Skip to contentSkip to conversation

resend: radeon backlight shutoff flashes momentarily

PATCH In review Robert T. Drury 3 participants 4 messages Apr 1, 2009 – Apr 10, 2009 Framebuffer layer

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 resend: radeon backlight shutoff flashes momentarily

Addresses http://bugzilla.kernel.org/show_bug.cgi?id=12845

Aluminum Powerbook G4 with Radeon 9600:

The backlight momentarily flashes brightly when turning off. If the backlight is originally dim, the flash is more noticeable. This momentary flash is stressful on the power circuit.

This patch removes the & ~LVDS_BL_MOD_EN above the first OUTREG to LVDS_GEN_CNTL to eliminate the flash on shutoff, only for powerbook5,3 and CHIP_FAMILY_RV350.

drivers/video/aty/radeon_backlight.c renamed from drivers/video/aty/radeon_backlight.c.orig +4 -1
@@ -101,7 +101,10 @@ static int radeon_bl_update_status(struc
 		tmpPixclksCntl = INPLL(PIXCLKS_CNTL);
 		if (rinfo->is_mobility || rinfo->is_IGP)
 			OUTPLLP(PIXCLKS_CNTL, 0, ~PIXCLK_LVDS_ALWAYS_ONb);
-		lvds_gen_cntl &= ~(LVDS_BL_MOD_LEVEL_MASK | LVDS_BL_MOD_EN);
+		lvds_gen_cntl &= ~(LVDS_BL_MOD_LEVEL_MASK);
+		if(rinfo->family != CHIP_FAMILY_RV350 
+			|| !machine_is_compatible("PowerBook5,3"))
+				lvds_gen_cntl &= ~LVDS_BL_MOD_EN;
 		lvds_gen_cntl |= (radeon_bl_get_level_brightness(pdata, 0) <<
 				  LVDS_BL_MOD_LEVEL_SHIFT);
 		lvds_gen_cntl |= LVDS_DISPLAY_DIS;

------------------------------------------------------------------------------

#1 lore raw
Quoted 11 lines · Robert T. Drury

Addresses http://bugzilla.kernel.org/show_bug.cgi?id=12845

Aluminum Powerbook G4 with Radeon 9600:

The backlight momentarily flashes brightly when turning off. If the backlight is originally dim, the flash is more noticeable. This momentary flash is stressful on the power circuit.

This patch removes the & ~LVDS_BL_MOD_EN above the first OUTREG to LVDS_GEN_CNTL to eliminate the flash on shutoff, only for powerbook5,3 and CHIP_FAMILY_RV350.

As you're checking for a specific Apple laptop model, there's no need to check the chip family as well.

(As mentioned before, I'm seeing the flash on a PowerBook5,8 as well, but I haven't got around to trying the patch, will do when I get time)

Signature · 8 lines
Earthling Michel Dänzer           |                http://www.vmware.com
Libre software enthusiast         |          Debian, X and DRI developer

------------------------------------------------------------------------------
_______________________________________________
Linux-fbdev-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-fbdev-devel
#2 lore raw
Quoted 13 lines · Michel Dänzer
Quoted 11 lines · Robert T. Drury

Addresses http://bugzilla.kernel.org/show_bug.cgi?id=12845

Aluminum Powerbook G4 with Radeon 9600:

The backlight momentarily flashes brightly when turning off. If the backlight is originally dim, the flash is more noticeable. This momentary flash is stressful on the power circuit.

This patch removes the & ~LVDS_BL_MOD_EN above the first OUTREG to LVDS_GEN_CNTL to eliminate the flash on shutoff, only for powerbook5,3 and CHIP_FAMILY_RV350.

Quoted 5 lines

As you're checking for a specific Apple laptop model, there's no need to check the chip family as well.

(As mentioned before, I'm seeing the flash on a PowerBook5,8 as well, but I haven't got around to trying the patch, will do when I get time)

Unfortunately, 2.6.29.1 with the patch below doesn't seem to get rid of the flash on my PowerBook5,8. So no objections to your patch from me.

drivers/video/aty/radeon_backlight.c +4 -1
index 1a056ad..361bb64 100644
@@ -101,7 +101,10 @@ static int radeon_bl_update_status(struct backlight_device *bd)
 		tmpPixclksCntl = INPLL(PIXCLKS_CNTL);
 		if (rinfo->is_mobility || rinfo->is_IGP)
 			OUTPLLP(PIXCLKS_CNTL, 0, ~PIXCLK_LVDS_ALWAYS_ONb);
-		lvds_gen_cntl &= ~(LVDS_BL_MOD_LEVEL_MASK | LVDS_BL_MOD_EN);
+		lvds_gen_cntl &= ~LVDS_BL_MOD_LEVEL_MASK;
+		if (!(machine_is_compatible("PowerBook5,3") ||
+		      machine_is_compatible("PowerBook5,8")))
+			lvds_gen_cntl &= ~LVDS_BL_MOD_EN;
 		lvds_gen_cntl |= (radeon_bl_get_level_brightness(pdata, 0) <<
 				  LVDS_BL_MOD_LEVEL_SHIFT);
 		lvds_gen_cntl |= LVDS_DISPLAY_DIS;
Signature · 8 lines
Earthling Michel Dänzer           |                http://www.vmware.com
Libre software enthusiast         |          Debian, X and DRI developer

------------------------------------------------------------------------------
_______________________________________________
Linux-fbdev-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-fbdev-devel
#3 lore raw
In reply to Robert T. Drury
Quoted 18 lines · On Wed, 01 Apr 2009 02:27:54 -0700 "Robert T. Drury" <[email protected]>…

Addresses http://bugzilla.kernel.org/show_bug.cgi?id=12845

Aluminum Powerbook G4 with Radeon 9600:

The backlight momentarily flashes brightly when turning off. If the backlight is originally dim, the flash is more noticeable. This momentary flash is stressful on the power circuit.

This patch removes the & ~LVDS_BL_MOD_EN above the first OUTREG to LVDS_GEN_CNTL to eliminate the flash on shutoff, only for powerbook5,3 and CHIP_FAMILY_RV350.

drivers/video/aty/radeon_backlight.c renamed from drivers/video/aty/radeon_backlight.c.orig +0 -0
@@ -101,7 +101,10 @@ static int radeon_bl_update_status(struc
 		tmpPixclksCntl = INPLL(PIXCLKS_CNTL);

drivers/video/aty/radeon_backlight.c static int radeon_bl_update_status(struc · line 110

Quoted 10 lines
drivers/video/aty/radeon_backlight.c renamed from drivers/video/aty/radeon_backlight.c.orig +4 -1
 		if (rinfo->is_mobility || rinfo->is_IGP)
 			OUTPLLP(PIXCLKS_CNTL, 0, ~PIXCLK_LVDS_ALWAYS_ONb);
-		lvds_gen_cntl &= ~(LVDS_BL_MOD_LEVEL_MASK | LVDS_BL_MOD_EN);
+		lvds_gen_cntl &= ~(LVDS_BL_MOD_LEVEL_MASK);
+		if(rinfo->family != CHIP_FAMILY_RV350 
+			|| !machine_is_compatible("PowerBook5,3"))
+				lvds_gen_cntl &= ~LVDS_BL_MOD_EN;
 		lvds_gen_cntl |= (radeon_bl_get_level_brightness(pdata, 0) <<
 				  LVDS_BL_MOD_LEVEL_SHIFT);
 		lvds_gen_cntl |= LVDS_DISPLAY_DIS;

machine_is_compatible() isn't defined on x86 so this breaks the build.

------------------------------------------------------------------------------
This SF.net email is sponsored by:
High Quality Requirements in a Collaborative Environment. Download a free trial of Rational Requirements Composer Now!
http://p.sf.net/sfu/www-ibm-com