- kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

How to invalidate cache in kernel driver?
by m silverstri Feb. 13, 2014

Feb. 13, 2014

I have an output buffer in my kernel driver. And that share with my hardware, which uses it for output. When I got an interrupt, the kernel driver will copy the data from the output buffer to user space buffer. How can I invalidate cache after my kernel driver copy the output buffer?

2 1

Verify checksum of IP header and TCP header
by Guibin(Bill) Tian Feb. 11, 2014

Feb. 11, 2014

Hi, I am looking for the code in linux kernel that verifies the checksum of TCP header and IP header for inbound packet. The packet supposes to be dropped if the checksum doesn't match. I tried tcp_input.c and ip_input.c, also with global search. But I couldn't find any code that does this job. Can anyone tell me where is the code? Thanks.

2 2

How can I get physical address from a struct dma_buf
by m silverstri Feb. 11, 2014

Feb. 11, 2014

Hi, If I have a struct dma_buf dma buffer, how can I the physical address?

2 1

cache coherency guidelines for DMA buffer
by m silverstri Feb. 11, 2014

Feb. 11, 2014

Hi, I create DMA buffer in my kernel driver. I would like to know what is cache coherency guidelines for DMA buffer referred by this post? http://stackoverflow.com/questions/5564326/linux-kernel-device-driver-to-dm… Both my kernel driver and HW writes to the DMA buffer. I wnat to know how to make sure cache coherency. Thank you.

3 2

spinlock lockup suspected reasons
by Ashish Sangwan Feb. 11, 2014

Feb. 11, 2014

What could be reasons for the following message: BUG: spinlock lockup suspected on CPU#0, sh/11786 lock: kmap_lock+0x0/0x40, .magic: dead4ead, .owner: sh/11787, .owner_cpu: 1

2 1

When to submit into drivers/staging
by Johannes Thumshirn Feb. 11, 2014

Feb. 11, 2014

Hi, I have a question regarding driver submission into mainline. I'm not exactly sure if my driver is ready to be submittet into normal drivers or if it would be better to submit it to staging first. Is there any recomendation when to submit to drivers/staging instead of drivers? I've searched the archives but haven't found an answer to it. If you can instead just point me to a document I'm fine. The driver(s) in question is a bus driver for multi function FPGAs made by my company and one first iio/adc driver using this bus. It was sucessfully tested on x86 but is known to crash on powerpc because of my parser code (which will of cause be fixed when submitting, even into staging). Thanks a lot, Johannes

2 2

GPIO Muxing
by Vignesh Radhakrishnan Feb. 9, 2014

Feb. 9, 2014

Hi , I have one driver that calls gpio_request and holds on to it. When another driver requests for the same gpio, it failes giving EBUSY. Anyway i can force the gpio request in 2nd driver? Thanks and regards, Vignesh -- http://vigneshradhakrishnan.blogspot.com/

4 3

How to identify a 'fresh' page from read_cache_page?
by Zameer Manji Feb. 9, 2014

Feb. 9, 2014

Hey, My colleague Will and I are working on improving eCryptfs, an encrypted file system that ships with linux. We are trying to add a new cipher mode and we have run into a problem [1]. When the user calls `ftruncate` on a file and increases the file size, eCryptfs attempts fetch new pages for the file by calling `read_mapping_page` and which calls `read_cache_page`. This calls eCryptfs' `readpage` implementation. We believe `read_cache_page` calls `readpage` with a page that we have not written to before (since the user is increasing the file size via `ftruncate`). What function can we use to identify when we are given a page to our `readpage` implementation that is a page we have never written to before? We need to do this so we know if should check the integrity of the data in the page (if we wrote to it before) or just ignore the contents (because it is a fresh page with garbage data). For more information our ecryptfs_readpage implementation is available on github [2]. Please note that I am not subscribed to kernelnewbies so please include me directly in any replies. [1]: http://marc.info/?l=ecryptfs-users&m=139007357027191&w=2 [2]: https://github.com/zmanji/ecryptfs/blob/next-patch/fs/ecryptfs/mmap.c#L193 -- Zameer Manji

2 1

How to test my patches before submitting them to LKML?
by Matthias Beyer Feb. 7, 2014

Feb. 7, 2014

Hi, I'm currently working on some stylefix-patches for drivers/cdrom/cdrom.c and I'm slowly getting to a point where they seem to be ready. How to test my patches? I know I should compile them at least, to be sure they build. But how can I test if the code (still) works (the way it is expected to)? Can I boot it in a VM to test it? If yes, how to (generally and explicitly for my patches)? (There should be no difference if I try to boot it with VirtualBox vs. KVM vs. Xen, right?) -- Mit freundlichen Grüßen, Kind regards, Matthias Beyer Proudly sent with mutt. Happily signed with gnupg.

4 6

timer_t invalid value?
by John Whitmore Feb. 7, 2014

Feb. 7, 2014

Hi all, I've got some code, a protocol I wrote, from an embedded processor which I'm trying to port to Linux User Space, (to run on a Raspberry Pi). Going to get my embedded device talking to the RPi. Fortunately I wrote the embedded code so can change it to better fit with the Linux User Space. In addition I can write a shim layer between my protocol code and Linux. My question is about timer_t. I've been searching through documentation but can't find mention of any value that is considered an invalid timer identifier. My embedded code's timer creation function would be something like: timer_id = start_timer(duration, expiry_function); The returned timer_id if set to the invalid value has not been started. In addition when an event occurs in the protocol if the timer_id is not that invalid value then stop the timer. So the timer_id stored tells you if it's valid and running. As soon as the timer expires the expiry function returns the value to the INVLAID_TIMER_ID value. >From what I can tell in Linux there is no such Invalid value. A timer_t is just an integer and all values are valid. This being the case in my protocol code I'd have to change it so all timer id's are valid and another variable would have to store the status/validity of the timer variable. Is this the case? Just wanted to confirm this before I go re-writing. Hope that isn't considered off topic, it is linux programming related-ish.

1 0