April 2011 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

Re: stimulate usb device
by Prashant Shah April 22, 2011

April 22, 2011

> Yes you just call this functions from another module. Just make sure you > exported the functions. > > greetings mosta Sounds good. Thanks a lot :)

1 0

stimulate usb device
by Prashant Shah April 22, 2011

April 22, 2011

Hi, I am trying to write a driver for a USB based Data Acquisition Device. I have limited access to the hardware device and I want to test the attach and probe calls. Is there anyway to stimulate or create a dummy USB device so that the kernel can make those calls. Any information will be helpful. Regards.

2 1

perhaps confused about operation of alloc_chrdev_region()
by Robert P. J. Day April 21, 2011

April 21, 2011

i'm in the midst of updating some basic kernel documentation and i just ran across something that makes me question my understanding of the alloc_chrdev_region() kernel routine. as i've understood it all this time, that routine is the preferred routine if you want to allocate some character devices *dynamically* -- that is, to accept whatever major number(s) the kernel is willing to hand back. with that routine, you're not supposed to *specify* a major number that you want, you just accept what you're given. at least that's what i always thought. i was perusing the kernel source looking for a simple driver example that used that routine, and found drivers/char/bsr.c, but the init routine looks weird, as in: ===== static int __init bsr_init(void) { struct device_node *np; dev_t bsr_dev = MKDEV(bsr_major, 0); <--- ???????? int ret = -ENODEV; int result; np = of_find_compatible_node(NULL, NULL, "ibm,bsr"); if (!np) goto out_err; bsr_class = class_create(THIS_MODULE, "bsr"); if (IS_ERR(bsr_class)) { printk(KERN_ERR "class_create() failed for bsr_class\n"); goto out_err_1; } bsr_class->dev_attrs = bsr_dev_attrs; result = alloc_chrdev_region(&bsr_dev, 0, BSR_MAX_DEVS, "bsr"); bsr_major = MAJOR(bsr_dev); ===== note that early declaration of "dev_t bsr_dev". why is it being initialized to what is an uninitialized static variable much earlier in that source file? the value of "bsr_major" certainly isn't used until it's *assigned* in the call: result = alloc_chrdev_region(&bsr_dev, 0, BSR_MAX_DEVS, "bsr"); bsr_major = MAJOR(bsr_dev); as best as i can tell, that initial declaration can simply be: dev_t bsr_dev; i see no reason to initialize it to anything since alloc_chrdev_region() is supposed to treat the dev_t parameter as output only, is it not? am i missing something here? rday -- ======================================================================== Robert P. J. Day Ottawa, Ontario, CANADA http://crashcourse.ca Tw itter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================

2 2

Now getting settled in Ottawa, in case you cared :-)
by Robert P. J. Day April 21, 2011

April 21, 2011

Just in case you were interested, I am no longer based in Kitchener-Waterloo. I just made the move to Ottawa and am still unpacking. I'll be a regular at Ottawa Linux User Group meetings, and if you want to chat offline about kernel or embedded stuff or Linux training in general in the Ottawa area, drop me a note. See you at the next OCLUG meeting: http://oclug.on.ca/meeting/74/. rday -- ======================================================================== Robert P. J. Day Ottawa, Ontario, CANADA http://crashcourse.ca Twitter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================

3 3

How to set work_struct data
by Pankaj B April 21, 2011

April 21, 2011

Hi, Current INIT_WORK() macro takes just work_struct pointer and function pointer. In previous kernels there used to be a third parameter for data. Is there a proper way to set data in work_struct? Thanks

2 2

Contents of CR3 register when a kernel thread is executed
by limp April 21, 2011

April 21, 2011

Hi all, I am trying to understand in what extend the PGD (page global directory) value stored in the CR3 register indicate the running process/thread by the Linux scheduler. I know that each process has its own PGD value but what I am confused about is the value of CR3 register when kernel threads are scheduled. AFAIK, a "kernel thread" can 'borrow' the page-directory that belongs to another process (i.e. its address space). Does this apply *only* to threads of the same process (i.e. all the threads of process A are using the PGD entry of process A) or could also apply to threads not relevant to process A? That is, if a thread is executed and the value of CR3 at the time is 0x1E107000, does this *necessary* mean that a thread of the process with PGD 0xDE107000 is executed *or* a thread not relevant with that process could have 'borrowed' its PGD entry? Any help will be greatly appreciated. John K.

4 5

Re: Architecture specific implementations for tickless kernel
by Vikram Narayanan April 19, 2011

April 19, 2011

Yes. My question is pointed towards uClinux. I am sorry. I forgot to mention that in the first post /neo On Tue, Apr 19, 2011 at 11:47 PM, Chaitannya Mahatme <chaitannya(a)gmail.com>wrote: > On Tuesday 19 April 2011 10:53 PM, Vikram Narayanan wrote: > > Hi, > I am developing a BSP for a ARM Cortex-M3 (no MMU) based board. In order to > have tickless kernel and deferrable timers, what are the architecture > specific implementations. Is the clocksource and clockevent data structures > enough? How to hook those timers to the tickless implementation of the > kernel/timer.c. Can some one point out any examples that already have this > feature enabled and also in the kernel tree. > > And also, are clocksoure and clockevents dependent on each other? I see in > some platforms, that they have used 2 different timers for the above. If it > can be used so, which ties up the clocksource and clockevent? > > Thanks in advance > neo > > > _______________________________________________ > Kernelnewbies mailing listKernelnewbies@kernelnewbies.orghttp://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > > Since you have mentioned that there is no MMU, you cannot port Linux > kernel onto it. You would be mostly be looking forward to uclinux or bare > metal programming. > Is your question specific to uclinux ? > > -- > Cheers !!! > Chaitannya > > follow me on twitter http://twitter.com/chaitannya > > I believe there's a hero in all of us > who keeps us honest, gives us strength, makes us noble, and finally allows > us to die with pride, > even though sometimes we have to be steady, and give up the thing we want > the most... even our dreams. > - Spider Man 2 >

1 0

Architecture specific implementations for tickless kernel
by Vikram Narayanan April 19, 2011

April 19, 2011

Hi, I am developing a BSP for a ARM Cortex-M3 (no MMU) based board. In order to have tickless kernel and deferrable timers, what are the architecture specific implementations. Is the clocksource and clockevent data structures enough? How to hook those timers to the tickless implementation of the kernel/timer.c. Can some one point out any examples that already have this feature enabled and also in the kernel tree. And also, are clocksoure and clockevents dependent on each other? I see in some platforms, that they have used 2 different timers for the above. If it can be used so, which ties up the clocksource and clockevent? Thanks in advance neo

1 0

Re: Distributed Computing
by Rigved Rakshit April 18, 2011

April 18, 2011

> > > 1. Process A has just been scheduled to execute for the first time on > > machine A. > > 2. I send a suspend signal to Process A, mostly a SIGSTOP. > > 3. Then I copy Process A's PCB to machine B. > > 4. On machine B, I send the continue signal to Process A, mostly a > SIGCONT. > > 5. Whenever Process A needs user input or needs to output something, I > cause > > a trap, mostly with a SIGTRAP (or I could use SIGSTOP here again). I > > transfer this I/O request to machine A and let the user interact. I send > the > > interaction back to Process A on machine B. I resume Process A on machine > B. > > 6. I continue these steps till Process A has completed execution. > > Sounds like just what MOSIX, OpenSSI or Kerrighed already do.... > Thanks for the information. It was really helpful.

1 0

Attr Groups in sysfs.
by Gasparini Andrea April 18, 2011

April 18, 2011

Hi, I've a pattern similar to this in sysfs: /sys/.../mykobj/myattr_group/attr. And I've some different attr groups, but I'm willing to group also the get/set functions and recognize at runtime which kobj,group,and attribute is called inside the functions. But, if I'm not wrong the attribute functions are called: static ssize_t some_attribute_set(struct kobject *kobj, struct container_attribute *attr, const char *buf, size_t count) And I can get the "mykobj" name with kobj->name, and attribute names by attr->attr.name. I'm missing only the myattr_group name... anyone has a clue of how can I access this? Just for completeness, I'm working on a 2.6.38 kernel, and I'm declaring the groups this way: struct container_attribute { struct attribute attr; ssize_t (*show)(struct kobject *, struct container_attribute *, char *); ssize_t (*store)(struct kobject *, struct container_attribute *, const char *, size_t); }; static struct attribute *lot_of_attrs[] = { first_attribute.attr, and_so_on.attr, NULL, }; static struct attribute_group my_attr_group_0 = { .name = "myattr_group_0", .attrs = lots_of_attrs, }; And adding to my kobject this way: retval = sysfs_create_group(mykobj, &my_attr_group_0); Thanks, bye! -- -gaspa-

2 3