Message ID | 20220124210319.325512-1-deller@gmx.de (mailing list archive) |
---|---|
Headers | show |
Series | Fix regression introduced by disabling accelerated scrolling in fbcon | expand |
On Mon, Jan 24, 2022 at 10:03:17PM +0100, Helge Deller wrote: > This series reverts two patches which disabled scrolling acceleration in > fbcon/fbdev. Those patches introduced a regression for fbdev-supported graphic > cards because of the performance penalty by doing screen scrolling by software > instead of using existing 2D hardware acceleration. > > This series has been discussed on linux-fbdev, dri-devel and linux-kernel > mailing lists. The consensus is that they will be reverted, but in addition > it's being worked on to find a solution which allows to disable parts of the > classic fbdev hardware acceleration code which isn't needed in DRM-only > configurations. That's not my take at least. > The patchset is planned to be kept in fbdev's for-next git branch and later > pushed into kernel v5.18 if no issues are found until then. Neither this. I don't think we've resovled maintainership of fbdev core code and fbcon yet. That is the part that drm folks very much still maintained, as a quick git shortlog will show. Maintaining these parts outside of drm trees just doesn't make sense, since none of the other graphics relevant development happens outside of drm. -Daniel > > Helge > > --- > Changes in v2 compared to v1: > - adjusted the stable tags (one v5.10+, the other v5.16+) > - moved the analysis and reasoning why those should be reverted into > the commit message. That simplifies to analyze later why they were > reverted. > - resorted the To an Cc mail recipient lists > > Helge Deller (2): > Revert "fbdev: Garbage collect fbdev scrolling acceleration, part 1 > (from TODO list)" > Revert "fbcon: Disable accelerated scrolling" > > Documentation/gpu/todo.rst | 24 -- > drivers/video/fbdev/core/bitblit.c | 16 + > drivers/video/fbdev/core/fbcon.c | 540 +++++++++++++++++++++++- > drivers/video/fbdev/core/fbcon.h | 59 +++ > drivers/video/fbdev/core/fbcon_ccw.c | 28 +- > drivers/video/fbdev/core/fbcon_cw.c | 28 +- > drivers/video/fbdev/core/fbcon_rotate.h | 9 + > drivers/video/fbdev/core/fbcon_ud.c | 37 +- > drivers/video/fbdev/core/tileblit.c | 16 + > drivers/video/fbdev/skeletonfb.c | 12 +- > include/linux/fb.h | 2 +- > 11 files changed, 703 insertions(+), 68 deletions(-) > > -- > 2.31.1 >
On 1/24/22 23:19, Daniel Vetter wrote: > On Mon, Jan 24, 2022 at 10:03:17PM +0100, Helge Deller wrote: >> This series reverts two patches which disabled scrolling acceleration in >> fbcon/fbdev. Those patches introduced a regression for fbdev-supported graphic >> cards because of the performance penalty by doing screen scrolling by software >> instead of using existing 2D hardware acceleration. >> >> This series has been discussed on linux-fbdev, dri-devel and linux-kernel >> mailing lists. The consensus is that they will be reverted, but in addition >> it's being worked on to find a solution which allows to disable parts of the >> classic fbdev hardware acceleration code which isn't needed in DRM-only >> configurations. > > That's not my take at least. Which is? Didn't we discussed about introducing something like CONFIG_FBCON_LEGACY_ACCELERATION ? >> The patchset is planned to be kept in fbdev's for-next git branch and later >> pushed into kernel v5.18 if no issues are found until then. > > Neither this. I don't think we've resovled maintainership of fbdev core > code and fbcon yet. That is the part that drm folks very much still > maintained, as a quick git shortlog will show. > > Maintaining these parts outside of drm trees just doesn't make sense, > since none of the other graphics relevant development happens outside of I have no objections if you are willing to take those two patches through the drm tree. Helge >> --- >> Changes in v2 compared to v1: >> - adjusted the stable tags (one v5.10+, the other v5.16+) >> - moved the analysis and reasoning why those should be reverted into >> the commit message. That simplifies to analyze later why they were >> reverted. >> - resorted the To an Cc mail recipient lists >> >> Helge Deller (2): >> Revert "fbdev: Garbage collect fbdev scrolling acceleration, part 1 >> (from TODO list)" >> Revert "fbcon: Disable accelerated scrolling" >> >> Documentation/gpu/todo.rst | 24 -- >> drivers/video/fbdev/core/bitblit.c | 16 + >> drivers/video/fbdev/core/fbcon.c | 540 +++++++++++++++++++++++- >> drivers/video/fbdev/core/fbcon.h | 59 +++ >> drivers/video/fbdev/core/fbcon_ccw.c | 28 +- >> drivers/video/fbdev/core/fbcon_cw.c | 28 +- >> drivers/video/fbdev/core/fbcon_rotate.h | 9 + >> drivers/video/fbdev/core/fbcon_ud.c | 37 +- >> drivers/video/fbdev/core/tileblit.c | 16 + >> drivers/video/fbdev/skeletonfb.c | 12 +- >> include/linux/fb.h | 2 +- >> 11 files changed, 703 insertions(+), 68 deletions(-) >> >> -- >> 2.31.1
On Mon, Jan 24, 2022 at 11:37:03PM +0100, Helge Deller wrote: > On 1/24/22 23:19, Daniel Vetter wrote: > > On Mon, Jan 24, 2022 at 10:03:17PM +0100, Helge Deller wrote: > >> This series reverts two patches which disabled scrolling acceleration in > >> fbcon/fbdev. Those patches introduced a regression for fbdev-supported graphic > >> cards because of the performance penalty by doing screen scrolling by software > >> instead of using existing 2D hardware acceleration. > >> > >> This series has been discussed on linux-fbdev, dri-devel and linux-kernel > >> mailing lists. The consensus is that they will be reverted, but in addition > >> it's being worked on to find a solution which allows to disable parts of the > >> classic fbdev hardware acceleration code which isn't needed in DRM-only > >> configurations. > > > > That's not my take at least. > > Which is? > Didn't we discussed about introducing something like > CONFIG_FBCON_LEGACY_ACCELERATION ? Yeah, as part of the reverts. Or at least that's what I meant to say. -Daniel > > >> The patchset is planned to be kept in fbdev's for-next git branch and later > >> pushed into kernel v5.18 if no issues are found until then. > > > > Neither this. I don't think we've resovled maintainership of fbdev core > > code and fbcon yet. That is the part that drm folks very much still > > maintained, as a quick git shortlog will show. > > > > Maintaining these parts outside of drm trees just doesn't make sense, > > since none of the other graphics relevant development happens outside of > > I have no objections if you are willing to take those two patches through > the drm tree. > > Helge > > >> --- > >> Changes in v2 compared to v1: > >> - adjusted the stable tags (one v5.10+, the other v5.16+) > >> - moved the analysis and reasoning why those should be reverted into > >> the commit message. That simplifies to analyze later why they were > >> reverted. > >> - resorted the To an Cc mail recipient lists > >> > >> Helge Deller (2): > >> Revert "fbdev: Garbage collect fbdev scrolling acceleration, part 1 > >> (from TODO list)" > >> Revert "fbcon: Disable accelerated scrolling" > >> > >> Documentation/gpu/todo.rst | 24 -- > >> drivers/video/fbdev/core/bitblit.c | 16 + > >> drivers/video/fbdev/core/fbcon.c | 540 +++++++++++++++++++++++- > >> drivers/video/fbdev/core/fbcon.h | 59 +++ > >> drivers/video/fbdev/core/fbcon_ccw.c | 28 +- > >> drivers/video/fbdev/core/fbcon_cw.c | 28 +- > >> drivers/video/fbdev/core/fbcon_rotate.h | 9 + > >> drivers/video/fbdev/core/fbcon_ud.c | 37 +- > >> drivers/video/fbdev/core/tileblit.c | 16 + > >> drivers/video/fbdev/skeletonfb.c | 12 +- > >> include/linux/fb.h | 2 +- > >> 11 files changed, 703 insertions(+), 68 deletions(-) > >> > >> -- > >> 2.31.1
Cc regressions On Mon, Jan 24, 2022 at 10:04 PM Helge Deller <deller@gmx.de> wrote: > This series reverts two patches which disabled scrolling acceleration in > fbcon/fbdev. Those patches introduced a regression for fbdev-supported graphic > cards because of the performance penalty by doing screen scrolling by software > instead of using existing 2D hardware acceleration. In the mean time, a userspace regression has been reported, too: https://lwn.net/Articles/882054/ https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=988039 Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds