All DSI PHY/PLL drivers were referencing their VCO parent clock by a global name, most of which don't exist or have been renamed. These clock drivers seem to function fine without that except the 14nm driver for the sdm6xx [1].
At the same time all DTs provide a "ref" clock as per the requirements of dsi-phy-common.yaml, but the clock is never used. This patchset puts that clock to use without relying on a global clock name, so that all dependencies are explicitly defined in DT (the firmware) in the end.
[1]: https://lore.kernel.org/linux-arm-msm/386db1a6-a1cd-3c7d-a88e-dc83f8a1be96@s...
Cc: Dmitry Baryshkov dmitry.baryshkov@linaro.org Cc: Bjorn Andersson bjorn.andersson@linaro.org Signed-off-by: Marijn Suijten marijn.suijten@somainline.org --- drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c | 4 +++- drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c | 4 +++- drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c | 4 +++- drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm_8960.c | 4 +++- drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c | 4 +++- 5 files changed, 15 insertions(+), 5 deletions(-)
diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c index e46b10fc793a..3cbb1f1475e8 100644 --- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c +++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c @@ -562,7 +562,9 @@ static int pll_10nm_register(struct dsi_pll_10nm *pll_10nm, struct clk_hw **prov char clk_name[32], parent[32], vco_name[32]; char parent2[32], parent3[32], parent4[32]; struct clk_init_data vco_init = { - .parent_names = (const char *[]){ "xo" }, + .parent_data = &(const struct clk_parent_data) { + .fw_name = "ref", + }, .num_parents = 1, .name = vco_name, .flags = CLK_IGNORE_UNUSED, diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c index bb31230721bd..406470265408 100644 --- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c +++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c @@ -804,7 +804,9 @@ static int pll_14nm_register(struct dsi_pll_14nm *pll_14nm, struct clk_hw **prov { char clk_name[32], parent[32], vco_name[32]; struct clk_init_data vco_init = { - .parent_names = (const char *[]){ "xo" }, + .parent_data = &(const struct clk_parent_data) { + .fw_name = "ref", + }, .num_parents = 1, .name = vco_name, .flags = CLK_IGNORE_UNUSED, diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c index 2da673a2add6..8ee9c9c0548d 100644 --- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c +++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c @@ -521,7 +521,9 @@ static int pll_28nm_register(struct dsi_pll_28nm *pll_28nm, struct clk_hw **prov { char clk_name[32], parent1[32], parent2[32], vco_name[32]; struct clk_init_data vco_init = { - .parent_names = (const char *[]){ "xo" }, + .parent_data = &(const struct clk_parent_data) { + .fw_name = "ref", + }, .num_parents = 1, .name = vco_name, .flags = CLK_IGNORE_UNUSED, diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm_8960.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm_8960.c index aaa37456f4ee..9662cb236468 100644 --- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm_8960.c +++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm_8960.c @@ -385,7 +385,9 @@ static int pll_28nm_register(struct dsi_pll_28nm *pll_28nm, struct clk_hw **prov { char *clk_name, *parent_name, *vco_name; struct clk_init_data vco_init = { - .parent_names = (const char *[]){ "pxo" }, + .parent_data = &(const struct clk_parent_data) { + .fw_name = "ref", + }, .num_parents = 1, .flags = CLK_IGNORE_UNUSED, .ops = &clk_ops_dsi_pll_28nm_vco, diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c index 7c23d4c47338..c77c30628cca 100644 --- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c +++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c @@ -590,7 +590,9 @@ static int pll_7nm_register(struct dsi_pll_7nm *pll_7nm, struct clk_hw **provide char clk_name[32], parent[32], vco_name[32]; char parent2[32], parent3[32], parent4[32]; struct clk_init_data vco_init = { - .parent_names = (const char *[]){ "bi_tcxo" }, + .parent_data = &(const struct clk_parent_data) { + .fw_name = "ref", + }, .num_parents = 1, .name = vco_name, .flags = CLK_IGNORE_UNUSED,
Hi,
On Sun, 29 Aug 2021 at 23:30, Marijn Suijten marijn.suijten@somainline.org wrote:
All DSI PHY/PLL drivers were referencing their VCO parent clock by a global name, most of which don't exist or have been renamed. These clock drivers seem to function fine without that except the 14nm driver for the sdm6xx [1].
At the same time all DTs provide a "ref" clock as per the requirements of dsi-phy-common.yaml, but the clock is never used. This patchset puts that clock to use without relying on a global clock name, so that all dependencies are explicitly defined in DT (the firmware) in the end.
msm8974 (28nm-hpm) does not define the "ref" clock. So you'd have to: 1) add ref clock to the dtsi (should come in a separate patch). 2) add .name = "xo" as a fallback to the 28nm driver (to be compatible with older devices)
Other than that this looks good to me.
Cc: Dmitry Baryshkov dmitry.baryshkov@linaro.org Cc: Bjorn Andersson bjorn.andersson@linaro.org Signed-off-by: Marijn Suijten marijn.suijten@somainline.org
drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c | 4 +++- drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c | 4 +++- drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c | 4 +++- drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm_8960.c | 4 +++- drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c | 4 +++- 5 files changed, 15 insertions(+), 5 deletions(-)
diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c index e46b10fc793a..3cbb1f1475e8 100644 --- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c +++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_10nm.c @@ -562,7 +562,9 @@ static int pll_10nm_register(struct dsi_pll_10nm *pll_10nm, struct clk_hw **prov char clk_name[32], parent[32], vco_name[32]; char parent2[32], parent3[32], parent4[32]; struct clk_init_data vco_init = {
.parent_names = (const char *[]){ "xo" },
.parent_data = &(const struct clk_parent_data) {.fw_name = "ref",}, .num_parents = 1, .name = vco_name, .flags = CLK_IGNORE_UNUSED,diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c index bb31230721bd..406470265408 100644 --- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c +++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_14nm.c @@ -804,7 +804,9 @@ static int pll_14nm_register(struct dsi_pll_14nm *pll_14nm, struct clk_hw **prov { char clk_name[32], parent[32], vco_name[32]; struct clk_init_data vco_init = {
.parent_names = (const char *[]){ "xo" },
.parent_data = &(const struct clk_parent_data) {.fw_name = "ref",}, .num_parents = 1, .name = vco_name, .flags = CLK_IGNORE_UNUSED,diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c index 2da673a2add6..8ee9c9c0548d 100644 --- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c +++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c @@ -521,7 +521,9 @@ static int pll_28nm_register(struct dsi_pll_28nm *pll_28nm, struct clk_hw **prov { char clk_name[32], parent1[32], parent2[32], vco_name[32]; struct clk_init_data vco_init = {
.parent_names = (const char *[]){ "xo" },
.parent_data = &(const struct clk_parent_data) {.fw_name = "ref",}, .num_parents = 1, .name = vco_name, .flags = CLK_IGNORE_UNUSED,diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm_8960.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm_8960.c index aaa37456f4ee..9662cb236468 100644 --- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm_8960.c +++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm_8960.c @@ -385,7 +385,9 @@ static int pll_28nm_register(struct dsi_pll_28nm *pll_28nm, struct clk_hw **prov { char *clk_name, *parent_name, *vco_name; struct clk_init_data vco_init = {
.parent_names = (const char *[]){ "pxo" },
.parent_data = &(const struct clk_parent_data) {.fw_name = "ref",}, .num_parents = 1, .flags = CLK_IGNORE_UNUSED, .ops = &clk_ops_dsi_pll_28nm_vco,diff --git a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c index 7c23d4c47338..c77c30628cca 100644 --- a/drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c +++ b/drivers/gpu/drm/msm/dsi/phy/dsi_phy_7nm.c @@ -590,7 +590,9 @@ static int pll_7nm_register(struct dsi_pll_7nm *pll_7nm, struct clk_hw **provide char clk_name[32], parent[32], vco_name[32]; char parent2[32], parent3[32], parent4[32]; struct clk_init_data vco_init = {
.parent_names = (const char *[]){ "bi_tcxo" },
.parent_data = &(const struct clk_parent_data) {.fw_name = "ref",}, .num_parents = 1, .name = vco_name, .flags = CLK_IGNORE_UNUSED,-- 2.33.0
Hi Dmitry,
On 8/29/21 10:39 PM, Dmitry Baryshkov wrote:
Hi,
On Sun, 29 Aug 2021 at 23:30, Marijn Suijten marijn.suijten@somainline.org wrote:
All DSI PHY/PLL drivers were referencing their VCO parent clock by a global name, most of which don't exist or have been renamed. These clock drivers seem to function fine without that except the 14nm driver for the sdm6xx [1].
At the same time all DTs provide a "ref" clock as per the requirements of dsi-phy-common.yaml, but the clock is never used. This patchset puts that clock to use without relying on a global clock name, so that all dependencies are explicitly defined in DT (the firmware) in the end.
msm8974 (28nm-hpm) does not define the "ref" clock. So you'd have to:
- add ref clock to the dtsi (should come in a separate patch).
Thanks for double-checking and noticing this! I've queued up this patch for v2.
- add .name = "xo" as a fallback to the 28nm driver (to be compatible
with older devices)
Are there msm8974 devices out there that might upgrade kernels, but not firmware (DT)? On other boards (sdm630) I'm removing these from various drivers as to not have any possibility of relying on global names, in favour of having the clock dependencies fully specified in the DT.
Other than that this looks good to me.
Any r-b/a-b/t-b I can pick up for the next round?
- Marijn
On Mon, 30 Aug 2021 at 00:53, Marijn Suijten marijn.suijten@somainline.org wrote:
Hi Dmitry,
On 8/29/21 10:39 PM, Dmitry Baryshkov wrote:
Hi,
On Sun, 29 Aug 2021 at 23:30, Marijn Suijten marijn.suijten@somainline.org wrote:
All DSI PHY/PLL drivers were referencing their VCO parent clock by a global name, most of which don't exist or have been renamed. These clock drivers seem to function fine without that except the 14nm driver for the sdm6xx [1].
At the same time all DTs provide a "ref" clock as per the requirements of dsi-phy-common.yaml, but the clock is never used. This patchset puts that clock to use without relying on a global clock name, so that all dependencies are explicitly defined in DT (the firmware) in the end.
msm8974 (28nm-hpm) does not define the "ref" clock. So you'd have to:
- add ref clock to the dtsi (should come in a separate patch).
Thanks for double-checking and noticing this! I've queued up this patch for v2.
- add .name = "xo" as a fallback to the 28nm driver (to be compatible
with older devices)
Are there msm8974 devices out there that might upgrade kernels, but not firmware (DT)? On other boards (sdm630) I'm removing these from various drivers as to not have any possibility of relying on global names, in favour of having the clock dependencies fully specified in the DT.
IIUC it is a general policy of trying to be (somewhat) backwards-compatible. For example because your dts might come from a different source/be a part of different build process/etc.
Other than that this looks good to me.
Any r-b/a-b/t-b I can pick up for the next round?
Let's get those issues fixed and I'll respond with R-B tags
Hi Dmitry,
On Mon, Aug 30, 2021 at 04:17:32AM +0300, Dmitry Baryshkov wrote:
On Mon, 30 Aug 2021 at 00:53, Marijn Suijten marijn.suijten@somainline.org wrote:
Hi Dmitry,
On 8/29/21 10:39 PM, Dmitry Baryshkov wrote:
Hi,
On Sun, 29 Aug 2021 at 23:30, Marijn Suijten marijn.suijten@somainline.org wrote:
All DSI PHY/PLL drivers were referencing their VCO parent clock by a global name, most of which don't exist or have been renamed. These clock drivers seem to function fine without that except the 14nm driver for the sdm6xx [1].
At the same time all DTs provide a "ref" clock as per the requirements of dsi-phy-common.yaml, but the clock is never used. This patchset puts that clock to use without relying on a global clock name, so that all dependencies are explicitly defined in DT (the firmware) in the end.
msm8974 (28nm-hpm) does not define the "ref" clock. So you'd have to:
- add ref clock to the dtsi (should come in a separate patch).
Thanks for double-checking and noticing this! I've queued up this patch for v2.
- add .name = "xo" as a fallback to the 28nm driver (to be compatible
with older devices)
Are there msm8974 devices out there that might upgrade kernels, but not firmware (DT)? On other boards (sdm630) I'm removing these from various drivers as to not have any possibility of relying on global names, in favour of having the clock dependencies fully specified in the DT.
IIUC it is a general policy of trying to be (somewhat) backwards-compatible. For example because your dts might come from a different source/be a part of different build process/etc.
Good thinking; DT was after all intended to be used as firmware shipping on the device, when we're usually modifying and shipping it with the kernel in the end.
Just to make sure other platforms aren't affected by these changes, every board currently providing a "ref" clock has done so since the DSI node was added, except these for these three patches that added them after the fact:
79e51645a1dd ("arm64: dts: qcom: msm8916: Set 'xo_board' as ref clock of the DSI PHY") 6969d1d9c615 ("ARM: dts: qcom-apq8064: Set 'cxo_board' as ref clock of the DSI PHY") 0c0e72705a33 ("arm64: dts: sdm845: Set 'bi_tcxo' as ref clock of the DSI PHYs")
Their commit-messages confuse me. They make it seem like the "ref" clock was previously used when this doesn't seem to be the case (hence my patch). Has there possibly been a patchset like mine that removed the mentioned hardcoded clock, but ended up never being merged?
Either way, perhaps it's worth mentioning those patches with Fixes: so that this commit can be backported (have to be careful that DT changes for the other drivers are also backported, or this patch is split per PHY file), and maybe it's worth cc-ing the original authors to ask for clarification or at least make them aware?
- Marijn
dri-devel@lists.freedesktop.org
-
Dmitry Baryshkov -
Marijn Suijten