Re: Has Anyone used Ftrace tool.
Hi, Please check the licks http://lwn.net/Articles/365835/ and http://lwn.net/Articles/370423/ This will help you in understanding the usage of ftrace. Regards Anjana ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Intel-IOMMU
I am trying to understand the intel-iommu. a) I find mention of three address types. Untranslated Requests, Translation Requests and Translated Requests. Translated Requests mean translation of host virtual to host physical. But I am unclear about the other two requests. One of them should be GVA to HPA. The VT-d spec does not clearly mention the difference. b) Secondly, there are two levels of Address Translations. I am working in a non VM environment but still using IOMMU through VFIO. I was curious to know whether we really need two levels of translations as in my case. Thanks Saptarshi ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
RE: How to diasble interrupt of current core in multicore cpu
-Original Message- From: kernelnewbies-boun...@kernelnewbies.org [mailto:kernelnewbies-boun...@kernelnewbies.org] On Behalf Of Pranay Srivastava Sent: Friday, May 30, 2014 4:26 PM To: J.Hwan Kim Cc: kernelnewbies Subject: Re: How to diasble interrupt of current core in multicore cpu are you looking for local_irq_disable? On Fri, May 30, 2014 at 1:42 PM, J.Hwan Kim wrote: > Hi, everyone > > How to disable interrupt for current core in multicore CPU? > > Thanks in advance > Kim. > local_irq_save/restore would be more suitable. you can use them without considering is interrupt enabled now. your code may looks like this: local_irq_save() /*your code comes here with interrupt disabled*/ local_irq_restore() in this way, if interrupt is in state A (may be enabled or disabled), before you call local_irq_save(), then after you call local_irq_restore, the interrupt will still be in state A. harmless to other code. > ___ > Kernelnewbies mailing list > Kernelnewbies@kernelnewbies.org > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies -- ---P.K.S ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Implementing a NFS proxy
Have a look at this: https://github.com/nfs-ganesha/nfs-ganesha/wiki/PROXY I believe NFS ganesha is quite mature and production ready implementation. On Thu, Jun 5, 2014 at 11:04 AM, Ramana Reddy wrote: > Thanks for your reply. This link does not solve my problem. > I need a proxy which act like a nfs client as well as nfs server. > It takes request from the clients and parse it and sends to the real > server. > In the second case it acts as a client. I am looking for minimal > implementation of nfs daemons. > > Thanks, > Ramana. > > > On Thu, Jun 5, 2014 at 7:18 PM, Peter Senna Tschudin < > peter.se...@gmail.com> wrote: > >> I would start here: >> >> http://serverfault.com/questions/401312/how-to-create-an-nfs-proxy-by-using-kernel-server-client >> >> On Thu, Jun 5, 2014 at 8:53 AM, Ramana Reddy wrote: >> > Hi all, >> > >> > I want to implement a minimal skeleton of NFS proxy. >> > Is there any place where I can look at and find some useful >> > info. What are the steps I should follow to implement this approach. >> > >> > Help in this regards is highly appreciated. >> > >> > Thanks, >> > Ramana. >> > >> > ___ >> > Kernelnewbies mailing list >> > Kernelnewbies@kernelnewbies.org >> > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies >> > >> >> >> >> -- >> Peter >> > > > ___ > Kernelnewbies mailing list > Kernelnewbies@kernelnewbies.org > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > > ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Implementing a NFS proxy
Thanks for your reply. This link does not solve my problem. I need a proxy which act like a nfs client as well as nfs server. It takes request from the clients and parse it and sends to the real server. In the second case it acts as a client. I am looking for minimal implementation of nfs daemons. Thanks, Ramana. On Thu, Jun 5, 2014 at 7:18 PM, Peter Senna Tschudin wrote: > I would start here: > > http://serverfault.com/questions/401312/how-to-create-an-nfs-proxy-by-using-kernel-server-client > > On Thu, Jun 5, 2014 at 8:53 AM, Ramana Reddy wrote: > > Hi all, > > > > I want to implement a minimal skeleton of NFS proxy. > > Is there any place where I can look at and find some useful > > info. What are the steps I should follow to implement this approach. > > > > Help in this regards is highly appreciated. > > > > Thanks, > > Ramana. > > > > ___ > > Kernelnewbies mailing list > > Kernelnewbies@kernelnewbies.org > > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > > > > > > -- > Peter > ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Has Anyone used Ftrace tool.
Hi, you might want to have a look at kernelshark or rt-tests: https://lwn.net/Articles/425583/ http://git.kernel.org/cgit/linux/kernel/git/clrkwllms/rt-tests.git they both interface with ftrace. andi On 06/05/2014 06:39 PM, Robert Clove wrote: > Hi All, > > I have made a simple client server program and want to trace the write > system call using Ftrace. > > Any idea how can we do that. > > Regards > > > > ___ > Kernelnewbies mailing list > Kernelnewbies@kernelnewbies.org > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Has Anyone used Ftrace tool.
Hi All, I have made a simple client server program and want to trace the write system call using Ftrace. Any idea how can we do that. Regards ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Eudyptula Challenge Submission and no response?
On Thu, Jun 5, 2014 at 9:58 AM, George E. Moore wrote: > Is there a POC (email contact for Eudyptula Challenge)? Submitted > response to Eudyptula Challenge, task 01 (05/23/14) and have yet to > receive response. This is due the high number of subscribers (much more than expected) in the challenge. I recommend this text in LWN: "Taking the Eudyptula Challenge" http://lwn.net/Articles/599231/ -- Augusto Mecking Caringi ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Man pages for Kernel API
ok . i figure it out . you should have have have kernel source tree ( which i think most of us had). any way if you dont : $git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git $cd linux $make mandocs #make installmandocs #mandb (and yes you should have xmlto and other package installed ) i hope this will help. best wishes, -navid On Thu, Jun 5, 2014 at 4:44 AM, navid Rahimi wrote: > thank you frank , for make this clear . > so frank in this case can you be more specific , what should i do ? > for example i installed kernel-doc , even downloaded kernel man page > from kernel.org and completed make install routine . but , for example > i don't have any kmalloc man page yet in my fedora . last time i was > in arch linux , i had it . > > best wishes, > -navid > > > On Thu, Jun 5, 2014 at 2:49 AM, Frank Ch. Eigler wrote: >> >> navid Rahimi writes: >> >>> probe::vm.kmalloc (3stap) - Fires when kmalloc is requested >>> probe::vm.kmalloc_node (3stap) - Fires when kmalloc_node is requested >> >> Those are systemtap man pages, not kernel API ones. >> >> - FChE ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Adding a syscall to specific kernel version
On Thu, 05 Jun 2014 18:10:06 +0530, Dipanjan Das said: > Why isn't there a consistent interface for addition of syscall across > kernel version? Please read Documentation/stable_api_nonsense.txt for the explanation. > How will I get to know the method for a specific kernel version officially? I'll be very blunt. Wiring up a new syscall is actually fairly easy and straightforward. If you are having trouble figuring out how to do it when each arch already has 200 to 300 already-done examples and 4 how-to tutorials off the net, you're probably going to have even more trouble writing the actual code of a syscall that does something more useful than doing a printk("Hello world"). Sometimes, you need to admit to yourself that you're never going to be a Formula One driver... pgpNhfnzHgu3y.pgp Description: PGP signature ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Adding a syscall to specific kernel version
On Thu, Jun 05, 2014 at 06:10:06PM +0530, Dipanjan Das wrote: > > Hi, > > I am trying to add a new syscall to kernel 3.4. So far I have tried at least > four different tutorials from Internet. To my utter frustration, whatever > source files they have referred in the tutorials, at least one of those is > missing in all the cases. > > Why isn't there a consistent interface for addition of syscall across kernel > version? Because code changes over time for valid reasons. If it didn't, the project would be dead. > How will I get to know the method for a specific kernel version officially? What do you mean by "officially"? You have the source to the kernel for all versions, that's the only "official" thing to use. Why are you using the 3.4 kernel, that one is obsolete for any new development? thanks, greg k-h ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Implementing a NFS proxy
I would start here: http://serverfault.com/questions/401312/how-to-create-an-nfs-proxy-by-using-kernel-server-client On Thu, Jun 5, 2014 at 8:53 AM, Ramana Reddy wrote: > Hi all, > > I want to implement a minimal skeleton of NFS proxy. > Is there any place where I can look at and find some useful > info. What are the steps I should follow to implement this approach. > > Help in this regards is highly appreciated. > > Thanks, > Ramana. > > ___ > Kernelnewbies mailing list > Kernelnewbies@kernelnewbies.org > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > -- Peter ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Process memory
> I am referring to the committed memory size not RSS. IMU - you could get the committed_as for the entire system easily (using "grep Committed_AS /proc/meminfo"), normally its the kernel responsibility to keep track of the over-committed memory, the process will not get to see an actual increase in RSS unless the process starts using the requested / malloc()-ed memory. It should be possibly to notice an increase in VIRT (from top) or VSZ (valgrind) when the committed_AS for a process goes up. On Thu, Jun 5, 2014 at 3:40 PM, Pratap kommula wrote: > I am referring to the committed memory size not RSS. > > Thanks, > Rammohan > > > On Thu, Jun 5, 2014 at 12:20 PM, Mulyadi Santosa > wrote: >> >> On Thu, Jun 5, 2014 at 1:35 PM, Pratap kommula >> wrote: >> > Hi, >> > >> > How do we know that how much memory allocated for a process and how to >> > set a >> > limit on it? >> >> Are you referring to resident set size or committed memory size? >> >> >> Mulyadi Santosa >> Freelance Linux trainer and consultant >> >> blog: the-hydra.blogspot.com >> training: mulyaditraining.blogspot.com > > > > ___ > Kernelnewbies mailing list > Kernelnewbies@kernelnewbies.org > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Eudyptula Challenge Submission and no response?
Sometimes they take a long time to answer, maybe they are stuck again, just be patient and resend :) On Thu, Jun 5, 2014 at 10:10 AM, Kristofer Hallin wrote: > Guess the scripts are somewhat busy, give them a few more days. > > On Thu, Jun 5, 2014 at 3:05 PM, Fernando Apesteguía > wrote: >> >> El 05/06/2014 14:59, "George E. Moore" escribió: >> >> >>> >>> Is there a POC (email contact for Eudyptula Challenge)? Submitted >>> response to Eudyptula Challenge, task 01 (05/23/14) and have yet to >>> receive response. >> >> Patience. A lot of people are taking the challenge. >> >> Cheers. >> >>> -- >>> >>> >>> >>> >>> ___ >>> Kernelnewbies mailing list >>> Kernelnewbies@kernelnewbies.org >>> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies >> >> >> ___ >> Kernelnewbies mailing list >> Kernelnewbies@kernelnewbies.org >> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies >> > > ___ > Kernelnewbies mailing list > Kernelnewbies@kernelnewbies.org > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies -- -- Leônidas S. Barbosa (Kirotawa) blog: corecode.wordpress.com ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Eudyptula Challenge Submission and no response?
Guess the scripts are somewhat busy, give them a few more days. On Thu, Jun 5, 2014 at 3:05 PM, Fernando Apesteguía wrote: > > El 05/06/2014 14:59, "George E. Moore" escribió: > > >> >> Is there a POC (email contact for Eudyptula Challenge)? Submitted >> response to Eudyptula Challenge, task 01 (05/23/14) and have yet to >> receive response. > > Patience. A lot of people are taking the challenge. > > Cheers. > >> -- >> >> >> >> >> ___ >> Kernelnewbies mailing list >> Kernelnewbies@kernelnewbies.org >> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > > > ___ > Kernelnewbies mailing list > Kernelnewbies@kernelnewbies.org > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Eudyptula Challenge Submission and no response?
El 05/06/2014 14:59, "George E. Moore" escribió: > > Is there a POC (email contact for Eudyptula Challenge)? Submitted > response to Eudyptula Challenge, task 01 (05/23/14) and have yet to > receive response. Patience. A lot of people are taking the challenge. Cheers. > -- > > > > > ___ > Kernelnewbies mailing list > Kernelnewbies@kernelnewbies.org > http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Eudyptula Challenge Submission and no response?
Is there a POC (email contact for Eudyptula Challenge)? Submitted response to Eudyptula Challenge, task 01 (05/23/14) and have yet to receive response. -- ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Adding a syscall to specific kernel version
Hi, I am trying to add a new syscall to kernel 3.4. So far I have tried at least four different tutorials from Internet. To my utter frustration, whatever source files they have referred in the tutorials, at least one of those is missing in all the cases. Why isn't there a consistent interface for addition of syscall across kernel version? How will I get to know the method for a specific kernel version officially? ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Implementing a NFS proxy
Hi all, I want to implement a minimal skeleton of NFS proxy. Is there any place where I can look at and find some useful info. What are the steps I should follow to implement this approach. Help in this regards is highly appreciated. Thanks, Ramana. ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Process memory
I am referring to the committed memory size not RSS. Thanks, Rammohan On Thu, Jun 5, 2014 at 12:20 PM, Mulyadi Santosa wrote: > On Thu, Jun 5, 2014 at 1:35 PM, Pratap kommula > wrote: > > Hi, > > > > How do we know that how much memory allocated for a process and how to > set a > > limit on it? > > Are you referring to resident set size or committed memory size? > > > Mulyadi Santosa > Freelance Linux trainer and consultant > > blog: the-hydra.blogspot.com > training: mulyaditraining.blogspot.com > ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies