Re: [GIT PULL] Ressurect the atomisp staging driver

[Patrik Gfeller <patrik.gfeller@gmail.com>]

On Mon, 4 May 2020 10:49:34 +0200
Mauro Carvalho Chehab <mchehab@kernel.org> wrote:

> Em Mon, 4 May 2020 10:16:28 +0200
> Patrik Gfeller <patrik.gfeller@gmail.com> escreveu:
> 
> > On Sun, 3 May 2020 18:07:51 +0200
> > Patrik Gfeller <patrik.gfeller@gmail.com> wrote:
> >    
>  [...]  
> > 
> > I was not able yet to compile the kernel 4.4 successfully yet.   
> 
> Yesterday, I imported only a subset of the yocto patches. I imported today
> all 24 patches from it, as some patches depend on hacks on other upstream
> stuff. You'll need to do a:
> 
> 	$ git remote update && git reset --hard origin/yocto_intel_atom
> 
> To update your local branch.
> 
> > I get quite
> > some warnings that are treated as errors. I assume that I can "resolve"
> > this by removing the respective compiler flag (-Werror).  
> 
> Yeah, you may need to to that. There's a Kernel option with disables
> -Werror. You may also change the Makefile to add things like:
> 
> 	# HACK! While this driver is in bad shape, don't enable several warnings
> 	#       that would be otherwise enabled with W=1
> 	ccflags-y += $(call cc-disable-warning, implicit-fallthrough)
> 	ccflags-y += $(call cc-disable-warning, missing-prototypes)
> 	ccflags-y += $(call cc-disable-warning, missing-declarations)
> 	ccflags-y += $(call cc-disable-warning, suggest-attribute=format)
> 	ccflags-y += $(call cc-disable-warning, unused-const-variable)
> 	ccflags-y += $(call cc-disable-warning, unused-but-set-variable)
> 
> (this is what we do with the upstream version, as this driver is on
> bad shape)
> 
> > But there are a few things that will take more time for me to investigate:
> > 
> > drivers/media/pci/atomisp//css2401a0_v21_build/../atomisp_driver/atomisp_cmd.c:36:10: fatal error: linux/kct.h: No such file or directory
> > include/media/v4l2-subdev.h:821:51: error: ‘const struct v4l2_subdev_core_ops’ has no member named ‘g_chip_ident’
> > include/media/v4l2-subdev.h:822:15: error: ‘const struct v4l2_subdev_core_ops’ has no member named ‘g_chip_ident’

There are still errors when I try to build. I've found code like the
one below:

const struct v4l2_ioctl_ops atomisp_ioctl_ops = {
	.vidioc_querycap = atomisp_querycap,
#ifndef CONFIG_INTEL_MID_ISP
	.vidioc_g_chip_ident = atomisp_g_chip_ident,
#endif
	.vidioc_enum_input = atomisp_enum_input,

Therefore I've tried with "CONFIG_INTEL_MID_ISP=y". Not sure if this is
a good idea function wise - but it compiles. But the last step of the
module build produces some scary warnings:

WARNING: "cpu_tss" [drivers/media/pci/atomisp//css2401a0_v21_build/atomisp-css2401a0_v21.ko] undefined!
WARNING: "pv_lock_ops" [drivers/media/pci/atomisp//css2401a0_v21_build/atomisp-css2401a0_v21.ko] undefined!
WARNING: "__alloc_workqueue_key" [drivers/media/pci/atomisp//css2401a0_v21_build/atomisp-css2401a0_v21.ko] undefined!
WARNING: "v4l2_subdev_querymenu" [drivers/media/pci/atomisp//css2401a0_v21_build/atomisp-css2401a0_v21.ko] undefined!
WARNING: "intel_mid_msgbus_write32" [drivers/media/pci/atomisp//css2401a0_v21_build/atomisp-css2401a0_v21.ko] undefined!
...

Maybe this is related to my partial build (only the module to test if
it builds). I'll now do a full build of the kernel.

> > 
> > I'll keep you updated.  
> 
> Maybe now with the other patches applied, this could have been solved.
> 
> Thanks,
> Mauro

with kind regards,
Patrik