- kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

net_device: limit rate ot tx packets
by christian+kn＠wwad.de April 17, 2013

April 17, 2013

Hi All, can someone please explain me, how the kernel handles different transfer rates of different net_devices? Or in other words: How does the systemcall send() know, when to block? An example: cat /dev/zero | pv | nc -u <someip> will show different throughput speeds depending on the network device, the packets are sent over (wlan0 will be slower than eth0). - Can someone point out the location in the linux kernel source, where this is handled? - If I register a net_device. How do I signal to the upper network layers that my driver can only accept packets at a certain rate? I tried stopping the egress queue by calling netif_stop_queue(), but this only has the effect that the queue overruns. I have the feeling that I'm missing out a very vital point on how the kernel's networking subsystem works. Unfortunately, Understanding Linux Network Internals couldn't help me out here. Thanks, Christian

3 5

Build error in linux-3.8.7
by Giridhara RP (grp) April 17, 2013

April 17, 2013

Hi, I am new to linux kernel buidling. I have download 3.8.7 latest (stable) kernel from https://www.kernel.org/. I want build kernel for X86_64. 1. make x86_64_defconfig (this created .config file) 2. make Make failed with below error, please help me in resolving this. CC kernel/power/user.o CC kernel/power/block_io.o CC kernel/power/poweroff.o LD kernel/power/built-in.o CC kernel/sched/core.o CC kernel/sched/clock.o CC kernel/sched/cputime.o CC kernel/sched/idle_task.o CC kernel/sched/fair.o kernel/sched/fair.c: In function `update_curr': kernel/sched/fair.c:428: sorry, unimplemented: inlining failed in call to 'account_cfs_rq_runtime': function body not available kernel/sched/fair.c:714: sorry, unimplemented: called from here make-3.79.1-p7[2]: *** [kernel/sched/fair.o] Error 1 make-3.79.1-p7[1]: *** [kernel/sched] Error 2 make-3.79.1-p7: *** [kernel] Error 2 bash-3.00$ pwd thanks Giri

2 7

Handling MSI interrupts
by Matti Vaittinen April 17, 2013

April 17, 2013

Hello! I am working with an asic that is controlled by CPU. With our current board this asic sends interrupts to processor via single interrupt line. CPU sees level active interrupt, and reads the reason(s) from asic's register. When handler(s) have executed, the read interrupts are acked to asic, and if no more interrupt(s) have ocurred on asic, this interrupt line is brought in inactive state. If new (different type of) interrupts have occurred on asic, the acking wont clean all interrupts from asic's register and asic keeps the interrupt line activated. Thus CPU is interrupted again && this process is done again && all interrupts are handled. Now there is plans to make this asic to use PCIe, and MSI interrupts. Also we hope to utilize the same driver(s) for this asic. So first idea would be to implement this as: 1. asic generates irq 2. MSI interrupt is sent (to PCIe controller) and CPU is interrupted 3. CPU acks interrupt to PCIe controller. 4. isr reads the interrupt reason(s) from asic and handles interrupt(s) 5. isr acks the interrupts. So is there a problem if new interrupt is generated between steps 4 and 5. Do we lose this new interrupt because the isr is still running? So when the isr acks interrupts to asic, it does not ack this new one because it had not happened when the reasons were read at step 4? I have read the MSI-HOWTO.txt and genirq guides, and even tried to read some sources. However I have not found out solid answer. Is there some de-facto method to do this so it works reliably? Following section may contain errors, it is what I have guessed by reading the sources - please kick me if it sounds stupid. By reading the sources I have figured that the "interrupt-controller" for MSI interrupts on processor does used handle_simple_irq as flow handler. If I am correct, this does not "queue" interrupts in same fashion the handle_edge_irq would do. Does this mean that my fear is correct? Br Matti Vaittinen

1 0

How to debug kernel on ARM with gdb
by Woody Wu April 17, 2013

April 17, 2013

Hi, I am working on an embedded ARM board (s3c2416) and I can already debug its boot loader (u-boot) with arm-linux-gdb, which talks to a GDB server provided from J-Link. Now I want to debug the kernel. The problem is that how can I get the correct symbol table for listing source lines in gdb. I checked the symbols in my compiled vmlinux, it starts from 0xC0008000. I think this is the final runtime address and should be a virtual address after MMU enabled. When kernel getting loaded into RAM, the virtual address probably not exists. On the other hand, the zImage loaded by boot loader is not vmlinux actually, it should be head.S which then carries vmlinux as piggy. So, with these complexities, how can I correctly debug Linux kernel starting from it's first instruction in head.S? Who have experience on this? Thanks in advance. -- woody I can't go back to yesterday - because I was a different person then.

1 0

Invitation to connect on LinkedIn
by Kumar Sukhani via LinkedIn April 17, 2013

April 17, 2013

LinkedIn ------------ Kumar Sukhani requested to add you as a connection on LinkedIn: ------------------------------------------ Deepak, I'd like to add you to my professional network on LinkedIn. - Kumar Accept invitation from Kumar Sukhani http://www.linkedin.com/e/8z2206-hflhs696-5i/L90NvsYL3GREvJdnM0f-HnfkWNbSD_… View profile of Kumar Sukhani http://www.linkedin.com/e/8z2206-hflhs696-5i/rso/123751041/F9xI/name/494894… ------------------------------------------ You are receiving Invitation emails. This email was intended for Deepak Shah. Learn why this is included: http://www.linkedin.com/e/8z2206-hflhs696-5i/plh/http%3A%2F%2Fhelp%2Elinked… (c) 2012, LinkedIn Corporation. 2029 Stierlin Ct, Mountain View, CA 94043, USA.

3 2

So I want to get some kernel routine called....
by Arlie Stephens April 16, 2013

April 16, 2013

Hi Folks, It's that BSD engineer again, trying to do something in linux and unable to see the _linux_ way to do it ;-) I have some kernel routine I'd like to get called, with the decision to call it made in user space. Obviously I could do this by making it into a full blown system call, but that's pretty much never the right answer, except for a quick hack. In FreeBSD, I'd make it into a sysctl, and have my user space script do something like "sysctl debug.call_my_hack=1" I'd then define call_my_hack to to be the type of sysctl that calls a procedure, using a convenient macro, and I'd be done. In linux I've determined that the /proc/sys interface is the preferred way to access sysctl's, but that even that is possibly deprecated in favour of sysfs. Documentation on implementing sysctl's is conspicuous by its absence, and much of what I've found applies only to the deprecated interface. In particular, the "strategy" routine appears designed for my purpose - but is never called if you come in to the kernel with read/write to /proc/sys/whatever. Probably I can get the effect I want by (ab)using the proc_handler function pointer - but is that what I "should be doing"? As for sysfs, it's tightly tied to the physical device hierarchy. I can't even find how one is supposed to deal with pseudo-devices. There do however seem to be a few interesting things in /sys/kernel which clearly have nothing to do with the device hierarchy. Probably checking their implementation will teach me how to use this ... but not whether it's the "right thing to do". For what it's worth, there is a pseudo-device driver involved, but no convenient /dev interface to issue ioctls on. (Probably calling it a pseudo-device exagerates its abilities.) As always, I'm looking primarily at 2.6.32, but want to write code that's good for the long haul, and potentially upstreamable. So, how would someone who thinks in linux go about doing this? Thanks for any enlightenment. -- Arlie (Arlie Stephens arlie(a)worldash.org)

2 1

Module needs to lookup its own symbols at runtime.
by Neil Baylis April 16, 2013

April 16, 2013

I have a kernel loadable module that needs to be able to lookup its own symbols at runtime. I.e., given a string that represents the name of the symbol, the module needs to find the address. Preferably, this should work for symbols that are not marked 'EXPORT_SYMBOL' as well as those that are. I used to be able to do this by calling parse_symbol() that's part of kdb, but now my module is required to run without depending on kdb. Is there an API that provides this functionality? Otherwise, any hints about how to proceed would be helpful.

2 1

SNORT/SURICATA LEARNing
by Robert Clove April 16, 2013

April 16, 2013

Dear All, I want to learn the snort and suricata whats the best way to learn it. I have read the architecture. Now i want to go into the code Whats the best way plz tell Thanks

3 2

File System Learning Approach
by Pietro Paolini April 15, 2013

April 15, 2013

Hello all, I have to compare different filesystems options for an embedded platform and I'd like understand what's the best way in your opinion, I use to read wikipedia but do you have any suggestion or articles or references that can help me ? I am not really expert about filesystem but I have to start! Thanks, Pietro.

3 4

Where is /dev/mtdX on Linux-3.8.0-rc7+
by bill4carson April 15, 2013

April 15, 2013

Hi, All It seems "MTD partitioning support" CONFIG_MTD_PARTITIONS has been deleted some time ago, at least not exist on Linux-3.8.0-rc7+. And there is no /dev/mtdx neither, so how do I make it happen? thanks in advance! -- 八百里秦川尘土飞扬,三千万老陕齐吼秦腔。 --bill

2 1