March 2017 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

Understanding get_maintainer.pl and MAINTAINERS
by Robin Krahl March 9, 2017

March 9, 2017

Hi again, I’m now preparing my patchset for bcm2835-audio. get_maintainer.pl returns: $ ./scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/staging/vc04_services/bcm2835-audio/ Stephen Warren <swarren(a)wwwdotorg.org> (maintainer:BROADCOM BCM2835 ARM ARCHITECTURE) Lee Jones <lee(a)kernel.org> (maintainer:BROADCOM BCM2835 ARM ARCHITECTURE) Eric Anholt <eric(a)anholt.net> (maintainer:BROADCOM BCM2835 ARM ARCHITECTURE) Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> (supporter:STAGING SUBSYSTEM) Florian Fainelli <f.fainelli(a)gmail.com> (maintainer:BROADCOM BCM281XX/BCM11XXX/BCM216XX ARM ARCHITE...) Ray Jui <rjui(a)broadcom.com> (maintainer:BROADCOM BCM281XX/BCM11XXX/BCM216XX ARM ARCHITE...) Scott Branden <sbranden(a)broadcom.com> (maintainer:BROADCOM BCM281XX/BCM11XXX/BCM216XX ARM ARCHITE...) bcm-kernel-feedback-list(a)broadcom.com (maintainer:BROADCOM BCM281XX/BCM11XXX/BCM216XX ARM ARCHITE...) linux-rpi-kernel(a)lists.infradead.org (moderated list:BROADCOM BCM2835 ARM ARCHITECTURE) linux-arm-kernel(a)lists.infradead.org (moderated list:BROADCOM BCM2835 ARM ARCHITECTURE) devel(a)driverdev.osuosl.org (open list:STAGING SUBSYSTEM) linux-kernel(a)vger.kernel.org (open list) Why are the BROADCOM BCM281XX/BCM11XXX/BCM216XX ARM ARCHITE guys added to the list? The MAINTAINERS file defines these rules: BROADCOM BCM281XX/BCM11XXX/BCM216XX ARM ARCHITECTURE M: Florian Fainelli <f.fainelli(a)gmail.com> M: Ray Jui <rjui(a)broadcom.com> M: Scott Branden <sbranden(a)broadcom.com> M: bcm-kernel-feedback-list(a)broadcom.com T: git git://github.com/broadcom/mach-bcm S: Maintained N: bcm281* N: bcm113* N: bcm216* N: kona F: arch/arm/mach-bcm/ As far as I see, none of these rules matches the given path. Am I missing something? Regards, Robin

2 3

Unable to get new root in new namespace
by Prabhunath G March 9, 2017

March 9, 2017

Hi All, Here is the complete code attempting to work in the new root of the new namespace. This program is built and executed successfully. But within the child context, my root is still the original root directory. My intention is to make */opt/lxc/ *as my new root. Required executables (bash and ls) along with the required libraries are placed in /opt/lxc. ** New PID Namespace is working right. # ls /proc is showing the processes that are executed only in the new namespace. Any help is appreciated. ----------------------------------------------------------------------------------------------------------------------------------------- #define STACK_SIZE (8 * 1024) unsigned char stack[STACK_SIZE]; char *args[] = { "/usr/bin/bash", NULL }; int child_function (void *arg) { ret = mount("", "/", "", MS_REC|MS_PRIVATE, NULL); ret = mount("/opt/lxc/", "/", "", MS_BIND|MS_REC, NULL); ret = mount("proc", "/proc", "proc", 0, NULL); execv(args[0],args); return 0; } int main() { unshare (CLONE_NEWNS | CLONE_NEWUTS | CLONE_NEWIPC | CLONE_NEWPID | CLONE_NEWNET); clone(child_function, stack+STACK_SIZE, CLONE_PARENT | SIGCHLD, NULL); return 0; } ----------------------------------------------------------------------------------------------------------------------------------------------------- Regards, Prabhu

1 0

Struggling with git (format-patch & send-email)
by Prasant J March 9, 2017

March 9, 2017

Hi, I want to send my 2 commits as patches in chain reply to my first email. My first email will have multiple files as attachment. How can I achieve it? first email subject: [mystring] Feature 01 patch 01: |_ [mystring] [PATCH 01/02] My commit msg 1 patch 02: |_ [mystring] [PATCH 02/02] My commit msg 2 My git send-email is working, but I want to achieve the above desired outcome which I'm unable to. I have tried so many combinations of git format-patch and send-email over the past days but I'm unable to get the above outcome. One combination that was close to working: => git format-patch -2 --numbered --start-number 01 -o patch-dir --thread => git send-email --in-reply-to="Message-Id" --to user(a)gmail.com --chain-reply-to --suppress-cc=self patch-dir/* I'm trying to see the desired effect in my web gmail, but I do not see. * What am I doing wrong? Any inputs will be of help! * How to add subject prefix? (subject-prefix of git format-patch is not working as above) * How to get 2 digit patch number? I'm getting PATCH 1/2 instead of PATCH 01/02 Awaiting help! Thanks & regards, Pj

3 4

how to get filename of execve() system call from kernel module which install hook to syscall table to intercept original syscall in kernels before 4.2 and atter 4.2 ? X86_64
by Lev Olshvang March 8, 2017

March 8, 2017

Hello all, In kernels 3.X up to 4.2 execve(|) system call was for x86_64 architecture the the system call was made through some magic ( I can't say I understand it ) assembly stub in arch/x86/kernel/entry_64.S so up to kernel 4.2 it was possble to patch this assembly to install the hook, ex. see http://stackoverflow.com/questions/8372912/hooking-sys-execve-on-linux-3-x/… But this hook still can't access in a proper way filename argument, althouth I tried to do it with in the same way as fs/exec.c does : using kernel's getname() function (which I was need to find through kallsyms_lookup_name() In kernels 4.2 and up, the arch/x86/kernel/entry_64.S is gone, and I still dont' have a clue what to do to get filename as a char string. Please advise. Lev.

3 5

bcm2835-audio: logging mechanism
by Robin Krahl March 8, 2017

March 8, 2017

Hi there, the TODO file for drivers/staging/vc04_services/bcm2835-audio contains the following item: > 4) Cleanup the logging mechanism. The driver should probably be using > the standard kernel logging mechanisms such as dev_info, dev_dbg, and > friends. Currently there are macros audio_debug, audio_warning, LOG_DBG, LOG_WARN etc. that call pr_* and use the function name and the line number as a prefix for the logged message. Refactoring this driver to use the dev_* methods would remove this additional information, and would also lead to some clutter as the driver structure is not directly accessible in most of the functions, for example dev_dbg(alsa_stream->chip->card->dev, "... instead of audio_debug("... Furthermore, the driver currently uses "IN" and "OUT" debug messages at the beginning and at the end of some functions, for example snd_bcm2835_ctl_info in bcm2835-ctl.c. This is useless if the function name is not included in the log message. But on the other hand, I’m not sure whether it is desirable to have this kind of messages in the kernel code at all. So my question is: Does it make sense to replace the audio_* and LOG_* macros with the standard dev_* functions? And is it okay to remove log messages like "IN" and "OUT" that are only used to track function entries and exits? Regards, Robin

2 2

patch protocol question
by Tobin C. Harding March 8, 2017

March 8, 2017

I would like to know the correct protocol in order to make the maintainers job as easy as possible please. Once a patch has been reviewed and the review makes good points that mean the patch is invalid/unnecessary what is the protocol from then? Assuming one replies to the reviewer with thanks and acknowledging their points. Is it then protocol to state that you are not going to pursue the patch further? How do maintainers know to not bother any more with a patch? Similar question; if the last patch of a patch series is not needed should one resend another version without the last patch or is there an accepted protocol to signal this so that the maintainer only looks at merging the initial patches in the series. thanks, Tobin.

2 2

sizeof
by Tobin C. Harding March 8, 2017

March 8, 2017

Question relating to the validity/usefulness of patching calls to sizeof. >From Documentation/process/coding-style.rst The preferred form for passing a size of a struct is the following: .. code-block:: c p = kmalloc(sizeof(*p), ...); The alternative form where struct name is spelled out hurts readability and introduces an opportunity for a bug when the pointer variable type is changed but the corresponding sizeof that is passed to a memory allocator is not.Is the distinction between Question: Is a the following a valid/useful patch or just code churn diff-- - sizeof(struct foo) + sizeof(*fp) ---

2 1

Modify efivars permisson denied
by Lucas Tanure March 7, 2017

March 7, 2017

Hi, I would like to modify efivars at /sys/firmware/efi/efivars/. But I can't modify it using root. The file : -rw-r--r-- 1 root root 31 Mar 4 19:04 /sys/firmware/efi/efivars/OilSetup-88d1911c-ab4a-4335-9e9a-26be9081acc3 the mount point : efivarfs on /sys/firmware/efi/efivars type efivarfs (rw,nosuid,nodev,noexec,relatime) Any ideas ? Thanks -- Lucas Tanure

2 5

log file created by git send-email?
by Tobin C. Harding March 7, 2017

March 7, 2017

Is there a log of emails sent using `git send-email`? The man page does not mention one and web search on 'log' returns commands related (unsurprisingly) to `git log`. thanks, Tobin.

2 2

simple block device - request processing in threads
by Andras Pal March 4, 2017

March 4, 2017

Dear All, with some experience in low-level C programming, I started to develop a kernel module which allows I2C slave EEPROM-alike devices (EEPROMs, FRAMs, BBRAMs, etc) to be accessed as block devices. Since the i/o transactions are relatively slow (say, 40k/sec), i'm using kernel threads. However, it seems that while the read (rq_data_dir(req)) == 0) operations are fine, write operations (rq_data_dir != 0) seems to be lost. For instance, if I do `dd bs=512 count=32 of=/dev/i2cblk0`, only the first 8 sectors are fetched by blk_fetch_request(). I copy the critical/relevant/interesting parts after this message. The code is based on drivers/block/nbd.c. In addition - according to printk()-based debugging - after each write request, the kernel starts to access sectors in the order of 0, 24, 8, 56, 120, 64, 32 (with 8 sectors == 1 page per request). Maybe it seeks for some filesystems...? Can this feature be turned off (like setting disk->queue->backing_dev_info.ra_pages to zero in order to turn off readaheaad)? thanks in advance, cheers, Andras /*****************************************************************************/ static int i2cblk_process_io_request(struct i2cblk_device *device,struct request *req) { int ret; /* i2cblk_transfer(): this function calls i2c_master_send/recv(), etc. */ ret=i2cblk_transfer(device, blk_rq_pos(req), blk_rq_cur_sectors(req), bio_data(req->bio), rq_data_dir(req)); return(ret); } static void i2cblk_request(struct request_queue *q) { struct request *req; while ( (req=blk_fetch_request(q)) != NULL ) { struct i2cblk_device *device; if ( req->cmd_type != REQ_TYPE_FS ) { __blk_end_request_all(req, -EIO); continue; } /* spin_unlock_irq(q->queue_lock); */ device=req->rq_disk->private_data; spin_lock_irq(&device->active_lock); list_add_tail(&req->queuelist,&device->active_queue); spin_unlock_irq(&device->active_lock); wake_up(&device->active_wq); /* spin_lock_irq(q->queue_lock); */ }; } static void i2cblk_end_request(struct request *req) { int error; struct request_queue *q = req->q; unsigned long flags; error=(req->errors ? -EIO : 0); spin_lock_irqsave(q->queue_lock, flags); __blk_end_request_all(req, error); spin_unlock_irqrestore(q->queue_lock, flags); } static void i2cblk_handle_request(struct i2cblk_device *device,struct request *req) { if ( i2cblk_process_io_request(device,req) ) req->errors++; i2cblk_end_request(req); return; } static int i2cblk_thread(void *data) { struct i2cblk_device *device = data; struct request *req; while ( ! kthread_should_stop() ) { wait_event_interruptible(device->active_wq,kthread_should_stop() || (!list_empty(&device->active_queue))); if ( list_empty(&device->active_queue) ) continue; spin_lock_irq(&device->active_lock); req = list_entry(device->active_queue.next,struct request,queuelist); list_del_init(&req->queuelist); spin_unlock_irq(&device->active_lock); i2cblk_handle_request(device,req); }; return(0); }

2 2