Add support for kernel 6.0 + a minor fix

[SimPilotAdamT]

Fixes https://github.com/DisplayLink/evdi/issues/376 by implementing @Crashdummyy's fix as posted here, + fixes the style used in evdi_painter.c

Before submitting the PR please make sure you have run ci scripts:

• ./ci/build_against_kernel (see --help for all options) (done) We need backward compatibility and this script is a handy way of testing build compliance with many kernel versions.
• ./ci/run_style_check (done) We want to be as style compliant as possible. Again, this is a handy way of checking it.

If you have more than one change consider creating separate PRs for them; it will make the review process much easier. Also provide some description (links to source code or mailing list are welcome - this might help to understand the change).

Thanks for the contribution!

[SimPilotAdamT]

This works for me. I am by no means a C programmer, but I think there needs to be a way to make sure the #include <drm/drm_framebuffer.h> line added to module/evdi_drm_drv.h is only done on Kernel versions 6.0.0 and newer. I have no clue how to do this... What about you guys?

[Crashdummyy]

This works for me. I am by no means a C programmer, but I think there needs to be a way to make sure the #include <drm/drm_framebuffer.h> line added to module/evdi_drm_drv.h is only done on Kernel versions 6.0.0 and newer. I have no clue how to do this... What about you guys?

Well of course it would break on every kernel < 6.0. You need some compiler flags ( look at the lines 28 through 32 ).

I think you need to do something like this

#if KERNEL_VERSION(6, 0, 0) >= LINUX_VERSION_CODE
# All kernel versions greather than 6.0.0 use the framebuffer of drm_framebuffer.h with the install method migrated to drm_legacy.h
#include <drm/drm_legacy.h>
#include <drm/drm_framebuffer.h>
#else
# Below 6.0.0 the framebuffer struct still resides in drm_irq.h
#include <drm/drm_irq.h>
#endif

Sorry Im on the phone right now cant test ist

[SimPilotAdamT]

Oh I think ik how to do it, commit coming soon

[SimPilotAdamT]

OK that's done... I do have an issue using the CI build script checking the build on Kernel versions older than 5.17...

Here is the error, which really stems from a warning using -Werror. This was run using Kernel version 5.16, but the same issue is present when I try on any other older version.

  CC      /home/adam/evdi/tmp/linux-5.16/tools/objtool/help.o
In file included from help.c:12:
In function ‘xrealloc’,
    inlined from ‘add_cmdname’ at help.c:24:2:
subcmd-util.h:56:23: error: pointer may be used after ‘realloc’ [-Werror=use-after-free]
   56 |                 ret = realloc(ptr, size);
      |                       ^~~~~~~~~~~~~~~~~~
subcmd-util.h:52:21: note: call to ‘realloc’ here
   52 |         void *ret = realloc(ptr, size);
      |                     ^~~~~~~~~~~~~~~~~~
subcmd-util.h:58:31: error: pointer may be used after ‘realloc’ [-Werror=use-after-free]
   58 |                         ret = realloc(ptr, 1);
      |                               ^~~~~~~~~~~~~~~
subcmd-util.h:52:21: note: call to ‘realloc’ here
   52 |         void *ret = realloc(ptr, size);
      |                     ^~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
make[4]: *** [/home/adam/evdi/tmp/linux-5.16/tools/build/Makefile.build:97: /home/adam/evdi/tmp/linux-5.16/tools/objtool/help.o] Error 1
make[3]: *** [Makefile:59: /home/adam/evdi/tmp/linux-5.16/tools/objtool/libsubcmd-in.o] Error 2
make[2]: *** [Makefile:63: /home/adam/evdi/tmp/linux-5.16/tools/objtool/libsubcmd.a] Error 2
make[1]: *** [Makefile:69: objtool] Error 2
make: *** [Makefile:1349: tools/objtool] Error 2

[SimPilotAdamT]

(Changed again bc I'm so dumb and couldn't fathom >= and <=. The error still happens.)

[SimPilotAdamT]

ok now it should properly work, don't quote me on that

[displaylink-emajewsk]

Thanks for the PR, looks good, but let me verify it

[Crashdummyy]

Thanks for the PR, looks good, but let me verify it

looks good to me but you should check if the dependency to drm_legacy.h is still required. The migrated methods were drm_irq_uninstall and drm_irq_install I dont see them being used in the implementation so maybe its not needed at all

[SimPilotAdamT]

Is anyone able to help me test smth? The CI build script passes this branch on my repo on kernel versions 5.16 and newer, but I'm not sure if it will actually install the module properly... I don't have a machine I can test on myself

[displaylink-emajewsk]

Is anyone able to help me test smth? The CI build script passes this branch on my repo on kernel versions 5.16 and newer, but I'm not sure if it will actually install the module properly... I don't have a machine I can test on myself

Seems fine to me. If you merge those branches, we can pull it in. We'll do a thorough testing before a release :)

[SimPilotAdamT]

Seems fine to me. If you merge those branches, we can pull it in. We'll do a thorough testing before a release :)

Merged

[dimitarsh1]

I just updated to kernel 6 on Debian testing

I get the following:

$ sudo dkms install -m evdi -v 1.12.0
Sign command: /usr/lib/linux-kbuild-6.0/scripts/sign-file
Signing key: /var/lib/dkms/mok.key
Public certificate (MOK): /var/lib/dkms/mok.pub

Building module:
Cleaning build area...
make -j8 KERNELRELEASE=6.0.0-2-amd64 all INCLUDEDIR=/lib/modules/6.0.0-2-amd64/build/include KVERSION=6.0.0-2-amd64 DKMS_BUILD=1...(bad exit status: 2)
Error! Bad return status for module build on kernel: 6.0.0-2-amd64 (x86_64)
Consult /var/lib/dkms/evdi/1.12.0/build/make.log for more information.

I have pulled the latest devel update.

Any suggestions

[SimPilotAdamT]

Any suggestions

Can you send that make.log file?

[dimitarsh1]

Yup, here it is.

make.log

[Crashdummyy]

Yup, here it is.

make.log

error: field ‘base’ has incomplete type Are you sure youre on the latest devel ? Because that is the exact error that is thrown before the fix.

Further please dont use a ancient distro

[dimitarsh1]

$ git status
On branch devel
Your branch is up to date with 'origin/devel'.

nothing to commit, working tree clean
$ git pull
Already up to date.

I don't get the last comment ;)

[SimPilotAdamT]

@dimitarsh1 it might be a good idea to make another issue... Might be a bug in this PR that wasn't spotted. I'd ignore the last comment.

[Crashdummyy]

$ git status
On branch devel
Your branch is up to date with 'origin/devel'.

nothing to commit, working tree clean
$ git pull
Already up to date.

I don't get the last comment ;)

Mhm that looks okay to me. I dont have my laptop with me ( vacation ) so I can test it. But someone here might try to compile it using distrobox or docker.

Regarding my last comment: The Problem with old distros ( eg Debian ) are old packages and an old kernel as well. Youll hit a lot of these obstacles if you run Debian on a machine of this decade and youll make your life a lot easier if you just install fedora or arch on your device. Im pretty sure the reason you cant compile is that either your compiler itself or some dependencies evdi relies on are too old on your system.

[dimitarsh1]

Dear all,

After updating my system it is all fine now. I assume some other libraries had to be updated...

Thanks for all the support.

Kind regards, Dimitar

[Skesov]

When you push tag on these changes?