http://blog.gmane.org/gmane.linux.kernel.kernelnewbies hourly 1 1901-01-01T00:00+00:00 http://gmane.org/img/gmane-25t.png http://gmane.org http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28209 Hi All, I am new to linux kernel. I was refering to an interesting article titled "Inside the Linux scheduler", published at http://www.ibm.com/developerworks/linux/library/l-scheduler/. I have a questions related to "priority imbalance". The article mentions that each CPU has a runqueue made up of 140 priority lists i.e. per CPU ready queue. Hence it is possible that a runnable high priority task waits for a CPU in ready queue while a lower priority task is running on a different CPU. This can lead to priority imbalance/priority inversion. How does linux SMP scheduler handle this priority imbalance? Thanks in advance for you answer. Regards Thirupathiah Annapureddy 2008-11-28T15:08:31 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28203 Hi all, 1) I have a driver code providing ioctl,read, write, mmap etc. entry points. Now, i want to find the execution time of each entry point for comparison. How this can be done ? for ex. take time_stamp1 call entry point take time_stamp2 The above code from a user application can measure the execution time ? Timing functions in user space are close to the real time ? 2) Which timing functions should be used to get the, close to correct time, in user & kernel space ? Mainly the purpose is to calculate timing interval. Kindly help me. Any link/reference will be helpful. With regards, Yogeshwar -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis< at >nl.linux.org Please read the FAQ at http://kernelnewbies.org/FAQ yogeshwar sonawane 2008-11-28T09:10:32 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28202 Hi all, I need some information on profilers which can profile driver codes for linux. Mainly, it should be able to pin-point the functions which are taking more time, plus some other info which can be helpful in making code efficient. Kindly help me. Any reference/link will be useful. With regards, Yogeshwar -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis< at >nl.linux.org Please read the FAQ at http://kernelnewbies.org/FAQ yogeshwar sonawane 2008-11-28T09:03:33 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28201 Hi all, Suppose, I have an application or library code containing following :- while (expected condition has not occurred) { handling of expected condition } Now, If the expected condition does not occur, the program goes into an infinite loop. Now, how to detect this ? For user of the code who is running, it will appear as the program got stuck somewhere, not moving ahead. But user may not make out that the code is in infinite loop. How such info can be derived using some tools like debuggers ? I am thinking, whether it is possible to attach a debugger to a running code & detect such conditions ? name the debugger also or any other tools. What about similar situation in driver code ? Any link/reference will be useful. Thanks in advance, Yogeshwar -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis< at >nl.linux.org Please read the FAQ at http://kernelnewbies.org/FAQ yogeshwar sonawane 2008-11-28T08:55:09 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28198 In this module, i want to get the phy address for kmalloc() through page table. i think the phy address of kmalloc() is __pa(va), but when i use page table, it does not seem the same. May be i am not fully understand the principle. Is any one can help me about this program? #include <linux/module.h> #include <linux/kernel.h> #include <linux/init.h> #include <linux/sched.h> #include <linux/mm.h> #include <asm/pgtable.h> #include <asm/page.h> #include <linux/highmem.h> #include <asm/kmap_types.h> MODULE_LICENSE("GPL"; u32 ka = 0; static int find_phy_init(void) { printk(KERN_ALERT "Hello in\n"; u32 kpa = 0; pgd_t *k_pgd = NULL; pud_t *k_pud = NULL; pmd_t *k_pmd = NULL; pte_t *k_pte = NULL; ka = (u32)kmalloc(4096, GFP_KERNEL); if(!ka) { printk(KERN_ALERT "kmalloc failed\n"; return 0; } memset(ka, 0x5a, 4096); printk(KERN_ALERT "ka = 0x%08x\n", ka); /* kmalloc */ k_pgd = pgd_offset_k(ka); printk(KERN_ALERT "k_pgd = 0x%p, *k_pgd = 0x%08x\n", (u32)k_pgd, *k_pgd); if(!pgd_none(*k_pgd)) { k_pud = pud_offset(k_pgd, ka); if(!pud_none(*k_pud)) { k_pmd = pmd_offset(k_pud, ka); printk(KERN_ALERT "k_pmd = 0x%08x, *k_pmd = 0x%08x\n", (u32)k_pmd, *k_pmd); if(!pmd_none(*k_pmd)) { k_pte = pte_offset_kernel(k_pmd, ka); printk(KERN_ALERT "k_pte = 0x%08x, *k_pte = 0x%08x\n", (u32)k_pte, *k_pte); if(!pte_none(*k_pte)) { kpa = (pte_val(*k_pte) & PAGE_MASK) | (ka & ~PAGE_MASK); if(!pte_present(*k_pte)) printk(KERN_ALERT "k-page not in RAM\n"; } else printk(KERN_ALERT "pte is none\n"; } } } if(kpa != 0) { printk(KERN_ALERT "k_addr 0x%08x in RAM is 0x%08x\n", ka, kpa); printk(KERN_ALERT "__pa(k_addr) = 0x%08x\n", __pa(ka)); } kfree((void *)ka); return 0; } static void find_phy_exit(void) { printk(KERN_ALERT "Exit!\n"; } module_init(find_phy_init); module_exit(find_phy_exit); -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis< at >nl.linux.org Please read the FAQ at http://kernelnewbies.org/FAQ 胡明 2008-11-27T13:51:34 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28196 I am looking into some scalability issues on a 4-way Xeon machine (4 separate CPUs, not cores). I believe I have tracked down the problem to bus contention: OProfile results suggest a strong correlation between instructions reporting a high number of global_power_events and FSB_data_activity events. Some of these events can be easily explained. However, what surprises me, is that certain lock operations seem to cause considerable lock activity. For example, a call to spin_unlock_irqsave() from e1000_xmit_frame(). The strange thing about it is that the experiments I am conducting strictly partition the NICs among the CPUs (interrupt and process affinity), so that there is no contention on the lock (verified with lockstat). My understanding suggests that the variable of a lock that is only accessed by a single CPU should be constantly in the CPU's cache in Modified mode, as no other CPU is ever invalidating it, and thus there should be little if any FSB activity due to access to this variable. It should be noted that the number of FSB events per-cpu increases considerably when moving from 3 to 4 CPUs, while the number of cache misses stays roughly the same. Is my understanding correct? Are there any other reasons for FSB activity related to locks? --Elad -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis< at >nl.linux.org Please read the FAQ at http://kernelnewbies.org/FAQ Elad Lahav 2008-11-26T22:22:13 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28193 hi all I am a newbie to kernel programming.I wish to do a research project to bring the core gridd services to the kernel so as to build a grid Operating System.As the first step to do,I wish to configure the TUX server in my linux kernel-2.6.9.5. Please can be tell where i can get the sys_tux system call Please reply as early as possible regards Prabha prabha nair 2008-11-26T08:47:40 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28186 Hi Team, I want to test ethernet driver for NIC, I have only one PC with two different NICs(for both drivers are different), So I disabled loopback and connected both the NICs using cross cable. Output of "ifconfig" is displaying both the interfaces and ip address is also set for both the interfaces. I executed following to set ip address : #ifconfig eth0 192.168.0.1 netmask 255.255.255.0 #ifconfig eth1 192.168.0.5 netmask 255.255.255.0 #ifconfig lo down After that I tried to execute following commands, #route add -net 192.168.0.5 netmask 255.255.255.0 gw 192.168.0.5 dev eth0 #route add -net 192.168.0.1 netmask 255.255.255.0 gw 192.168.0.1 dev eth1 but those comands are giving error like : " route: netmask doesn't match route address" I also put "1" in ip_forward file. I am not able to ping anyone of those interfaces using ping command. Can anyone give some inputs on this? Thanks, Bhavesh. Bhavesh Kamani 2008-11-25T09:49:39 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28177 Hello, I am used to run "ARCH=powerpc make menuconfig" or similar, but I have always been wondering: Why isn't ARCH a CONFIG_ARCH inside the kbuild system? Regards, Leon Woestenberg 2008-11-24T15:17:35 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28174 Hello, as I promised - here's the announcement ;) My interactive filesystem is working and available at http://sourceforge.net/projects/iafs Keep in mind, that's my first try, so it will not at all be bugfree. Currently it's not possible, to mount from fstab - I'll have to dive a bit deeper into fuse to find my error. But it works as planned and expected and I already stressed the fs with xine and krusader - so I'm happy with the result. Comments and bugreports are always welcome. kind regards Santiago -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis< at >nl.linux.org Please read the FAQ at http://kernelnewbies.org/FAQ Santiago 2008-11-24T12:57:46 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28173 Hi All, I have created a /dev/ node using class_create() and class_device_create() APIs, by default the created node is having root only permisson as below crw-rw---- 1 root root 251, 0 2008-11-24 16:56 /dev/sp how can I give acces permission so every one can access that node I dont want to use chmod in user space intead it should be done in the kernel mode itsel .Is the any way to this stuff. Mohamed Thalib .H 2008-11-24T11:32:58 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28162 Hi, everyone, I'm an college student , and now I'm trying to analysis the old version of linux, does anybody do this ? -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis< at >nl.linux.org Please read the FAQ at http://kernelnewbies.org/FAQ leafan 2008-11-22T08:03:55 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28161 when a process mmap() a section of a file into its own process memory, the process memory will maintain a copy of the data of that section of the file. but then this is pagecache which cache the physically opened file, right? so...does there exists duplicated buffering? (one in kernel - pagecache, and one in userspace - for mmap() content of the file in process memory) Correct? So so...then it looks like a lot of duplicated stuff ..... Peter Teoh 2008-11-22T07:49:24 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28145 Hi to all, i am having linux installed on my laptop .but whenever i used to connect Projector to it it never used to show me new device has been detected to laptop.why? is there any way to make log in kernel buffer about such external VGA device like projector. or how to know that projector has connected or removed by using kernel log or /proc ,/sys etc. Thanks & Regards, Verule amol BOSS Team member amol verule 2008-11-20T12:48:30 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28136 Hello, X86 has 4 hardware debug registers, which could be used to implement watchpoints in debuggers. I checked gdb maillist and some gdb code to find out that gdb uses ptrace to set up hardware debug registers. I'm not familiar with details of ptrace syscall, but I know that a process can not ptrace itself. So there has to be a third-part program such as gdb to call ptrace. My question is: is there any approach to use debugger registers directly without using a third-party program? I mean to implement gdb's watchpoint in user program, so that I can set up hardware watchpoints manually and conveniently without gdb, any suggestions? thx Leo -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis< at >nl.linux.org Please read the FAQ at http://kernelnewbies.org/FAQ Wang Yi 2008-11-20T01:05:21 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28132 Hi, I am porting some test codes from Fedora core 2 to Fedora core 8.it works fine in FC2(2.6.5-1.358) but problem in FC8(2.6.23.1-42.fc8). Pgm.c --------- #include<stdio.h> int main void(int) { fprintf(stderr,"hello world\n",); } $ gcc -o Pgm Pgm.c $ vim myscript.sh ./Pgm & $ /usr/bin/ssh -x -n -o BatchMode=yes 127.0.0.1 cd /home/tom/Test_Dir/ '&&' ./myscript.sh If i run this command hello world is not printed. If i run this same in FC2 hello world in printed. Whether there is solution to this or it is a bug of FC8? -------------------------------------------------------------------------------------------------- Below solutions are correct?. -------------------------------------------------------------------------------------------- Solution 1: ---------- If i do these changes i.e if i re-direct to stdout and run in background. $ vim myscript.sh ./Pgm 2>&1 & $ /usr/bin/ssh -x -n -o BatchMode=yes 127.0.0.1 cd /home/tom/Test_Dir/ '&&' ./myscript.sh If i run this command it works fine. Solution 2: ---------- vim myscript.sh ./Pgm </dev/null >&/dev/null & $ /usr/bin/ssh -x -n -o BatchMode=yes 127.0.0.1 cd /home/tom/Test_Dir/ '&&' ./myscript.sh Regards, Hari Krishna New Email addresses available on Yahoo! Get the Email name you&#39;ve always wanted on the new < at >ymail and < at >rocketmail. Hurry before someone else does! http://mail.promotions.yahoo.com/newdomains/aa/ hari krishna angadi 2008-11-19T05:20:53 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28114 Hi All, I am writing an user space application which gets the statistics from the NIC drivers. To do this, I am using Ethtool ioctl with Ioctl command SIOCETHTOOL and with ethtool command ETHTOOL_GSTATS along with the address of struct ethtool_stats variable. My question is: What would be the format of the statistics copied from the kernel space (to the address pointed by the data field of struct ethtool_stats) ? Do all the drivers implement the same data format when they copy the statistics to user space in response to this Ethtool ioctl? If not, how can I know what format used by the driver I am working on ( I am working on Atheros driver)? Thanks in advance, Paraneetharan C -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis< at >nl.linux.org Please read the FAQ at http://kernelnewbies.org/FAQ Paraneetharan Chandrasekaran 2008-11-18T14:00:57 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28103 Hi, In the file Documentation/IO-mapping.txt it has been written as follows "---------- Now, just as an example, on the PReP (PowerPC Reference Platform), the CPU sees a memory map something like this (this is from memory): 0-2 GB"real memory" 2 GB-3 GB"system IO" (inb/out and similar accesses on x86) 3 GB-4 GB "IO memory" (shared memory over the IO bus) Now, that looks simple enough. However, when you look at the same thing from the viewpoint of the devices, you have the reverse, and the physical memory address 0 actually shows up as address 2 GB for any IO master. -----" Would someone please clarify how does an IO master sees physical memory address 0 as 2GB?. Is it specific to PowerPC? Thanks in advance, Sudheer sudheer.divakaran< at >wipro.com 2008-11-18T02:50:57 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28102 What test suite is ran when new Linux kernel is released. I am perticularly interested in socket api and networking subsystem testing. Thanks Usman -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis< at >nl.linux.org Please read the FAQ at http://kernelnewbies.org/FAQ Usman S. Ansari 2008-11-18T01:22:24 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28099 hi, all as I understand, ->readpage() will be invoked on address space object at some point when you do a syscall on read. for ext3, it will be ext3_readpage, which ask mpage_readpage() to do the job. what I don't follow is why the function doc right before mapge_readpage() claims "This isn't called much at all". is this not on the primary read I/O path anymore? thanks Ruby Ruby 2008-11-17T17:06:07 http://comments.gmane.org/gmane.linux.kernel.kernelnewbies/28094 Hi All, As I read from Understanding The Linux Kernel for each process, Linux packs two different data structures in a single per-process memory area: a small data structure linked to the process descriptor, namely the thread_info structure, and the Kernel Mode process stack. Here I am not understanding significance of thread info structure. shyam Burkule 2008-11-17T15:22:13 Search the mailing list at Gmane query http://search.gmane.org/?group=$group=gmane.linux.kernel.kernelnewbies