June 2017 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

My network device don't work
by wiktoria.lewicka June 20, 2017

June 20, 2017

Hello. I write simple network device, but its don't work. Module is loading, loading, loading... Code: #include <linux/kernel.h> #include <linux/module.h> #include <linux/init.h> #include <linux/netdevice.h> #include <linux/string.h> #define DEV_NAME "chwdp" struct net_device my_netdev; int init_my_net_dev(struct net_device *dev); struct net_device_ops nops = { .ndo_init = init_my_net_dev, //.ndo_uninit = uninit_my_net_dev, }; static int __init init_dev(void) { //my_netdev.netdev_ops = &nops; int result; if((netdev_boot_setup_check(&my_netdev))){ printk(KERN_ERR "NETDEV: setup error"); return 0; } strncpy(my_netdev.name, DEV_NAME, 5); if((result = register_netdev(&my_netdev))) printk(KERN_ERR "NETDEV: Error registering device"); printk("NETDEV: Device registered successfully"); return 0; } static void __exit remove_dev(void) { unregister_netdev(&my_netdev); } int init_my_net_dev(struct net_device *dev) { printk("INIT"); return 0; } module_init(init_dev); module_exit(remove_dev);

4 3

TUN/TAP device is stuck on read
by Ramon Fried June 20, 2017

June 20, 2017

Hi ! I've written a TUN device in user space and it works well. Sometimes however, it appears that connectivity stops. no packets are transferred. I manage to produce the issue only when I run iperf for a long time for testing the throughput. Digging deeper, it looks that the application is blocked on reading from the tun device. I opened the verbosity of the logs in the kernel tun driver (drivers/net/tun.c) and it doesn't seem that packets arrive to it's queue. How can I debug it further ? I'm new to kernel network stack, and I'm looking for a way to track the packet (for instance ping) through the stack until it reaches the queue in the tun driver. IE. $ ping 10.0.0.2 I should be able to trace that ICMP packet all along the way to understand where it was dropped. BTW - The network statics in ifconfig don't show any drop of packets. Thanks A lot.

1 1

Is there any way to add an attribute under sysfs at the “platform” level?
by Roman Storozhenko June 19, 2017

June 19, 2017

Hello everybody! I am trying to modify "pcspkr" driver undere linux kernel source tree. I added a new line there: sample_id_dir = kobject_create_and_add("devices/platform/pcspkr/sample", NULL); It works and creates kobject without errors. I checked this in the driver code and in the dmesg output. But "sample" catalog isn't appear under "/sys/devices/platform/pcspkr/". What is wrong with that code ? Thanks in advance, Roman Storozhenko

2 3

VFS: mount_bdev and fill_super
by Shrikant Giridhar June 19, 2017

June 19, 2017

I'm a little confused about how the mount_bdev() call works. The VFS documentation lists it as a generic method to mount a block device. Based on what I understood from the source, it seems to open a block device whose path is passed via the mount() syscall and later associates it with the superblock, finally returning the corresponding dentry to the caller. The documentation also mentions that mount() can return a subtree of an existing filesystem on the block device. However, in the very next line it mentions that the caller also receives a pointer to the dentry of the root node. I can see from the source (fs/super.c:1099) that the mount_bdev call does check if the superblock's s_root == NULL or not and proceeds to set a new superblock if it is. I know my question sounds a little vague but I'm having trouble understanding how mount() can return a dentry for something other than the root? --- Shrikant

2 5

gdb not loading kernel module symbols at the correct offset
by Kamran Khan June 18, 2017

June 18, 2017

On the guest VM, I'm parsing kernel module's .text address with: > # cat /sys/module/dm_crypt/sections/{.text,.data} > 0xffffffffa04ee000 > 0xffffffffa04f3000 > root@ubuntu-zesty:~# cat /proc/kallsyms | grep dm_crypt_exit > ffffffffa04f14d2 t dm_crypt_exit [dm_crypt] When I load the module symbols at the .text address, it prompts me for the correct address: > (gdb) add-symbol-file drivers/md/dm-crypt.ko 0xffffffffa04ee000 -s .data 0xffffffffa04f3000 > add symbol table from file "drivers/md/dm-crypt.ko" at > .text_addr = 0xffffffffa04ee000 > .data_addr = 0xffffffffa04f3000 > (y or n) y > Reading symbols from drivers/md/dm-crypt.ko...done. But when I query for the module's symbols, they're not loaded at the appropriate offset, making it impossible to set breakpoints. > (gdb) info address dm_crypt_exit > Symbol "dm_crypt_exit" is a function at address 0x4e. > (gdb) b dm_crypt_exit > Cannot access memory at address 0x4e Any ideas what's going wrong? Thanks, Kamran.

1 0

Virtualbox + kgdb: Can't see vmlinux symbols
by Kamran Khan June 17, 2017

June 17, 2017

I have an Ubuntu 16.04 VM running in Virtualbox. I'm building kernel from git using make deb-pkg. After I install all the deb packages (including debug symbols) on the VM and restart it, I can see in uname -r output that the appropriate kernel was booted from. I'm passing the following parameters to the kernel: > kgdboc=ttyS0,115200 kgdbwait Once the kernel boots, I am able to connect to it via socat + gdb from host. I'm using the full uncompressed vmlinux file to start the gdb which contains all the debugging symbols. However, I can't see any of the symbols in backtrace etc. > $ du -sh vmlinux > 425M vmlinux > $ gdb vmlinux > GNU gdb (Ubuntu 7.11.1-0ubuntu1~16.04) 7.11.1 > Copyright (C) 2016 Free Software Foundation, Inc. > License GPLv3+: GNU GPL version 3 or later [gnu.org] > [...] > For help, type "help". > Type "apropos word" to search for commands related to "word"... > Reading symbols from vmlinux...done. > (gdb) target remote /dev/pts/2 > Remote debugging using /dev/pts/2 > 0xffffffffb113c5b4 in ?? () > (gdb) bt > #0 0xffffffffb113c5b4 in ?? () > #1 0xffffb035421c7dd0 in ?? () > #2 0xffffffffb113c60c in ?? () > #3 0xffffb035421c7e00 in ?? () > #4 0xffffffffb153035a in ?? () > #5 0x0000000000000002 in irq_stack_union () > #6 0x00000000013a7408 in ?? () > #7 0xffff8fb8bd0d3b00 in ?? () > #8 0xffffb035421c7e18 in ?? () > #9 0xffffffffb15307df in ?? () > #10 0xffff8fb8b8779900 in ?? () > #11 0xffffb035421c7e38 in ?? () > #12 0xffffffffb12ab412 in ?? () > #13 0xffff8fb8bd0d3b00 in ?? () Any ideas what am I doing wrong? Thanks, Kamran.

2 4

Struct sk_buff_head
by wiktoria.lewicka June 16, 2017

June 16, 2017

Hello! I read code from file include/linux/skbuff.h and I have question about sk_buff_head structure. This structure is pointing to next and prev sk_buff structure on list. But I found the same pointers in first field structure sk_buff- it is union, which has pointers to next and prev sk_buff structures on list. Is this union replace sk_buff_head structure? I didn`t found any field in sk_buff, which may use sk_buff_head, but I found some functions which uses this structure.

2 1

How to implement virtual hardware in QEMU?
by Douglas Su June 14, 2017

June 14, 2017

Hello everyone, I want to implement some virtual hardware in QEMU for driver testing. For example, to implement a timer which sends interrupt to CPU periodically. However, I can't find any documentation which details the interfaces and the framework. The only clue is the `hw` directory in the source tree of QEMU, but it is so complex for a beginners to start up. So, are there helpful videos or articles? Regards!

3 2

Edit permissions to help improve the wiki
by Jose Javier Segura June 14, 2017

June 14, 2017

Hello everybody, Reading the Documentation in the wiki I found a typo and I thought I would like to be added to the EditorsGroup list so that I can make corrections to the wiki when I find something wrong or to add more information in the future. Could someone who's in EditorsGroup add me in, please? Thank you. Jose

1 0

network driver tx/rx path buffers
by Tobin C. Harding June 13, 2017

June 13, 2017

Hi, I am implementing the rx/tx paths for a network driver. My current understanding is; On the tx path the driver processes the sk_buff, producing a frame, and writes this to the tx buffer. At interrupt time (given that the device produces interrupts for rx data received and tx data transmitted) the driver either fills the rx buffer with the received frame read from the device or empties the tx buffer and writes the frame to the device. Once rx data has been written to the buffer the driver should empty the rx buffer, process the frame and pass it upstairs. Questions --------- 1. Is it ok to process the rx frame in a tasklet or should the tasklet schedule a work thread to do it so that we are not in software interrupt context for the entire processing of the frame? 2. Can you us kfifo.h for varying size discreet chunks of data? The macros in kfifo.h are quite complex for a kernelnewbie, but it seems that kfifo is better suited to storing fixed size items (i.e stucts)? 3. If the above buffer handling is correct then the data being written to the rx and tx buffers is of varying size. Is circ_buf.h combined with buffer in the device data structure a better option? For reference, paraphrasing LDD3: For interrupt driven I/O the input buffer should be filled at interrupt time and emptied by processes that read the device; the output buffer should be filled by processes that write to the device and emptied at interrupt time. thanks, Tobin.

1 0