- kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

Does FreeRTOS works on Linux?
by Raul Piper May 11, 2021

May 11, 2021

Hello, I am having an embedded linux system and wanted to use any RTOS on it. Can I use Free RTOS? My project has 2 tasks , one is producer and another is consumer working parallely. I have seen Free RTOS used in microcontrollers handling this , how can this be done in Linux(which such APIs if not Free RTOS). Regards, R

4 6

kernel module parameter API ala drm.debug
by jim.cromie＠gmail.com May 11, 2021

May 11, 2021

so as a proof of concept, Ive converted drm to use dynamic-debug drm has ~dozen categories of debug messages, mapped to bits in drm.debug, aka /sys/module/drm/parameters/debug these bits are checked at runtime by drm_debug_enabled() to do drm debug printing. my patchset updates users of drm_debug_enabled() to call pr_debug instead, avoiding lots of bit-checking. it maps bits of drm.debug using dynamic_debug_exec_queries(), a recently exported function to support this sort of thing. you can see it here https://github.com/jimc/linux/tree/dd-drm A narrower interface than dynamic_debug_exec_queries() is possible: +static char *format_prefix_classes[] = { + "gvt: cmd: ", + "gvt: core: ", + "gvt: dpy: ", + "gvt: el: ", + "gvt: irq: ", + "gvt: mm: ", + "gvt: mmio: ", + "gvt: render: ", + "gvt: sched: " +}; +static const struct kernel_param_ops param_ops_dyndbg = { + .set = param_set_dyndbg, + .get = param_get_dyndbg, +}; + +#define info_ln(hexi, prefix) "\n\t0x" __stringify(hexi) "\t" prefix + +MODULE_PARM_DESC(debug_gvt, " gvt debug categories:" + info_ln(1, "gvt: cmd:") + info_ln(2, "gvt: core:") + info_ln(4, "gvt: dpy:") + info_ln(8, "gvt: el:") + info_ln(10, "gvt: irq:") + info_ln(20, "gvt: mm:") + info_ln(40, "gvt: mmio:") + info_ln(80, "gvt: render:") + info_ln(100, "gvt: sched:")); + +module_param_cb(debug_gvt, &param_ops_dyndbg, &__gvt_debug, 0644); this is useful in that it shows up in modinfo output but there could be better presentation, maybe /sys/module/drm/parameters/debug.help param_set_dyndbg could be moved into dynamic-debug proper, instead of reimplemented everywhere debug bits control debug prints (currently drm, maybe others, could be i915/gvt) the point is that this integer parameter maps consecutive bits to "classes" named in an input vector. these "classes" are just format ^prefix queries clearly we shouldnt need the 1,2,4... whats a good interface design for this module-param-bitmap ?

1 0

GPIO Driver Design
by Robert Joslyn May 8, 2021

May 8, 2021

I'm working on writing a driver for custom hardware, but I'm not sure the best way to expose the features of the hardware to userspace. The hardware in question is basically a configurable industrial GPIO board, but it has additional features that don't seem to map well to the libgpio subsystem. The board has a set of output pins that can be treated like normal GPIO pins. The input pins, however, have additional configuration that can be set by a user. For example, the inputs can be configured for different input voltage levels, such as 24 Vdc, 110 Vac at 60 Hz, 220 Vac at 50 Hz, etc. Debounce times can also be configured. The hardware also has a FIFO buffer that records the state of all input pins, as well as a free running counter (FRC), when any input value changes. What is a reasonable way to expose the voltage and debounce configuration to userspace? Or the FIFO? Presently, I have an initial driver that creates a character device in /dev and uses ioctls so a user can provide configuration, read input/output state, and read the FIFO. There is also a custom sysfs interface to access the same configuration values and data.The sysfs interface is manily to provide something easier to use from shell, all the functionality is also duplicated in the ioctl interface. I plan on adding access to the input/output states through libgpio, but have not finished this yet. I'd like to get this driver in a state where it could be merged upstream, but I'm not sure that a custom sysfs class or ioctls are the right way to go. Is there a better way for userspace to provide this configuration, or access to the FIFO? This is the first kernel driver I've worked on, so it's possible there is an obvious way to implement this that I just haven't seen yet. Or maybe this hardware is just too specific and doesn't make a good candidate for pushing upstream. I'd appreciate any suggestions or ideas! Thanks, Robert

1 0

Re: Does FreeRTOS works on Linux?
by Raul Piper May 8, 2021

May 8, 2021

On Fri, May 7, 2021 at 12:07 PM Greg KH <greg(a)kroah.com> wrote: > > On Thu, May 06, 2021 at 08:12:48PM +0530, Raul Piper wrote: > > On Thu, May 6, 2021 at 1:01 AM loďc tourlonias > > <loic.tourlonias(a)gmail.com> wrote: > > > > > > On Wed, May 5, 2021 at 8:07 PM Raul Piper <raulpblooper(a)gmail.com> wrote: > > > > > > > > Hello, > > > Hi > > > > I am having an embedded linux system and wanted to use any RTOS on > > > > it. Can I use Free RTOS? > > > > My project has 2 tasks , one is producer and another is consumer > > > > working parallely. > > > I'm afraid I don't understand what you are requesting and I hope that > > > someone else understand more clearly. But I'll still explain my > > > understanding. > > > You want to run a RTOS upon Linux OS, doesn't make it too much OSes? > > > Then I have several questions: > > > Do you really need a Real Time Operating System, can't you run your > > > project in Linux userspace? > > I can do that but Linux is not an ROS afaik. Even though i dont have > > strict timing requirements , its like a sensor on i2c(producer) and > > then another consumer tasks which not > > only feeds on this data but also tell the i2c sensor by sending some > > commands to it. > > That should be fine, Linux can be a "Real Time Operating system" if you > use the RT patchset if your workloads require it. > OK I need to check on this RT Patchset if at all I need real time features. > If you do NOT have strict timing requirements, then there is no need to > worry about any of this, so I do not understand your wish to slow your > system down by running a "ROS". Why? > Typo this is RTOS (not R OS - Robot operating system ;) ) > > > Have you tried Linux RTOS extension (again I'm no expert and haven't > > > tried them but you can look at Xenomai)? > > I will explore this.I was thinking if i can use Posix thread to have > > two tasks one reading sensors and another sending commands to it.But > > then again i am confused > > if using a posix thread full fills the requirement of RTOS. The reason > > I want to use RTOS is , I am porting a Microcontroller code using RTOS > > (free RTOS) to Linux and > > afaik FreeRTOS doesnt work on Linux. > > What is your "requirement" exactly? Read the data from the sensor and also change the type of data sent by the sensor by analysing the received data by sending the command to the sensor periodically based on trigger.There is some batch of data which a "simultaneously" running algorithm is evaluating and send the command based on calculated trigger, to sensor to change the data stream.This batch of data is of size 40-60 bytes length and may vary each time a batch is being read. So in summary i wany two tasks to run together. RTOS is one option Posix thread is Another Based on comments i recieved i will go for posix threads. For future i think there is less clarity among developers (especially me) when to use rtos and when not to use. Does using threads gurantees real time behaviour? If not why to use posix threads, debate is endless . What are those timing requirements which draws a border between using rtos and not using rtos. @GKH if you have some good “practical”references to decide on this subject please do share. > > thanks, > > greg k-h

2 1

deleting timer that re-registers itself
by Hyeonggon Yoo May 5, 2021

May 5, 2021

Does del_timer work well for timers that re-registers itself? what if the timer is currently running, and del_timer is called, and the running timer re-registers itself? how should I handle it? Below is my simple kernel timer example. timer.c #include <linux/kernel.h> #include <linux/module.h> #include <linux/time.h> #define MODULE_NAME "TIMER" #define DELAY (1 * HZ) struct timer_data { int value; struct timer_list timer; }; struct timer_data data = { .value = 0 }; void timer_callback(struct timer_list *timer) { struct timer_data *data = from_timer(data, timer, timer); data->value++; printk(KERN_INFO "[%s] value is = %d\n", __func__, data->value); mod_timer(timer, jiffies + DELAY); } int __init timer_init(void) { printk("[%s] creating timer...\n", __func__); timer_setup(&data.timer, timer_callback, 0); mod_timer(&data.timer, jiffies + DELAY); return 0; } void __exit timer_exit(void) { int ret = del_timer(&data.timer); printk("[%s] deleting timer..., ret = %d\n", __func__, ret); } module_init(timer_init); module_exit(timer_exit); MODULE_LICENSE("GPL");

3 8

Test if a socket accept is from external network
by John Wood May 5, 2021

May 5, 2021

Hi, I'm working in a LSM to detect and mitigate fork brute force attacks against vulnerable userspace applications. Now, to fine tuning the detection I want to detect a network activity. To do so, I can use the following code in the "socket_sock_rcv_skb" hook: static int brute_sock_rcv_skb(struct sock *sk, struct sk_buff *skb) { if (!skb->dev || (skb->dev->flags & IFF_LOOPBACK)) return 0; network_activity = true; return 0; } This way, only external connections are taken into account. Or in other words, the communication using local sockets are skipped. The drawback with this approach is that the commented hook is call with every packet received. So, I have decided to use the hook that is called only when a connection is accepted: "socket_accept". static int brute_socket_accept(struct socket *sock, struct socket *newsock) { /* I need to detect external connections */ return 0; } But now I don't be able to detect only external connections. Now, I don't have access to the device (or I don't know how to do it). I have tried with the "sock->sk->sk_bound_dev_if" member of the sock struct but its value is always 0 for internal and external connections (at least in my tests). How can I detect that an external connection (using a net device) is accepted and avoid internal network communication? Any help would be greatly appreciated. Thanks in advance. John Wood

3 3

high precision timers
by Hyeonggon Yoo May 3, 2021

May 3, 2021

the 'timer wheel' - which has low resolution works at jiffy resolution. so usually it has precision of 1~10ms. and there's high resolution timers in kernel to have nanosecond resolution, but it doesn't seem to have nanosecond precision is there a way to deal with high precision timers in kernel? does it require specific hardware?

4 5

Re: Scheduler benchmarks
by Muni Sekhar April 29, 2021

April 29, 2021

On Tue, Aug 18, 2020 at 11:45 PM peter enderborg <peter.enderborg(a)sony.com> wrote: > > On 8/18/20 7:53 PM, Muni Sekhar wrote: > > On Tue, Aug 18, 2020 at 11:06 PM Greg KH <greg(a)kroah.com> wrote: > >> On Tue, Aug 18, 2020 at 11:01:35PM +0530, Muni Sekhar wrote: > >>> On Tue, Aug 18, 2020 at 10:44 PM Greg KH <greg(a)kroah.com> wrote: > >>>> On Tue, Aug 18, 2020 at 10:24:13PM +0530, Muni Sekhar wrote: > >>>>> On Tue, Aug 18, 2020 at 8:06 PM Greg KH <greg(a)kroah.com> wrote: > >>>>>> On Tue, Aug 18, 2020 at 08:00:11PM +0530, Muni Sekhar wrote: > >>>>>>> Hi all, > >>>>>>> > >>>>>>> I’ve two identical Linux systems with only kernel differences. > >>>>>> What are the differences in the kernels? > >>>> You didn't answer this question, is this the same kernel source being > >>>> compared here? Same version? Same compiler? Everything identical? > >>> Both systems are having exactly the same hardware configuration. > >>> Compiler and kernel versions are different. One system has Ubuntu > >>> 16.04.4 LTS(4.4.0-66-generic kernel with gcc version 5.4.0) kernel and > >>> the other one has Ubuntu 18.04.4 LTS(4.15.0-91-generic kernel with gcc > >>> version 7.5.0). > >> Those are _very_ different kernel versions, with many years and tens of > >> thousands of different changes between them. > >> > >> Hopefully the newer kernel is faster, so just stick with that :) > > But unfortunately the newer kernel is very slow, that is the reason > > for starting this investigation :) > > Any type of help, and guidelines to dive deeper will be highly appreciated. > > On the 4.4 kernel you dont have > > +CONFIG_RETPOLINE=y > +CONFIG_INTEL_RDT=y Thanks! That is helpful. Yes, I see 4.4 kernel don't have the above two config options. What analysis can be done to narrow down the root cause? Any example of reference could be helpful to understand. > > And your base is very different two. > > Try to use mainline on both system and see. > > You can also use the same base kernel version from ubuntu and > > run your test. > > > >> greg k-h > > > > > -- Thanks, Sekhar

6 6

[PATCH] MAINTAINERS: mark isapnp as obsolete
by Anupama K Patil April 29, 2021

April 29, 2021

isapnp code is very old and according to this link https://en.wikipedia.org/wiki/Legacy_Plug_and_Play#Specifications from Wikipedia, even Windows Vista disabled ISA PnP by default. This change is in follow up to https://lore.kernel.org/lkml/20210422180322.7wlyg63kv3n2k6id@ubuntu/T/#u and https://lore.kernel.org/lkml/20210424194301.jmsqpycvsm7izbk3@ubuntu/T/#u Suggested-by: B K Karthik <bkkarthik(a)pesu.pes.edu> Signed-off-by: Anupama K Patil <anupamakpatil123(a)gmail.com> --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 04e7de8c95be..215e05d93c60 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9534,7 +9534,7 @@ F: drivers/media/radio/radio-isa* ISAPNP M: Jaroslav Kysela <perex(a)perex.cz> -S: Maintained +S: Obsolete F: Documentation/driver-api/isapnp.rst F: drivers/pnp/isapnp/ F: include/linux/isapnp.h -- 2.25.1

5 8

can't find device_create funtion in a driver containing cdev (habanalabs)
by ckim＠etri.re.kr April 29, 2021

April 29, 2021

Hello list members, In habanalabs' goya chip driver, (https://elixir.bootlin.com/linux/v5.4.21/source/drivers/misc/habanalabs/dev ice.c#L1057) I can see cdev_init, cdev_add and class_create functions. Cdev_init is called insde device_init_cdev and cdev_add is called from device_cdev_sysfs_add -> cdev_device_add. I can see class_create being called too. But I couldn't find device_create function (which is used to make sysfs entry?) Can anyone tell me where device_create is called? (or where alternative method is being used?) Thank you! Chan Kim

3 3