux has drivers). Hold on a moment; you certainly need drivers for the whole lot of devices that interface with the system, and Linux does have drivers for them. However, their byte-oriented cessibility puts all of them under the character vertical — this is, in reality, the majority bucket. In fact, because of the vast number of drivers in this vertical, character drivers have been further sub-classified — so you have tty drivers, input drivers, console drivers, frame-buffer drivers, sound drivers, etc. The typical horizontals here would be RS232, PS/2, VGA, I2C, I2S, SPI, etc.

Multiple-vertical drivers

One final note on the complete picture (placement of all the drivers in the Linux driver ecosystem): the horizontals like USB, PCI, etc, span below multiple verticals. Why is that?

Simple — you already know that you can have a USB Wi-Fi dongle, a USB pen drive, and a USB-to-serial converter — all are USB, but come under three different verticals!

In Linux, bus drivers or the horizontals, are often split into two parts, or even two drivers: a) device controller-specific, and b) an abstraction layer over that for the verticals to interface, commonly called cores. A classic example would be the USB controller drivers ohci, ehci, etc., and the USB abstraction, usbcore.

Summing up

So, to conclude, a device driver is a piece of software that drives a device, though there are so many classifications. In case it drives only another piece of software, we call it just a driver. Examples are file-system drivers, usbcore, etc. Hence, all device drivers are drivers, but all drivers are not device drivers.

“Hey, Pugs, hold on; we’re getting late for class, and you know what kind of trouble we can get into. Let’s continue from here, later,” exclaimed Shweta.

Jumping up, Pugs finished his explanation: “Okay. This is the basic theory about device drivers. If you’re interested, later, I can show you the code, and all that we have been doing for the various kinds of drivers.” And they hurried towards their classroom.

78 COMMENTS

  1. Thanks for this article. The Related Posts section at the end of the article doesn’t really show related device driver posts. I hope there is some way you can display links to the other articles of this series in related posts so we can continue on with it. Also I know this will take time but ultimately we wish to see all the articles of this series published online on this website. Thank you once again.

  2. Hi Anil,

    I have worked on MCU ranging 8051 to cortex M3. but all on window with c/c++.

    But now I want to work on linux & mcu.Can you guide how/where to start

  3. Being a guy, as my girlfriend works on GPUs and knows all of this much better than me or most of you, I think the title should be “Device Drivers, Part 1: Linux Device Drivers for Your Boy Friend”.

    In other words, this title is utterly sexist and should be changed.

    • Writing an article doesn’t mean that world’s all guys & gals fall into this bucket. That’s just *a* perspective. There could be hundred & thousands of other perspectives, which we mere mortals would never be able to put down all together. So, what a big deal dude – all this sexist and … stuff. Moreover, these articles are not meant for sexism but for learning the fun way.

      • “Moreover, these articles are not meant for sexism but for learning the fun way.”

        Exactly! Then this one does not have to be sexist :), yet it sadly contributes to the overall nasty atmosphere concerning women and computers. Sorry for my rough first comment, but ordinary (and, in many cases, not intended) sexism is a real problem is the computer science community.

        • I hope you do believe, that there are people existing around of all varieties. It is just that you have come across a different set than what I have generally come across. And, a writer typically is bound to write from his own experiences in his own perspective – that doesn’t mean that everyone in this world are like that – it is just one of the perspectives. And hence, I do not even see a need or reason to talk about or bring out sexism in such technical scenarios.

    • The author assumes that the first article being introductory in nature is for the beginners. So there is no question of differentiating between the people as men or women. He has lined up other advanced driver explanations for those who want to learn more.

      And author himself has received several awards in his erstwhile organization Intel(at least 11 awards) for his contributions to the embedded domain.
      And he is an Entrepreneur too. The author is constantly engaged in sharing thoughts with the other experts and together they keep developing innovative solutions.

      Author explains the concepts of drivers with the narration of pugs and sweta in the backdrop just to make drivers interesting to learn for even those who are total novices to kernel. (BTW Pugs is the college nickname of the author !)

      Author is enthusiastic in imparting knowledge to all. He mixes fun with teaching and makes the complicated concepts available for consumption for the beginners and then the complicated concepts begin to build on the candidates and they are hugely benefited in the process.

      I can vouch for the above fact as I have attended his training sessions.

  4. Thanks for the tutorial. Today I just have started the part I. And this really started clearing some of the basic doubts that I had.Still I am getting confused on some points.

    -> Bus driver[H/W specific] and device driver[OS specific]. Still couldn’t understand the bus driver part. May be it is bcoz of the words ‘hardware protocols’ ‘hardware-specific interfaces ‘ etc. Could you clarify me with an example if possible.

    And do device driver programmers program bus driver as well? or there are two separate group of device drivers for bus and device driver.

    • Bus driver is also a driver – just that instead of a particular device, it manages a complete bus, on which the devices sit. Example: Pen Drive is a USB device needing a USB device driver (namely the usb-storage); USB itself is a bus, needing a bus driver (namely the usbcore).

      Device driver programmers write device specific drivers more often than bus specific drivers, but may need to write bus drivers, in case of new board portings.

  5. Hello,

    I am using RHEL-5, and install “kernel-2.6.15-1.2054_FC5.026test006.src.rpm”

    after writing the sample code ofd.c from

    “https://www.opensourceforu.com/2010/12/writing-your-first-linux-driver/”

    and then writing command “make” won’t respond any thing rather it says.”No target specified & no makefile found. Stop”

    • Where did you do make? I assume it was in the folder with ofd.c and I believe that you have not created the Makefile in that folder, as mentioned in the article mentioned by you. Please do so, and then change the kernel source path in the Makefile to your installed kernel source, before you do make.

  6. So in the case of a USB ethernet adapter, would it be correct to say that the actual network traffic, as in: ethernet frames, are “encapsulated” in USB-specific messages?
    Or is it more a case like the following:
    1. The driver and controller/device do some kind of USB handshake,
    2. The driver then says: “I’m now going to send some end device specific data”
    3. The driver sends a datastream which is meaningless to the host (USB) controller, but just blindly passes it through to the end device.
    4. The driver then says “now I’m finished sending device specific data”
    Just wondering what this would look like on the bus/controller level at that point in time…

      • Ah I see, thanks and thanks for the excellent work you’ve done on these articles! :)

        One more question though, if I may, because this is another thing that isn’t quite clear to me. You speak of this abstraction layer sitting “on top” of the controller drivers. What does it specifically abstract for the higher level (peripheral) driver?
        Is it there to abstract the underlying bus technologie? Like say, in case of the evolution of USB2.0 to USB3.0 (or PCI-e 2.0 to PCI-e 3.0 for that matter)? So the higher level driver doesn’t have to be concerned about whether the peripheral device is connected to a USB2 or USB3 port?
        And is the Windows equivalent of this the HAL (Hardware Abstraction Layer)?

        • Yes, it is a sort of HAL. More than abstracting bus technology, it is abstracting the architecture. For example, USB host controller may be on PCI on x86 but on AMBA on ARM. So, the controller driver itself is architecture-specific bus dependent driver, so that all the layers above it could be independent of it.

  7. #here i write the ANKIT in the 1 quantum of the scull and by this i am able to write in the kernel layer….
    RCS file: application.c,v
    Working file: application.c
    head: 1.7
    branch:
    locks: strict
    access list:
    symbolic names:
    keyword substitution: kv
    total revisions: 7; selected revisions: 7
    description:
    make the open call.
    —————————-
    revision 1.7
    date: 2014/05/18 09:45:03; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.6
    date: 2014/05/18 09:10:37; author: root; state: Exp; lines: +8 -8
    *** empty log message ***
    —————————-
    revision 1.5
    date: 2014/05/18 04:26:31; author: root; state: Exp; lines: +8 -7
    *** empty log message ***
    —————————-
    revision 1.4
    date: 2014/05/13 15:18:27; author: root; state: Exp; lines: +1 -0
    include the string.h
    —————————-
    revision 1.3
    date: 2014/05/13 15:17:20; author: root; state: Exp; lines: +5 -1
    use the write system call.
    —————————-
    revision 1.2
    date: 2014/05/08 06:06:12; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.1
    date: 2014/05/08 06:05:16; author: root; state: Exp;
    Initial revision
    =============================================================================

    RCS file: declaration.h,v
    Working file: declaration.h
    head: 1.4
    branch:
    locks: strict
    access list:
    symbolic names:
    keyword substitution: kv
    total revisions: 4; selected revisions: 4
    description:
    declare the extern variable as dev, nod, minorno and majorno.
    —————————-
    revision 1.4
    date: 2014/05/07 03:36:11; author: root; state: Exp; lines: +5 -0
    define extern for new_dev quantum_size qset_size data_size and device_size
    —————————-
    revision 1.3
    date: 2014/05/06 06:25:08; author: root; state: Exp; lines: +2 -2
    *** empty log message ***
    —————————-
    revision 1.2
    date: 2014/05/06 03:29:21; author: root; state: Exp; lines: +6 -0
    declare a structure SkullDev with a structure cdev define in cdev,h
    —————————-
    revision 1.1
    date: 2014/05/06 02:07:46; author: root; state: Exp;
    Initial revision
    =============================================================================

    RCS file: devrelease.c,v
    Working file: devrelease.c
    head: 1.1
    branch:
    locks: strict
    access list:
    symbolic names:
    keyword substitution: kv
    total revisions: 1; selected revisions: 1
    description:
    —————————-
    revision 1.1
    date: 2014/05/08 04:56:14; author: root; state: Exp;
    Initial revision
    =============================================================================

    RCS file: header.h,v
    Working file: header.h
    head: 1.16
    branch:
    locks: strict
    access list:
    symbolic names:
    keyword substitution: kv
    total revisions: 16; selected revisions: 16
    description:
    including the linux/modules.
    includding the linux/init.
    define GPL.
    —————————-
    revision 1.16
    date: 2014/05/18 04:25:51; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.15
    date: 2014/05/15 10:20:14; author: root; state: Exp; lines: +1 -0
    include asm-generic uaccess.h header file for the copy_to_user.
    —————————-
    revision 1.14
    date: 2014/05/13 13:27:30; author: root; state: Exp; lines: +2 -2
    *** empty log message ***
    —————————-
    revision 1.13
    date: 2014/05/11 14:39:30; author: root; state: Exp; lines: +4 -0
    make the macro for the header because its included multiple times in the program…..and we have to use __header as macro otherwise it leads to a error.
    —————————-
    revision 1.12
    date: 2014/05/07 03:41:07; author: root; state: Exp; lines: +4 -4
    define MACROS
    —————————-
    revision 1.11
    date: 2014/05/07 03:31:31; author: root; state: Exp; lines: +27 -1
    define macro for QSET_SIZE, QUANTUM_SIZE, DATA_SIZE and DEVICE_SIZE.
    define structure ScullQset.
    —————————-
    revision 1.10
    date: 2014/05/06 07:17:41; author: root; state: Exp; lines: +2 -1
    include the moduleparam.h for the module_param i.e. passing the arguments in the kernel layer.
    —————————-
    revision 1.9
    date: 2014/05/06 06:36:59; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.8
    date: 2014/05/06 06:33:38; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.7
    date: 2014/05/06 06:31:04; author: root; state: Exp; lines: +2 -0
    include slab.h for the kmalloc and kdev_t for the Macros MINOR and MAJOR.
    —————————-
    revision 1.6
    date: 2014/05/06 06:26:43; author: root; state: Exp; lines: +1 -0
    include cdev.h header file for struct cdev
    —————————-
    revision 1.5
    date: 2014/05/06 06:24:27; author: root; state: Exp; lines: +4 -0
    declare the struct SkullDev in header instead of in declaration.h.
    —————————-
    revision 1.4
    date: 2014/05/06 02:12:04; author: root; state: Exp; lines: +4 -4
    redfine the macro in header file again.
    —————————-
    revision 1.3
    date: 2014/05/06 02:03:21; author: root; state: Exp; lines: +17 -0
    include the linux/fs.h
    difine the macros DEVNAME N
    —————————-
    revision 1.2
    date: 2014/05/06 01:04:44; author: root; state: Exp; lines: +1 -1
    correct the license GPL
    —————————-
    revision 1.1
    date: 2014/05/06 00:58:23; author: root; state: Exp;
    Initial revision
    =============================================================================

    RCS file: cleanup.c,v
    Working file: cleanup.c
    head: 1.6
    branch:
    locks: strict
    access list:
    symbolic names:
    keyword substitution: kv
    total revisions: 6; selected revisions: 6
    description:
    use the module_exit(exitfn) and define the declaration of exitfn
    —————————-
    revision 1.6
    date: 2014/05/06 03:28:56; author: root; state: Exp; lines: +1 -0
    kfree the memory allocated during kmalloc.
    —————————-
    revision 1.5
    date: 2014/05/06 02:35:09; author: root; state: Exp; lines: +2 -1
    include the declaration.h in this.
    —————————-
    revision 1.4
    date: 2014/05/06 02:32:49; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.3
    date: 2014/05/06 02:30:41; author: root; state: Exp; lines: +1 -1
    the function name we use as __
    —————————-
    revision 1.2
    date: 2014/05/06 02:06:30; author: root; state: Exp; lines: +3 -1
    call unregister_chrdev_region for the unregistration.
    —————————-
    revision 1.1
    date: 2014/05/06 01:00:24; author: root; state: Exp;
    Initial revision
    =============================================================================

    RCS file: devopen.c,v
    Working file: devopen.c
    head: 1.4
    branch:
    locks: strict
    access list:
    symbolic names:
    keyword substitution: kv
    total revisions: 4; selected revisions: 4
    description:
    write the defination of the devopen
    —————————-
    revision 1.4
    date: 2014/05/13 14:35:37; author: root; state: Exp; lines: +22 -0
    call the trim function.
    write the defination of the trim function.
    —————————-
    revision 1.3
    date: 2014/05/11 15:53:59; author: root; state: Exp; lines: +3 -2
    include declaration.h
    —————————-
    revision 1.2
    date: 2014/05/11 15:26:48; author: root; state: Exp; lines: +12 -0
    introduce container of which is used to point out the memory of external device.
    —————————-
    revision 1.1
    date: 2014/05/08 04:56:14; author: root; state: Exp;
    Initial revision
    =============================================================================

    RCS file: devwrite.c,v
    Working file: devwrite.c
    head: 1.12
    branch:
    locks: strict
    access list:
    symbolic names:
    keyword substitution: kv
    total revisions: 12; selected revisions: 12
    description:
    write the devwrite function.
    map the devwrite function.
    —————————-
    revision 1.12
    date: 2014/05/18 09:42:30; author: root; state: Exp; lines: +1 -1
    return the size-ret to the write system call of the application layer i.e application.c program.
    —————————-
    revision 1.11
    date: 2014/05/18 09:16:13; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.10
    date: 2014/05/18 05:28:36; author: root; state: Exp; lines: +7 -7
    make the scullqset_creat qset_create and quantum_creat
    make the 1 quantum for the world emblogic.
    and make the 1 scullqset for the world emblogic
    —————————-
    revision 1.9
    date: 2014/05/18 04:26:34; author: root; state: Exp; lines: +31 -15
    *** empty log message ***
    —————————-
    revision 1.8
    date: 2014/05/15 14:20:59; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.7
    date: 2014/05/15 11:56:09; author: root; state: Exp; lines: +1 -0
    *** empty log message ***
    —————————-
    revision 1.6
    date: 2014/05/15 11:53:16; author: root; state: Exp; lines: +6 -0
    *** empty log message ***
    —————————-
    revision 1.5
    date: 2014/05/15 11:37:23; author: root; state: Exp; lines: +7 -5
    *** empty log message ***
    —————————-
    revision 1.4
    date: 2014/05/15 10:21:47; author: root; state: Exp; lines: +28 -20
    write the defination for the create_scullqset, create_qset, and create_quantum.
    —————————-
    revision 1.3
    date: 2014/05/13 17:03:30; author: root; state: Exp; lines: +73 -0
    *** empty log message ***
    —————————-
    revision 1.2
    date: 2014/05/13 15:12:20; author: root; state: Exp; lines: +4 -1
    give the defination of devwrite()
    —————————-
    revision 1.1
    date: 2014/05/13 15:07:20; author: root; state: Exp;
    Initial revision
    =============================================================================

    RCS file: init.c,v
    Working file: init.c
    head: 1.31
    branch:
    locks: strict
    access list:
    symbolic names:
    keyword substitution: kv
    total revisions: 31; selected revisions: 31
    description:
    use the module_init(initfn) and define the declaration of initfn
    —————————-
    revision 1.31
    date: 2014/05/18 04:26:28; author: root; state: Exp; lines: +9 -0
    *** empty log message ***
    —————————-
    revision 1.30
    date: 2014/05/13 13:27:32; author: root; state: Exp; lines: +1 -0
    *** empty log message ***
    —————————-
    revision 1.29
    date: 2014/05/08 03:47:42; author: root; state: Exp; lines: +4 -4
    *** empty log message ***
    —————————-
    revision 1.28
    date: 2014/05/08 03:31:38; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.27
    date: 2014/05/08 03:30:27; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.26
    date: 2014/05/08 03:28:42; author: root; state: Exp; lines: +4 -2
    *** empty log message ***
    —————————-
    revision 1.25
    date: 2014/05/08 03:20:39; author: root; state: Exp; lines: +2 -1
    *** empty log message ***
    —————————-
    revision 1.24
    date: 2014/05/08 03:15:11; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.23
    date: 2014/05/08 03:05:46; author: root; state: Exp; lines: +3 -3
    *** empty log message ***
    —————————-
    revision 1.22
    date: 2014/05/08 03:00:33; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.21
    date: 2014/05/08 02:59:54; author: root; state: Exp; lines: +2 -2
    *** empty log message ***
    —————————-
    revision 1.20
    date: 2014/05/08 02:13:15; author: root; state: Exp; lines: +4 -3
    replace by [] and -> by .
    —————————-
    revision 1.19
    date: 2014/05/08 02:06:31; author: root; state: Exp; lines: +11 -7
    print the major and minor number for all nodes.
    —————————-
    revision 1.18
    date: 2014/05/07 03:41:18; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.17
    date: 2014/05/07 03:34:19; author: root; state: Exp; lines: +29 -8
    call MKDEV
    call cdev_add
    —————————-
    revision 1.16
    date: 2014/05/06 07:34:43; author: root; state: Exp; lines: +2 -1
    *** empty log message ***
    —————————-
    revision 1.15
    date: 2014/05/06 07:32:45; author: root; state: Exp; lines: +2 -2
    *** empty log message ***
    —————————-
    revision 1.14
    date: 2014/05/06 07:23:46; author: root; state: Exp; lines: +2 -2
    define the module_param above the init_func
    —————————-
    revision 1.13
    date: 2014/05/06 07:18:42; author: root; state: Exp; lines: +2 -0
    define module_param call
    —————————-
    revision 1.12
    date: 2014/05/06 06:44:23; author: root; state: Exp; lines: +2 -1
    initilize the struct ScullDev above the function name
    —————————-
    revision 1.11
    date: 2014/05/06 06:24:59; author: root; state: Exp; lines: +5 -5
    *** empty log message ***
    —————————-
    revision 1.10
    date: 2014/05/06 03:34:40; author: root; state: Exp; lines: +5 -5
    *** empty log message ***
    —————————-
    revision 1.9
    date: 2014/05/06 03:31:46; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.8
    date: 2014/05/06 03:26:36; author: root; state: Exp; lines: +9 -0
    call kmalloc and allocate the space uptp the structure SkullDev.
    SkullDev is used to map the memory of the Device Driver.
    defint the *skulldev which is extern and define in the struct SkullDev
    —————————-
    revision 1.7
    date: 2014/05/06 02:55:59; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.6
    date: 2014/05/06 02:31:58; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.5
    date: 2014/05/06 02:30:20; author: root; state: Exp; lines: +2 -2
    the fuction name we use __
    —————————-
    revision 1.4
    date: 2014/05/06 02:24:27; author: root; state: Exp; lines: +2 -2
    *** empty log message ***
    —————————-
    revision 1.3
    date: 2014/05/06 02:17:40; author: root; state: Exp; lines: +5 -2
    initialize the extern variable to MACRO within the function itself
    —————————-
    revision 1.2
    date: 2014/05/06 02:05:12; author: root; state: Exp; lines: +21 -1
    call alloc chr_dev region to register the DD into kernel and then call MAJOR and MINOR to display ita majorno and mno.
    —————————-
    revision 1.1
    date: 2014/05/06 00:59:15; author: root; state: Exp;
    Initial revision
    =============================================================================

    RCS file: prototype.h,v
    Working file: prototype.h
    head: 1.6
    branch:
    locks: strict
    access list:
    symbolic names:
    keyword substitution: kv
    total revisions: 6; selected revisions: 6
    description:
    mention the prototype of devopen and devrelease
    —————————-
    revision 1.6
    date: 2014/05/15 11:37:21; author: root; state: Exp; lines: +1 -1
    *** empty log message ***
    —————————-
    revision 1.5
    date: 2014/05/15 10:20:55; author: root; state: Exp; lines: +3 -5
    write the prototype fot the create_scullqset, create_qset, create_quantun.
    —————————-
    revision 1.4
    date: 2014/05/13 17:03:23; author: root; state: Exp; lines: +4 -0
    *** empty log message ***
    —————————-
    revision 1.3
    date: 2014/05/13 15:11:44; author: root; state: Exp; lines: +1 -0
    write the prototype of the devwrite()
    —————————-
    revision 1.2
    date: 2014/05/13 14:34:33; author: root; state: Exp; lines: +1 -0
    give the prototype of trim function used in the devopen and this trim function is used to flush the data, when its open in WR mode.
    —————————-
    revision 1.1
    date: 2014/05/08 04:56:14; author: root; state: Exp;
    Initial revision
    =============================================================================

    RCS file: fileopr.h,v
    Working file: fileopr.h
    head: 1.1
    branch:
    locks: strict
    access list:
    symbolic names:
    keyword substitution: kv
    total revisions: 1; selected revisions: 1
    description:
    this is the mapping of the open close and write system call from the application layer to the kernel system call…
    —————————-
    revision 1.1
    date: 2014/05/18 09:50:00; author: root; state: Exp;
    Initial revision
    =============================================================================

      • Hi Anil

        Thank you for the tutorial. I saw the prompt response you have given to people’s queries. This excites me to ask my query:

        I have worked on linux for few years but didn’t get chance to dig in to linux kernel device driver. To build my confidence in the same, I would like to define an exercise for device driver on beaglebone black. Can you please guide me in defining the exercise?

        Thank you in advance !

  8. Hi,
    Thanks for the wonderful tutorial.
    But i have a doubt, may be a silly even. Want to know whether the actual Bus IP and the Device controller(Bus controller) are one and the same. For eg: The microcontroller which i am working with, has I2C interface IP, SPI interface IP supplied by Synopsys. Does that mean it has the I2C device controller(Bus driver) and SPI device controller(Bus driver) as well by default ? From your text, it appears both are different.If so, does not all compliant device require device controllers to have connection with their interface buses ? Can a I2C compliant A/D converter directly talk with I2C interface IP itself, without a I2C device controller ?

  9. can you tell me that what is the use of device controllers and bus drivers? why device drivers are not directly interact with devices? as in fig 1 and 2. plz reply…

    • Paras: this is what we call a layered architecture of Linux Kernel. No, program can access the devices directly. They need to access the device through layers. This is for the security of the kernel.

  10. What is the exact difference between Firmwares and Device drivers and Writing the Program directly on device registers????

  11. Hello sir, I need your help. I couldn’t be able to compile the C file which is having basic “Hello_world.c” kind of program for Kernel Module Programming.

    Whenever i tried to compile using cc, gcc even make, I’m getting the error like shown below

    ” fatal error linux/nodules.h: No such file or directory”

    In my /src directory i can find two files

    linux-headers-3.13.0-24
    linux-headers-3.13.0-24-generic

    Inside the include i can see the files init.h, modules.h……..

    But while compiling the program it is showing the above error

    After installing Linux on my Virtual machine I never made any manual updates or installation.
    I’m running Ubuntu 14.04 on VMware
    My kernel version is 3.13.0.24-generic

    Do i need to install something ????

    Am i doing any mistake in Makefile???? I dont know much about makefile..

    Please help me with this. Thank you in advance.

  12. This article is a good introduction for beginners in driver development. But, the first part is obnoxious. I have seen girls who know about device drivers, modules, etc. much in depth than their guys. So, the time is right to stop stereotyping people.

  13. You hit the nail on the head. I had a good experience filling forms online and happy to share it with you. I’ve found PDFfiller – online service for forms filling. It’s pretty easy to use and pretty cheap. You can find fillable IRS 2290 here http://goo.gl/wMFeVH

  14. I doubt you have a girlfriend. I believe you’re just imagining a situation where a girl will come and ask you to teach her linux device drivers. It takes a lot more than fancy coding to impress a girl. I’m have an engineer/model/entrepreneur girlfriend and not once has she asked me about code.

    • Then you might have not intentionally mingled with female programmers. They are just as fun the as the ones in other fields.

      And I did dated one through fancy coding. :D

    • The goal of the tutorial is to understand the theory behind Linux device drivers. So, stop cribbing about who has a gf or who doesn’t and show some appreciation, Duh!

  15. Thanks, Pug! Eight years later and your write up is still helping others like me! BTW, is technical writing still your forte? I’d be interested in your other works.

LEAVE A REPLY

Please enter your comment!
Please enter your name here