video: fbdev: udlfb: drop log level for blanking
PATCH 1/1 video: fbdev: udlfb: drop log level for blanking
Log blanking that goes from FB_BLANK_UNBLANK to itself at debug. It turns out Xorg likes to habitually unblank when already unblanked and this can fill up logs over a long period of time.
--- drivers/video/fbdev/udlfb.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-)
drivers/video/fbdev/udlfb.c +11 -2
@@ -1105,8 +1105,17 @@ static int dlfb_ops_blank(int blank_mode, struct fb_info *info)char *bufptr;struct urb *urb; -pr_info ("/dev/fb%d FB_BLANK mode %d --> %d\n" , - info->node, dev->blank_mode, blank_mode); +/* +* Xorg likes to habitually unblank when already unblanked so log these +* kinds of events at debug. +*/ +if (dev->blank_mode ==FB_BLANK_UNBLANK && + blank_mode ==FB_BLANK_UNBLANK ) +pr_debug ("/dev/fb%d FB_BLANK mode %d --> %d (FB_BLANK_UNBLANK --> FB_BLANK_UNBLANK)\n" , + info->node,FB_BLANK_UNBLANK ,FB_BLANK_UNBLANK ); +else +pr_info ("/dev/fb%d FB_BLANK mode %d --> %d\n" , + info->node, dev->blank_mode, blank_mode);if ((dev->blank_mode ==FB_BLANK_POWERDOWN ) && (blank_mode !=FB_BLANK_POWERDOWN )) {
Signature · 6 lines
2.13.0.219.gdb65acc882-goog -- Mike Gerow [email protected]
It would also be far simpler (and perhaps better) to just drop the existing log to debug instead of this conditional nonsense. I'm mostly interested in being able to check a machine running for a few weeks and seeing the logs not be 95% "udlfb: /dev/fb1 FB_BLANK mode 0 --> 0".
Signature · 2 lines
Mike Gerow [email protected]
Hi,
Quoted 4 lines · Mike Gerow
It would also be far simpler (and perhaps better) to just drop the existing log to debug instead of this conditional nonsense. I'm mostly interested in being able to check a machine running for a few weeks and seeing the logs not be 95% "udlfb:
/dev/fb1FB_BLANKmode 0 --> 0".
Fine with me, could you please redo your patch to do this?
Best regards,
--
Bartlomiej Zolnierkiewicz
Samsung R&D Institute Poland
Samsung Electronics