2016年12月4日 星期日

gmp-4.1.4 build error "automatic de-ANSI-fication support has been removed"


porting某些package時會出現"automatic de-ANSI-fication support has been removed",這是因為automake1.12之後,就不再支援該功能了,也就是不再支援ansi2knr這個選項了。 而且建議將AM_C_PROTOTYPES改成AC_C_PROTOTYPES的寫法,其範例如下
--- a/Makefile.am
+++ b/Makefile.am
@@ -20,7 +20,7 @@
# MA 02111-1307, USA.
-AUTOMAKE_OPTIONS = gnu no-dependencies $(top_builddir)/ansi2knr
+AUTOMAKE_OPTIONS = gnu no-dependencies

修改configure.in
--- a/configure.in
+++ b/configure.in
@@ -1501,7 +1501,9 @@ echo "      MPN_PATH=\"$path\""
# Automake ansi2knr support.
-AM_C_PROTOTYPES
+AC_C_PROTOTYPES
+AC_HEADER_STDC
+AC_CHECK_HEADERS("string.h")


    參考資料:
  1. https://autotools.io/forwardporting/automake.html




2016年10月9日 星期日

Install Node.js on Openembedded


基本上Node.js在Openembedded上的recipes都已經寫好了,只要clone下來,並且加入IMAGE_INSTALL列表即可
brook@vista:~/projects/poky$ git clone https://github.com/imyller/meta-nodejs.git
Cloning into 'meta-nodejs'...
remote: Counting objects: 1575, done.
remote: Compressing objects: 100% (14/14), done.
remote: Total 1575 (delta 9), reused 0 (delta 0), pack-reused 1561
Receiving objects: 100% (1575/1575), 241.91 KiB | 368.00 KiB/s, done.
Resolving deltas: 100% (846/846), done.
Checking connectivity... done.
brook@vista:~/projects/poky$ . oe-init-build-env
You had no conf/local.conf file. This configuration file has therefore been
created for you with some default values. You may wish to edit it to, for
example, select a different MACHINE (target hardware). See conf/local.conf
for more information as common configuration options are commented.

You had no conf/bblayers.conf file. This configuration file has therefore been
created for you with some default values. To add additional metadata layers
into your configuration please add entries to conf/bblayers.conf.

The Yocto Project has extensive documentation about OE including a reference
manual which can be found at:
    http://yoctoproject.org/documentation

For more information about OpenEmbedded see their website:
    http://www.openembedded.org/


### Shell environment set up for builds. ###

You can now run 'bitbake <target>'

Common targets are:
    core-image-minimal
    core-image-sato
    meta-toolchain
    meta-ide-support

You can also run generated qemu images with a command like 'runqemu qemux86'

brook@vista:~/projects/poky/build$ vim conf/bblayers.conf
...下段說明
brook@vista:~/projects/poky/build$ vim conf/local.conf
...下段說明
brook@vista:~/projects/poky/build$ bitbake core-image-minimal
WARNING: Host distribution "Ubuntu-16.04" has not been validated with this version of the build system; you may possibly experience unexpected failures. It is recommended that you use a tested distribution.
Parsing recipes: 100% |#############################################################################################| Time: 00:00:28
Parsing of 876 .bb files complete (0 cached, 876 parsed). 1316 targets, 49 skipped, 0 masked, 0 errors.
NOTE: Resolving any missing task queue dependencies

Build Configuration:
BB_VERSION        = "1.30.0"
BUILD_SYS         = "x86_64-linux"
NATIVELSBSTRING   = "Ubuntu-16.04"
TARGET_SYS        = "i586-poky-linux"
MACHINE           = "qemux86"
DISTRO            = "poky"
DISTRO_VERSION    = "2.1.1"
TUNE_FEATURES     = "m32 i586"
TARGET_FPU        = ""
meta
meta-poky
meta-yocto-bsp    = "krogoth:8c69f7d56cbd496aa01ba0738675a170826a536b"
meta-nodejs       = "master:848b0defe8eba6e7ffa97b66e4316c17c92be9d4"
...
brook@vista:~/projects/poky/build$ ./tmp/sysroots/x86_64-linux/usr/bin/qemu-system-i386 -nographic -kernel ./tmp/deploy/images/qemux86/bzImage-qemux86.bin -cpu qemu32 -drive file=./tmp/deploy/images/qemux86/core-image-minimal-qemux86.ext4,if=virtio,format=raw -show-cursor -usb -usbdevice tablet -vga vmware -no-reboot -m 256 -append "vga=0 uvesafb.mode_option=640x480-32 root=/dev/vda rw mem=256M ip=192.168.7.2::192.168.7.1:255.255.255.0 oprofile.timer=1 rootfstype=ext4 "

Poky (Yocto Project Reference Distro) 2.1.1 qemux86 /dev/ttyS0

qemux86 login: root
root@qemux86:~# uname -a
Linux qemux86 4.4.11-yocto-standard #1 SMP PREEMPT Sun Oct 9 19:29:24 CST 2016 i686 GNU/Linux
root@qemux86:~# echo 'console.log("hello world");' > node.js
root@qemux86:~# node node.js
hello world


conf/bblayers.conf修改的內容
BBLAYERS ?= " \
  /home/brook/projects/poky/meta \
  /home/brook/projects/poky/meta-poky \
  /home/brook/projects/poky/meta-yocto-bsp \
  ${TOPDIR}/../meta-nodejs \
  "


conf/local.conf修改的內容
# This sets the default machine to be qemux86 if no other machine is selected:
MACHINE ??= "qemux86"
CORE_IMAGE_EXTRA_INSTALL += "nodejs"


    參考資料:
  1. meta-nodejs, Node.js的openembedded recipes
  2. JavaScript for IoT: Blinking LED on Raspberry Pi with Node.js , Node.js的demo影片。
  3. 「Node.js & IoT: Zero to One」 是一本 Node.js 的入門電子書,內容定位為基礎教學,目標是介紹 Node.js 以及 Node.js + IoT 相關技術主題,每個主題都從基本觀念(Zero)開始,介紹到能撰寫簡單的程式為止(One)。不過目前只有介紹JS的基本語法。
  4. Addons教你如何擴充JS,用C/C++寫一個Node.js module。




2016年10月8日 星期六

Using openembedded to build QEMU


這篇文章主要來說明如何使用Openembedded/Krogoth來build出QEMU for ARM。
步驟大概就是:
先下載yp-krogoth之後,source目錄下的oe-init-build-env,接著再編輯conf/local.conf,最後就是執行bitbake core-image-minimal。
我的Krogoth最後一個commit ID是e93596f。
brook@vista:~/projects$ git clone -b krogoth git://git.yoctoproject.org/poky.git yp-krogoth
Cloning into 'yp-krogoth'...
remote: Counting objects: 339411, done.
remote: Compressing objects: 100% (82247/82247), done.
remote: Total 339411 (delta 251486), reused 338934 (delta 251035)
Receiving objects: 100% (339411/339411), 127.65 MiB | 4.49 MiB/s, done.
Resolving deltas: 100% (251486/251486), done.
Checking connectivity... done.
brook@vista:~/projects$ cd yp-krogoth/
brook@vista:~/projects/yp-krogoth$ git log --oneline -1
e93596f binutils: fix AR issue when opkg is unpacking IPKs containing empty entries
brook@vista:~/projects/yp-krogoth$ . oe-init-build-env
You had no conf/local.conf file. This configuration file has therefore been
created for you with some default values. You may wish to edit it to, for
example, select a different MACHINE (target hardware). See conf/local.conf
for more information as common configuration options are commented.

You had no conf/bblayers.conf file. This configuration file has therefore been
created for you with some default values. To add additional metadata layers
into your configuration please add entries to conf/bblayers.conf.

The Yocto Project has extensive documentation about OE including a reference
manual which can be found at:
    http://yoctoproject.org/documentation

For more information about OpenEmbedded see their website:
    http://www.openembedded.org/


### Shell environment set up for builds. ###

You can now run 'bitbake <target>'

Common targets are:
    core-image-minimal
    core-image-sato
    meta-toolchain
    meta-ide-support

You can also run generated qemu images with a command like 'runqemu qemux86'
brook@vista:~/projects/yp-krogoth/build$ vim conf/local.conf
...內容請看一段
brook@vista:~/projects/yp-krogoth/build$ bitbake core-image-minimal
WARNING: Host distribution "Ubuntu-16.04" has not been validated with this version of the build system; you may possibly experience unexpected failures. It is recommended that you use a tested distribution.
Parsing recipes...done.
Parsing of 871 .bb files complete (0 cached, 871 parsed). 1301 targets, 66 skipped, 0 masked, 0 errors.
NOTE: Resolving any missing task queue dependencies

Build Configuration:
BB_VERSION        = "1.30.0"
BUILD_SYS         = "x86_64-linux"
NATIVELSBSTRING   = "Ubuntu-16.04"
TARGET_SYS        = "arm-poky-linux-gnueabi"
MACHINE           = "qemuarm"
DISTRO            = "poky"
DISTRO_VERSION    = "2.1.1"
TUNE_FEATURES     = "arm armv5 thumb dsp"
TARGET_FPU        = "soft"
meta
meta-poky
meta-yocto-bsp    = "krogoth:e93596fe74927e2e2f4dd7f671994ccb9744cff8"

NOTE: Fetching uninative binary shim from http://downloads.yoctoproject.org/releases/uninative/1.0.1/x86_64-nativesdk-libc.tar.bz2;sha256sum=acf1e44a0ac2e855e81da6426197d36358bf7b4e88e552ef933128498c8910f8
NOTE: Preparing RunQueue
NOTE: Executing SetScene Tasks
...略
brook@vista:~/projects/yp-krogoth/build$ ./tmp/sysroots/x86_64-linux/usr/bin/qemu-system-arm -kernel ./tmp/deploy/images/qemuarm/zImage-qemuarm.bin -M versatilepb -drive file=/home/jpr/projects/yp-krogoth/build/tmp/deploy/images/qemuarm/core-image-minimal-qemuarm.ext4,if=virtio,format=raw -no-reboot -show-cursor -usb -usbdevice tablet -m 128 -nographic -append 'root=/dev/vda rw console=ttyAMA0,115200 console=tty ip=192.168.7.2::192.168.7.1:255.255.255.0 mem=128M highres=off rootfstype=ext4 ' 
audio: Could not init `oss' audio driver
ALSA lib ../../alsa-lib-1.1.0/src/confmisc.c:768:(parse_card) cannot find card '0'
...略
qemuarm login: root
root@qemuarm:~# uname -a
Linux qemuarm 4.4.11-yocto-standard #1 PREEMPT Sun Oct 9 11:36:44 CST 2016 armv5tejl GNU/Linux


conf/local.conf相關修改
change MACHINE for "qemux86" to "qemuarm"
MACHINE ?= "qemuarm"
#MACHINE ?= "qemuarm64"
#MACHINE ?= "qemumips"
#MACHINE ?= "qemumips64"
#MACHINE ?= "qemuppc"
#MACHINE ?= "qemux86"
#MACHINE ?= "qemux86-64"
#
# There are also the following hardware board target machines included for
# demonstration purposes:
#
#MACHINE ?= "beaglebone"
#MACHINE ?= "genericx86"
#MACHINE ?= "genericx86-64"
#MACHINE ?= "mpc8315e-rdb"
#MACHINE ?= "edgerouter"
#
# This sets the default machine to be qemux86 if no other machine is selected:
#MACHINE ??= "qemux86"
...
commout out these 2 lines
#PACKAGECONFIG_append_pn-qemu-native = " sdl"
#PACKAGECONFIG_append_pn-nativesdk-qemu = " sdl"



    參考資料:
  1. Yocto Project Quick Start
  2. Yocto Project Build System/Download




2016年10月1日 星期六

QC的partition table大小設定與實際大小對應關係


這裡記錄一下QC的9x40 partition table configure file(common/build/partition_nand.xml)size與實際的大小怎麼換算。 大概就兩類tag,一個tag是size_blks,以eraseblock為單位。另一個是tag是size_kb,所以實際大小還要轉成eraseblock。 如:
/ # cat /proc/mtd
dev:    size   erasesize  name
mtd0: 00280000 00040000 "sbl"
mtd1: 00280000 00040000 "mibib"
mtd2: 00c00000 00040000 "efs2"
mtd3: 00140000 00040000 "tz"
mtd4: 00100000 00040000 "rpm"
mtd5: 00140000 00040000 "aboot"
...

partition table configure file內容
<partition>
    <name length="16" type="string">0:SBL</name>
    <size_blks length="4">0x8</size_blks>
    <pad_blks length="4">0x2</pad_blks>
    ...
</partition>
...
<partition>
    <name length="16" type="string">0:EFS2</name>
    <size_kb length="4">11264</size_kb>
    <pad_kb length="4">1024</pad_kb>
    ....
</partition>
SBL的設定為0x8+0x2個eraseblock,所以/proc/mtd大小為 (8 + 2) * 256 * 1024 = 0x280000。 EFS2的設定為11264+1024 KB,所以/proc/mtd大小為ceiling(11264 / 256)+ceiling(1024 / 256) = 48個eraseblock,所以/proc/mtd大小為48 * 256 * 1024 = 0xC0000。 eraseblock就是上面的erasesize,0x40000 / 1024 = 256kb。


2016年9月17日 星期六

build 32bit app on 64bit Machine -- /usr/bin/ld: skipping incompatible


我將Ubuntu從14.04換到16.04後,系統好像移除了一些package,導致build 32bit的程式會failed。 安裝lib32gcc-4.7-dev之後就解了。
brook@vista:~$ uname -a
Linux vista 4.4.0-36-generic #55-Ubuntu SMP Thu Aug 11 18:01:55 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
brook@vista:~$ cat /etc/issue
Ubuntu 16.04.1 LTS \n \l
brook@vista:~$ gcc -m32 x.c
/usr/bin/ld: skipping incompatible /usr/lib/gcc/x86_64-linux-gnu/4.7/libgcc.a when searching for -lgcc
/usr/bin/ld: cannot find -lgcc
/usr/bin/ld: skipping incompatible /usr/lib/gcc/x86_64-linux-gnu/4.7/libgcc_s.so when searching for -lgcc_s
/usr/bin/ld: cannot find -lgcc_s
collect2: error: ld returned 1 exit status
brook@vista:~$ gcc -v
Using built-in specs.
COLLECT_GCC=gcc
COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-linux-gnu/4.7/lto-wrapper
Target: x86_64-linux-gnu
Configured with: ../src/configure -v --with-pkgversion='Ubuntu/Linaro 4.7.4-3ubuntu12' --with-bugurl=file:///usr/share/doc/gcc-4.7/README.Bugs --enable-languages=c,c++,go,fortran,objc,obj-c++ --prefix=/usr --program-suffix=-4.7 --enable-shared --enable-linker-build-id --libexecdir=/usr/lib --without-included-gettext --enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.7 --libdir=/usr/lib --enable-nls --with-sysroot=/ --enable-clocale=gnu --enable-libstdcxx-debug --enable-gnu-unique-object --disable-libmudflap --enable-plugin --with-system-zlib --enable-objc-gc --with-cloog --enable-cloog-backend=ppl --disable-cloog-version-check --disable-ppl-version-check --enable-multiarch --disable-werror --with-arch-32=i686 --with-abi=m64 --with-multilib-list=m32,m64,mx32 --with-tune=generic --enable-checking=release --build=x86_64-linux-gnu --host=x86_64-linux-gnu --target=x86_64-linux-gnu
Thread model: posix
gcc version 4.7.4 (Ubuntu/Linaro 4.7.4-3ubuntu12)
brook@vista:~$ sudo apt-get install lib32gcc-4.7-dev





2016年4月16日 星期六

修改MAC的hostname


在中文版的 Mac OS裡,系統會自動依照安裝時給的中文姓名翻譯成英文作為主機名稱。對我來說因為常常會用shell底下操作,就超想改hostname,方法如下:
直接編輯"系統偏好設定"/"共享"/"電腦名稱"即可,後面的".local"不用理他。


打開終端機之後就可以看到hostname修改好了。


    參考資料:
  1. 更改 Mac 主機名稱




2016年2月21日 星期日

Linux Kernel(16.1)- Network Device Driver, simple snull.


這一篇藉由LDD(Linux Device Drivers)中的SNULL來了解最基本的Network Device Driver的架構,本章的sample code比原本的SNULL更為簡化,但是Network Topology是相同的,讓interface sn0/sn1可以透過遠方虛擬的remote0/remote1彼此溝通。

最基本的Network Device Driver的寫法就是allocate network device, "struct net_device"並且賦予hook function, "struct net_device_ops",然後將該network device註冊到kernel中,Kernel就可以調用該Network device,最基本的net_device_ops包含
  • ndo_open() and ndo_validate_addr() are called, when the NIC is bring up.
  • ndo_stop() is called, when the NIC is shut down.
  • ndo_start_xmit() is called, when a packet is sent from the NIC.
  • ndo_change_mtu() is called, when the MTU of the NIC is changed.
  • ndo_set_mac_address() is called, when the MAC address of the NIC is changed.
以下是demo code的net_device_ops部分
static const struct net_device_ops nic_netdev_ops = {
    /* Kernel calls ndo_open() and ndo_validate_addr()
     * when you bring up the NIC
     */
    .ndo_open               = nic_open,
    .ndo_validate_addr      = nic_validate_addr,

    /* when you shut down the NIC, kernel call the .ndo_stop() */
    .ndo_stop               = nic_close,

    /* Kernel calls ndo_start_xmit() when it wants to 
     *   transmit a packet. 
     */
    .ndo_start_xmit         = nic_start_xmit,

    /* ndo_change_mtu() is called, when you change MTU */
    .ndo_change_mtu         = nic_change_mtu,

    /* ndo_set_mac_address() is called,
     *   when you change the MAC addr
     */
    .ndo_set_mac_address    = nic_set_mac_addr,
};


我們是模擬ethernet,而ethernet有一些hook function可以用,如下
  • ndo_validate_addr() -> eth_validate_addr().
  • ndo_change_mtu() -> eth_change_mtu().
  • ndo_set_mac_address() -> eth_mac_addr().
在demo code中,ndo_validate_addr()/ndo_change_mtu()/ndo_set_mac_address()我都是將其轉成ethernet的default hook function,我沒直接掛,是因為我想印出訊息來看
static int nic_validate_addr(struct net_device *netdev)
{
    struct nic_priv *priv = netdev_priv(netdev);
    netif_info(priv, drv, netdev, "%s(#%d), priv:%p\n",
                __func__, __LINE__, priv);
    return eth_validate_addr(netdev);
}

static int nic_change_mtu(struct net_device *netdev, int new_mtu)
{
    struct nic_priv *priv = netdev_priv(netdev);
    netif_info(priv, drv, netdev, "%s(#%d), priv:%p\n",
                __func__, __LINE__, priv);
    return eth_change_mtu(netdev, new_mtu);
}

static int nic_set_mac_addr(struct net_device *netdev, void *addr)
{
    struct nic_priv *priv = netdev_priv(netdev);
    netif_info(priv, drv, netdev, "%s(#%d), priv:%p\n",
                __func__, __LINE__, priv);
    return eth_mac_addr(netdev, addr);
}

另外幾個比較重要的function是netif_start_queue()/netif_stop_queue()
  • netif_start_queue()是通知上層,可以將資料送到該網卡,通常放在ndo_open()裡面
  • netif_stop_queue()是通知上層,停止將資料送到該網卡,通常放在ndo_stop()裡面


由於我們沒有真的remote0/remote1可以回應,所以必須設定flag/IFF_NOARP在sn0跟sn1,並且自己要處理L2的header,所以必須在額外掛上"struct header_ops"。

以下為完整的driver,主要code都是印訊息觀察driver的call flow
/* reference ldd3, snull.c */
#include <linux/module.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>

/* for in_device, in_ifaddr */
#include <linux/inetdevice.h>

MODULE_AUTHOR("Brook");
MODULE_DESCRIPTION("Kernel module for demo");
MODULE_LICENSE("GPL");

#define MAX_ETH_FRAME_SIZE   1792
struct nic_priv {
    /* you can use array to queue more packet */
    unsigned char *tx_buf;
    unsigned int  tx_len;
    u32           msg_enable;
};

static struct net_device *nic_dev[2];
/* netif msg type, defined in netdevice.h
    NETIF_MSG_DRV           = 0x0001,
    NETIF_MSG_PROBE         = 0x0002,
    NETIF_MSG_LINK          = 0x0004,
    NETIF_MSG_TIMER         = 0x0008,
    NETIF_MSG_IFDOWN        = 0x0010,
    NETIF_MSG_IFUP          = 0x0020,
    NETIF_MSG_RX_ERR        = 0x0040,
    NETIF_MSG_TX_ERR        = 0x0080,
    NETIF_MSG_TX_QUEUED     = 0x0100,
    NETIF_MSG_INTR          = 0x0200,
    NETIF_MSG_TX_DONE       = 0x0400,
    NETIF_MSG_RX_STATUS     = 0x0800,
    NETIF_MSG_PKTDATA       = 0x1000,
    NETIF_MSG_HW            = 0x2000,
    NETIF_MSG_WOL           = 0x4000,
*/
#define DEF_MSG_ENABLE 0xffff

static void dump(unsigned char *buf)
{
    unsigned char *p, sbuf[2*(sizeof(struct ethhdr) + sizeof(struct iphdr))];
    int i;
    p = sbuf;

    for(i = 0; i < sizeof(struct ethhdr); i++) {
        p += sprintf(p, "%02X ", buf[i]);
    }
    printk("eth %s\n", sbuf);

    p = sbuf;
    for(i = 0; i < sizeof(struct iphdr); i++) {
        p += sprintf(p, "%02X ", buf[sizeof(struct ethhdr) + i]);
    }
    printk("iph %s\n", sbuf);

    p = sbuf;
    for(i = 0; i < 4; i++) {
        p += sprintf(p, "%02X ", buf[sizeof(struct ethhdr) + sizeof(struct iphdr) + i]);
    }
    printk("payload %s\n", sbuf);
}


static void
nic_rx(struct net_device *netdev, int len, unsigned char *buf)
{
    struct sk_buff *skb;
    struct nic_priv *priv = netdev_priv(netdev);

    netif_info(priv, hw, netdev, "%s(#%d), rx:%d\n",
                __func__, __LINE__, len);
    /*
     * The packet has been retrieved from the transmission
     * medium. Build an skb around it, so upper layers can handle it
     */
    skb = dev_alloc_skb(len + 2);
    if (!skb) {
        netif_err(priv, rx_err, netdev,
                  "%s(#%d), rx: low on mem - packet dropped\n",
                  __func__, __LINE__);
        netdev->stats.rx_dropped++;
        return;
    }
    skb_reserve(skb, 2); /* align IP on 16B boundary */
    memcpy(skb_put(skb, len), buf, len);

    /* Write metadata, and then pass to the receive level */
    skb->dev = netdev;
    skb->protocol = eth_type_trans(skb, netdev);
    skb->ip_summed = CHECKSUM_UNNECESSARY; /* don't check it */
    netdev->stats.rx_packets++;
    netdev->stats.rx_bytes += len;
    netif_rx(skb);
}

static int nic_open(struct net_device *netdev)
{
    struct nic_priv *priv = netdev_priv(netdev);
    netif_info(priv, ifup, netdev, "%s(#%d), priv:%p\n",
                __func__, __LINE__, priv);
    /* may be using DMA */
    priv->tx_buf = kmalloc(MAX_ETH_FRAME_SIZE, GFP_KERNEL);
    if (priv->tx_buf == NULL) {
        netif_info(priv, ifup, netdev, "%s(#%d), cannot alloc tx buf\n",
                    __func__, __LINE__);
        return -ENOMEM;
    }
    netif_start_queue(netdev);
    return 0;
}

static int nic_close(struct net_device *netdev)
{
    struct nic_priv *priv = netdev_priv(netdev);
    netif_info(priv, ifdown, netdev, "%s(#%d), priv:%p\n",
                __func__, __LINE__, priv);
    netif_stop_queue(netdev);
    return 0;
}

static void nic_hw_xmit(struct net_device *netdev)
{
    struct nic_priv *priv = netdev_priv(netdev);
    struct iphdr *iph;
    u32 *saddr, *daddr;
    struct in_device* in_dev;
    struct in_ifaddr* if_info;

    if (priv->tx_len < sizeof(struct ethhdr) + sizeof(struct iphdr)) {
        netif_info(priv, hw, netdev, "%s(#%d), too short\n",
                   __func__, __LINE__);
        return;
    }
    dump(priv->tx_buf);
    iph = (struct iphdr *)(priv->tx_buf + sizeof(struct ethhdr));
    saddr = &iph->saddr;
    daddr = &iph->daddr;

    netif_info(priv, hw, netdev, "%s(#%d), orig, src:%pI4, dst:%pI4, len:%d\n",
                __func__, __LINE__, saddr, daddr, priv->tx_len);

    in_dev = nic_dev[(netdev == nic_dev[0] ? 1 : 0)]->ip_ptr;
    if (in_dev) {
        if_info = in_dev->ifa_list;
        for (if_info = in_dev->ifa_list; if_info; if_info=if_info->ifa_next) {
#if 0
            printk("label:%s, address=%pI4\n",
               if_info->ifa_label, &if_info->ifa_address);
#endif
            *saddr = *daddr = if_info->ifa_address;
            ((u8 *)saddr)[3]++;
            netif_info(priv, hw, netdev, "%s(#%d), new, src:%pI4, dst:%pI4\n",
                        __func__, __LINE__, saddr, daddr);
            break;
        }
        if (!if_info) {
            /* drop packet */
            netdev->stats.tx_dropped++;
            netif_info(priv, hw, netdev, "%s(#%d), drop packet\n",
                        __func__, __LINE__);
            return;
        }
    }

    iph->check = 0;         /* and rebuild the checksum (ip needs it) */
    iph->check = ip_fast_csum((unsigned char *)iph, iph->ihl);

    netdev->stats.tx_packets++;
    netdev->stats.tx_bytes += priv->tx_len;

    nic_rx(nic_dev[(netdev == nic_dev[0] ? 1 : 0)], priv->tx_len, priv->tx_buf);
}

static netdev_tx_t nic_start_xmit(struct sk_buff *skb,
                                  struct net_device *netdev)
{
    struct nic_priv *priv = netdev_priv(netdev);
    netif_info(priv, drv, netdev, "%s(#%d), orig, src:%pI4, dst:%pI4\n",
                __func__, __LINE__, &(ip_hdr(skb)->saddr), &(ip_hdr(skb)->daddr));
    priv->tx_len = skb->len;
    if (likely(priv->tx_len < MAX_ETH_FRAME_SIZE)) {
        if (priv->tx_len < ETH_ZLEN) {
            memset(priv->tx_buf, 0, ETH_ZLEN);
            priv->tx_len = ETH_ZLEN;
        }
        skb_copy_and_csum_dev(skb, priv->tx_buf);
        dev_kfree_skb_any(skb);
    } else {
        dev_kfree_skb_any(skb);
        netdev->stats.tx_dropped++;
        return NETDEV_TX_OK;
    }

    nic_hw_xmit(netdev);
    return NETDEV_TX_OK;
}

static int nic_validate_addr(struct net_device *netdev)
{
    struct nic_priv *priv = netdev_priv(netdev);
    netif_info(priv, drv, netdev, "%s(#%d), priv:%p\n",
                __func__, __LINE__, priv);
    return eth_validate_addr(netdev);
}

static int nic_change_mtu(struct net_device *netdev, int new_mtu)
{
    struct nic_priv *priv = netdev_priv(netdev);
    netif_info(priv, drv, netdev, "%s(#%d), priv:%p\n",
                __func__, __LINE__, priv);
    return eth_change_mtu(netdev, new_mtu);
}

static int nic_set_mac_addr(struct net_device *netdev, void *addr)
{
    struct nic_priv *priv = netdev_priv(netdev);
    netif_info(priv, drv, netdev, "%s(#%d), priv:%p\n",
                __func__, __LINE__, priv);
    return eth_mac_addr(netdev, addr);
}

/*
 * This function is called to fill up an eth header, since arp is not
 * available on the interface
 */
int snull_header(struct sk_buff *skb, struct net_device *netdev,
                unsigned short type, const void *daddr, const void *saddr,
                unsigned len)
{
    struct nic_priv *priv = netdev_priv(netdev);
    struct ethhdr *eth = (struct ethhdr *)skb_push(skb, ETH_HLEN);
    struct net_device *dst_netdev;

    netif_info(priv, drv, netdev, "%s(#%d)\n",
                __func__, __LINE__);
    dst_netdev = nic_dev[(netdev == nic_dev[0] ? 1 : 0)];
    eth->h_proto = htons(type);
    memcpy(eth->h_source, saddr ? saddr : netdev->dev_addr, netdev->addr_len);
    memcpy(eth->h_dest, dst_netdev->dev_addr, dst_netdev->addr_len);
    return (netdev->hard_header_len);
}

static const struct header_ops snull_header_ops = {
        .create  = snull_header,
};

static const struct net_device_ops nic_netdev_ops = {
    /* Kernel calls ndo_open() and ndo_validate_addr()
     * when you bring up the NIC
     */
    .ndo_open               = nic_open,
    .ndo_validate_addr      = nic_validate_addr,

    /* when you shut down the NIC, kernel call the .ndo_stop() */
    .ndo_stop               = nic_close,

    /* Kernel calls ndo_start_xmit() when it wants to 
     *   transmit a packet. 
     */
    .ndo_start_xmit         = nic_start_xmit,

    /* ndo_change_mtu() is called, when you change MTU */
    .ndo_change_mtu         = nic_change_mtu,

    /* ndo_set_mac_address() is called,
     *   when you change the MAC addr
     */
    .ndo_set_mac_address    = nic_set_mac_addr,
};

static struct net_device* nic_alloc_netdev(void)
{
    struct net_device *netdev;

    netdev = alloc_etherdev(sizeof(struct nic_priv));
    if (!netdev) {
        pr_err("%s(#%d): alloc dev failed",
               __func__, __LINE__);
        return NULL;
    }
    eth_hw_addr_random(netdev);
    netdev->netdev_ops = &nic_netdev_ops;

    /* keep the default flags, just add NOARP */
    netdev->flags |= IFF_NOARP;

    /* There are no explicit users, so this is 
     *     now equivalent to NETIF_F_HW_CSUM. */
    netdev->features |= NETIF_F_HW_CSUM;

    netdev->header_ops = &snull_header_ops;

    return netdev;
}

static int __init brook_init(void)
{
    int ret;
    struct nic_priv *priv;

    nic_dev[0] = nic_alloc_netdev();
    if (!nic_dev[0]) {
        pr_err("%s(#%d): alloc netdev[0] failed", __func__, __LINE__);
        return -ENOMEM;
    }

    nic_dev[1] = nic_alloc_netdev();
    if (!nic_dev[1]) {
        pr_err("%s(#%d): alloc netdev[1] failed", __func__, __LINE__);
        ret = -ENOMEM;
        goto alloc_2nd_failed;
    }

    ret = register_netdev(nic_dev[0]);
    if (ret) {
        pr_err("%s(#%d): reg net driver failed. ret:%d",
               __func__, __LINE__, ret);
        goto reg1_failed;
    }

    ret = register_netdev(nic_dev[1]);
    if (ret) {
        pr_err("%s(#%d): reg net driver failed. ret:%d",
               __func__, __LINE__, ret);
        goto reg2_failed;
    }

    priv = netdev_priv(nic_dev[0]);
    priv->msg_enable = DEF_MSG_ENABLE;
    priv = netdev_priv(nic_dev[1]);
    priv->msg_enable = DEF_MSG_ENABLE;
    return 0;

reg2_failed:
    unregister_netdev(nic_dev[0]);
reg1_failed:
    free_netdev(nic_dev[1]);
alloc_2nd_failed:
    free_netdev(nic_dev[0]);
    return ret;
}
module_init(brook_init);

static void __exit brook_exit(void)
{
    int i;
    pr_info("%s(#%d): remove module", __func__, __LINE__);
    for (i = 0; i < ARRAY_SIZE(nic_dev); i++) {
        unregister_netdev(nic_dev[i]);
        free_netdev(nic_dev[i]);
    }
}
module_exit(brook_exit);





參考資料:
  1. Linux Device Drivers, Third Edition, Chapter 17: Network Drivers, https://lwn.net/Kernel/LDD3/,
  2. Linux Networking and Network Devices APIs, https://www.kernel.org/doc/htmldocs/networking/index.html




2016年1月3日 星期日

OpenEmbedded - Writing a recipes/package inherit cmake


基本上inherit cmake跟inherit autotools是差不多的,其實cmake class裡面也是有inherit autotool,比較大的差異是,source file在cmake下是放在${PN}目錄下,即package name下,檔案結構如下:
brook@vista:~/oe-core$ tree meta/recipes-support/brook-cmake
meta/recipes-support/brook-cmake
|-- brook-cmake
|   `-- brook-cmake.patch
`-- brook-cmake_1.0.bb
1 directory, 2 files


brook-cmake.patch就是放置CMakeLists.txt、main.c跟LICENSE,其內容如下:

CMakeLists.txt

cmake_minimum_required (VERSION 2.6)
project(brook-cmake-demo)
add_executable(brook-cmake main.c)
install(TARGETS brook-cmake DESTINATION bin)


main.c

#include <stdio.h>

int main(int argc, char *argv[])
{
    printf("hello, cmake\n");
    return 0;
}


brook-cmake_1.0.bb

DESCRIPTION = "Brook CMake Project"
LICENSE = "GPLv2"
LIC_FILES_CHKSUM = "file://LICENSE;md5=5ff2bd2dd80c7cc542a6b8de3ee4ff87"

SRC_URI = "file://brook-cmake.patch \
"

inherit cmake

EXTRA_OEMAKE = "-Wall"


下面這裡是多個版本的寫法

其實寫法都一樣,只是patch file的目錄要包含版本號,檔案結構如下:
brook@vista:~/oe-core/meta/recipes-support$ tree brook-cmake
brook-cmake
|-- brook-cmake-1.0
|   `-- brook-cmake.patch
|-- brook-cmake-2.0
|   `-- brook-cmake.patch
|-- brook-cmake_1.0.bb
`-- brook-cmake_2.0.bb


BB檔與patch的內容跟上面都是一樣的
brook@vista:~/oe-core$ cat meta/recipes-support/brook-cmake/brook-cmake_1.0.bb
DESCRIPTION = "Brook CMake Project"
LICENSE = "GPLv2"
LIC_FILES_CHKSUM = "file://LICENSE;md5=7e4bf6f59b6da8e57ed5425a10a7310b"

SRC_URI = "file://brook-cmake.patch \
"

inherit cmake

BBCLASSEXTEND = "native"

EXTRA_OEMAKE = "-Wall"

brook@vista:~/oe-core$ cat meta/recipes-support/brook-cmake/brook-cmake_2.0.bb
DESCRIPTION = "Brook CMake Project"
LICENSE = "GPLv2"
LIC_FILES_CHKSUM = "file://LICENSE;md5=7e4bf6f59b6da8e57ed5425a10a7310b"

SRC_URI = "file://brook-cmake.patch \
"

inherit cmake

BBCLASSEXTEND = "native"

EXTRA_OEMAKE = "-Wall"

brook@vista:~/oe-core$ cat meta/recipes-support/brook-cmake/brook-cmake-1.0/brook-cmake.patch
From 037c40faba39a11e54f282c3eab5e47a4cf1c7ca Mon Sep 17 00:00:00 2001
From: Brook <rene3210 at gmail.com>
Date: Tue, 27 Sep 2016 20:42:11 +0800
Subject: [PATCH] brook oe demo code

---
 CMakeLists.txt | 5 +++++
 LICENSE        | 1 +
 main.c         | 8 ++++++++
 3 files changed, 14 insertions(+)
 create mode 100644 CMakeLists.txt
 create mode 100644 LICENSE
 create mode 100644 main.c

diff --git a/CMakeLists.txt b/CMakeLists.txt
new file mode 100644
index 0000000..ee4d6d9
--- /dev/null
+++ b/CMakeLists.txt
@@ -0,0 +1,5 @@
+cmake_minimum_required (VERSION 2.6)
+project(brook-cmake-demo)
+add_executable(brook-cmake main.c)
+install(TARGETS brook-cmake DESTINATION bin)
+
diff --git a/LICENSE b/LICENSE
new file mode 100644
index 0000000..0664320
--- /dev/null
+++ b/LICENSE
@@ -0,0 +1 @@
+"GPLv2"
diff --git a/main.c b/main.c
new file mode 100644
index 0000000..d089aad
--- /dev/null
+++ b/main.c
@@ -0,0 +1,8 @@
+#include 
+
+int main(int argc, char *argv[])
+{
+    printf("hello, cmake 1.0\n");
+    return 0;
+}
+
--
2.7.4

brook@vista:~/oe-core$ cat meta/recipes-support/brook-cmake/brook-cmake-2.0/brook-cmake.patch
From 037c40faba39a11e54f282c3eab5e47a4cf1c7ca Mon Sep 17 00:00:00 2001
From: Brook <rene3210 at gmail.com>
Date: Tue, 27 Sep 2016 20:42:11 +0800
Subject: [PATCH] brook oe demo code

---
 CMakeLists.txt | 5 +++++
 LICENSE        | 1 +
 main.c         | 8 ++++++++
 3 files changed, 14 insertions(+)
 create mode 100644 CMakeLists.txt
 create mode 100644 LICENSE
 create mode 100644 main.c

diff --git a/CMakeLists.txt b/CMakeLists.txt
new file mode 100644
index 0000000..ee4d6d9
--- /dev/null
+++ b/CMakeLists.txt
@@ -0,0 +1,5 @@
+cmake_minimum_required (VERSION 2.6)
+project(brook-cmake-demo)
+add_executable(brook-cmake main.c)
+install(TARGETS brook-cmake DESTINATION bin)
+
diff --git a/LICENSE b/LICENSE
new file mode 100644
index 0000000..0664320
--- /dev/null
+++ b/LICENSE
@@ -0,0 +1 @@
+"GPLv2"
diff --git a/main.c b/main.c
new file mode 100644
index 0000000..d089aad
--- /dev/null
+++ b/main.c
@@ -0,0 +1,8 @@
+#include 
+
+int main(int argc, char *argv[])
+{
+    printf("hello, cmake 2.0\n");
+    return 0;
+}
+
--
2.7.4


如果沒有特別指定版本的話,會使用最新的,可以透過設定PREFERRED_VERSION_package_name設定特定版本,如
brook@vista:~/oe-core$ grep -H PREFERRED_VERSION build/conf/local.conf
build/conf/local.conf:PREFERRED_VERSION_brook-cmake = "1.0"
相關資訊可以參考Bitbake User Manual/2.3. Preferences and Providers

    參考資料:
  1. CMake 入門




BitBake User Manual - CH3.3. Sharing Functionality


BitBake 有metadata sharing的機制,基本上分為兩類,include file(.inc)跟class file(.bbclass),你可以使用include或require來include .inc檔案,其行為就是會將.inc內容插入include那行的位置中。
include跟require相似,差異是,如果檔案不存在include不會error,但是requre就會,如:
include foo_no_exist.inc
require foo_need_exist.inc

你可以使用iherit來使用class file,使用這指令會去尋找對應的class來使用,如:
inherit autotools

如果每個recipes都需要inherit的話,可以在.conf使用INHERIT指令達成這目的,如:
INHERIT += sanity

example:meta/conf/sanity.conf
# Sanity checks for common user misconfigurations
#
# See sanity.bbclass
#
# Expert users can confirm their sanity with "touch conf/sanity.conf"
BB_MIN_VERSION = "1.27.1"

SANITY_ABIFILE = "${TMPDIR}/abi_version"

SANITY_VERSION ?= "1"
LOCALCONF_VERSION  ?= "1"
LAYER_CONF_VERSION ?= "6"
SITE_CONF_VERSION  ?= "1"

INHERIT += "sanity"


example:meta/classes/sanity.bbclass


    參考資料:
  1. BitBake User Manual




Linux Kernel(15.2)- platform_device_register()之如何调用driver.probe()


Linux Kernel(16.1)- platform_driver_register()之如何调用driver.probe()之後,我們來看如何由platform_device_register调用driver.probe()

int platform_device_register(struct platform_device *pdev)
{
    device_initialize(&pdev->dev);
    arch_setup_pdev_archdata(pdev);
    return platform_device_add(pdev);
}
EXPORT_SYMBOL_GPL(platform_device_register);


int platform_device_add(struct platform_device *pdev)
{
    int i, ret;

    if (!pdev)
        return -EINVAL;

    if (!pdev->dev.parent)
        pdev->dev.parent = &platform_bus;

    pdev->dev.bus = &platform_bus_type;

    switch (pdev->id) {
    default:
        dev_set_name(&pdev->dev, "%s.%d", pdev->name,  pdev->id);
        break;
    case PLATFORM_DEVID_NONE:
        dev_set_name(&pdev->dev, "%s", pdev->name);
        break;
    case PLATFORM_DEVID_AUTO:
        /*
         * Automatically allocated device ID. We mark it as such so
         * that we remember it must be freed, and we append a suffix
         * to avoid namespace collision with explicit IDs.
         */
        ret = ida_simple_get(&platform_devid_ida, 0, 0, GFP_KERNEL);
        if (ret < 0)
            goto err_out;
        pdev->id = ret;
        pdev->id_auto = true;
        dev_set_name(&pdev->dev, "%s.%d.auto", pdev->name, pdev->id);
        break;
    }

    for (i = 0; i < pdev->num_resources; i++) {
        struct resource *p, *r = &pdev->resource[i];

        if (r->name == NULL)
            r->name = dev_name(&pdev->dev);

        p = r->parent;
        if (!p) {
            if (resource_type(r) == IORESOURCE_MEM)
                p = &iomem_resource;
            else if (resource_type(r) == IORESOURCE_IO)
                p = &ioport_resource;
        }

        if (p && insert_resource(p, r)) {
            dev_err(&pdev->dev, "failed to claim resource %d\n", i);
            ret = -EBUSY;
            goto failed;
        }
    }

    pr_debug("Registering platform device '%s'. Parent at %s\n",
         dev_name(&pdev->dev), dev_name(pdev->dev.parent));

    ret = device_add(&pdev->dev);
    if (ret == 0)
        return ret;

 failed:
    if (pdev->id_auto) {
        ida_simple_remove(&platform_devid_ida, pdev->id);
        pdev->id = PLATFORM_DEVID_AUTO;
    }

    while (--i >= 0) {
        struct resource *r = &pdev->resource[i];
        if (r->parent)
            release_resource(r);
    }

 err_out:
    return ret;
}


int device_add(struct device *dev) 
{
    struct device *parent = NULL;
    struct kobject *kobj;
    struct class_interface *class_intf;
    int error = -EINVAL;

    dev = get_device(dev);
    if (!dev)
        goto done;

    if (!dev->p) {
        error = device_private_init(dev);
        if (error)
            goto done;
    }

    /*
     * for statically allocated devices, which should all be converted
     * some day, we need to initialize the name. We prevent reading back
     * the name, and force the use of dev_name()
     */
    if (dev->init_name) {
        dev_set_name(dev, "%s", dev->init_name);
        dev->init_name = NULL;
    }

    /* subsystems can specify simple device enumeration */
    if (!dev_name(dev) && dev->bus && dev->bus->dev_name)
        dev_set_name(dev, "%s%u", dev->bus->dev_name, dev->id);

    if (!dev_name(dev)) {
        error = -EINVAL;
        goto name_error;
    }

    pr_debug("device: '%s': %s\n", dev_name(dev), __func__);

    parent = get_device(dev->parent);
    kobj = get_device_parent(dev, parent);
    if (kobj)
        dev->kobj.parent = kobj;

    /* use parent numa_node */
    if (parent)
        set_dev_node(dev, dev_to_node(parent));

    /* first, register with generic layer. */
    /* we require the name to be set before, and pass NULL */
    error = kobject_add(&dev->kobj, dev->kobj.parent, NULL);
    if (error)
        goto Error;

    /* notify platform of device entry */
    if (platform_notify)
        platform_notify(dev);

    error = device_create_file(dev, &dev_attr_uevent);
    if (error)
        goto attrError;

    error = device_add_class_symlinks(dev);
    if (error)
        goto SymlinkError;
    error = device_add_attrs(dev);
    if (error)
        goto AttrsError;
    error = bus_add_device(dev);
    if (error)
        goto BusError;
    error = dpm_sysfs_add(dev);
    if (error)
        goto DPMError;
    device_pm_add(dev);

    if (MAJOR(dev->devt)) {
        error = device_create_file(dev, &dev_attr_dev);
        if (error)
            goto DevAttrError;

        error = device_create_sys_dev_entry(dev);
        if (error)
            goto SysEntryError;

        devtmpfs_create_node(dev);
    }

    /* Notify clients of device addition.  This call must come
     * after dpm_sysfs_add() and before kobject_uevent().
     */
    if (dev->bus)
        blocking_notifier_call_chain(&dev->bus->p->bus_notifier,
                         BUS_NOTIFY_ADD_DEVICE, dev);

    kobject_uevent(&dev->kobj, KOBJ_ADD);
    bus_probe_device(dev);
    if (parent)
        klist_add_tail(&dev->p->knode_parent,
                   &parent->p->klist_children);

    if (dev->class) {
        mutex_lock(&dev->class->p->mutex);
        /* tie the class to the device */
        klist_add_tail(&dev->knode_class,
                   &dev->class->p->klist_devices);

        /* notify any interfaces that the device is here */
        list_for_each_entry(class_intf,
                    &dev->class->p->interfaces, node)
            if (class_intf->add_dev)
                class_intf->add_dev(dev, class_intf);
        mutex_unlock(&dev->class->p->mutex);
    }
done:
    put_device(dev);
    return error;
 SysEntryError:
    if (MAJOR(dev->devt))
        device_remove_file(dev, &dev_attr_dev);
 DevAttrError:
    device_pm_remove(dev);
    dpm_sysfs_remove(dev);
 DPMError:
    bus_remove_device(dev);
 BusError:
    device_remove_attrs(dev);
 AttrsError:
    device_remove_class_symlinks(dev);
 SymlinkError:
    device_remove_file(dev, &dev_attr_uevent);
 attrError:
    kobject_uevent(&dev->kobj, KOBJ_REMOVE);
    kobject_del(&dev->kobj);
 Error:
    cleanup_device_parent(dev);
    put_device(parent);
name_error:
    kfree(dev->p);
    dev->p = NULL;
    goto done;
}


void bus_probe_device(struct device *dev)
{
    struct bus_type *bus = dev->bus;
    struct subsys_interface *sif;

    if (!bus)
        return;

    if (bus->p->drivers_autoprobe)
        device_initial_probe(dev);

    mutex_lock(&bus->p->mutex);
    list_for_each_entry(sif, &bus->p->interfaces, node)
        if (sif->add_dev)
            sif->add_dev(dev, sif);
    mutex_unlock(&bus->p->mutex);
}


void device_initial_probe(struct device *dev)
{
    __device_attach(dev, true);
}


static int __device_attach(struct device *dev, bool allow_async)
{
    int ret = 0;

    device_lock(dev);
    if (dev->driver) {
        if (klist_node_attached(&dev->p->knode_driver)) {
            ret = 1;
            goto out_unlock;
        }
        ret = device_bind_driver(dev);
        if (ret == 0)
            ret = 1;
        else {
            dev->driver = NULL;
            ret = 0;
        }
    } else {
        struct device_attach_data data = {
            .dev = dev,
            .check_async = allow_async,
            .want_async = false,
        };

        if (dev->parent)
            pm_runtime_get_sync(dev->parent);

        ret = bus_for_each_drv(dev->bus, NULL, &data,
                    __device_attach_driver);
        if (!ret && allow_async && data.have_async) {
            /*
             * If we could not find appropriate driver
             * synchronously and we are allowed to do
             * async probes and there are drivers that
             * want to probe asynchronously, we'll
             * try them.
             */
            dev_dbg(dev, "scheduling asynchronous probe\n");
            get_device(dev);
            async_schedule(__device_attach_async_helper, dev);
        } else {
            pm_request_idle(dev);
        }

        if (dev->parent)
            pm_runtime_put(dev->parent);
    }
out_unlock:
    device_unlock(dev);
    return ret;
}


static int __device_attach_driver(struct device_driver *drv, void *_data)
{
    struct device_attach_data *data = _data;
    struct device *dev = data->dev;
    bool async_allowed;

    /*
     * Check if device has already been claimed. This may
     * happen with driver loading, device discovery/registration,
     * and deferred probe processing happens all at once with
     * multiple threads.
     */
    if (dev->driver)
        return -EBUSY;

    if (!driver_match_device(drv, dev))
        return 0;

    async_allowed = driver_allows_async_probing(drv);

    if (async_allowed)
        data->have_async = true;

    if (data->check_async && async_allowed != data->want_async)
        return 0;

    return driver_probe_device(drv, dev);
}


static inline int driver_match_device(struct device_driver *drv,
                      struct device *dev)
{
    return drv->bus->match ? drv->bus->match(dev, drv) : 1;
}





Linux Kernel(15.1)- platform_driver_register()之如何调用driver.probe()



轉自platform_driver_register()--如何match之后调用probe

int platform_driver_register(struct platform_driver *drv)
{
    drv->driver.bus = &platform_bus_type;/*关联总线*/
    /*关联driver的设备方法*/
    if (drv->probe)
        drv->driver.probe = platform_drv_probe;
    if (drv->remove)
        drv->driver.remove = platform_drv_remove;
    if (drv->shutdown)
        drv->driver.shutdown = platform_drv_shutdown;

    return driver_register(&drv->driver);/*注册驱动*/
}

/******************************************************************************/
struct platform_driver {
    int (*probe)(struct platform_device *);/*匹配到设备后调用,下面分析内核代码怎么调用的*/
    int (*remove)(struct platform_device *);
    void (*shutdown)(struct platform_device *);
    int (*suspend)(struct platform_device *, pm_message_t state);
    int (*resume)(struct platform_device *);
    struct device_driver driver;
    const struct platform_device_id *id_table;
};

struct bus_type platform_bus_type = {
    .name        = "platform",
    .dev_attrs    = platform_dev_attrs,
    .match        = platform_match,
    .uevent        = platform_uevent,
    .pm        = &platform_dev_pm_ops,
};
/********************************************************************************/

int driver_register(struct device_driver *drv)
{
    int ret;
    struct device_driver *other;

    BUG_ON(!drv->bus->p);

    if ((drv->bus->probe && drv->probe) ||
        (drv->bus->remove && drv->remove) ||
        (drv->bus->shutdown && drv->shutdown))
        printk(KERN_WARNING "Driver '%s' needs updating - please use "
            "bus_type methods\n", drv->name);

    other = driver_find(drv->name, drv->bus);
    if (other) {
        put_driver(other);
        printk(KERN_ERR "Error: Driver '%s' is already registered, "
            "aborting...\n", drv->name);
        return -EBUSY;
    }

    ret = bus_add_driver(drv);
    if (ret)
        return ret;
    ret = driver_add_groups(drv, drv->groups);
    if (ret)
        bus_remove_driver(drv);
    return ret;
}

int bus_add_driver(struct device_driver *drv)
{
    struct bus_type *bus;
    struct driver_private *priv;
    int error = 0;

    bus = bus_get(drv->bus);
    if (!bus)
        return -EINVAL;

    pr_debug("bus: '%s': add driver %s\n", bus->name, drv->name);

    priv = kzalloc(sizeof(*priv), GFP_KERNEL);
    if (!priv) {
        error = -ENOMEM;
        goto out_put_bus;
    }
    klist_init(&priv->klist_devices, NULL, NULL);
    priv->driver = drv;
    drv->p = priv;
    priv->kobj.kset = bus->p->drivers_kset;
    error = kobject_init_and_add(&priv->kobj, &driver_ktype, NULL,
                     "%s", drv->name);
    if (error)
        goto out_unregister;

    if (drv->bus->p->drivers_autoprobe) {
        error = driver_attach(drv);
        if (error)
            goto out_unregister;
    }
    klist_add_tail(&priv->knode_bus, &bus->p->klist_drivers);
    module_add_driver(drv->owner, drv);

    error = driver_create_file(drv, &driver_attr_uevent);
    if (error) {
        printk(KERN_ERR "%s: uevent attr (%s) failed\n",
            __func__, drv->name);
    }
    error = driver_add_attrs(bus, drv);
    if (error) {
        /* How the hell do we get out of this pickle? Give up */
        printk(KERN_ERR "%s: driver_add_attrs(%s) failed\n",
            __func__, drv->name);
    }

    if (!drv->suppress_bind_attrs) {
        error = add_bind_files(drv);
        if (error) {
            /* Ditto */
            printk(KERN_ERR "%s: add_bind_files(%s) failed\n",
                __func__, drv->name);
        }
    }

    kobject_uevent(&priv->kobj, KOBJ_ADD);
    return 0;

out_unregister:
    kobject_put(&priv->kobj);
    kfree(drv->p);
    drv->p = NULL;
out_put_bus:
    bus_put(bus);
    return error;
}

int driver_attach(struct device_driver *drv)
{
    /*对总线上的每一个设备都调用__driver_attach*/
    return bus_for_each_dev(drv->bus, NULL, drv, __driver_attach);
}

static int __driver_attach(struct device *dev, void *data)
{
    struct device_driver *drv = data;

    /*
     * Lock device and try to bind to it. We drop the error
     * here and always return 0, because we need to keep trying
     * to bind to devices and some drivers will return an error
     * simply if it didn't support the device.
     *
     * driver_probe_device() will spit a warning if there
     * is an error.
     */

    if (!driver_match_device(drv, dev))
        return 0;

    if (dev->parent)    /* Needed for USB */
        device_lock(dev->parent);
    device_lock(dev);
    if (!dev->driver)
        driver_probe_device(drv, dev);
    device_unlock(dev);
    if (dev->parent)
        device_unlock(dev->parent);

    return 0;
}

static inline int driver_match_device(struct device_driver *drv,
                      struct device *dev)
{
    /*调用总线的match去匹配设备和驱动*/
    return drv->bus->match ? drv->bus->match(dev, drv) : 1;
}

int driver_probe_device(struct device_driver *drv, struct device *dev)
{
    int ret = 0;

    if (!device_is_registered(dev))
        return -ENODEV;

    pr_debug("bus: '%s': %s: matched device %s with driver %s\n",
         drv->bus->name, __func__, dev_name(dev), drv->name);

    pm_runtime_get_noresume(dev);
    pm_runtime_barrier(dev);
    ret = really_probe(dev, drv);
    pm_runtime_put_sync(dev);

    return ret;
}
static int really_probe(struct device *dev, struct device_driver *drv)
{
    int ret = 0;

    atomic_inc(&probe_count);
    pr_debug("bus: '%s': %s: probing driver %s with device %s\n",
         drv->bus->name, __func__, drv->name, dev_name(dev));
    WARN_ON(!list_empty(&dev->devres_head));

    dev->driver = drv;
    if (driver_sysfs_add(dev)) {
        printk(KERN_ERR "%s: driver_sysfs_add(%s) failed\n",
            __func__, dev_name(dev));
        goto probe_failed;
    }
/**********************************************************************************/
    if (dev->bus->probe) {/*首先看总线有没有probe函数,若有则调用,而平台总线没有probe*/
        ret = dev->bus->probe(dev);
        if (ret)
            goto probe_failed;
    } else if (drv->probe) {/*然后看驱动有没有probe函数,若有则调用,*/
        ret = drv->probe(dev);
        if (ret)
            goto probe_failed;
    }
/************************************************************************************/

    driver_bound(dev);
    ret = 1;
    pr_debug("bus: '%s': %s: bound device %s to driver %s\n",
         drv->bus->name, __func__, dev_name(dev), drv->name);
    goto done;

probe_failed:
    devres_release_all(dev);
    driver_sysfs_remove(dev);
    dev->driver = NULL;

    if (ret != -ENODEV && ret != -ENXIO) {
        /* driver matched but the probe failed */
        printk(KERN_WARNING
               "%s: probe of %s failed with error %d\n",
               drv->name, dev_name(dev), ret);
    }
    /*
     * Ignore errors returned by ->probe so that the next driver can try
     * its luck.
     */
    ret = 0;
done:
    atomic_dec(&probe_count);
    wake_up(&probe_waitqueue);
    return ret;
}


/*平台总线的match逻辑*/
static int platform_match(struct device *dev, struct device_driver *drv)
{
    struct platform_device *pdev = to_platform_device(dev);
    struct platform_driver *pdrv = to_platform_driver(drv);

    /* match against the id table first */
    if (pdrv->id_table)
        return platform_match_id(pdrv->id_table, pdev) != NULL;

    /* fall-back to driver name match */
    return (strcmp(pdev->name, drv->name) == 0);/*驱动名字与设备名字要匹配*/
}




2016年1月2日 星期六

Linux Kernel(15)- Platform Devices


很多人心中都有過一個問題What is the difference between Platform driver and normal device driver?,簡單的來說Platform devices就non-discoverable,也就是device本身沒辦法跟系統說"我在這裡",典型的就是I2C device,它不會通知kernel"我在這裡",通常是預先知道有個I2C device在那裡,再由software設定好,這類non-discoverable device就適用Platform devices架構來寫。
platform device會被connect在platform bus上,而platform bus是一個虛擬的bus(pseudo-bus),這樣可以讓整個架構platform driver符合Linux的標準driver model。

這篇會根據The platform device API教導如何寫一個簡單的Platform devices,基本上最基本的platform device只需要name,因為platform bus會根據platform device與platform driver的name是否match執行driver的probe(),而最簡單的platform driver只需要name,跟probe()與remove()即可。
#include <linux/module.h>
#include <linux/platform_device.h>

MODULE_AUTHOR("Brook");
MODULE_DESCRIPTION("Kernel module for demo");
MODULE_LICENSE("GPL");

#define DEVNAME "brook"

#define DYN_ALLOC 1

static struct platform_device brook_device = {
    .name = DEVNAME,
};

static int brook_probe(struct platform_device *pdev)
{
    pr_info("%s(#%d)\n", __func__, __LINE__);
    return 0;
}

static int brook_remove(struct platform_device *pdev)
{
    pr_info("%s(#%d)\n", __func__, __LINE__);
    return 0;
}

static struct platform_driver brook_driver = {
    .driver = {
        .name  = DEVNAME,
        .owner = THIS_MODULE,
    },
    .probe  = brook_probe,
    .remove = brook_remove,
};

static int __init brook_init(void)
{
    int err;
    pr_info("%s(#%d)\n", __func__, __LINE__);

    err = platform_device_register(&brook_device);
    if (err) {
        pr_err("%s(#%d): platform_device_register failed(%d)\n",
                __func__, __LINE__, err);
        return err;
    }

    err = platform_driver_register(&brook_driver);
    if (err) {
        dev_err(&(brook_device.dev), "%s(#%d): platform_driver_register fail(%d)\n",
                __func__, __LINE__, err);
        goto dev_reg_failed;
    }
    return err;

dev_reg_failed:
    platform_device_unregister(&brook_device);

    return err;
}
module_init(brook_init);

static void __exit brook_exit(void)
{
    dev_info(&(brook_device.dev), "%s(#%d)\n", __func__, __LINE__);
    platform_device_unregister(&brook_device);
    platform_driver_unregister(&brook_driver);
}
module_exit(brook_exit);


使用platform_device_register()會導致"brook.0" does not have a release() function, it is broken and must be fixed.的OOPS,可以改用platform_device_alloc() + platform_device_add(),platform_device_alloc()裡面就會做pa->pdev.dev.release = platform_device_release。
#include <linux/module.h>
#include <linux/platform_device.h>

MODULE_AUTHOR("Brook");
MODULE_DESCRIPTION("Kernel module for demo");
MODULE_LICENSE("GPL");

#define DEVNAME "brook"

#define DYN_ALLOC 1

static struct platform_device *brook_device;

static int brook_probe(struct platform_device *pdev)
{
    pr_info("%s(#%d)\n", __func__, __LINE__);
    return 0;
}

static int brook_remove(struct platform_device *pdev)
{
    pr_info("%s(#%d)\n", __func__, __LINE__);
    return 0;
}

static struct platform_driver brook_driver = {
    .driver = {
        .name  = DEVNAME,
        .owner = THIS_MODULE,
    },
    .probe  = brook_probe,
    .remove = brook_remove,
};

static int __init brook_init(void)
{
    int err;
    pr_info("%s(#%d)\n", __func__, __LINE__);

    /* using platform_device_alloc() + platform_device_add() 
     * instead of platform_device_register() to avoid the OOPS, 
     *     "Device 'brook.0' does not have a release() function,
     *      it is broken and must be fixed."
     */
    brook_device = platform_device_alloc(DEVNAME, 0);
    if (!brook_device) {
        pr_err("%s(#%d): platform_device_alloc fail\n",
               __func__, __LINE__);
        return -ENOMEM;
    }

    err = platform_device_add(brook_device);
    if (err) {
        pr_err("%s(#%d): platform_device_add failed\n",
               __func__, __LINE__);
        goto dev_add_failed;
    }

    err = platform_driver_register(&brook_driver);
    if (err) {
        dev_err(&(brook_device->dev), "%s(#%d): platform_driver_register fail(%d)\n",
                __func__, __LINE__, err);
        goto dev_reg_failed;
    }
    return err;

dev_add_failed:
    platform_device_put(brook_device);
dev_reg_failed:
    platform_device_unregister(brook_device);

    return err;
}
module_init(brook_init);

static void __exit brook_exit(void)
{
    dev_info(&(brook_device->dev), "%s(#%d)\n", __func__, __LINE__);
    platform_device_unregister(brook_device);
    platform_driver_unregister(&brook_driver);
}
module_exit(brook_exit);



    參考資料:
  1. Documentation/driver-model
  2. What is the difference between Platform driver and normal device driver?
  3. The platform device API
  4. Linux Kernel architecture for device drivers
  5. platform_driver_register()--如何match之后调用probe
  6. Improved dynamically allocated platform_device interface