[00:50] *** BrianG61UK has joined #linux-media [01:54] *** lexano has quit IRC (Ping timeout: 480 seconds) [02:58] *** larunbe has joined #linux-media [03:03] *** alarumbe has quit IRC (Ping timeout: 480 seconds) [03:03] *** mrpops2ko_ has quit IRC () [03:04] *** mrpops2ko has joined #linux-media [03:59] *** eelstrebor has quit IRC (Quit: Ex-Chat) [06:12] *** GBenji has joined #linux-media [06:18] *** djrscally has joined #linux-media [06:46] *** ao2 has joined #linux-media [07:01] *** rauji____ has quit IRC () [07:36] *** frieder has joined #linux-media [08:02] *** mripard has joined #linux-media [08:15] *** frieder has quit IRC (Ping timeout: 480 seconds) [08:25] *** frieder has joined #linux-media [08:31] *** frieder has quit IRC (Quit: Leaving) [08:31] *** frieder has joined #linux-media [09:37] *** hansg has joined #linux-media [11:01] *** lexano has joined #linux-media [11:20] *** hansg has quit IRC (Quit: Leaving) [11:25] <bemug> how is a driver chosen to be builded by default in the kernel ? [11:28] <javierm> bemug: you mean when there should be a "default y" directive in a Kconfig symbol definition or ? [11:28] <bemug> javierm exactly, or "default m" [11:28] <bemug> is there a word for this by the way ? [11:29] <bemug> "defaulting" ? [11:34] <javierm> bemug: https://www.kernel.org/doc/html/next/kbuild/kconfig-language.html has some words about this [11:34] <javierm> bemug: I never understood the value on default m because AFAIK default y will also set as a module if some of the Kconfig dependencies are enabled as a module [11:35] <javierm> *value of [11:45] <bemug> javierm thank you for your link I just read it. As I understand in Kconfig we can put "default y" for something that is supposed to be used a lot in the kernel, makes sense. [11:46] <bemug> javierm that makes me think adding a driver as default is done via the defconfig. Are there requirements to add "CONFIG_MY_DRIVER=y" in a defconfig ? [11:47] <javierm> bemug: yes, in general for drivers I think that the "defualt {y,m}" is not usually needed [11:47] <javierm> distros will decide whether they want to enable it or not in their kernel configs [11:48] <javierm> and whether they will built-in or build as a module [11:48] <bemug> javierm makes sense. Thank you for your answers [11:49] <javierm> there may be exceptions, for example clock drivers that could be required for a platform, it makes sense to have it "default y" I guess [11:49] <javierm> bemug: but media drivers probably don't fall into that category. This is just my opinion though, not sure what others think :) [12:11] *** hansg has joined #linux-media [12:20] *** hansg has quit IRC (Quit: Leaving) [12:26] *** frieder has quit IRC (Ping timeout: 480 seconds) [12:43] *** ten15723774324630506 has quit IRC (Remote host closed the connection) [12:44] *** ten15723774324630506 has joined #linux-media [12:55] *** frieder has joined #linux-media [14:22] <tmerciai> Hi all, I have a question: for the alvium-csi2 driver I need to implements 2 custom ioctl functions to transfer objects of several sizes between userspace and the alvium. Right now I'm taking as reference ADV7842_CMD_RAM_TEST. This could be a valid starting point or you have some others reference function to suggest? Thanks in advance. [14:22] <pinchartl> tmerciai: what are those "objects" ? [14:23] <tmerciai> piece of fw update [14:24] <pinchartl> then I recommend using the firmware update API [14:24] <tmerciai> https://github.com/alliedvision/alvium-csi2-driver/blob/1e7a87962acbe0a39211ed76e785fb9831897de3/avt-csi2-3.c#L4636 [14:24] <pinchartl> is it a firmware that is stored in the camera ? [14:24] <pinchartl> or a firmware that is loaded on every boot ? [14:24] <tmerciai> yes fw is stored into the camera and right now alliedvision is using the following: VIDIOC_R_I2C, VIDIOC_W_I2C [14:25] <tmerciai> They needs that because the userspace tools need these ioctl [14:26] <pinchartl> look at the thp7312.c driver to see how it handles firmware updates [14:26] <pinchartl> they will need to change their userspace tool [14:26] <tmerciai> :'( [14:26] <pinchartl> is the tool open-source ? [14:27] *** frieder has quit IRC (Remote host closed the connection) [14:28] <tmerciai> no fw update is not open source. This is based on vimba. [14:29] <tmerciai> for that I think firmware update API not fit my case. [14:29] <tmerciai> *don't [14:30] <pinchartl> then you won't be able to upstream this [14:30] <pinchartl> we don't allow vendor-specific undocumented APIs used by closed-source tools only [14:30] <tmerciai> Oks thanks for the clarification. [14:30] <pinchartl> if they don't want to document how firmware update works, to enable an open-source implementation, then this won't be accepted [14:47] <tmerciai> Oks, thanks again. :) [15:38] *** rz has quit IRC (Remote host closed the connection) [15:38] *** rz has joined #linux-media [15:50] *** GBenji has left [16:43] *** BrianG61UK has quit IRC (Read error: Connection reset by peer) [18:13] *** gouchi has joined #linux-media [18:18] *** eelstrebor has joined #linux-media [20:08] *** gouchi has quit IRC (Remote host closed the connection) [20:45] *** BrianG61UK has joined #linux-media [21:31] *** ao2 has quit IRC (Quit: Leaving) [22:49] *** djrscally has quit IRC (Ping timeout: 480 seconds)