- kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

Writing a driver for the Corsair K90 keyboard
by Clément Vuchener Aug. 10, 2015

Aug. 10, 2015

Hello. I first sent this email to kernel-mentors, but I got no response. So I am sending it again here. I wrote a driver for the Corsair K90 keyboard, it is my first kernel development and I have a few questions about how to properly do a few things (mainly interactions with the user space). The Corsair Vengeance K90 is a mechanical keyboard with a backlight, macro keys (G1 to G18) and some other special keys (macro recording, profile switching, changing the light level or disabling the "windows logo" keys). The macro playback can be switched between hardware or software (from the driver not a key). Using the generic HID driver, the macro keys and special keys send wrong key code or no key code at all (because of the HID usage code used by the keyboard). The source for my driver can be found at this address: https://github.com/cvuchener/k90-linux-driver . It solves the problem of the generic driver and add some handling of the other features (see the README for more details). So, my questions are about: - Macro keys For handling the macro keys from a software, they need to send some keycodes but I have no idea what to use. Currently I use F keys above 12 and some MISC buttons and a parameter for changing that (at load time). Are there 18 keycodes reserved for custom uses? It would also be possible to send no keycode at all and use a specific software listening to the hidraw or hiddev devices, but I think it would be better to be able to use a generic software. - Backlight I chose to use a led class device to manage the backlight. It seems to work fine but when unregistering the driver, the light is set to 0. The keyboard remembers the previous light level and so nothing should be done when registering or unregistering the driver. Is the led class the right class for that? or am I using it wrong? - Sending the macro profiles to the keyboard I created a "profile" branch on my git repo for adding binary attributes used for sending macro data to the keyboard. But I am not sure it is a good idea, it may be better to leave that to the user space. Any other comment is welcome. Thanks, Clément

2 3

Re: What is my i2c device slave address?
by Woody Wu Aug. 10, 2015

Aug. 10, 2015

Hi Raul, Thanks for the manual. Now I believe the client address of this chip should be set to either 0x4a or 0x4b. The only problem left is how do I know what is the ADDR_SEL status? Since I see in the driver code there is no a line that is setting up this pin, so can I assume that the ADDR_SEL is hard-wired on my board and I just need to ask my hardware engineer to do a measure on the pin, right? Another not so related thing is that I also noticed that there are still other two addresses that can be used to do firmware upgrading. I just read this from the driver code and also want to know if it is necessary in my case to concern about downloading a firmware to the chip? I really hope things won't go such complicated and the chip should simplify work as long as I correctly setup and load the linux driver for it. Do you also has an idea about this? Many thanks. -woody On Saturday, 8 August 2015, Raul Piper <raulpblooper(a)gmail.com> wrote: > Hi Woody, > Form the attached datasheet section 4.3.2 the address seems to be : 0x4A > and 0x4B depending upon the ADDR_SEL line(low or high respectively). > Please check in the driver code is the address is 7 bit or 8 bit before > using the above address. > Also you can check the i2c utilities : > http://www.lm-sensors.org/wiki/I2CTools to check the address of the > devices connected with your SoC.on running i2c detect (with appropriate > arguments) you will come to know about the address of the devices which you > can verify with the address you put in the driver code. > Thanks and Rgds, > Rp > > On Sat, Aug 8, 2015 at 10:09 AM, Woody Wu <narkewoody(a)gmail.com > <javascript:_e(%7B%7D,'cvml','narkewoody(a)gmail.com');>> wrote: > >> I am recently porting an i2c touchscreen to my board which runs ARM linux >> 3.1. I found I need to setup some device platform data for the i2c chip, >> among which there is an 'addr' field that I dont understand. I search the >> device driver code, then I know the address will be used when compositing >> some i2c messages. My question is that how do I know what the address I >> should fill in the field? The datasheet of the device doesn't provide an >> answer. By the way, the touchscreen is actually an Atmel mXT224 i2c >> multi-touch screen panel, which is wired to a Samsung S3C i2c controller in >> my SoC. >> >> Thanks in advance! >> -woody >> >> >> -- >> Sent from Gmail Mobile >> >> _______________________________________________ >> Kernelnewbies mailing list >> Kernelnewbies(a)kernelnewbies.org >> <javascript:_e(%7B%7D,'cvml','Kernelnewbies(a)kernelnewbies.org');> >> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies >> >> > -- Sent from Gmail Mobile

2 2

[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

Unable to detect / control CPU fan
by Albert Coder Aug. 8, 2015

Aug. 8, 2015

Hi friends, By all means I have been unable to ensure that my CPU fan works well because it cannot be detected by /proc/cpuinfo, dmidecode, sensors, various ACPI fan scripts etc. My laptop sometimes hangs and then shuts down and then gets stuck at boot with black screen. But this happens sometimes. I am of the view that my CPU fan does not function properly and on detecting high temperature it shuts down. So far I have come to the conclusion that there is lack of support from ACPI on my Acer Aspire V5 571G. Another problem I face is my laptop gets stuck in shutdown-reboot loop be it Kali Linux, be it Ubuntu. Any idea to get around this folks? I am using Ubuntu 14.04.02 on Acer Aspire V5 571G. I have followed millions of links like: https://www.linux.com/learn/tutorials/620416:discovering-and-monitoring-har… https://help.ubuntu.com/community/AspireOne/AOA150 http://ubuntuforums.org/showthread.php?t=1612786 -- Thanks Albert Christ www.coderalbert.wordpress.com https://github.com/albertcoder/ || If I had eight hours to chop down a tree, I’d spend the first six of them sharpening my axe. ||

1 0

Checkpatch.pl warning
by Shraddha Barke Aug. 8, 2015

Aug. 8, 2015

I have come across the following checkpatch.pl warning a number of times: WARNING: Prefer ether_addr_equal() or ether_addr_equal_unaligned() over memcmp() What exactly does it mean? What is the fix for this warning? Any help or links to appropriate Documentation will be good. Thanks in advance!

2 1

What is my i2c device slave address?
by Woody Wu Aug. 8, 2015

Aug. 8, 2015

I am recently porting an i2c touchscreen to my board which runs ARM linux 3.1. I found I need to setup some device platform data for the i2c chip, among which there is an 'addr' field that I dont understand. I search the device driver code, then I know the address will be used when compositing some i2c messages. My question is that how do I know what the address I should fill in the field? The datasheet of the device doesn't provide an answer. By the way, the touchscreen is actually an Atmel mXT224 i2c multi-touch screen panel, which is wired to a Samsung S3C i2c controller in my SoC. Thanks in advance! -woody -- Sent from Gmail Mobile

2 1

Kernel options to boot root over NFS (kernel-4.1.fc22/linux-4.1.3-200.fc22.x86_64)
by Nestor Waldyd Aug. 7, 2015

Aug. 7, 2015

Hello, I would like to compile a kernel bzImage in order to boot root over NFS. I am using the fedora kernel source package (kernel-4.1.fc22/linux-4.1.3-200.fc22.x86_64). What kernel options should i modify in Kernel Configuration in order to do so? (I have already configured both tftp and dhcp server, and PXELINUX) Best Regards, Nestor Waldyd

3 2

Re: Make sure the searching header file path when building Linux kernel
by Nan Xiao Aug. 6, 2015

Aug. 6, 2015

"gcc -xc -E -v -" can show the include search paths. Best Regards Nan Xiao On Wed, Aug 5, 2015 at 6:34 PM, Abhilash Mhaisne <abhilashmhaisne(a)gmail.com> wrote: > By default gcc does search under /usr/include. I am not sure about > /usr/local/include though. > If header files at the top are #included as > "your_build_direc/include/some_header_file.h" then gcc will look into that > path. > Otherwise if it is #include <some_header_file.h> , it will search under > /usr/include and anything that is under C_INCLUDE_PATH variable. See man > page of gcc for more. > > Regards, > Abhilash Mhaisne > > On Wed, Aug 5, 2015 at 3:13 PM, Nan Xiao <xiaonan830818(a)gmail.com> wrote: > >> Hi all, >> >> I want to make sure the searching header file path when building Linux >> kernel: >> Per my understanding, the gcc only find header files in current building >> linux >> directory: include/*, won't search system header file path: such as >> /usr/inclue, >> /usr/local/include, etc. >> >> Is my understanding right? Thanks very much in advance! >> >> Best Regards >> Nan Xiao >> >> _______________________________________________ >> Kernelnewbies mailing list >> Kernelnewbies(a)kernelnewbies.org >> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies >> >> >

1 0

Requirement to customize Linux Network Interface naming Convention
by Akash Bhatnagar Aug. 5, 2015

Aug. 5, 2015

Hi, We have a requirement to tweak the Linux Network Interface Naming conventions so that, it can support names of interfaces with "/" character. For example, ip tuntap add dev fe/1/1 mode tap With iproute2 it simply shows error as INVALID Argument. What are the changes we supposed to do inside Kernel and iproute2 package? Thanks Akash Bhatnagar

2 3