July 2014 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

More than 9 days on task 06 Eudyptula Challenge
by Lucas Tanure July 11, 2014

July 11, 2014

Sorry to bug this again. I'm waiting for task 06 more than 9 days. There are more guys with the same delay time in this task ? Thanks -- Lucas Tanure +55 (19) 988176559

6 6

memory dump problem
by Lucas Tanure July 11, 2014

July 11, 2014

Hi, I have a linear raw file as a dump of memory ram. There is a way to convert that to elf core dump, so gdb can read ? Thanks -- Lucas Tanure +55 (19) 988176559

2 3

Re: Do per cpu variables need to be atomic?
by Pranith Kumar July 10, 2014

July 10, 2014

Hi Christoph, Thank you for answering. On Thu, Jul 10, 2014 at 4:25 PM, Christoph Lameter <cl(a)gentwo.org> wrote: > On Thu, 10 Jul 2014, Pranith Kumar wrote: > >> Hello, >> >> Let us say we have a per-cpu data structure as follows: >> >> struct data { >> atomic_t val; >> }; >> >> Since this is a per-cpu data structure, do I need to have 'val' as >> atomic? Or can I just use a normal 'int' for val? > > Per cpu data structures are for the use of one cpu exclusively. If you use > the variable as intended then there is no atomic_t required. If you want > to update the percpu variable from other processors then atomic_t is > required. But then this is not a proper percpu variable anymroe. > That also makes using barriers while updating such variables meaningless I guess? like smp_mb(); val++; smp_mb(); since val here is supposed to be per-cpu. > >> > Also are the following the same? If yes, which is preferable? > >> DEFINE_PER_CPU(struct data, datap); > > datap is now a fixed offset into the percpu area. The processor can encode > a load using a segment prefix and the fixed offset if the this_cpu > operations are being used for accesses. > > >> struct data __percpu *p = &datap; > > This wont work without the definition above and will assign the offset of > datap in the percpu area to p. p can be used with this_cpu operations but > is not a generally usable pointer. Its an offset into the per cpu area. > >> struct data *p = this_cpu_ptr(&datap); > > This converts the offset into an address that can be generally used as > a pointer (but no longer with this_cpu operations). > So we can pass around this pointer to other cpus but not the previous __percpu pointer? I think adding this information to the documentation would be a great idea. Especially the "__percpu" annotation. Is it ok if I try? -- Pranith

2 1

Do per cpu variables need to be atomic?
by Pranith Kumar July 10, 2014

July 10, 2014

Hello, Let us say we have a per-cpu data structure as follows: struct data { atomic_t val; }; Since this is a per-cpu data structure, do I need to have 'val' as atomic? Or can I just use a normal 'int' for val? Also are the following the same? If yes, which is preferable? DEFINE_PER_CPU(struct data, datap); struct data __percpu *p = &datap; struct data *p = this_cpu_ptr(&datap); Thanks, -- Pranith

2 1

Working on NoMoreModuleVmalloc Project
by shashank karkare July 10, 2014

July 10, 2014

Hi, I am currently working on the NoMoreModuleVmalloc project under the guidance of Dave Hansen. Can you add me as the developer to the project? Name :Shashank Karkare mail id: karkareshashank(a)gmail.com Date Taken: June 5th 2014 -- With regards, Shashank Karkare Computer Science Graduate Student Ira A Fulton School Of Engineering Arizona State University

1 0

Make queues in kernel space
by Robert Clove July 9, 2014

July 9, 2014

Hi All, How we create queues in kernel space and what are the size limit ? Regards

3 4

USB forceful removal
by AYAN KUMAR HALDER July 9, 2014

July 9, 2014

Hi All, When a usb mass-storage is attached for the first time, it gets a device (/dev/sda). When it is mounted and the device is manually removed and then when it is attached again, it gets a different name ie /dev/sdb. This is because earlier the mass-storage device was removed (manually) when it was mounted, so scsi_disk_release() did not get called as the device (sda) had a reference count > 1. So, if my understanding is correct, I would like to know if the above mentioned behaviour is a bug or a limitation of the kernel? Regards, Ayan Kumar Halder

3 4

View device <=> IOMMU mapping table
by Ahmed A July 9, 2014

July 9, 2014

Hello, I have a server, and on the motherboard there are two CPU sockets, with a CPU in each. There is a x16 PCIe slot connected to each CPU socket. There are two different cards in each PCIe slot. Is there a utility or some kernel data structure I can look at to find the device to IOMMU (socket 0 or socket 1) mapping? Thank you, Ahmed.

2 3

Identifying neighbors within the LAN
by Hettiarachchige Hasitha Shan July 8, 2014

July 8, 2014

Dear Sirs/ Ma'ams, I am an undergraduate student following the B Eng (Hons) Software Engineering degree program at University of Westminster, UK. As for my final year research project I am implementing a packet security enhancement module. In the concept, it is required to obtain the list of neighbors within the LAN. I did some research online and I found out that I can obtain IPv4/ IPv6 neighbors by typing in "ip -4 neigh show" or "ip -6 neigh show" in the terminal where this feature is implemented by the module neighbour.c. My query is , If I am to obtain these information programatically through my kernel module, which methods should I call. As in, i tried doing a #include <linux/neighbour.h> and tried to call `neightbl_dump_info' but that did not work. I would be really grateful if you experts could advice me on how to implement the function to obtain the list of IPv4/ IPv6 neighbours programatically? Thank you, Best Regards, H. Hasitha Shan P.S. I apologize in advance if this question has an obvious answer, but as I am a novice in this area, my research lead me to dead ends for weeks and your help will be very much appreciated.

6 11

Re: ptrace and signal handlers.
by manty kuma July 8, 2014

July 8, 2014

Changed the subject to a better relevant name. Sorry for spamming. Looking forward for response. On Tue, Jul 8, 2014 at 7:53 PM, manty kuma <mantykuma(a)gmail.com> wrote: > Hi, > Consider a scenario that a signal handler is getting executed and also > ptrace is attached to the process in parallel. > I have following queries : > 1. If a signal is sent to the process from signal handler, Will the signal > be delivered and handled(With ptrace already attached)? > > 2. what is current->exit_code here > <https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/kernel…> > ? > If this is the exit code of the application, is it properly initialized at > this point in time(while the signal is being delivered)? If yes, could > someone help me spotting it in source? > > Looking forward for responses. Thank you > > Best Regards, > Manty >

1 0