Skip to contentSkip to conversation

firmware: sysfb: Unorphan and fix header

PATCH In review Andy Shevchenko 3 participants 13 messages Jun 26, 2025 – Jun 27, 2025 DRM driver for firmware framebuffers

lore
Share & tools
lore
#0 lore raw

cover firmware: sysfb: Unorphan and fix header

It appears that sysfb files became orphaned, reverse that to make the changes visible for the maintainers. Second patch just to drop kernel.h which is not supposed to be included (in the headers).

Andy Shevchenko (2):
firmware: sysfb: Unorphan sysfb files
firmware: sysfb: Don't use "proxy" headers

 MAINTAINERS           | 2 ++
 include/linux/sysfb.h | 6 +++++-
 2 files changed, 7 insertions(+), 1 deletion(-)
Signature · 1 line
2.47.2
#1 lore raw

PATCH 1/2 firmware: sysfb: Unorphan sysfb files

The commit d391c5827107 ("drivers/firmware: move x86 Generic System Framebuffers support") moved some code to the common folders and effectively orphaned it without any reason. Put it back under DRM MISC record.

---
 MAINTAINERS | 2 ++
 1 file changed, 2 insertions(+)
MAINTAINERS +2 -0
index 8dc0f6609d1f..c2c7aa594160 100644
@@ -7973,10 +7973,12 @@ T:	git https://gitlab.freedesktop.org/drm/misc/kernel.git
 F:	Documentation/devicetree/bindings/display/
 F:	Documentation/devicetree/bindings/gpu/
 F:	Documentation/gpu/
+F:	drivers/firmware/sysfb*.c
 F:	drivers/gpu/drm/
 F:	drivers/gpu/vga/
 F:	rust/kernel/drm/
 F:	include/drm/drm
+F:	include/linux/sysfb.h
 F:	include/linux/vga*
 F:	include/uapi/drm/
 X:	drivers/gpu/drm/amd/
Signature · 1 line
2.47.2
#2 lore raw

Re: [PATCH v1 1/2] firmware: sysfb: Unorphan sysfb files

Andy Shevchenko <[email protected]> writes:

Hello Andy,

Quoted 5 lines

The commit d391c5827107 ("drivers/firmware: move x86 Generic System Framebuffers support") moved some code to the common folders and effectively orphaned it without any reason. Put it back under DRM MISC record.

What do you mean that it was "orphaned without any reason" ? There were no regex matchs for the old file paths in MAINTAINERS either before that commit.

Quoted 10 lines
---
 MAINTAINERS | 2 ++
 1 file changed, 2 insertions(+)
MAINTAINERS +0 -0
index 8dc0f6609d1f..c2c7aa594160 100644

MAINTAINERS T: git https://gitlab.freedesktop.org/drm/misc/kernel.git · line 7976

Quoted 4 lines
MAINTAINERS +1 -0
@@ -7973,10 +7973,12 @@ T:	git https://gitlab.freedesktop.org/drm/misc/kernel.git
 F:	Documentation/devicetree/bindings/display/
 F:	Documentation/devicetree/bindings/gpu/
 F:	Documentation/gpu/
+F:	drivers/firmware/sysfb*.c

I would prefer these to be in the "DRM DRIVER FOR FIRMWARE FRAMEBUFFERS" entry instead of "DRM DRIVERS" since the former is what has most of the code for the sysfb infrastructure.

Signature · 5 lines
Best regards,

Javier Martinez Canillas
Core Platforms
Red Hat
#3 lore raw

Re: [PATCH v1 1/2] firmware: sysfb: Unorphan sysfb files

Quoted 9 lines · Javier Martinez Canillas
Quoted 4 lines · Andy Shevchenko

The commit d391c5827107 ("drivers/firmware: move x86 Generic System Framebuffers support") moved some code to the common folders and effectively orphaned it without any reason. Put it back under DRM MISC record.

What do you mean that it was "orphaned without any reason" ? There were no regex matchs for the old file paths in MAINTAINERS either before that commit.

There were maintainers for that code before. The change in question dropped that.

[((bf44e8cecc03...))]$ scripts/get_maintainer.pl -f arch/x86/kernel/sysfb* Thomas Gleixner <[email protected]> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) Ingo Molnar <[email protected]> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) Borislav Petkov <[email protected]> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) [email protected] (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT))
"H. Peter Anvin" <[email protected]> (reviewer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) [email protected] (open list:X86 ARCHITECTURE (32-BIT AND 64-BIT))

[((d391c5827107...))]$ scripts/get_maintainer.pl -f drivers/firmware/sysfb* [email protected] (open list)

See the difference?

...

Quoted 5 lines
Quoted 1 line

+F: drivers/firmware/sysfb*.c

I would prefer these to be in the "DRM DRIVER FOR FIRMWARE FRAMEBUFFERS" entry instead of "DRM DRIVERS" since the former is what has most of the code for the sysfb infrastructure.

Then do it, please, fix the above.

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#4 lore raw

Re: [PATCH v1 1/2] firmware: sysfb: Unorphan sysfb files

Quoted 21 lines · Andy Shevchenko
Quoted 9 lines · Javier Martinez Canillas
Quoted 4 lines · Andy Shevchenko

The commit d391c5827107 ("drivers/firmware: move x86 Generic System Framebuffers support") moved some code to the common folders and effectively orphaned it without any reason. Put it back under DRM MISC record.

What do you mean that it was "orphaned without any reason" ? There were no regex matchs for the old file paths in MAINTAINERS either before that commit.

There were maintainers for that code before. The change in question dropped that.

[((bf44e8cecc03...))]$ scripts/get_maintainer.pl -f arch/x86/kernel/sysfb* Thomas Gleixner <[email protected]> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) Ingo Molnar <[email protected]> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) Borislav Petkov <[email protected]> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) [email protected] (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT))
"H. Peter Anvin" <[email protected]> (reviewer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) [email protected] (open list:X86 ARCHITECTURE (32-BIT AND 64-BIT))

Quoted 3 lines

[((d391c5827107...))]$ scripts/get_maintainer.pl -f drivers/firmware/sysfb* [email protected] (open list)

That's just because there is an entry for arch/x86/. The problem then is that there isn't an entry for drivers/firmware. It was orphaned then just because it was moved to a directory that has no entry in MAINTAINERS.

Quoted 2 lines

See the difference?

There is no need to have such a rude tone.

Quoted 10 lines

...

Quoted 5 lines
Quoted 1 line

+F: drivers/firmware/sysfb*.c

I would prefer these to be in the "DRM DRIVER FOR FIRMWARE FRAMEBUFFERS" entry instead of "DRM DRIVERS" since the former is what has most of the code for the sysfb infrastructure.

Then do it, please, fix the above.

Part of the review process is to give feedback to patch authors. I don't understand why you expect me to fix an issue you brought up just because I ask you to rework your patch a little.

Quoted 5 lines
Signature · 2 lines
With Best Regards,
Andy Shevchenko
Signature · 5 lines
Best regards,

Javier Martinez Canillas
Core Platforms
Red Hat
#5 lore raw

Re: [PATCH v1 1/2] firmware: sysfb: Unorphan sysfb files

Quoted 25 lines · Javier Martinez Canillas
Quoted 21 lines · Andy Shevchenko
Quoted 9 lines · Javier Martinez Canillas
Quoted 4 lines · Andy Shevchenko

The commit d391c5827107 ("drivers/firmware: move x86 Generic System Framebuffers support") moved some code to the common folders and effectively orphaned it without any reason. Put it back under DRM MISC record.

What do you mean that it was "orphaned without any reason" ? There were no regex matchs for the old file paths in MAINTAINERS either before that commit.

There were maintainers for that code before. The change in question dropped that.

[((bf44e8cecc03...))]$ scripts/get_maintainer.pl -f arch/x86/kernel/sysfb* Thomas Gleixner <[email protected]> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) Ingo Molnar <[email protected]> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) Borislav Petkov <[email protected]> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) [email protected] (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT))
"H. Peter Anvin" <[email protected]> (reviewer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) [email protected] (open list:X86 ARCHITECTURE (32-BIT AND 64-BIT))

Quoted 2 lines

[((d391c5827107...))]$ scripts/get_maintainer.pl -f drivers/firmware/sysfb* [email protected] (open list)

Quoted 7 lines

That's just because there is an entry for arch/x86/. The problem then is that there isn't an entry for drivers/firmware. It was orphaned then just because it was moved to a directory that has no entry in MAINTAINERS.

Quoted 1 line

See the difference?

There is no need to have such a rude tone.

It wasn't meant to be rude, sorry. The point is that any change in drivers/firmware/sysfb* and respective include are not visible to (any) maintainers, they just might be sent for a luck of somebody to pick them up by browsing the LKML for such things.

...

Quoted 11 lines
Quoted 7 lines
Quoted 5 lines
Quoted 1 line

+F: drivers/firmware/sysfb*.c

I would prefer these to be in the "DRM DRIVER FOR FIRMWARE FRAMEBUFFERS" entry instead of "DRM DRIVERS" since the former is what has most of the code for the sysfb infrastructure.

Then do it, please, fix the above.

Part of the review process is to give feedback to patch authors. I don't understand why you expect me to fix an issue you brought up just because I ask you to rework your patch a little.

In my humble opinion, the author of the patch that makes the problem appear can help to fix that as well. Are my expectations too high?

In any case, this was an ad-hoc patch due to the second one, so this one may be considered as a administrative bug report.

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#6 lore raw

Re: [PATCH v1 1/2] firmware: sysfb: Unorphan sysfb files

Andy Shevchenko <[email protected]> writes:

[...]

Quoted 10 lines
Quoted 8 lines

That's just because there is an entry for arch/x86/. The problem then is that there isn't an entry for drivers/firmware. It was orphaned then just because it was moved to a directory that has no entry in MAINTAINERS.

Quoted 1 line

See the difference?

There is no need to have such a rude tone.

It wasn't meant to be rude, sorry. The point is that any change in

No worries.

Quoted 4 lines

drivers/firmware/sysfb* and respective include are not visible to (any) maintainers, they just might be sent for a luck of somebody to pick them up by browsing the LKML for such things.

Right. But get_maintainer.pl still reports I think the correct people to Cc:

./scripts/get_maintainer.pl -f -- drivers/firmware/sysfb*
Thomas Zimmermann <[email protected]> (commit_signer:4/4=100%,authored:2/4=50%,added_lines:11/43=26%,removed_lines:5/11=45%,commit_signer:1/1=100%,authored:1/1=100%,added_lines:1/1=100%,removed_lines:30/30=100%) Javier Martinez Canillas <[email protected]> (commit_signer:4/4=100%,authored:1/4=25%,added_lines:19/43=44%,commit_signer:1/1=100%)
Alex Deucher <[email protected]> (commit_signer:3/4=75%,authored:1/4=25%,added_lines:13/43=30%,removed_lines:6/11=55%)
Tzung-Bi Shih <[email protected]> (commit_signer:1/4=25%)
[email protected] (open list)

In my opinion both Thomas and me have much more context and knowledge of the sysfb codebase than the x86 maintainers. It was just for historical reasons that the sysfb code ended in the arch/x86/ sub-directory.

But you are correct that dri-devel at least should also be in the Cc list.

Quoted 14 lines

...

Quoted 11 lines
Quoted 7 lines
Quoted 5 lines
Quoted 1 line

+F: drivers/firmware/sysfb*.c

I would prefer these to be in the "DRM DRIVER FOR FIRMWARE FRAMEBUFFERS" entry instead of "DRM DRIVERS" since the former is what has most of the code for the sysfb infrastructure.

Then do it, please, fix the above.

Part of the review process is to give feedback to patch authors. I don't understand why you expect me to fix an issue you brought up just because I ask you to rework your patch a little.

Quoted 6 lines

In my humble opinion, the author of the patch that makes the problem appear can help to fix that as well. Are my expectations too high?

In any case, this was an ad-hoc patch due to the second one, so this one may be considered as a administrative bug report.

That's OK, but it wasn't framed as a bug report but as a patch and that's why I gave my feedback. But I'll post a patch and add a Reported-by tag from you.

Thomas, I think we can then only merge patch #2 and I will take care of #1.

Signature · 5 lines
Best regards,

Javier Martinez Canillas
Core Platforms
Red Hat
#7 lore raw

Re: [PATCH v1 1/2] firmware: sysfb: Unorphan sysfb files

Quoted 1 line · Javier Martinez Canillas

Andy Shevchenko <[email protected]> writes:

[...]

Quoted 17 lines
Quoted 9 lines
Quoted 7 lines

That's just because there is an entry for arch/x86/. The problem then is that there isn't an entry for drivers/firmware. It was orphaned then just because it was moved to a directory that has no entry in MAINTAINERS.

Quoted 1 line

See the difference?

There is no need to have such a rude tone.

It wasn't meant to be rude, sorry. The point is that any change in

No worries.

Quoted 4 lines

drivers/firmware/sysfb* and respective include are not visible to (any) maintainers, they just might be sent for a luck of somebody to pick them up by browsing the LKML for such things.

Quoted 8 lines

Right. But get_maintainer.pl still reports I think the correct people to Cc:

./scripts/get_maintainer.pl -f -- drivers/firmware/sysfb*
Thomas Zimmermann <[email protected]> (commit_signer:4/4=100%,authored:2/4=50%,added_lines:11/43=26%,removed_lines:5/11=45%,commit_signer:1/1=100%,authored:1/1=100%,added_lines:1/1=100%,removed_lines:30/30=100%) Javier Martinez Canillas <[email protected]> (commit_signer:4/4=100%,authored:1/4=25%,added_lines:19/43=44%,commit_signer:1/1=100%)
Alex Deucher <[email protected]> (commit_signer:3/4=75%,authored:1/4=25%,added_lines:13/43=30%,removed_lines:6/11=55%)
Tzung-Bi Shih <[email protected]> (commit_signer:1/4=25%)
[email protected] (open list)

The problem is deeper. This is default behaviour of get_maintainer digging into the Git history. In the past people were complaining (a lot in some cases) that they were included in the threads by a mistake because they made cosmetic patches or the treewide change while not being interested _at all_ in looking after the certain file / driver. So, with --no-git-fallback it gives nothing, expect LKML.

Quoted 5 lines

In my opinion both Thomas and me have much more context and knowledge of the sysfb codebase than the x86 maintainers. It was just for historical reasons that the sysfb code ended in the arch/x86/ sub-directory.

But you are correct that dri-devel at least should also be in the Cc list.

See also above. Depending on the options it may still give bad result w.o. explicit mention in the MAINTAINERS (or via glob).

...

Quoted 12 lines
Quoted 12 lines
Quoted 11 lines
Quoted 7 lines
Quoted 5 lines
Quoted 1 line

+F: drivers/firmware/sysfb*.c

I would prefer these to be in the "DRM DRIVER FOR FIRMWARE FRAMEBUFFERS" entry instead of "DRM DRIVERS" since the former is what has most of the code for the sysfb infrastructure.

Then do it, please, fix the above.

Part of the review process is to give feedback to patch authors. I don't understand why you expect me to fix an issue you brought up just because I ask you to rework your patch a little.

Quoted 8 lines
Quoted 5 lines

In my humble opinion, the author of the patch that makes the problem appear can help to fix that as well. Are my expectations too high?

In any case, this was an ad-hoc patch due to the second one, so this one may be considered as a administrative bug report.

That's OK, but it wasn't framed as a bug report but as a patch and that's why I gave my feedback. But I'll post a patch and add a Reported-by tag from you.

Sure, thanks ahead!

Quoted 1 line

Thomas, I think we can then only merge patch #2 and I will take care of #1.

I just sent a v2 without the first patch.

Thanks for review!

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#8 lore raw

Re: [PATCH v1 1/2] firmware: sysfb: Unorphan sysfb files

Quoted 1 line · Andy Shevchenko

On Fri, Jun 27, 2025 at 12:33:31PM +0200, Javier Martinez Canillas wrote:

[...]

Quoted 8 lines

The problem is deeper. This is default behaviour of get_maintainer digging into the Git history. In the past people were complaining (a lot in some cases) that they were included in the threads by a mistake because they made cosmetic patches or the treewide change while not being interested _at all_ in looking after the certain file / driver. So, with --no-git-fallback it gives nothing, expect LKML.

Quoted 9 lines
Quoted 5 lines

In my opinion both Thomas and me have much more context and knowledge of the sysfb codebase than the x86 maintainers. It was just for historical reasons that the sysfb code ended in the arch/x86/ sub-directory.

But you are correct that dri-devel at least should also be in the Cc list.

See also above. Depending on the options it may still give bad result w.o. explicit mention in the MAINTAINERS (or via glob).

Sure, I was not saying that is not worth it.

Quoted 17 lines

...

Quoted 15 lines
Quoted 15 lines
Quoted 11 lines
Quoted 7 lines
Quoted 5 lines
Quoted 1 line

+F: drivers/firmware/sysfb*.c

I would prefer these to be in the "DRM DRIVER FOR FIRMWARE FRAMEBUFFERS" entry instead of "DRM DRIVERS" since the former is what has most of the code for the sysfb infrastructure.

Then do it, please, fix the above.

Part of the review process is to give feedback to patch authors. I don't understand why you expect me to fix an issue you brought up just because I ask you to rework your patch a little.

In my humble opinion, the author of the patch that makes the problem appear can help to fix that as well. Are my expectations too high?

Quoted 8 lines
Quoted 5 lines
Quoted 2 lines

In any case, this was an ad-hoc patch due to the second one, so this one may be considered as a administrative bug report.

That's OK, but it wasn't framed as a bug report but as a patch and that's why I gave my feedback. But I'll post a patch and add a Reported-by tag from you.

Sure, thanks ahead!

Patch sent:

https://lore.kernel.org/dri-devel/[email protected]/T/#u

Quoted 6 lines
Quoted 1 line

Thomas, I think we can then only merge patch #2 and I will take care of #1.

I just sent a v2 without the first patch.

Thanks for review!

You are welcome.

Quoted 5 lines
Signature · 2 lines
With Best Regards,
Andy Shevchenko
Signature · 5 lines
Best regards,

Javier Martinez Canillas
Core Platforms
Red Hat
#9 lore raw

PATCH 2/2 firmware: sysfb: Don't use "proxy" headers

In reply to Andy Shevchenko

Update header inclusions to follow IWYU (Include What You Use) principle.

Note that kernel.h is discouraged to be included as it's written at the top of that file.

---
 include/linux/sysfb.h | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
include/linux/sysfb.h +5 -1
index 07cbab516942..b449665c686a 100644
@@ -7,9 +7,13 @@
  * Copyright (c) 2012-2013 David Herrmann <[email protected]>
  */
 
-#include <linux/kernel.h>
+#include <linux/err.h>
+#include <linux/types.h>
+
 #include <linux/platform_data/simplefb.h>
 
+struct device;
+struct platform_device;
 struct screen_info;
 
 enum {
Signature · 1 line
2.47.2
maintainer Reviewed-by on v1 2/2
#10 lore raw

Re: [PATCH v1 2/2] firmware: sysfb: Don't use "proxy" headers

Quoted 8 lines · Andy Shevchenko

Update header inclusions to follow IWYU (Include What You Use) principle.

Note that kernel.h is discouraged to be included as it's written at the top of that file.

---

Signature · 5 lines
Best regards,

Javier Martinez Canillas
Core Platforms
Red Hat
maintainer Reviewed-by on the v1 series
#12 lore raw
In reply to Andy Shevchenko

Hi,

for the series

Best regards
Thomas

Am 26.06.25 um 19:18 schrieb Andy Shevchenko:

Quoted 12 lines

It appears that sysfb files became orphaned, reverse that to make the changes visible for the maintainers. Second patch just to drop kernel.h which is not supposed to be included (in the headers).

Andy Shevchenko (2):
firmware: sysfb: Unorphan sysfb files
firmware: sysfb: Don't use "proxy" headers

  MAINTAINERS           | 2 ++
  include/linux/sysfb.h | 6 +++++-
  2 files changed, 7 insertions(+), 1 deletion(-)
Signature · 7 lines
--
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Frankenstrasse 146, 90461 Nuernberg, Germany
GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman
HRB 36809 (AG Nuernberg)