On 3/9/25 05:10, Michael Kelley wrote:

From: Helge Deller <deller@gmx.de> Sent: Saturday, March 8, 2025 6:59 PM

quoted

On 2/19/25 00:01, mhkelley58@gmail.com wrote:

quoted

From: Michael Kelley <redacted>

[snip]

quoted

quoted

Reported-by: Thomas Tai <redacted>
Fixes: c25a19afb81c ("fbdev/hyperv_fb: Do not clear global screen_info")
Signed-off-by: Michael Kelley <redacted>
---
The "Fixes" tag uses commit c25a19afb81c because that's where the problem
was re-exposed, and how far back a stable backport is needed. But I've
taken a completely different, and hopefully better, approach in the
solution that isn't related to the code changes in c25a19afb81c.

   drivers/video/fbdev/hyperv_fb.c | 20 +++++++++++++-------
   1 file changed, 13 insertions(+), 7 deletions(-)

applied to fbdev tree.

Thank you!

Related, I noticed the patch "fbdev: hyperv_fb: iounmap() the correct
memory when removing a device" is also in the fbdev for-next branch.
Wei Liu previously applied this patch to the hyperv-fixes tree (see [1])
and it's already in linux-next. Won't having it also in fbdev produce a
merge conflict?
[1] https://lore.kernel.org/linux-hyperv/Z6wHDw8BssJyQHiM@liuwe-devbox-debian-v2/ (local)

Thanks Michael!
I now dropped that patch from the fbdev tree to avoid collisions.

Btw, I'm fine if we agree that all hyperv-fbdev fixes & patches go through
hyperv or other trees. Just let me know.

Helge