October 2011 - kernelnewbies@kernelnewbies.org

Question on Memory Leak: very tough problem
by Sri Ram Vemulpali Oct. 27, 2011

Oct. 27, 2011

Hi All, I have very difficult problem I am working on to figure out where the memory leak is in my system. Sometimes it shows up and sometimes it reveals very slowly (I mean increments very slowly). This is dependent on system load, if the system is in very high load, then leak is very fast. We tried solving this problem from many dimensions, but no success. After quite sometime, we started using valgrind to capture the memory leak. The funny thing is the leak disappears once we start running our application on top of valgrind. So, we cannot capture leak using valgrind. We came to conclusion that, since valgrind serializes the allocations (leak disappears), we thought this might be result of race conditions, that resulted from system high load (by the way ours is multi threaded system). We run system without valgrind we see the memory leak. The question is, is it true that valgrind serializes the mem allocations. Does this happen to anyone disappearing the leak when used valgrind. Any thoughts and input that can help me solve this problem. Thanks to all in advance. -- Regards, Sri.

3 2

how to check kernel is configured with preemption or not
by sri Oct. 27, 2011

Oct. 27, 2011

Hi, Am using kernel 2.6.18-195(centos 5.5). My kernel configs have CONFIG_PREEMPT_NONE=7 and "CONFIG_PREEMPT_VOLUNTERY is not set". How to check that preemption is really in place? Is there any way to check my kernel is configured with what preemption levels? Thanks inadvance. -- Sri -- -- Krishna Mohan B

6 8

Workqueue and kernel panic on device disconnect
by Frank Schäfer Oct. 26, 2011

Oct. 26, 2011

Hi, I'm trying to use a workqueue in a usb driver for polling the states of the device buttons. That works fine, but when I disconnect the device, the kernel sometimes explodes. The machine turns off immediately, without saving a backtrace (I will try to make a photo tomorrow) The code: static struct delayed_work poll_work; static const struct usb_device_id device_table[] = { {USB_DEVICE(0x1234, 0x5678)}, {} }; static void poll(struct work_struct *work) { schedule_delayed_work(&poll_work, msecs_to_jiffies(100)); } MODULE_DEVICE_TABLE(usb, device_table); static int test_probe(struct usb_interface *intf, const struct usb_device_id *id) { INIT_DELAYED_WORK(&poll_work, poll); schedule_delayed_work(&poll_work, msecs_to_jiffies(100)); return 0; } static void test_disconnect(struct usb_interface *intf) { cancel_delayed_work_sync(&poll_work); } static struct usb_driver test_driver = { .name = MODULE_NAME, .id_table = device_table, .probe = test_probe, .disconnect = test_disconnect, }; static int __init test_mod_init(void) { return usb_register(&test_driver); } static void __exit test_mod_exit(void) { usb_deregister(&test_driver); } module_init(test_mod_init); module_exit(test_mod_exit); What am I doing wrong ? Is there anything else I have to do in addition to calling cancel_delayed_work_sync() in the disconnect fcn ? Thanks, Frank Schaefer

1 1

Fix procmail rule in ML wikipage
by chandanrmail＠gmail.com Oct. 25, 2011

Oct. 25, 2011

The procmail rule for "netdev:Linux network development list" has a typo. The "condition" should be as shown below: * ^X-Mailing-List:.*netdev@vger\.kernel\.org Regards, chandan r

1 0

Purpose of using __be16 inside a data structure?
by 陳國成 Oct. 25, 2011

Oct. 25, 2011

Hi all, Please tell me the difference between using unsigned short and __be16 to declear a variable in a data structure. For example, in include/net/inet_sock.h, struct inet_sock is decleared as struct inet_sock { __be16 inet_dport; }; not struct inet_sock { unsigned short inet_dport; }; I think it has something to do with endianness. But I see that both __le16 and __be16 are unsigned short when using gcc with -E option for little-endian and big-endian platform. Can someone give me more information? * little-endian typedef unsigned short __u16; typedef __u16 __le16; typedef __u16 __be16; * big-endian typedef unsigned short __u16; typedef __u16 __le16; typedef __u16 __be16; Regards, GC

4 4

Re: Linux 0.02 version download
by Javier Martinez Canillas Oct. 25, 2011

Oct. 25, 2011

On Tue, Oct 25, 2011 at 9:36 AM, bob <ilikepie420(a)live.com> wrote: > He's probably doing os dev and wants something simple to learn. Stick with > 0.01 as 0.02 source is lost. Sad face. Yes I know, but I recommend either use a newer kernel or use an operating system that was written for teaching like: http://pdos.csail.mit.edu/6.828/2011/xv6.html I don't want to imagine how hard could be compiling a 0.01 kernel with GCC 4.5 and find a rootfs that is suitable for it. Best regards, -- Javier Martínez Canillas (+34) 682 39 81 69 Barcelona, Spain

4 3

Linux 0.02 version download
by Venkateswarlu P Oct. 25, 2011

Oct. 25, 2011

I want to download Linux 0.02 version I anybody knows plese send the link I already tried in the google -- ** P VENKATESWARLU M.Tech (CSE) II year NIT Calicut ,Kerala* ** *+91-9995115752

3 2

Is there a tool which can generate a small and simple source tree after config?
by jiangtao.jit Oct. 25, 2011

Oct. 25, 2011

Hi: While reading the kernel code 1. the huge amount of files make me scared there are too many alike functions in different files sometimes I can't figure out which one was really compiled 2. too many macros in the definition of a struct or functions declaration confused me a lot I tried to follow the generated file autoconf.h to guess the final face but it's really a difficult thing Is there a tool which can generate a small and simple source tree after config? not pre-processor just generate a small source tree contains the files and dirs which really will be compiled and no macros like CONFIG_SMP etc. according to the configuration or some other way to understand the architecture of the final working source tree? Thanks. -- 2011-10-23 jiangtao.jit

4 3

Please help me about the default rootfs
by Renjun Qu Oct. 25, 2011

Oct. 25, 2011

Hello everyone: I am building a small embedded linux system right now. I confused that why the default rootfs must has a "dev/console" node. I have made a experiment as follow: 1st, configure the kernel(2.6.32) to support the initramfs and initrd, and set the source of the initramfs to a directory which only has a "root" node. 2nd, configure the kernel to support the root file system on nfs, and rebuild the kernel. I have created all necessary files and directories in the directory exported by the nfs server. 3rd, set the appropriate boot parameter, and boot my kernel. The result is, the kernel can successfully mount the root file system, but the "/sbin/init" program can not print any information to me. My "sbin/init" program is just a hello world program which only print some greeting information.There is a "dev/console" device node in the nfs exported directory. But if i set the source of the initramfs to empty, and redo my experiment from the 2nd step,everything will be ok. So, i think there must be a “dev/console” node in the rootfs. The real root file system already has a "dev/console" node, why default rootfs must have a dev/console node. Please help me. Best regards, Ren jun Qu

1 0

How to interpret cat /proc/meminfo
by sandeep kumar Oct. 25, 2011

Oct. 25, 2011

Hi all, Here in my device cat /proc/meminfo is showing the following result, MemTotal: 767960 kB MemFree: 104252 kB Buffers: 14696 kB Cached: 183676 kB SwapCached: 0 kB Active: 304596 kB Inactive: 282020 kB Active(anon): 250560 kB Inactive(anon): 139488 kB Active(file): 54036 kB Inactive(file): 142532 kB Unevictable: 988 kB Mlocked: 0 kB HighTotal: 536576 kB HighFree: 23408 kB LowTotal: 231384 kB LowFree: 80844 kB SwapTotal: 0 kB SwapFree: 0 kB Dirty: 0 kB Writeback: 0 kB AnonPages: 389252 kB Mapped: 94368 kB Shmem: 816 kB Slab: 17932 kB SReclaimable: 7388 kB SUnreclaim: 10544 kB KernelStack: 5912 kB PageTables: 20064 kB NFS_Unstable: 0 kB Bounce: 0 kB WritebackTmp: 0 kB CommitLimit: 383980 kB Committed_AS: 3685728 kB VmallocTotal: 434176 kB VmallocUsed: 261584 kB VmallocChunk: 115324 kB && My kernel virtual memory lay out is like this, <5>[ 0.000000] vector : 0xffff0000 - 0xffff1000 ( 4 kB) <5>[ 0.000000] fixmap : 0xfff00000 - 0xfffe0000 ( 896 kB) <5>[ 0.000000] DMA : 0xff000000 - 0xffe00000 ( 14 MB) <5>[ 0.000000] vmalloc : 0xdf800000 - 0xfa000000 ( 424 MB) <5>[ 0.000000] lowmem : 0xc0000000 - 0xdf000000 ( 496 MB) <5>[ 0.000000] pkmap : 0xbfe00000 - 0xc0000000 ( 2 MB) <5>[ 0.000000] modules : 0xbf000000 - 0xbfe00000 ( 14 MB) <5>[ 0.000000] .init : 0xc0008000 - 0xc0049000 ( 260 kB) <5>[ 0.000000] .text : 0xc0100000 - 0xc0a26000 (9368 kB) <5>[ 0.000000] .data : 0xc0b00000 - 0xc0c464c0 (1306 kB) && In the kernel command line i had put the "vmalloc=432MB" , to reserve the vmalloc area. My question is, While in the kernel memory layout, lowmem is shown as 496MB, in /proc/meminfo LowMEMtotal is only 231MB. Why is the difference between the two? How to increase the Lowmem size with reserving my vmalloc region memory(432MB) -- With regards, Sandeep Kumar Anantapalli,

2 1