Commit ac89af1e1010640db072416c786f97391b85790f caused one of the monitors attached to a dual head radeon gpu to have inverted colors (until the first suspend/resume). Initializing dig phy a bit later fixes the problem.
--- drivers/gpu/drm/radeon/radeon_display.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index ae247ee..ddff2cf 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -1346,10 +1346,6 @@ int radeon_modeset_init(struct radeon_device *rdev) return ret; }
- /* init dig PHYs */ - if (rdev->is_atom_bios) - radeon_atom_encoder_init(rdev); - /* initialize hpd */ radeon_hpd_init(rdev);
@@ -1359,6 +1355,10 @@ int radeon_modeset_init(struct radeon_device *rdev) radeon_fbdev_init(rdev); drm_kms_helper_poll_init(rdev->ddev);
+ /* init dig PHYs */ + if (rdev->is_atom_bios) + radeon_atom_encoder_init(rdev); + return 0; }
On Thu, Jun 2, 2011 at 5:20 PM, Ari Savolainen ari.m.savolainen@gmail.com wrote:
Commit ac89af1e1010640db072416c786f97391b85790f caused one of the monitors attached to a dual head radeon gpu to have inverted colors (until the first suspend/resume). Initializing dig phy a bit later fixes the problem.
Strange, I don't see why that would make a difference, I guess perhaps there's some strange interaction between the hpd setup or the initial clock/voltage setup on DCE5 hw. What chip are you using?
Anyway, should be fine.
Acked-by: Alex Deucher alexdeucher@gmail.com
drivers/gpu/drm/radeon/radeon_display.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index ae247ee..ddff2cf 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -1346,10 +1346,6 @@ int radeon_modeset_init(struct radeon_device *rdev) return ret; }
- /* init dig PHYs */
- if (rdev->is_atom_bios)
- radeon_atom_encoder_init(rdev);
/* initialize hpd */ radeon_hpd_init(rdev);
@@ -1359,6 +1355,10 @@ int radeon_modeset_init(struct radeon_device *rdev) radeon_fbdev_init(rdev); drm_kms_helper_poll_init(rdev->ddev);
- /* init dig PHYs */
- if (rdev->is_atom_bios)
- radeon_atom_encoder_init(rdev);
return 0; }
-- 1.7.4.1
I'm using ASUS EAH3650SILENTMG/HTDP.
I got the feeling that in my case, radeon_atom_encoder_prepare needed to be called before radeon_atom_encoder_init (the call occurred during radeon_fbdev_init). That ensured radeon_encoder->enc_priv->dig_encoder to have a value other than -1 in the first call of atombios_dig_transmitter_setup (with action=ATOM_TRANSMITTER_ACTION_INIT). But I don't know the code well enough to be sure about that.
Ari
2011/6/3 Alex Deucher alexdeucher@gmail.com:
On Thu, Jun 2, 2011 at 5:20 PM, Ari Savolainen ari.m.savolainen@gmail.com wrote:
Commit ac89af1e1010640db072416c786f97391b85790f caused one of the monitors attached to a dual head radeon gpu to have inverted colors (until the first suspend/resume). Initializing dig phy a bit later fixes the problem.
Strange, I don't see why that would make a difference, I guess perhaps there's some strange interaction between the hpd setup or the initial clock/voltage setup on DCE5 hw. What chip are you using?
Anyway, should be fine.
Acked-by: Alex Deucher alexdeucher@gmail.com
drivers/gpu/drm/radeon/radeon_display.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index ae247ee..ddff2cf 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -1346,10 +1346,6 @@ int radeon_modeset_init(struct radeon_device *rdev) return ret; }
- /* init dig PHYs */
- if (rdev->is_atom_bios)
- radeon_atom_encoder_init(rdev);
/* initialize hpd */ radeon_hpd_init(rdev);
@@ -1359,6 +1355,10 @@ int radeon_modeset_init(struct radeon_device *rdev) radeon_fbdev_init(rdev); drm_kms_helper_poll_init(rdev->ddev);
- /* init dig PHYs */
- if (rdev->is_atom_bios)
- radeon_atom_encoder_init(rdev);
return 0; }
-- 1.7.4.1
On Fri, Jun 3, 2011 at 4:03 PM, Ari Savolainen ari.m.savolainen@gmail.com wrote:
I'm using ASUS EAH3650SILENTMG/HTDP.
I got the feeling that in my case, radeon_atom_encoder_prepare needed to be called before radeon_atom_encoder_init (the call occurred during radeon_fbdev_init). That ensured radeon_encoder->enc_priv->dig_encoder to have a value other than -1 in the first call of atombios_dig_transmitter_setup (with action=ATOM_TRANSMITTER_ACTION_INIT). But I don't know the code well enough to be sure about that.
NACK on the patch for now. Let's try and sort this out. INIT needs to be called before the mode is set, so this needs to come before fbdev sets the mode. I'll send out a patch soon.
Alex
Ari
2011/6/3 Alex Deucher alexdeucher@gmail.com:
On Thu, Jun 2, 2011 at 5:20 PM, Ari Savolainen ari.m.savolainen@gmail.com wrote:
Commit ac89af1e1010640db072416c786f97391b85790f caused one of the monitors attached to a dual head radeon gpu to have inverted colors (until the first suspend/resume). Initializing dig phy a bit later fixes the problem.
Strange, I don't see why that would make a difference, I guess perhaps there's some strange interaction between the hpd setup or the initial clock/voltage setup on DCE5 hw. What chip are you using?
Anyway, should be fine.
Acked-by: Alex Deucher alexdeucher@gmail.com
drivers/gpu/drm/radeon/radeon_display.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index ae247ee..ddff2cf 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -1346,10 +1346,6 @@ int radeon_modeset_init(struct radeon_device *rdev) return ret; }
- /* init dig PHYs */
- if (rdev->is_atom_bios)
- radeon_atom_encoder_init(rdev);
/* initialize hpd */ radeon_hpd_init(rdev);
@@ -1359,6 +1355,10 @@ int radeon_modeset_init(struct radeon_device *rdev) radeon_fbdev_init(rdev); drm_kms_helper_poll_init(rdev->ddev);
- /* init dig PHYs */
- if (rdev->is_atom_bios)
- radeon_atom_encoder_init(rdev);
return 0; }
-- 1.7.4.1
The PHY was not initialized correctly after ac89af1e1010640db072416c786f97391b85790f since the function bailed early as an encoder was not assigned. The encoder isn't necessary for PHY init so just assign to 0 for init so that the table is executed.
Reported-by: Ari Savolainen ari.m.savolainen@gmail.com
Cc: Ari Savolainen ari.m.savolainen@gmail.com Signed-off-by: Alex Deucher alexdeucher@gmail.com --- drivers/gpu/drm/radeon/radeon_encoders.c | 17 +++++++++++------ 1 files changed, 11 insertions(+), 6 deletions(-)
diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c index 13e4fa0..302248b 100644 --- a/drivers/gpu/drm/radeon/radeon_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_encoders.c @@ -949,10 +949,15 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t int dp_lane_count = 0; int connector_object_id = 0; int igp_lane_info = 0; + int dig_encoder = dig->dig_encoder;
- if (action == ATOM_TRANSMITTER_ACTION_INIT) + if (action == ATOM_TRANSMITTER_ACTION_INIT) { connector = radeon_get_connector_for_encoder_init(encoder); - else + /* just needed to avoid bailing in the encoder check. the encoder + * isn't used for init + */ + dig_encoder = 0; + } else connector = radeon_get_connector_for_encoder(encoder);
if (connector) { @@ -968,7 +973,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t }
/* no dig encoder assigned */ - if (dig->dig_encoder == -1) + if (dig_encoder == -1) return;
if (atombios_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_DP) @@ -1018,7 +1023,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t
if (dig->linkb) args.v3.acConfig.ucLinkSel = 1; - if (dig->dig_encoder & 1) + if (dig_encoder & 1) args.v3.acConfig.ucEncoderSel = 1;
/* Select the PLL for the PHY @@ -1068,7 +1073,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t args.v3.acConfig.fDualLinkConnector = 1; } } else if (ASIC_IS_DCE32(rdev)) { - args.v2.acConfig.ucEncoderSel = dig->dig_encoder; + args.v2.acConfig.ucEncoderSel = dig_encoder; if (dig->linkb) args.v2.acConfig.ucLinkSel = 1;
@@ -1095,7 +1100,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t } else { args.v1.ucConfig = ATOM_TRANSMITTER_CONFIG_CLKSRC_PPLL;
- if (dig->dig_encoder) + if (dig_encoder) args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_DIG2_ENCODER; else args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_DIG1_ENCODER;
I tested the patch and it fixed the problem.
Thanks, Ari
2011/6/3 Alex Deucher alexdeucher@gmail.com:
The PHY was not initialized correctly after ac89af1e1010640db072416c786f97391b85790f since the function bailed early as an encoder was not assigned. The encoder isn't necessary for PHY init so just assign to 0 for init so that the table is executed.
Reported-by: Ari Savolainen ari.m.savolainen@gmail.com
Cc: Ari Savolainen ari.m.savolainen@gmail.com Signed-off-by: Alex Deucher alexdeucher@gmail.com
drivers/gpu/drm/radeon/radeon_encoders.c | 17 +++++++++++------ 1 files changed, 11 insertions(+), 6 deletions(-)
diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c index 13e4fa0..302248b 100644 --- a/drivers/gpu/drm/radeon/radeon_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_encoders.c @@ -949,10 +949,15 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t int dp_lane_count = 0; int connector_object_id = 0; int igp_lane_info = 0;
- int dig_encoder = dig->dig_encoder;
- if (action == ATOM_TRANSMITTER_ACTION_INIT)
- if (action == ATOM_TRANSMITTER_ACTION_INIT) {
connector = radeon_get_connector_for_encoder_init(encoder);
- else
- /* just needed to avoid bailing in the encoder check. the encoder
- * isn't used for init
- */
- dig_encoder = 0;
- } else
connector = radeon_get_connector_for_encoder(encoder);
if (connector) { @@ -968,7 +973,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t }
/* no dig encoder assigned */
- if (dig->dig_encoder == -1)
- if (dig_encoder == -1)
return;
if (atombios_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_DP) @@ -1018,7 +1023,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t
if (dig->linkb) args.v3.acConfig.ucLinkSel = 1;
- if (dig->dig_encoder & 1)
- if (dig_encoder & 1)
args.v3.acConfig.ucEncoderSel = 1;
/* Select the PLL for the PHY @@ -1068,7 +1073,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t args.v3.acConfig.fDualLinkConnector = 1; } } else if (ASIC_IS_DCE32(rdev)) {
- args.v2.acConfig.ucEncoderSel = dig->dig_encoder;
- args.v2.acConfig.ucEncoderSel = dig_encoder;
if (dig->linkb) args.v2.acConfig.ucLinkSel = 1;
@@ -1095,7 +1100,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t } else { args.v1.ucConfig = ATOM_TRANSMITTER_CONFIG_CLKSRC_PPLL;
- if (dig->dig_encoder)
- if (dig_encoder)
args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_DIG2_ENCODER; else args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_DIG1_ENCODER; -- 1.7.1.1
dri-devel@lists.freedesktop.org
-
Alex Deucher -
Ari Savolainen