linux-stable.git/drivers/phy/rockchip, branch linux-6.5.y

phy/rockchip: inno-hdmi: do not power on rk3328 post pll on reg write

2023-09-13T07:53:46+00:00

[ Upstream commit 19a1d46bd699940a496d3b0d4e142ef99834988c ]

inno_write is used to configure 0xaa reg, that also hold the
POST_PLL_POWER_DOWN bit.
When POST_PLL_REFCLK_SEL_TMDS is configured the power down bit is not
taken into consideration.

Fix this by keeping the power down bit until configuration is complete.
Also reorder the reg write order for consistency.

Fixes: 53706a116863 ("phy: add Rockchip Innosilicon hdmi phy")
Signed-off-by: Jonas Karlman 
Link: https://lore.kernel.org/r/20230615171005.2251032-5-jonas@kwiboo.se
Signed-off-by: Vinod Koul 
Signed-off-by: Sasha Levin

phy/rockchip: inno-hdmi: round fractal pixclock in rk3328 recalc_rate

2023-09-13T07:53:46+00:00

[ Upstream commit d5ef343c1d62bc4c4c2c393af654a41cb34b449f ]

inno_hdmi_phy_rk3328_clk_recalc_rate() is returning a rate not found
in the pre pll config table when the fractal divider is used.
This can prevent proper power_on because a tmdsclock for the new rate
is not found in the pre pll config table.

Fix this by saving and returning a rounded pixel rate that exist
in the pre pll config table.

Fixes: 53706a116863 ("phy: add Rockchip Innosilicon hdmi phy")
Signed-off-by: Zheng Yang 
Signed-off-by: Jonas Karlman 
Link: https://lore.kernel.org/r/20230615171005.2251032-3-jonas@kwiboo.se
Signed-off-by: Vinod Koul 
Signed-off-by: Sasha Levin

phy/rockchip: inno-hdmi: use correct vco_div_5 macro on rk3328

2023-09-13T07:53:46+00:00

[ Upstream commit 644c06dfbd0da713f772abf0a8f8581ac78e6264 ]

inno_hdmi_phy_rk3328_clk_set_rate() is using the RK3228 macro
when configuring vco_div_5 on RK3328.

Fix this by using correct vco_div_5 macro for RK3328.

Fixes: 53706a116863 ("phy: add Rockchip Innosilicon hdmi phy")
Signed-off-by: Jonas Karlman 
Link: https://lore.kernel.org/r/20230615171005.2251032-2-jonas@kwiboo.se
Signed-off-by: Vinod Koul 
Signed-off-by: Sasha Levin

phy: rockchip: remove unused hw_to_inno function

2023-04-12T16:17:50+00:00

clang with W=1 reports
drivers/phy/rockchip/phy-rockchip-inno-dsidphy.c:284:36: error:
  unused function 'hw_to_inno' [-Werror,-Wunused-function]
static inline struct inno_dsidphy *hw_to_inno(struct clk_hw *hw)
                                   ^
This function is not used so remove it.

Signed-off-by: Tom Rix 
Reviewed-by: Nick Desaulniers 
Link: https://lore.kernel.org/r/20230324132649.2649166-1-trix@redhat.com
Signed-off-by: Vinod Koul

phy: rockchip: Add naneng combo phy support for RK3588

2023-03-31T13:55:37+00:00

Add support for RK3588 combo phy

This is based on prior work from XiaoDong Huang and
Peter Geis fixing this issue specifically for Rockchip 356x.

Co-developed-by: Andrew Powers-Holmes 
Signed-off-by: Andrew Powers-Holmes 
Signed-off-by: Lucas Tanure 
Link: https://lore.kernel.org/r/20230314135555.44162-4-lucas.tanure@collabora.com
Signed-off-by: Vinod Koul

phy: rockchip-pcie: remove unused phy_rd_cfg function

2023-03-31T13:53:13+00:00

clang with W=1 reports
drivers/phy/rockchip/phy-rockchip-pcie.c:122:19: error:
  unused function 'phy_rd_cfg' [-Werror,-Wunused-function]
static inline u32 phy_rd_cfg(struct rockchip_pcie_phy *rk_phy,
                  ^
This function is not used, so remove it.

Signed-off-by: Tom Rix 
Link: https://lore.kernel.org/r/20230321122503.1783311-1-trix@redhat.com
Signed-off-by: Vinod Koul

phy: rockchip: phy-rockchip-typec: Convert to platform remove callback returning void

2023-03-20T12:44:58+00:00

The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is (mostly) ignored
and this typically results in resource leaks. To improve here there is a
quest to make the remove callback return void. In the first step of this
quest all drivers are converted to .remove_new() which already returns
void.

Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.

Signed-off-by: Uwe Kleine-König 
Reviewed-by: Heiko Stuebner 
Link: https://lore.kernel.org/r/20230307115900.2293120-23-u.kleine-koenig@pengutronix.de
Signed-off-by: Vinod Koul

phy: rockchip: phy-rockchip-inno-hdmi: Convert to platform remove callback returning void

2023-03-20T12:44:57+00:00

The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is (mostly) ignored
and this typically results in resource leaks. To improve here there is a
quest to make the remove callback return void. In the first step of this
quest all drivers are converted to .remove_new() which already returns
void.

Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.

Signed-off-by: Uwe Kleine-König 
Reviewed-by: Heiko Stuebner 
Link: https://lore.kernel.org/r/20230307115900.2293120-22-u.kleine-koenig@pengutronix.de
Signed-off-by: Vinod Koul

phy: rockchip: phy-rockchip-inno-dsidphy: Convert to platform remove callback returning void

2023-03-20T12:44:57+00:00

The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is (mostly) ignored
and this typically results in resource leaks. To improve here there is a
quest to make the remove callback return void. In the first step of this
quest all drivers are converted to .remove_new() which already returns
void.

Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.

Signed-off-by: Uwe Kleine-König 
Reviewed-by: Heiko Stuebner 
Link: https://lore.kernel.org/r/20230307115900.2293120-21-u.kleine-koenig@pengutronix.de
Signed-off-by: Vinod Koul

phy: rockchip: phy-rockchip-inno-csidphy: Convert to platform remove callback returning void

2023-03-20T12:44:57+00:00

The .remove() callback for a platform driver returns an int which makes
many driver authors wrongly assume it's possible to do error handling by
returning an error code. However the value returned is (mostly) ignored
and this typically results in resource leaks. To improve here there is a
quest to make the remove callback return void. In the first step of this
quest all drivers are converted to .remove_new() which already returns
void.

Trivially convert this driver from always returning zero in the remove
callback to the void returning variant.

Signed-off-by: Uwe Kleine-König 
Reviewed-by: Heiko Stuebner 
Link: https://lore.kernel.org/r/20230307115900.2293120-20-u.kleine-koenig@pengutronix.de
Signed-off-by: Vinod Koul