After recent changes HDMI code is ready to be enabled on DCE5. This patch just changes conditions to execute already present code on DCE5.
Signed-off-by: Rafał Miłecki zajec5@gmail.com --- Dave: I know it's common to accept patches adding IDs even while merge window's closed. Is this OK for you to take this patch as it only enables existing code for more hardware? DCE5 has same HDMI engine/code as DCE4.
This was tested for regressions on R6xx and Evergreen. It makes audio work on my NI Caicos card. --- drivers/gpu/drm/radeon/atombios_encoders.c | 4 +++- drivers/gpu/drm/radeon/evergreen_hdmi.c | 3 --- drivers/gpu/drm/radeon/r600_audio.c | 2 +- drivers/gpu/drm/radeon/r600_hdmi.c | 7 ++----- 4 files changed, 6 insertions(+), 10 deletions(-)
diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index e7b1ec5..486ccdf 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -1926,7 +1926,9 @@ radeon_atom_encoder_mode_set(struct drm_encoder *encoder,
if (atombios_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_HDMI) { r600_hdmi_enable(encoder); - if (ASIC_IS_DCE4(rdev)) + if (ASIC_IS_DCE6(rdev)) + ; /* TODO (use pointers instead of if-s?) */ + else if (ASIC_IS_DCE4(rdev)) evergreen_hdmi_setmode(encoder, adjusted_mode); else r600_hdmi_setmode(encoder, adjusted_mode); diff --git a/drivers/gpu/drm/radeon/evergreen_hdmi.c b/drivers/gpu/drm/radeon/evergreen_hdmi.c index a51f880..65c5416 100644 --- a/drivers/gpu/drm/radeon/evergreen_hdmi.c +++ b/drivers/gpu/drm/radeon/evergreen_hdmi.c @@ -156,9 +156,6 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; uint32_t offset;
- if (ASIC_IS_DCE5(rdev)) - return; - /* Silent, r600_hdmi_enable will raise WARN for us */ if (!dig->afmt->enabled) return; diff --git a/drivers/gpu/drm/radeon/r600_audio.c b/drivers/gpu/drm/radeon/r600_audio.c index 7479a5c..79b5591 100644 --- a/drivers/gpu/drm/radeon/r600_audio.c +++ b/drivers/gpu/drm/radeon/r600_audio.c @@ -57,7 +57,7 @@ static bool radeon_dig_encoder(struct drm_encoder *encoder) */ static int r600_audio_chipset_supported(struct radeon_device *rdev) { - return (rdev->family >= CHIP_R600 && !ASIC_IS_DCE5(rdev)) + return (rdev->family >= CHIP_R600 && !ASIC_IS_DCE6(rdev)) || rdev->family == CHIP_RS600 || rdev->family == CHIP_RS690 || rdev->family == CHIP_RS740; diff --git a/drivers/gpu/drm/radeon/r600_hdmi.c b/drivers/gpu/drm/radeon/r600_hdmi.c index 969c275..82a0a4c 100644 --- a/drivers/gpu/drm/radeon/r600_hdmi.c +++ b/drivers/gpu/drm/radeon/r600_hdmi.c @@ -322,9 +322,6 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; uint32_t offset;
- if (ASIC_IS_DCE5(rdev)) - return; - /* Silent, r600_hdmi_enable will raise WARN for us */ if (!dig->afmt->enabled) return; @@ -483,7 +480,7 @@ void r600_hdmi_enable(struct drm_encoder *encoder) uint32_t offset; u32 hdmi;
- if (ASIC_IS_DCE5(rdev)) + if (ASIC_IS_DCE6(rdev)) return;
/* Silent, r600_hdmi_enable will raise WARN for us */ @@ -543,7 +540,7 @@ void r600_hdmi_disable(struct drm_encoder *encoder) struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; uint32_t offset;
- if (ASIC_IS_DCE5(rdev)) + if (ASIC_IS_DCE6(rdev)) return;
/* Called for ATOM_ENCODER_MODE_HDMI only */
2012/6/10 Rafał Miłecki zajec5@gmail.com:
After recent changes HDMI code is ready to be enabled on DCE5. This patch just changes conditions to execute already present code on DCE5.
Signed-off-by: Rafał Miłecki zajec5@gmail.com
Dave: I know it's common to accept patches adding IDs even while merge window's closed. Is this OK for you to take this patch as it only enables existing code for more hardware? DCE5 has same HDMI engine/code as DCE4.
This was tested for regressions on R6xx and Evergreen. It makes audio work on my NI Caicos card.
I wish to say THANKS here for AMD. They donated me with that Caicos card that allowed me to write this code and fix few minor things in audio support.
On Sun, Jun 10, 2012 at 11:59 AM, Rafał Miłecki zajec5@gmail.com wrote:
After recent changes HDMI code is ready to be enabled on DCE5. This patch just changes conditions to execute already present code on DCE5.
Signed-off-by: Rafał Miłecki zajec5@gmail.com
Dave: I know it's common to accept patches adding IDs even while merge window's closed. Is this OK for you to take this patch as it only enables existing code for more hardware? DCE5 has same HDMI engine/code as DCE4.
This was tested for regressions on R6xx and Evergreen. It makes audio work on my NI Caicos card.
Reviewed-by: Alex Deucher alexander.deucher@amd.com
Thanks for all of your hard work on this Rafał!
Alex
drivers/gpu/drm/radeon/atombios_encoders.c | 4 +++- drivers/gpu/drm/radeon/evergreen_hdmi.c | 3 --- drivers/gpu/drm/radeon/r600_audio.c | 2 +- drivers/gpu/drm/radeon/r600_hdmi.c | 7 ++----- 4 files changed, 6 insertions(+), 10 deletions(-)
diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index e7b1ec5..486ccdf 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -1926,7 +1926,9 @@ radeon_atom_encoder_mode_set(struct drm_encoder *encoder,
if (atombios_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_HDMI) { r600_hdmi_enable(encoder);
- if (ASIC_IS_DCE4(rdev))
- if (ASIC_IS_DCE6(rdev))
- ; /* TODO (use pointers instead of if-s?) */
- else if (ASIC_IS_DCE4(rdev))
evergreen_hdmi_setmode(encoder, adjusted_mode); else r600_hdmi_setmode(encoder, adjusted_mode); diff --git a/drivers/gpu/drm/radeon/evergreen_hdmi.c b/drivers/gpu/drm/radeon/evergreen_hdmi.c index a51f880..65c5416 100644 --- a/drivers/gpu/drm/radeon/evergreen_hdmi.c +++ b/drivers/gpu/drm/radeon/evergreen_hdmi.c @@ -156,9 +156,6 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; uint32_t offset;
- if (ASIC_IS_DCE5(rdev))
- return;
/* Silent, r600_hdmi_enable will raise WARN for us */ if (!dig->afmt->enabled) return; diff --git a/drivers/gpu/drm/radeon/r600_audio.c b/drivers/gpu/drm/radeon/r600_audio.c index 7479a5c..79b5591 100644 --- a/drivers/gpu/drm/radeon/r600_audio.c +++ b/drivers/gpu/drm/radeon/r600_audio.c @@ -57,7 +57,7 @@ static bool radeon_dig_encoder(struct drm_encoder *encoder) */ static int r600_audio_chipset_supported(struct radeon_device *rdev) {
- return (rdev->family >= CHIP_R600 && !ASIC_IS_DCE5(rdev))
- return (rdev->family >= CHIP_R600 && !ASIC_IS_DCE6(rdev))
|| rdev->family == CHIP_RS600 || rdev->family == CHIP_RS690 || rdev->family == CHIP_RS740; diff --git a/drivers/gpu/drm/radeon/r600_hdmi.c b/drivers/gpu/drm/radeon/r600_hdmi.c index 969c275..82a0a4c 100644 --- a/drivers/gpu/drm/radeon/r600_hdmi.c +++ b/drivers/gpu/drm/radeon/r600_hdmi.c @@ -322,9 +322,6 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; uint32_t offset;
- if (ASIC_IS_DCE5(rdev))
- return;
/* Silent, r600_hdmi_enable will raise WARN for us */ if (!dig->afmt->enabled) return; @@ -483,7 +480,7 @@ void r600_hdmi_enable(struct drm_encoder *encoder) uint32_t offset; u32 hdmi;
- if (ASIC_IS_DCE5(rdev))
- if (ASIC_IS_DCE6(rdev))
return;
/* Silent, r600_hdmi_enable will raise WARN for us */ @@ -543,7 +540,7 @@ void r600_hdmi_disable(struct drm_encoder *encoder) struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; uint32_t offset;
- if (ASIC_IS_DCE5(rdev))
- if (ASIC_IS_DCE6(rdev))
return;
/* Called for ATOM_ENCODER_MODE_HDMI only */
1.7.7
On 10.06.2012 21:22, Alex Deucher wrote:
On Sun, Jun 10, 2012 at 11:59 AM, Rafał Miłeckizajec5@gmail.com wrote:
After recent changes HDMI code is ready to be enabled on DCE5. This patch just changes conditions to execute already present code on DCE5.
Signed-off-by: Rafał Miłeckizajec5@gmail.com
Dave: I know it's common to accept patches adding IDs even while merge window's closed. Is this OK for you to take this patch as it only enables existing code for more hardware? DCE5 has same HDMI engine/code as DCE4.
This was tested for regressions on R6xx and Evergreen. It makes audio work on my NI Caicos card.
Reviewed-by: Alex Deucheralexander.deucher@amd.com
Thanks for all of your hard work on this Rafał!
Yeah indeed, keep sticking with it! We really need more people working on the driver.
Unfortunately testing it with my Cayman based card shows that it plays with the correct speed, audio registers seems to be ok, but unfortunately I don't hear any sound at all :(
Going to dig into it tomorrow, Christian.
Alex
drivers/gpu/drm/radeon/atombios_encoders.c | 4 +++- drivers/gpu/drm/radeon/evergreen_hdmi.c | 3 --- drivers/gpu/drm/radeon/r600_audio.c | 2 +- drivers/gpu/drm/radeon/r600_hdmi.c | 7 ++----- 4 files changed, 6 insertions(+), 10 deletions(-)
diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index e7b1ec5..486ccdf 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -1926,7 +1926,9 @@ radeon_atom_encoder_mode_set(struct drm_encoder *encoder,
if (atombios_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_HDMI) { r600_hdmi_enable(encoder);
if (ASIC_IS_DCE4(rdev))
if (ASIC_IS_DCE6(rdev)); /* TODO (use pointers instead of if-s?) */else if (ASIC_IS_DCE4(rdev)) evergreen_hdmi_setmode(encoder, adjusted_mode); else r600_hdmi_setmode(encoder, adjusted_mode);diff --git a/drivers/gpu/drm/radeon/evergreen_hdmi.c b/drivers/gpu/drm/radeon/evergreen_hdmi.c index a51f880..65c5416 100644 --- a/drivers/gpu/drm/radeon/evergreen_hdmi.c +++ b/drivers/gpu/drm/radeon/evergreen_hdmi.c @@ -156,9 +156,6 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; uint32_t offset;
if (ASIC_IS_DCE5(rdev))return;/* Silent, r600_hdmi_enable will raise WARN for us */ if (!dig->afmt->enabled) return;diff --git a/drivers/gpu/drm/radeon/r600_audio.c b/drivers/gpu/drm/radeon/r600_audio.c index 7479a5c..79b5591 100644 --- a/drivers/gpu/drm/radeon/r600_audio.c +++ b/drivers/gpu/drm/radeon/r600_audio.c @@ -57,7 +57,7 @@ static bool radeon_dig_encoder(struct drm_encoder *encoder) */ static int r600_audio_chipset_supported(struct radeon_device *rdev) {
return (rdev->family>= CHIP_R600&& !ASIC_IS_DCE5(rdev))
return (rdev->family>= CHIP_R600&& !ASIC_IS_DCE6(rdev)) || rdev->family == CHIP_RS600 || rdev->family == CHIP_RS690 || rdev->family == CHIP_RS740;diff --git a/drivers/gpu/drm/radeon/r600_hdmi.c b/drivers/gpu/drm/radeon/r600_hdmi.c index 969c275..82a0a4c 100644 --- a/drivers/gpu/drm/radeon/r600_hdmi.c +++ b/drivers/gpu/drm/radeon/r600_hdmi.c @@ -322,9 +322,6 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; uint32_t offset;
if (ASIC_IS_DCE5(rdev))return;/* Silent, r600_hdmi_enable will raise WARN for us */ if (!dig->afmt->enabled) return;@@ -483,7 +480,7 @@ void r600_hdmi_enable(struct drm_encoder *encoder) uint32_t offset; u32 hdmi;
if (ASIC_IS_DCE5(rdev))
if (ASIC_IS_DCE6(rdev)) return; /* Silent, r600_hdmi_enable will raise WARN for us */@@ -543,7 +540,7 @@ void r600_hdmi_disable(struct drm_encoder *encoder) struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; uint32_t offset;
if (ASIC_IS_DCE5(rdev))
if (ASIC_IS_DCE6(rdev)) return; /* Called for ATOM_ENCODER_MODE_HDMI only */-- 1.7.7
2012/6/10 Christian König deathsimple@vodafone.de:
On 10.06.2012 21:22, Alex Deucher wrote:
On Sun, Jun 10, 2012 at 11:59 AM, Rafał Miłeckizajec5@gmail.com wrote:
After recent changes HDMI code is ready to be enabled on DCE5. This patch just changes conditions to execute already present code on DCE5.
Signed-off-by: Rafał Miłeckizajec5@gmail.com
Dave: I know it's common to accept patches adding IDs even while merge window's closed. Is this OK for you to take this patch as it only enables existing code for more hardware? DCE5 has same HDMI engine/code as DCE4.
This was tested for regressions on R6xx and Evergreen. It makes audio work on my NI Caicos card.
Reviewed-by: Alex Deucheralexander.deucher@amd.com
Thanks for all of your hard work on this Rafał!
Yeah indeed, keep sticking with it! We really need more people working on the driver.
Thanks :)
Unfortunately testing it with my Cayman based card shows that it plays with the correct speed, audio registers seems to be ok, but unfortunately I don't hear any sound at all :(
If you can easily switch between fglrx and radeon, comparing "avivotool regs hdmi" should give a quick hint. Unless you have better debugging tools at AMD :)
2012/6/11 Rafał Miłecki zajec5@gmail.com:
2012/6/10 Christian König deathsimple@vodafone.de:
Unfortunately testing it with my Cayman based card shows that it plays with the correct speed, audio registers seems to be ok, but unfortunately I don't hear any sound at all :(
If you can easily switch between fglrx and radeon, comparing "avivotool regs hdmi" should give a quick hint. Unless you have better debugging tools at AMD :)
Cayman is the latest DCE5, it uses different init code (see ni.c). You probably just need to put audio init in ni.c :)
dri-devel@lists.freedesktop.org
-
Alex Deucher -
Christian König -
Rafał Miłecki