October 2012 - kernelnewbies@kernelnewbies.org

Advice on first patch
by matthew＠walster.org Oct. 4, 2012

Oct. 4, 2012

I'm hopefully going to be submitting the attached patch to the mainline kernel tree, and as it's my first patch, I figured it would be wise to run it past KN first in case I'd done something monumentally stupid! I found it through a checkpatch run, from the excellent talk by GregKH at FOSDEM a year or so back. Any input would be greatly appreciated! Matthew Walster

5 10

when to return -EOPNOTSUPP in _ioctl in a driver
by devendra.aaru Oct. 4, 2012

Oct. 4, 2012

Hi all, While going through some code, i have found a place in which we do return, -EOPNOTSUPP in an ioctl and the code of ioctl does just this one. static int drv_ioctl(.. *fops, ...) { return -EOPNOTSUPP; } do we actually need to specify a pointer to the ioctl callback, if the driver doesn't support ioctl?

2 4

Why "lsusb" return nothing?
by Peter Teoh Oct. 3, 2012

Oct. 3, 2012

I entered "lsusb" at the command line (as root) and nothing is return, not even any error message. Doing a strace the last few lines are: open("/dev/bus/usb", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = -1 ENOENT (No such file or directory) open("/proc/bus/usb", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = -1 ENOENT (No such file or directory) What happened? This is Ubuntu 10.04 (it used NOT to be like that, not sure I what did wrong last time). But running a VirtualBox INSIDE this same OS, I was able to get result from "lsusb" (after enabling the USB devices in VirtualBox interface) and strace gives result: open("/dev/bus/usb/001/002", O_RDWR) = 3 ioctl(3, USBDEVFS_IOCTL, 0xbff6f75c) = -1 ENOTTY (Inappropriate ioctl for device) close(3) = 0 open("/dev/bus/usb/001/001", O_RDWR) = 3 Why the difference? -- Regards, Peter Teoh

4 8

Facing trouble in creating a packet in kernel space
by Rifat Rahman Oct. 2, 2012

Oct. 2, 2012

Hello there, I need to mangle rtp packets in kernel space. So far I am new in kernel module programming. I am trying to implement a module for netfilter hooks. For the first time as exercise, I am trying to write smaller modules. Let me explain what I am actually doing now. I have an echo client and server. The server runs on port 6000. Both are on different machines (May be VMs in bridge filter mode). The client sends udp message and the server just echoes it back. Let us suppose the client sends "some message" as data. Then now I am trying to write a module for the client machine that will append "12345" after the data so that the server will get "some message12345" and echo it back. Now there are various things I did faced. I relied on the NF_IP_POST_ROUTING hook. At first, I copied the data to a temporary storage, and then add 12345 with that. Then I increase skb->tail using skb_put(). Then I memset() 0 to the packet data, and copy the temporary storage with that. Then as the procedure, NF_ACCEPT is returned. There are certain checking points like the udph->dest == 6000 etc. etc. When I use skb_put(), my system hangs out after two or three minutes. When I dmesg to be certain that everything goes right, I find it OK. But, suppose once I send a message like "This is a pretty big message" and another time I send "small message" then I get just "small message12345g message" that means, the bigger message is stored somewhere I don't understand. I tried with skb_add_data() but that works incorrectly here, I understand it's my fault. I just can't figure it out. Now, one thing came in my mind, if it's not possible, should I create new packets for that data appending? I find skb->end - skb->tail is not so big. But ultimately I have to merge two or three packets into one packet and then skb_put() will not suffice for me. Then the point comes, I can use alloc_skb(), skb_reserve(), skb_header_pointer() and other skb manipulation functions, but I don't understand how can I drop the packet got (should I return NF_DROP?) and how can I route my created packets in the packet flowing path? I saw in the xtables_addons for help (file xt_ECHO.c) but found they used ip_route_me_harder() for it. But the function they used is quite different, it is their own. Then I need suggestions about how to route the newly created packet in kernel space. This is the point I am stuck for a few days. If anyone can help me, I will be highly grateful. I understand my knowledge is poor about Linux kernel, but I need a definite way to understand, otherwise the code is not helping me much. Thanks in advance. -- Rifat Rahman

3 2

Re: How to automatically get subsystem name for a file?
by Ezequiel Garcia Oct. 2, 2012

Oct. 2, 2012

On Mon, Oct 1, 2012 at 5:51 PM, Dan Carpenter <dan.carpenter(a)oracle.com> wrote: > On Mon, Oct 01, 2012 at 04:49:13PM -0300, Ezequiel Garcia wrote: >> On Mon, Oct 1, 2012 at 3:49 PM, Dan Carpenter <dan.carpenter(a)oracle.com> wrote: >> > On Mon, Oct 01, 2012 at 06:57:44PM +0200, Peter Senna Tschudin wrote: >> >> I'm trying to figure it out how to automatically get the correct >> >> subsystem string for putting on the first line of the commit message / >> >> subject of the patch message. For example: >> >> >> >> Subject: [PATCH 001/142] arch/x86: Replace memcpy with struct assignment >> >> ^^^^^^^^^^ >> >> arch/x86 is only the first two levels of directories from Kernel >> >> source. This may not be smart enough... >> >> >> > >> > You have to do it manually. Here are the relevant lines from my >> > patch script. >> > >> >> Manually? Sure there must be some way of auto do it. > > There really isn't an automatic way. For example [media] and [SCSI] > are non-standard. > > I look at it as a good thing that people have to do it manually. It > hopefully makes them slow down a bit. > Agreed. Slow down, less maintainers head. Thanks, Ezequiel.

1 0

Re: How to automatically get subsystem name for a file?
by Ezequiel Garcia Oct. 1, 2012

Oct. 1, 2012

On Mon, Oct 1, 2012 at 3:49 PM, Dan Carpenter <dan.carpenter(a)oracle.com> wrote: > On Mon, Oct 01, 2012 at 06:57:44PM +0200, Peter Senna Tschudin wrote: >> I'm trying to figure it out how to automatically get the correct >> subsystem string for putting on the first line of the commit message / >> subject of the patch message. For example: >> >> Subject: [PATCH 001/142] arch/x86: Replace memcpy with struct assignment >> ^^^^^^^^^^ >> arch/x86 is only the first two levels of directories from Kernel >> source. This may not be smart enough... >> > > You have to do it manually. Here are the relevant lines from my > patch script. > Manually? Sure there must be some way of auto do it. Perhaps, parsing Makefile or (if we have access to a build) use foo.ko files. I wonder why we don't have a standardized way of naming drivers, yet. Something like MODULE_AUTHOR, called MODULE_NAME. (Yes, I realize it won't work if the driver is built-in) > git log --oneline $fullname | head -n 10 > echo "Copy and paste one of these subjects?" > read unused > > You should be reading through the patch manually anyway and the > reviewer needs to read it manually. It's not like the 20 seconds > it takes to consider which prefix to use is a big deal. > Sorry Dan, I couldn't catch what you meant by this. I guess my english slang is not *that* sharp. Thanks, Ezequiel.

1 0

How to automatically get subsystem name for a file?
by Peter Senna Tschudin Oct. 1, 2012

Oct. 1, 2012

I'm trying to figure it out how to automatically get the correct subsystem string for putting on the first line of the commit message / subject of the patch message. For example: Subject: [PATCH 001/142] arch/x86: Replace memcpy with struct assignment ^^^^^^^^^^ arch/x86 is only the first two levels of directories from Kernel source. This may not be smart enough... What should be the subsystem for the files: arch/x86/kernel/e820.c drivers/acpi/thermal.c drivers/atm/idt77105.c drivers/block/xen-blkfront.c drivers/media/dvb-core/dvb_frontend.c drivers/media/usb/pvrusb2/pvrusb2-v4l2.c drivers/misc/phantom.c drivers/misc/sgi-gru/grufault.c drivers/net/usb/hso.c drivers/net/wireless/at76c50x-usb.c drivers/scsi/ncr53c8xx.c drivers/sh/intc/core.c drivers/staging/iio/light/tsl2x7x_core.c fs/btrfs/tree-defrag.c fs/cifs/connect.c fs/devpts/inode.c net/bridge/netfilter/ebtables.c net/ceph/auth_x.c net/rds/ib_send.c sound/pci/bt87x.c sound/soc/codecs/wm8994.c sound/usb/caiaq/audio.c get_maintainers.pl helps but not a lot. Here is it's output for the files above: arch/x86/kernel/e820.c:X86 ARCHITECTURE (32-BIT AND 64-BIT) drivers/acpi/thermal.c:ACPI THERMAL DRIVER drivers/atm/idt77105.c:ATM drivers/block/xen-blkfront.c:XEN HYPERVISOR INTERFACE drivers/media/dvb-core/dvb_frontend.c:MEDIA INPUT INFRASTRUCTURE (V4L/DVB) drivers/media/usb/pvrusb2/pvrusb2-v4l2.c:PVRUSB2 VIDEO4LINUX DRIVER drivers/misc/phantom.c:SENSABLE PHANTOM drivers/misc/sgi-gru/grufault.c:SGI GRU DRIVER drivers/net/usb/hso.c:HSO 3G MODEM DRIVER drivers/net/wireless/at76c50x-usb.c:NETWORKING [WIRELESS] drivers/scsi/ncr53c8xx.c:SCSI SUBSYSTEM drivers/sh/intc/core.c:ARM/SHMOBILE ARM ARCHITECTURE drivers/staging/iio/light/tsl2x7x_core.c:STAGING - INDUSTRIAL IO fs/btrfs/tree-defrag.c:BTRFS FILE SYSTEM fs/cifs/connect.c:COMMON INTERNET FILE SYSTEM (CIFS) fs/devpts/inode.c:THE REST net/bridge/netfilter/ebtables.c:EBTABLES net/ceph/auth_x.c:CEPH DISTRIBUTED FILE SYSTEM CLIENT net/rds/ib_send.c:RDS - RELIABLE DATAGRAM SOCKETS sound/pci/bt87x.c:BT87X AUDIO DRIVER sound/soc/codecs/wm8994.c:WOLFSON MICROELECTRONICS DRIVERS sound/usb/caiaq/audio.c:NATIVE INSTRUMENTS USB SOUND INTERFACE DRIVER Subject: [PATCH] NATIVE INSTRUMENTS USB SOUND INTERFACE DRIVER: Replace memcpy with struct assignment Do not look as good option too. It is possible to use full path of the file but leads to Subject over 80 chars. Any ideas? -- Peter

1 0

Difference between IP_IP_ADD_MEMBERSHIP and MCAST_JOIN_GROUP
by Pietro Paolini Oct. 1, 2012

Oct. 1, 2012

Hi all, I am working with multicast socket and I saw in the source of the kernel different define for identical operation, what should I use in a setsockopt() ? The file is /incude/linux/in.h #define IP_MULTICAST_IF<http://lxr.linux.no/linux+v3.2.9/+code=IP_MULTICAST_IF> 32 #define IP_MULTICAST_TTL<http://lxr.linux.no/linux+v3.2.9/+code=IP_MULTICAST_TTL> 33 #define IP_MULTICAST_LOOP<http://lxr.linux.no/linux+v3.2.9/+code=IP_MULTICAST_LOOP> 34 #define IP_ADD_MEMBERSHIP<http://lxr.linux.no/linux+v3.2.9/+code=IP_ADD_MEMBERSHIP> 35 #define IP_DROP_MEMBERSHIP<http://lxr.linux.no/linux+v3.2.9/+code=IP_DROP_MEMBERSHIP> 36 #define IP_UNBLOCK_SOURCE<http://lxr.linux.no/linux+v3.2.9/+code=IP_UNBLOCK_SOURCE> 37 #define IP_BLOCK_SOURCE<http://lxr.linux.no/linux+v3.2.9/+code=IP_BLOCK_SOURCE> 38 #define IP_ADD_SOURCE_MEMBERSHIP<http://lxr.linux.no/linux+v3.2.9/+code=IP_ADD_SOURCE_MEMBERSHIP> 39 #define IP_DROP_SOURCE_MEMBERSHIP<http://lxr.linux.no/linux+v3.2.9/+code=IP_DROP_SOURCE_MEMBERSHIP> 40 #define IP_MSFILTER<http://lxr.linux.no/linux+v3.2.9/+code=IP_MSFILTER> 41 #define MCAST_JOIN_GROUP<http://lxr.linux.no/linux+v3.2.9/+code=MCAST_JOIN_GROUP> 42 #define MCAST_BLOCK_SOURCE<http://lxr.linux.no/linux+v3.2.9/+code=MCAST_BLOCK_SOURCE> 43 #define MCAST_UNBLOCK_SOURCE<http://lxr.linux.no/linux+v3.2.9/+code=MCAST_UNBLOCK_SOURCE> 44 #define MCAST_LEAVE_GROUP<http://lxr.linux.no/linux+v3.2.9/+code=MCAST_LEAVE_GROUP> 45 #define MCAST_JOIN_SOURCE_GROUP<http://lxr.linux.no/linux+v3.2.9/+code=MCAST_JOIN_SOURCE_GROUP> 46 #define MCAST_LEAVE_SOURCE_GROUP<http://lxr.linux.no/linux+v3.2.9/+code=MCAST_LEAVE_SOURCE_GROUP> 47 #define MCAST_MSFILTER<http://lxr.linux.no/linux+v3.2.9/+code=MCAST_MSFILTER> 48 I am working with IGMP V3 and I need to handle sources . Thanks in advance, Pietro Paolini

2 2