March 2015 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

(no subject)
by Woody Wu March 30, 2015

March 30, 2015

Hi, I have a kernel already run on production, but I then realized that I need to add one or two driver to it. But I hope I can avoid to upgrade the kernel image for those already running products, I hope I can only extend the kernel by add the driver modules to the root file system. Is that possible? The current kernel has already compiled with the loadable modules options, but for the drivers that I want now the old config is 'no'. Thanks in advance. -woody -- Sent from Gmail Mobile

3 2

Sysfs requirement for running generic_buffer.c application - IIO sensors
by s.rawat March 30, 2015

March 30, 2015

I have two system having different kernel version and drivers(IIO) on which i can run the generic_buffer.c application .On one system it works fine but on the other it does not give the input reports data. I have checked that the system on which it is working has the following sysfs under /sys/bus/iio/devices/iio:deviceX (where X is 1,2,3...) buffer dev in_accel_hysteresis in_accel_offset in_accel_sampling_frequency in_accel_scale *in_accel_x_raw* *in_accel_y_raw* *in_accel_z_raw* name power scan_elements subsystem trigger uevent This is for accelerometer similiarly for magnetometer (in_magn_x_raw,_y_raw,z_raw etc) and for Gyrometer ( in_anglvel_x_raw,in_anglvel_y_raw,in_anglvel_z_raw ) and the one on which it is not working has the below sysfs under /sys/bus/iio/devices/iio:deviceX (where X is 1,2,3...) : buffer dev in_accel_hysteresis in_accel_offset in_accel_sampling_frequency in_accel_scale name power scan_elements subsystem trigger uevent Is the missing _x_raw,_y_raw and _z_raw component main reason for not working of the app,.Where I can enable them in the driver? Thanks and Rgds, Saurabh

2 3

Time stamp of Kernel log prints is not including the suspend time
by manty kuma March 30, 2015

March 30, 2015

When CPU entered into power collapse, the time in the kernel log prints is not counting further. I mean to say that the time stamps of all the prints(end stage prints) during the power collapse are same and after resuming from power collapse, the time is continuing from here. The duration of suspend is not added to the kernel log's time stamps. I would like to know 1. So basically can I consider the wakeup time of the system until a particular print in kernel log is equal to its time stamp? 2. I would like to see in code how these time stamps are generated? Like, before going into power collapse, does it pause a timer or something? Best Regards, Manty

1 0

usage of dma_common_mmap
by Malte Vesper March 29, 2015

March 29, 2015

Hello, I am trying to use dma_mmap_attrs . However I keep getting the following error: mmap:6420 map pfn RAM range req uncached-minus for [mem 0xfd0000000-0xfd0000fff], got write-back Assuming that dma_common_map is used (http://lxr.free-electrons.com/source/drivers/base/dma-mapping.c#L246) I tracked the error source to be comming from: remap_pfn_range After struding the codeflow I think that this macro is the culprit: 12 #define pgprot_noncached(prot) \ 13 ((boot_cpu_data.x86 > 3) \ 14 ? (__pgprot(pgprot_val(prot) | \ 15 cachemode2protval(_PAGE_CACHE_MODE_UC_MINUS))) \ 16 : (prot)) since the "got write-back" is calculated from the memory address I assume I can not change that bit. Could anyone explain to me, why the page protection depends on boot_cpu_data? Googling turned up further hints http://comments.gmane.org/gmane.linux.kernel.mm/125059. Is this a Bug or am I missing something? How do I correctly use dma_mmap_attrs? Currently I try: mmap function: return dma_mmap_attrs(&pcidev->dev, vma, buffer->memoryAddress, buffer->dmaAddress, BUF_SIZE, 0); testfile: char* fpath = "/proc/myDevice"; uint64_t* buffer; int filepointer = open(fpath, O_RDWR); if(filepointer < 0) { printf("Could not open %s.\n", fpath); return -1; } buffer = (uint64_t*) mmap(0, 4096, PROT_READ | PROT_WRITE, MAP_PRIVATE, filepointer, 0); Thanks for any help, Malte

1 0

Atomics and memory barriers
by Malte Vesper March 26, 2015

March 26, 2015

Hello, I have been reading up on atomics and struggle to grasp when exactly I need explicit memory barriers. While the documentation (Documentat/atomic_ops.txt), talks about operations needing explicit barriers both sides, I assume this is only half true. If I for instance only want to use an atomic as a flag to show that some operation is complete I reckon that a barrier before is sufficient. Furthermore I found the following lines: 288 <http://users.sosdg.org/%7Eqiyong/lxr/source/Documentation/atomic_ops.txt#L2…> If a caller requires memory barrier semantics around an atomic_t 289 <http://users.sosdg.org/%7Eqiyong/lxr/source/Documentation/atomic_ops.txt#L2…> operation which does not return a value, a set of interfaces are 290 <http://users.sosdg.org/%7Eqiyong/lxr/source/Documentation/atomic_ops.txt#L2…> defined which accomplish this: 291 <http://users.sosdg.org/%7Eqiyong/lxr/source/Documentation/atomic_ops.txt#L2…> 292 <http://users.sosdg.org/%7Eqiyong/lxr/source/Documentation/atomic_ops.txt#L2…> void smp_mb__before_atomic(void); 293 <http://users.sosdg.org/%7Eqiyong/lxr/source/Documentation/atomic_ops.txt#L2…> void smp_mb__after_atomic(void); note how it refers to "operation which does not return a value", why can't I use these for atomic operations that do return a value? What should I use instead, normal barriers like mb? Please enlighten me Malte

2 2

scatterlist copy to buffer and copy back
by leo kirotawa March 26, 2015

March 26, 2015

Hi there, I'm facing with an issue that from a module the user sends a scatterlist which has sizes not power of two and because that I can not handle it properly in my module in other side. An example is: - scatterlist entries size 3 and data size 4096. - in other side module needs to build blocks in power of two sizes, but since scatterlist are size 3 or any size not power of two it's not possible. >From my understanding, entries gives a segment's size of the buffer passed. And each entries has an address to this piece. So if buffer is something like this: | | <- buffer scatter list will be spread in memory like this: | entry | entry | entry | ..., of course spread in memory. As a solution I'm thinking in copy those scatterlist to a buffer and process this buffer as a power of two sizes pieces, and so put the result in the output buffer and into the output scatterlist size not power of two. I know it seems a stupid solution, since scatterlist were done to help spread data from a buffer in not continual space and improve performance. So... *Question is*, is there any way I can copy a scatterlist size 3 to another scatterlist size power of two or any size? I didn't see any function in scatterlist.c api for this, just those solutions to copy from to buffer or buffer to scatterlist. Any suggestions or tips are very welcome :) Thanks in advance, -- ---------------------------------------------- Leônidas S. Barbosa (Kirotawa) blog: corecode.wordpress.com

1 0

mapping driver allocated memory to process
by Malte Vesper March 26, 2015

March 26, 2015

Hello, I am working on a DMA driver and noticed that consistent dma performs rapidly better than streaming dma. Since I want to employ zero copy dma I want to map the memory the driver allocates with dmam_alloc_coherent into the user process. I would like to implement this using IOCTL rather than mmap. I had a look at dma_mmap_attrs, which expects a vm_area struct. Is it ok to simply allocate one and fill it with apropriate data like in mmap_region(lxr.free-electrons.com/source/mm/mmap.c#L1594), if so I noticed that the kernel uses a memory pool (vm_area_cachep), which I apparently can't access (dmesg: Unknown symbol vm_area_cachep (err 0)), can I just wildly allocate or is there some black magic connected that relies on vma structs beeing in the area of the pool? I also looked at install_special_mapping which seems even closer to what I want to do (I do not need to provide a vma), however it is not exported as well. Is there a way to use it in a dynamic module anyway? I saw it used mainly in arch specific code. do_mmap, seems to have disappeared (at least it is not in any sane place in the 3.19 lxr), do_mmap_pgoff seems to be wrong for the job since I do not see how to tell that function that I have struct** page, rather than a file to back the mapping. I would greatly appreciate pointers on up to date summaries on the memory system, http://linux-mm.org/ seems slightly stale, https://www.kernel.org/doc/gorman/pdf/understand.pdf was helpful, although I could not entirly read it (yet). Any pointers to helpful functions, or gotchas are much appreciated. (I hope I managed to make it clear, that I tried to do my research + trial and lots of error...). Thanks, Malte

1 0

Kernel Linked Lists (list_splice)
by Leandro M Barbosa March 25, 2015

March 25, 2015

Hello all! I was reading the __list_splice () function code and I had a doubt. The code is: 274 static inline void __list_splice(const struct list_head *list, 275 struct list_head *prev, 276 struct list_head *next) 277 { 278 struct list_head *first = list->next; 279 struct list_head *last = list->prev; 280 281 first->prev = prev; 282 prev->next = first; 283 284 last->next = next; 285 next->prev = last; 286 } What happens with the *list head? As I understood, when you call list_splice (list_a, list_b, list_b->next), the code joins the two lists together such that the list_a is put before list_b. The code grabs list->next and list->prev but what about *list itself? -- Leandro Moreira Barbosa

2 2

Re: Question about switch_mm function
by Rajat Sharma March 25, 2015

March 25, 2015

On Mar 25, 2015 10:31 AM, "Sreejith M M" <sreejith.mm(a)gmail.com> wrote: > > On Wed, Mar 25, 2015 at 10:55 PM, <Valdis.Kletnieks(a)vt.edu> wrote: > > On Wed, 25 Mar 2015 21:35:22 +0530, Sreejith M M said: > > > >> > This code is handling context switch from a kernel thread back to user mode > >> > thread so TLB entries are invalid translation for user mode thread and do > >> > not correspond to user process pgd. It is Master kernel page table > >> > translation as a result of kernel thread execution. > >> > > >> > -Rajat > >> Hi Rajat, > >> > >> If that is the case, why this code is put under CONFIG_SMP switch? > > > > Vastly simplified because I'm lazy :) > > > > If you look at the code, it's poking the status on *other* CPUs. That's why > > the cpumask() stuff. > > > > If you're on a single execution unit, you don't have to tell the other > > CPU about the change in state, because there isn't an other CPU. > > can you come out of this lazy mode explain this a bit more because I > am a newbie ?or tell me what else I should know before I have to > understand this code > > -- > Regards, > Sreejith Valdis is talking about lazy tlb flush, not him being lazy. Otherwise he wouldn't have replied at all :)

2 3

Question about switch_mm function
by Sreejith M M March 25, 2015

March 25, 2015

Hi, I was trying to understand the difference in scheduling between processes and threads(belong to same process). I was thinking that, when kernel has to switch to a task which belong to the same process, it does not have to clear / replace page global directories and other memory related information. But in switch_mm function some code is put under CONFIG_SMP function. What is its signigicance? Code is below(http://lxr.free-electrons.com/source/arch/x86/include/asm/mmu_context… . What I infer is that the code is doing flush tlb, reload page table directories etc in multiprocessor mode(obviously) but I believe this code may never be executed . Can anyone help to understand what this part of the function supposed to do? 60 #ifdef CONFIG_SMP 61 else { 62 this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK); 63 BUG_ON(this_cpu_read(cpu_tlbstate.active_mm) != next); 64 65 if (!cpumask_test_cpu(cpu, mm_cpumask(next))) { 66 /* 67 * On established mms, the mm_cpumask is only changed 68 * from irq context, from ptep_clear_flush() while in 69 * lazy tlb mode, and here. Irqs are blocked during 70 * schedule, protecting us from simultaneous changes. 71 */ 72 cpumask_set_cpu(cpu, mm_cpumask(next)); 73 /* 74 * We were in lazy tlb mode and leave_mm disabled 75 * tlb flush IPI delivery. We must reload CR3 76 * to make sure to use no freed page tables. 77 */ 78 load_cr3(next->pgd); 79 trace_tlb_flush(TLB_FLUSH_ON_TASK_SWITCH, TLB_FLUSH_ALL); 80 load_LDT_nolock(&next->context); 81 } 82 } 83 #endif -- Regards, Sreejith

3 5