July 2013 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

Need help to understand system calls
by Nitisha Jain July 25, 2013

July 25, 2013

Hello all, I am running the Iozone benchmark commands to find disk characteristics on my system and trying to construct a call graph for the iozone functions. I am using perf tools to generate a call graph , but I am having trouble understanding the output from the perf commands as follows-- # Events: 62K cpu-clock # # Overhead Command Shared Object Symbol # ........ .............. .......................... .............................................................................................................................. # 79.23% swapper [kernel.kallsyms] [k] native_safe_halt | --- native_safe_halt default_idle cpu_idle start_kernel x86_64_start_kernel 4.99% iozone [kernel.kallsyms] [k] copy_user_generic_unrolled | --- copy_user_generic_unrolled iov_iter_copy_from_user_atomic generic_perform_write generic_file_buffered_write __generic_file_aio_write generic_file_aio_write ext4_file_write do_sync_write vfs_write sys_write ia32_sysret 0xf77904be __write_nocancel I need to understand these kernel commands , kindly give some pointers for the same. Are there any resources available online ?? I am attaching the complete file along with the mail. Thanks and regards, *Nitisha*

3 3

pulling in patches mentioned in mails (linux-kernel@vger.kernel.org)
by kernel neophyte July 25, 2013

July 25, 2013

Hi, I am sorry if this has been asked before, I tried to look around but did not get an solution and hence the mail. In this mail from Mel Gorman: https://lkml.org/lkml/2013/7/15/362 He mentions a set of patches: https://lkml.org/lkml/2013/7/15/292 https://lkml.org/lkml/2013/7/15/383 https://lkml.org/lkml/2013/7/15/322 Could someone please teach me 1) how do I find out where is the tree he is working on (i.e. is he working on linux-next or ... ) 2) Where do I find these patches mentioned ? Currently I cut copy paste in a vim manually and its really frustrating, could some one please point out the right way of pulling these patches. via git ? Thanks a lot for your time in advance. -Neophyte PS: This is my first ever mail in public, please pardon my grammatical errors or choice of words.

1 0

Re: preemptive kernels and the use of smp_processor_id()
by Valdis.Kletnieks＠vt.edu July 24, 2013

July 24, 2013

On Wed, 24 Jul 2013 18:52:40 -0000, Rajat Jain said: > I though preemptible only means that a process can only be preempted (not > moved to another core) under following situations: > 1) An interrupt happens (In case of timer, it could put the process back in queue if its time slice expired). Actually, in case of any interrupt, be it timer or I/O or NMI or whatever. > 2) The process goes to sleep (can be a result of a call that can sleep). And there's no actual guarantee that when the process gets rescheduled, that you'll return to the same core you were on. > Can some one shed light on what conditions would a running process be moved to another core? Pretty much anytime the scheduler comes to that decision. :)

2 1

preemptive kernels and the use of smp_processor_id()
by Christian Parpart July 24, 2013

July 24, 2013

Hey guys, while attempting to port a relatively small kernel module that was written for Linux kernel 3.1.4 to git master, I was encountering a few issues. First to note, I am really new to kernel development. So while knowing C, I don't know much about the kernel APIs yet :) Things that have changed mostly has been visibility changes, like symbols got unexported and such, but the thing that worries me are the stack traces regarding the use of smp_processor_id() when using a PREEMPTible kernel. So it seems the kernel module I am porting assumed to always run on the same CPU core once entered kernelspace, but on a preemptive system you cannot guarantee that. So my question is, how can I force certain calls not to be preempted by the scheduler and/or not changing the core it was running on? AFAIK, the userspace has already defined a cpu affinity mask, and that's why that kernel module assumes that. 1.) If that stack trace is non-fatal, can I make it less spammy in the log files, meaning, that the use of smp_processor_id() is only used for logging / informational purposes? 2.) there is one part in the kernel where it actually uses the smp_processor_id() value for further tests, so here it really shouldn't reschedule during the call - how do I ensure not being preempted there then? Many thanks, and best regards, Christian Parpart.

3 2

On pointer type and the return type of ioremap
by Grissiom July 24, 2013

July 24, 2013

Hi all, I'm writing my little module now a days and want to ioremap some space. I learnt from LDD3 that "Therefore, generic memory addresses in the kernel are usually unsigned long"(ch11). So I declared a pointer like "usigned long p2dev;" But when I feed it with ioremap, it yields a warning: warning: assignment makes integer from pointer without a cast [enabled by default] Do I really have to add a cast? What is the reason that while "generic pointers are unsigned long" but the return type of ioremap is void *? Thanks in advance. -- Cheers, Grissiom

1 0

Notifier chains
by PV Juliet July 23, 2013

July 23, 2013

Hi , I am new to kernel and i would like to know where i can get sample programs to use notifier chains .How to subscribe the notifier chains to some application so that i will get notified. Is there any way make this as service in linux system?? Thanks and Regards Juliet

4 4

kernel panic while doing dma_alloc_coherent() function
by sandeep kumar July 22, 2013

July 22, 2013

Hi guys I am trying to port opensource kernel 3.0.11 onto MSM chipset. While porting the SD card controller driver, i am seeing a kernel panic. While tracing with printk, i am struck at in arch/arm/mm/dma_mapping.c , Function, static struct page *__dma_alloc_buffer(struct device *dev, size_t size, gfp_t gfp) { <snip> ..... ptr = page_address(page); memset(ptr, 0, size); dmac_flush_range(ptr, ptr + size); --> struck here <snip> } And i am seeing the following call stack, 1>Unable to handle kernel paging request at virtual address 60600000 <1>pgd = c0004000 <1>[60600000] *pgd=00000000 <0>Internal error: Oops: 5 [#1] PREEMPT CPU: 0 Not tainted (3.0.31-00022-gb0c6bf3-dirty #2) PC is at copy_page+0x10/0x60 LR is at v6_copy_user_highpage_nonaliasing+0x8c/0xec pc : [<c010fb30>] lr : [<c002b5e8>] psr: 30000013 sp : ef82de28 ip : 00000000 fp : c029b19c r10: c029ace8 r9 : c0d02000 r8 : 000000d0 r7 : 00017fff r6 : 60600000 r5 : c030af34 r4 : 5fa00000 r3 : c0314000 r2 : 0000003f r1 : 60600000 r0 : 5fa00000 Flags: nzCV IRQs on FIQs on Mode SVC_32 ISA ARM Segment kernel Control: 10c57c7d Table: 00404059 DAC: 00000015 <0>Process swapper (pid: 1, stack limit = 0xef82c268) <0>Stack: (0xef82de28 to 0xef82e000) [<c010fb30>] (copy_page+0x10/0x60) from [<c002b5e8>] (v6_copy_user_highpage_nonaliasing+0x8c/0xec) [<c002b5e8>] (v6_copy_user_highpage_nonaliasing+0x8c/0xec) from [<c0028578>] (__dma_alloc+0x114/0x2e8) [<c0028578>] (__dma_alloc+0x114/0x2e8) from [<c00287cc>] (dma_alloc_coherent+0x54/0x60) [<c00287cc>] (dma_alloc_coherent+0x54/0x60) from [<c0194da8>] (msmsdcc_probe+0x534/0x1100) [<c0194da8>] (msmsdcc_probe+0x534/0x1100) from [<c01606d0>] (platform_drv_probe+0x18/0x1c) [<c01606d0>] (platform_drv_probe+0x18/0x1c) from [<c015f4a0>] (driver_probe_device+0xf8/0x184) [<c015f4a0>] (driver_probe_device+0xf8/0x184) from [<c015f5b8>] (__driver_attach+0x8c/0x90) [<c015f5b8>] (__driver_attach+0x8c/0x90) from [<c015ec7c>] (bus_for_each_dev+0x60/0x8c) [<c015ec7c>] (bus_for_each_dev+0x60/0x8c) from [<c015e558>] (bus_add_driver+0xa0/0x228) [<c015e558>] (bus_add_driver+0xa0/0x228) from [<c015fb8c>] (driver_register+0x78/0x144) [<c015fb8c>] (driver_register+0x78/0x144) from [<c00083d0>] (do_one_initcall+0x94/0x164) [<c00083d0>] (do_one_initcall+0x94/0x164) from [<c0008560>] (kernel_init+0x8c/0x278) [<c0008560>] (kernel_init+0x8c/0x278) from [<c002467c>] (kernel_thread_exit+0x0/0x8) <0>Code: e92d4010 f5d1f000 f5d1f020 e3a0203f (e8b15018) <4>---[ end trace ce4f1a328aa6a271 ]--- <6>note: swapper[1] exited with preempt_count 2 <0>Kernel panic - not syncing: Attempted to kill init! [<c0027ab4>] (unwind_backtrace+0x0/0xfc) from [<c0213054>] (panic+0x64/0x188) [<c0213054>] (panic+0x64/0x188) from [<c0045204>] (do_exit+0x664/0x6c8) [<c0045204>] (do_exit+0x664/0x6c8) from [<c0026e78>] (die+0x238/0x2b4) [<c0026e78>] (die+0x238/0x2b4) from [<c0028b10>] (__do_kernel_fault+0x64/0x84) [<c0028b10>] (__do_kernel_fault+0x64/0x84) from [<c0028c70>] (do_page_fault+0x140/0x1f4) [<c0028c70>] (do_page_fault+0x140/0x1f4) from [<c0023234>] (do_DataAbort+0x34/0x98) [<c0023234>] (do_DataAbort+0x34/0x98) from [<c002392c>] (__dabt_svc+0x4c/0x60) Exception stack(0xef82dde0 to 0xef82de28) dde0: 5fa00000 60600000 0000003f c0314000 5fa00000 c030af34 60600000 00017fff de00: 000000d0 c0d02000 c029ace8 c029b19c 00000000 ef82de28 c002b5e8 c010fb30 de20: 30000013 ffffffff [<c002392c>] (__dabt_svc+0x4c/0x60) from [<c010fb30>] (copy_page+0x10/0x60) [<c010fb30>] (copy_page+0x10/0x60) from [<c002b5e8>] (v6_copy_user_highpage_nonaliasing+0x8c/0xec) [<c002b5e8>] (v6_copy_user_highpage_nonaliasing+0x8c/0xec) from [<c0028578>] (__dma_alloc+0x114/0x2e8) [<c0028578>] (__dma_alloc+0x114/0x2e8) from [<c00287cc>] (dma_alloc_coherent+0x54/0x60) [<c00287cc>] (dma_alloc_coherent+0x54/0x60) from [<c0194da8>] (msmsdcc_probe+0x534/0x1100) [<c0194da8>] (msmsdcc_probe+0x534/0x1100) from [<c01606d0>] (platform_drv_probe+0x18/0x1c) [<c01606d0>] (platform_drv_probe+0x18/0x1c) from [<c015f4a0>] (driver_probe_device+0xf8/0x184) [<c015f4a0>] (driver_probe_device+0xf8/0x184) from [<c015f5b8>] (__driver_attach+0x8c/0x90) [<c015f5b8>] (__driver_attach+0x8c/0x90) from [<c015ec7c>] (bus_for_each_dev+0x60/0x8c) [<c015ec7c>] (bus_for_each_dev+0x60/0x8c) from [<c015e558>] (bus_add_driver+0xa0/0x228) [<c015e558>] (bus_add_driver+0xa0/0x228) from [<c015fb8c>] (driver_register+0x78/0x144) [<c015fb8c>] (driver_register+0x78/0x144) from [<c00083d0>] (do_one_initcall+0x94/0x164) [<c00083d0>] (do_one_initcall+0x94/0x164) from [<c0008560>] (kernel_init+0x8c/0x278) [<c0008560>] (kernel_init+0x8c/0x278) from [<c002467c>] (kernel_thread_exit+0x0/0x8) I had verified the following things from my side, --> There is enough virtual addresses available for dma_alloc() (14MB given) Is there any other possibility why dma_alloc() function fails??? With regards, Sandeep Kumar Anantapalli,

2 1

read_proc_t and file_operations in procfs
by Hua Siyuan July 21, 2013

July 21, 2013

Hi All, I'm now studying proc file system. I now know that read_proc_t function is called when the proofs file is read, and so for write_proc_t function when the proofs file is written to. But I also find the file_operaitons* field in the definition of proc_dir_entry, and this example code (http://linux.die.net/lkmpg/x810.html) So I'm confused that what happened if I provide both read_proc_t function, and the read function in file_operations structure? Thanks very much. Best, Siyuan

2 1

sys_read buffer too large
by Don Raikes July 21, 2013

July 21, 2013

Hello all, I am very new to kernel programming. In fact, I have been working on it for a week now. I am taking a computer science class, and one of our assignments is to hook some of the system calls in a 2.6.28 kernel. I have the basic module created, and I am hooking into the sys_read function. The assignment is to print to the log what is being read using sys_read. I have some checks in my function to limit my printing to only a certain file, so I don't get the contents of every file being read on the system. OnceI know I have the right file, I decided to print the size of the buffer. The function signature is: Asmlinkage long sys_read(unsigned int fd, char __user *buf, size_t count); So I do: Printk(KERN_INFO "Read buffer is [%d] bytes.\n",count); The particular file I am reading has 2 lines the first is 14 bytes long and the second is 30 bytes long. The output from my printk statements are 8192 and 4096 respectively. Why is the count so large? Also I do a strlen(buf) and print the value as well, and it is 29 and 16 respectively. Next I did copy_from_user(tbuf,buf,strlen(buf)) And tried to print the contents of the buffer but it came out garbage. Does anyone have any ideas how I can get this to work? BTW: the assignment is due on Sunday so any help would be aappreciated. -- Best Regards, Donald HYPERLINK "http://www.oracle.com/" \nOracle Donald raikes | Accessibility Specialist/ QA Engineer Phone: HYPERLINK "tel:+15202717608"+15202717608 | Mobile: HYPERLINK "tel:+15202717608"+15202717608 Oracle Quality Assurance | Tucson, Arizona HYPERLINK "http://www.oracle.com/commitment" \nGreen Oracle Oracle is committed to developing practices and products that help protect the environment

3 2

Re:Mag Gam
by Mag Gam July 21, 2013

July 21, 2013

http://dma.ma/wzbuk/yfxsnzlta.byqjungzaceqemttkjn Mag Gam 7/21/2013 7:24:36 AM

1 0