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

How to handle Hotplug with UIO userspace driver
by Mandeep Sandhu Sept. 28, 2017

Sept. 28, 2017

Hi All, We're working on a custom PCIe based hardware, which can be hotplugged into the system. We're using the UIO approach of having a minimal kernel driver which basically handles the interrupt and provides a mmap of the device memory to userspace and userspace driver which mmap's the provided memory and manages the device. We're using systemd and udev rules to spawn off a user process whenever the hardware is plugged into the system. Now my question is, how can we gracefully inform the userspace process if someone pulls out the hardware (sitting on a PCI slot). Even with full hotplug support, there can always be a case of "surprise removal", where someone yanks the board without going through the whole "attention button" routine. While the kernel driver has no problem with hot-removal, I see that if I "unregister" our UIO driver in response to the removal, then the mapping for the hardware's physical memory would become "invalid" (done by UIO on unregsiter). This could possibly cause the user-space process to crash due to illegal memory access. Is there a "standard" way to handle this scenario, i.e for hotpluggable hardware using UIO? Basically, we'd want to "unregister" the UIO device only after the userspace process is informed of the removal so that it could unmap/close the device and exit gracefully. We could try doing it using some systemd trick, where we put a dependency on one of the UIO created files (/dev/uioX), so that when the file is destroyed, systemd could kill/shutdown the process. Although, there's no guarantee that the userspace process won't try to access the mmaped space during that interval, so this isn't a reliable solution either. Any hints, thoughts? Thanks., -mandeep

3 7

How to get the inode - no path_lookup
by Rishi Agrawal Sept. 17, 2015

Sept. 17, 2015

Hi All, I had a module which used the path_lookup function to print the details of any file's inode. I now want to rewrite that module in order to show some juniors how to write some code in kernel. I am using 3.4.6 kernel, I tried finding out path_lookup but google showed that it has been removed. I tried the following code then which did not work . . . dentry = kern_path_create(AT_FDCWD, filename, &path, 1); if (IS_ERR(dentry)) { printk("Failed to obtain the dentry"); return; } its not returning dentry I again tried after seeing the implementation of vfs_stat function user_path_at(AT_FDCWD, filename, lookup_flags, &path); but this also fails. I am using a proc interface to pass the filename, and copying the filename into a kernel buffer. How can I get a copy of vfs inode for a file name. -- Regards, Rishi Agrawal

4 8

[PATCH] drivers: staging: lustre: lnet: klnds: o2iblnd: o2iblnd.c
by Bharadwaaj Aug. 9, 2015

Aug. 9, 2015

From: "bhaaradwaaj(a)gmail.com" <bhaaradwaaj(a)gmail.com> Fixed spelling mistake. Signed-off-by: Bharadwaaj Ramakrishnan <bhaaradwaaj(a)gmail.com> --- drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.c b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.c index 4eb24a1..463940b 100644 --- a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.c +++ b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.c @@ -1424,7 +1424,7 @@ static int kiblnd_create_fmr_pool(kib_fmr_poolset_t *fps, static void kiblnd_fail_fmr_poolset(kib_fmr_poolset_t *fps, struct list_head *zombies) { - if (fps->fps_net == NULL) /* intialized? */ + if (fps->fps_net == NULL) /* initialized? */ return; spin_lock(&fps->fps_lock); @@ -1628,7 +1628,7 @@ static void kiblnd_destroy_pool_list(struct list_head *head) static void kiblnd_fail_poolset(kib_poolset_t *ps, struct list_head *zombies) { - if (ps->ps_net == NULL) /* intialized? */ + if (ps->ps_net == NULL) /* initialized? */ return; spin_lock(&ps->ps_lock); -- 2.4.6

2 1

[PATCH 2/2] drivers: staging : media : davinci_vpfe : dma365_resizer.c drivers: staging : media : davinci_vpfe : dma365_resizer.h
by Bharadwaaj Aug. 9, 2015

Aug. 9, 2015

From: "bhaaradwaaj(a)gmail.com" <bhaaradwaaj(a)gmail.com> Fixed some spelling mistake. Signed-off-by: Bharadwaaj Ramakrishnan<bhaaradwaaj(a)gmail.com> --- drivers/staging/media/davinci_vpfe/dm365_resizer.c | 22 +++++++++++----------- drivers/staging/media/davinci_vpfe/dm365_resizer.h | 2 +- 2 files changed, 12 insertions(+), 12 deletions(-) diff --git a/drivers/staging/media/davinci_vpfe/dm365_resizer.c b/drivers/staging/media/davinci_vpfe/dm365_resizer.c index acb293e..10f51f4 100644 --- a/drivers/staging/media/davinci_vpfe/dm365_resizer.c +++ b/drivers/staging/media/davinci_vpfe/dm365_resizer.c @@ -495,7 +495,7 @@ resizer_configure_in_continious_mode(struct vpfe_resizer_device *resizer) int line_len; int ret; - if (resizer->resizer_a.output != RESIZER_OUPUT_MEMORY) { + if (resizer->resizer_a.output != RESIZER_OUTPUT_MEMORY) { dev_err(dev, "enable resizer - Resizer-A\n"); return -EINVAL; } @@ -507,7 +507,7 @@ resizer_configure_in_continious_mode(struct vpfe_resizer_device *resizer) param->rsz_en[RSZ_B] = DISABLE; param->oper_mode = RESIZER_MODE_CONTINIOUS; - if (resizer->resizer_b.output == RESIZER_OUPUT_MEMORY) { + if (resizer->resizer_b.output == RESIZER_OUTPUT_MEMORY) { struct v4l2_mbus_framefmt *outformat2; param->rsz_en[RSZ_B] = ENABLE; @@ -1048,13 +1048,13 @@ static void resizer_ss_isr(struct vpfe_resizer_device *resizer) if (ipipeif_sink != IPIPEIF_INPUT_MEMORY) return; - if (resizer->resizer_a.output == RESIZER_OUPUT_MEMORY) { + if (resizer->resizer_a.output == RESIZER_OUTPUT_MEMORY) { val = vpss_dma_complete_interrupt(); if (val != 0 && val != 2) return; } - if (resizer->resizer_a.output == RESIZER_OUPUT_MEMORY) { + if (resizer->resizer_a.output == RESIZER_OUTPUT_MEMORY) { spin_lock(&video_out->dma_queue_lock); vpfe_video_process_buffer_complete(video_out); video_out->state = VPFE_VIDEO_BUFFER_NOT_QUEUED; @@ -1064,7 +1064,7 @@ static void resizer_ss_isr(struct vpfe_resizer_device *resizer) /* If resizer B is enabled */ if (pipe->output_num > 1 && resizer->resizer_b.output == - RESIZER_OUPUT_MEMORY) { + RESIZER_OUTPUT_MEMORY) { spin_lock(&video_out->dma_queue_lock); vpfe_video_process_buffer_complete(video_out2); video_out2->state = VPFE_VIDEO_BUFFER_NOT_QUEUED; @@ -1074,7 +1074,7 @@ static void resizer_ss_isr(struct vpfe_resizer_device *resizer) /* start HW if buffers are queued */ if (vpfe_video_is_pipe_ready(pipe) && - resizer->resizer_a.output == RESIZER_OUPUT_MEMORY) { + resizer->resizer_a.output == RESIZER_OUTPUT_MEMORY) { resizer_enable(resizer, 1); vpfe_ipipe_enable(vpfe_dev, 1); vpfe_ipipeif_enable(vpfe_dev); @@ -1242,8 +1242,8 @@ static int resizer_do_hw_setup(struct vpfe_resizer_device *resizer) struct resizer_params *param = &resizer->config; int ret = 0; - if (resizer->resizer_a.output == RESIZER_OUPUT_MEMORY || - resizer->resizer_b.output == RESIZER_OUPUT_MEMORY) { + if (resizer->resizer_a.output == RESIZER_OUTPUT_MEMORY || + resizer->resizer_b.output == RESIZER_OUTPUT_MEMORY) { if (ipipeif_sink == IPIPEIF_INPUT_MEMORY && ipipeif_source == IPIPEIF_OUTPUT_RESIZER) ret = resizer_configure_in_single_shot_mode(resizer); @@ -1268,7 +1268,7 @@ static int resizer_set_stream(struct v4l2_subdev *sd, int enable) if (&resizer->crop_resizer.subdev != sd) return 0; - if (resizer->resizer_a.output != RESIZER_OUPUT_MEMORY) + if (resizer->resizer_a.output != RESIZER_OUTPUT_MEMORY) return 0; switch (enable) { @@ -1722,7 +1722,7 @@ static int resizer_link_setup(struct media_entity *entity, } if (resizer->resizer_a.output != RESIZER_OUTPUT_NONE) return -EBUSY; - resizer->resizer_a.output = RESIZER_OUPUT_MEMORY; + resizer->resizer_a.output = RESIZER_OUTPUT_MEMORY; break; default: @@ -1747,7 +1747,7 @@ static int resizer_link_setup(struct media_entity *entity, } if (resizer->resizer_b.output != RESIZER_OUTPUT_NONE) return -EBUSY; - resizer->resizer_b.output = RESIZER_OUPUT_MEMORY; + resizer->resizer_b.output = RESIZER_OUTPUT_MEMORY; break; default: diff --git a/drivers/staging/media/davinci_vpfe/dm365_resizer.h b/drivers/staging/media/davinci_vpfe/dm365_resizer.h index 93b0f44..00e64b0 100644 --- a/drivers/staging/media/davinci_vpfe/dm365_resizer.h +++ b/drivers/staging/media/davinci_vpfe/dm365_resizer.h @@ -210,7 +210,7 @@ enum resizer_input_entity { enum resizer_output_entity { RESIZER_OUTPUT_NONE = 0, - RESIZER_OUPUT_MEMORY = 1, + RESIZER_OUTPUT_MEMORY = 1, }; struct dm365_resizer_device { -- 2.4.6

2 1

shared memory via /dev/shm?
by Nicholas Murphy Aug. 1, 2015

Aug. 1, 2015

Apologies if this has been asked before: Yes, I’ve read all the disclaimers against accessing files in the kernel, but I want to set up a shared memory segment between a kernel module and one or more user processes, and /dev/shm seems like a good way to do that. Given that, what’s the preferred way to access files within the kernel (specifically within a kernel module)? At the moment, I’m using filp_open() on module load to open the file (which seems to work), then: filp_close() to close the file * kern_path() to get the entry for the /dev/shm file vfs_unlink() to delete the /dev/shm file dput() to free the dentry path_put() to free the path …on module unload, and it hangs. Any suggestions what I’m doing wrong? Thanks, Nick

2 3

Re: Kernel development using linux containers (LXC ) ?
by Grzegorz Dwornicki July 30, 2015

July 30, 2015

As I said I use kvm and this question was just for sake of argument. Since you ask I work with sockets ATM. 30 lip 2015 16:03 "nick" <xerofoify(a)gmail.com> napisał(a): > > > On 2015-07-30 09:57 AM, Grzegorz Dwornicki wrote: > > Yeah I've thought that hardware can be hard to program on UML > > 30 lip 2015 14:48 "nick" <xerofoify(a)gmail.com> napisał(a): > > > Just so I known what area are you working in as this may help > me find a better solution for you. > Nick > >> > >> > >> On 2015-07-30 08:47 AM, Grzegorz Dwornicki wrote: > >>> What about UML? I'm using KVM but asking for the sake of argument. > >>> > >> I already stated that it works fine if your not doing hardware > >> exact development like file systems or networking core. > >> Nick > >>> 2015-07-29 23:03 GMT+02:00 Manish Katiyar <mkatiyar(a)gmail.com>: > >>>> On Wed, Jul 29, 2015 at 1:49 PM, Greg Freemyer < > greg.freemyer(a)gmail.com> > >> wrote: > >>>>> On Wed, Jul 29, 2015 at 4:23 PM, Manish Katiyar <mkatiyar(a)gmail.com> > >> wrote: > >>>>>> On Wed, Jul 29, 2015 at 1:07 PM, Jeff Haran <Jeff.Haran(a)citrix.com> > >> wrote: > >>>>>>>> -----Original Message----- > >>>>>>>> From: kernelnewbies-bounces(a)kernelnewbies.org [mailto: > >> kernelnewbies- > >>>>>>>> bounces(a)kernelnewbies.org] On Behalf Of Manish Katiyar > >>>>>>>> Sent: Wednesday, July 29, 2015 11:51 AM > >>>>>>>> To: kernelnewbies > >>>>>>>> Subject: Kernel development using linux containers (LXC ) ? > >>>>>>>> > >>>>>>>> Hi, > >>>>>>>> > >>>>>>>> I've been playing and reading about control groups and linux > >> containers > >>>>>>>> recently and was wondering if there are any existing recipes on > how > >> to setup > >>>>>>>> a kernel environment in containers. Google hasn't been helpful so > >> far (or > >>>>>>>> maybe I'm not searching properly). > >>>>>>>> > >>>>>>>> I've used VMs for dev in past (Qemu, uml etc.), but looks like it > >> may be > >>>>>>>> interesting to have it in containers. Given that they share they > >> same OS image > >>>>>>>> as host, I'm not sure if its possible without making the host OS > >> crash. > >>>>>>>> > >>>>>>>> Any suggestions. > >>>>>>>> > >>>>>>>> Thanks - > >>>>>>>> Manish > >>>>>>> > >>>>>>> I'm not sure what your goal is here, but it sounds to me like you > >> might want to be googling for "linux namespaces". > >>>>>> > >>>>>> I'm sorry. I see that my previous mail had a key word "development" > >>>>>> missing. What I was trying to find out was that is it possible to > >>>>>> setup and use linux containers/cgroups to do kernel development. > >>>>>> Things like writing and test kernel modules, debugging kernel, > >>>>>> attaching gdb etc. etc. which normally require Qemu, busybox or > other > >>>>>> VM techniques since containers are much lightweight and if anyone > has > >>>>>> recipes for that setup to do development without crashing the host > OS. > >>>>>> > >>>>>> Any hints appreciated. I looked for "linux namespaces" but it > doesn't > >>>>>> give me what I want. > >>>>>> > >>>>> > >>>>> I have no first hand experience, but I don't think containers have > the > >>>>> flexibility you need. Docker in particular shares the kernel with > the > >>>>> host OS as far as I know, so it would not be of any value that I can > >>>>> see. > >>>>> > >>>>> I think a unikernel may be what you are looking for: > >>>>> > >>>>> > >> > http://www.linux.com/news/enterprise/cloud-computing/821243-unikernel-use-c… > >>>>> > >>>>> My understanding is that with a unikernel you can move targeted > >>>>> portions of the Hypervisor kernel up into the unikernel. Thus if you > >>>>> wanted to work on a network driver, you could implement it in a > >>>>> unikernel. Then if it blew up you would fall back to the hypervisor > >>>>> level, kill the unikernel and try again. > >>>>> > >>>>> The Rump kernel (a unikernel) in particular might be a good option: > >>>>> > >>>>> == > >>>>> Rump Kernels — provide free, portable, componentized, kernel quality > >>>>> drivers such as file systems, POSIX system call handlers, PCI device > >>>>> drivers, a SCSI protocol stack, virtio and a TCP/IP stack. These > >>>>> drivers may be integrated into existing systems, or run as > stand-alone > >>>>> unikernels on cloud hypervisors and embedded systems. > >>>>> == > >>>>> > >>>>> I have no first hand experience with the Rump Kernel, so I don't know > >>>>> if it would work as a way to do linux kernel development or not. It > >>>>> certainly seems like a great environment for generic kernel > >>>>> development. > >>>> > >>>> Thanks a lot GregF/GregKH, > >>>> > >>>> That's what I suspected, that it might not be possible because of > >>>> both sharing the same OS. I'll have a look at Rump Kernels and see how > >>>> far it goes. Thanks for the pointers ! > >>>> > >>>> _______________________________________________ > >>>> Kernelnewbies mailing list > >>>> Kernelnewbies(a)kernelnewbies.org > >>>> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > >>> > >>> _______________________________________________ > >>> Kernelnewbies mailing list > >>> Kernelnewbies(a)kernelnewbies.org > >>> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > >>> > >> > > >

1 0

Kernel development using linux containers (LXC ) ?
by Manish Katiyar July 30, 2015

July 30, 2015

Hi, I've been playing and reading about control groups and linux containers recently and was wondering if there are any existing recipes on how to setup a kernel environment in containers. Google hasn't been helpful so far (or maybe I'm not searching properly). I've used VMs for dev in past (Qemu, uml etc.), but looks like it may be interesting to have it in containers. Given that they share they same OS image as host, I'm not sure if its possible without making the host OS crash. Any suggestions. Thanks - Manish

5 6

USB Driver - Device No Response standard compliant
by Christian N July 29, 2015

July 29, 2015

Hi All According to USB On The Go and Embedded Host Automated Compliacne Plan, item 6.7.22 A-UUT “Device No Response” for connection timeout, the "Device not responding” error message should be displayed when a non-responsive device is attached. I know that the presentation of the message is the responsibility of the application (user space), but to do this is necesary an event from usb driver (I think it). Nowdays when a non responsive device is attached, "unable to enumerate USB device on port... " is logged a in dmesg, but not event is dispatched (by example in usb_uevent). Without an event application should poll the log (loading the system unnecessarily) searching new "not responsive devices". To comply with the standard usb driver should be modified? I'm probably confusing responsibilities of user and kernel space. I appreciate your opinion/answer. Thanks in advance Christian

1 0

HOSTAPD Error log
by Amit Gupta July 29, 2015

July 29, 2015

Hi All, I am executing below mentioned command on linux kernel 4.0.4 and observing below mentioned error log. 'Failed to update rate sets in kernel module'. # hostapd -B /etc/hostapd.conf Configuration file: /etc/hostapd.conf [ 199.672712] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready Failed to update rate sets in kernel module [ 199.687566] IPv6: ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready Using interface wlan0 with hwaddr 00:0e:8e:38:29:e6 and ssid 'test_wifi_2' Eventhough I am getting this error log from user space application(hostapd) code, my wifi device is working fine. But I want to remove this error log and want to know the root cause this error log. Please let me know the approach to solve this issue. Thanks, Amit Gupta

3 4

how can I find where is the macro or function deifined
by Ahmed Soliman July 29, 2015

July 29, 2015

Hi; when I used libc I never memorized where each function is located beyond the basics (for example I never knew if strlen is in string.h or ctype.h or strlib.h ) I just check out the functions man pages by man strlen is their such functionality for the kernel where I dont need to memorize the dependencies for example if MODULE_LICENSE is in kernel.h or module.h with out using the internel to google them .

6 6