Register a simple clock provider for the PHY pipe clock sources so that
device tree users can point at these clocks via phandles to the lane
nodes.
Signed-off-by: Evan Green <evgreen@chromium.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Tested-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Utilize the newly fixed up DT bindings to get the tx2 and rx2 register
regions for the second lane of dual-lane PHYs. Before this change,
the driver was simply using lane one's register region and adding
0x400, which reached well beyond the DT-specified register
allocation. This would have been a crash were it not for the page size
on ARM64. Fix the driver not to rely on the magic of virtual memory by
using the newly specified DT register regions for tx2 and rx2.
In order to support existing device trees, this change also contains a
fallback mode for when those new register regions don't exist, which
reverts to the original behavior of overreaching and prints a complaint.
Signed-off-by: Evan Green <evgreen@chromium.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Currently the attempt to add support for Ethernet interface mode PHY
(MII/GMII/RGMII) will lead to the necessity of extending enum phy_mode and
duplicate there values from phy_interface_t enum (or introduce more PHY
callbacks) [1]. Both approaches are ineffective and would lead to fast
bloating of enum phy_mode or struct phy_ops in the process of adding more
PHYs for different subsystems which will make them unmaintainable.
As discussed in [1] the solution could be to introduce dual level PHYs mode
configuration - PHY mode and PHY submode. The PHY mode will define generic
PHY type (subsystem - PCIE/ETHERNET/USB_) while the PHY submode - subsystem
specific interface mode. The last is usually already defined in
corresponding subsystem headers (phy_interface_t for Ethernet, enum
usb_device_speed for USB).
This patch is cumulative change which refactors PHY framework code to
support dual level PHYs mode configuration - PHY mode and PHY submode. It
extends .set_mode() callback to support additional parameter "int submode"
and converts all corresponding PHY drivers to support new .set_mode()
callback declaration.
The new extended PHY API
int phy_set_mode_ext(struct phy *phy, enum phy_mode mode, int submode)
is introduced to support dual level PHYs mode configuration and existing
phy_set_mode() API is converted to macros, so PHY framework consumers do
not need to be changed (~21 matches).
[1] http://lkml.kernel.org/r/d63588f6-9ab0-848a-5ad4-8073143bd95d@ti.com
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Add UFS PHY support to make SDM845 UFS work with common PHY framework.
Signed-off-by: Can Guo <cang@codeaurora.org>
Reviewed-by: Evan Green <evgreen@chromium.org>
Reviewed-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Move MSM8996 specific PHY vreg list struct name to a genernal one as it is
used by all PHYs. Add a specific field to handle dual lane situation.
Signed-off-by: Can Guo <cang@codeaurora.org>
Reviewed-by: Evan Green <evgreen@chromium.org>
Reviewed-by: Manu Gautam <mgautam@codeaurora.org>
Reviewed-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
All PHYs should be powered on before register configuration starts. And
only PCIe PHYs need an extra power control before deasserts reset state.
Signed-off-by: Can Guo <cang@codeaurora.org>
Reviewed-by: Manu Gautam <mgautam@codeaurora.org>
Reviewed-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Reviewed-by: Evan Green <evgreen@chromium.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
In preparation to remove the node name pointer from struct device_node,
convert printf users to use the %pOFn format specifier.
Cc: Kishon Vijay Abraham I <kishon@ti.com>
Cc: linux-arm-kernel@lists.infradead.org
Signed-off-by: Rob Herring <robh@kernel.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
The -EPROBE_DEFER virus demands special case code to avoid printing
error messages when the error is only -EPROBE_DEFER. Spread the virus
to a new host: qcom_qmp_phy_probe(). Specifically handle when our
regulators might not be ready yet.
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
QMP V3 UNI PHY is a single lane USB3 PHY without support
for DisplayPort (DP).
Main difference from DP combo QMPv3 PHY is that UNI PHY
doesn't have dual RX/TX lanes and no separate DP_COM
block for configuration related to type-c or DP.
Also remove "qcom,qmp-v3-usb3-phy" compatible string which
was earlier added for sdm845 only as there wouldn't be
any user of same.
While at it, fix has_pwrdn_delay attribute for USB-DP
PHY configuration and.
Reviewed-by: Evan Green <evgreen@chromium.org>
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
QMP PHY for USB/PCIE requires pipe_clk for locking of
retime buffers at the pipe interface. Driver checks for
PHY_STATUS without enabling pipe_clk due to which
phy_init() fails with initialization timeout.
Though pipe_clk is output from PHY (after PLL is programmed
during initialization sequence) to GCC clock_ctl and then fed
back to PHY but for PHY_STATUS register to reflect successful
initialization pipe_clk from GCC must be present.
Since, clock driver now ignores status_check for pipe_clk on
clk_enable/disable, driver can safely enable/disable pipe_clk
from phy_init/exit.
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
The SPDX identifier is a legally binding shorthand, which
can be used instead of the full boiler plate text.
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Disable clocks and enable PHY autonomous mode to detect
wakeup events when PHY is suspended.
Core driver should notify speed to PHY driver to enable
LFPS and/or RX_DET interrupts.
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
QMP V3 USB3 PHY is a DisplayPort (DP) and USB combo PHY
with dual RX/TX lanes to support type-c. There is a
separate block DP_COM for configuration related to type-c
or DP. Add support for dp_com region and secondary rx/tx
lanes initialization.
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
New revision (v3) of QMP PHY uses different offsets
for almost all of the registers. Hence, move these
definitions to header file so that updated offsets
can be added for QMP v3.
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Reviewed-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Driver is currently performing PHY reset after starting
SERDES/PCS. As per hardware datasheet reset must be done
before starting PHY. Hence, update the sequence.
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
PHY block or asynchronous reset requires signal
to be asserted before de-asserting. Driver is only
de-asserting signal which is already low, hence
reset operation is a no-op. Fix this by asserting
signal first. Also, resetting requires PHY clocks
to be turned ON only after reset is finished. Fix
that as well.
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Reviewed-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
PHY regulators which are enabled from power_on() must be ON
before turning-on clocks and initializing it as part of init().
As most of the core drivers perform power_on() after init(), move
PHY regulators enable to com_init() and use power_on() to
only enable pipe_clk. This pipe_clk is output from PHY and some
core drivers e.g. PCIe follow specific sequence after phy_init()
that mandates pipe_clk to be enabled from power_on() only.
On similar lines move clk_enable from init() to com_init() which
executes once for multi lane PHYs.
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Move from using array of clocks to clk_bulk_* APIs that
are available now.
Signed-off-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Pipe clock comes out of the phy and is available as long as
the phy is turned on. Clock controller fails to gate this
clock after the phy is turned off and generates a warning.
/ # [ 33.048561] gcc_usb3_phy_pipe_clk status stuck at 'on'
[ 33.048585] ------------[ cut here ]------------
[ 33.052621] WARNING: CPU: 1 PID: 18 at ../drivers/clk/qcom/clk-branch.c:97 clk_branch_wait+0xf0/0x108
[ 33.057384] Modules linked in:
[ 33.066497] CPU: 1 PID: 18 Comm: kworker/1:0 Tainted: G W 4.12.0-rc7-00024-gfe926e34c36d-dirty #96
[ 33.069451] Hardware name: Qualcomm Technologies, Inc. DB820c (DT)
...
[ 33.278565] [<ffff00000849b27c>] clk_branch_wait+0xf0/0x108
[ 33.286375] [<ffff00000849b2f4>] clk_branch2_disable+0x28/0x34
[ 33.291761] [<ffff0000084868dc>] clk_core_disable+0x5c/0x88
[ 33.297660] [<ffff000008487d68>] clk_core_disable_lock+0x20/0x34
[ 33.303129] [<ffff000008487d98>] clk_disable+0x1c/0x24
[ 33.309384] [<ffff0000083ccd78>] qcom_qmp_phy_poweroff+0x20/0x48
[ 33.314328] [<ffff0000083c53f4>] phy_power_off+0x80/0xdc
[ 33.320492] [<ffff00000875c950>] dwc3_core_exit+0x94/0xa0
[ 33.325784] [<ffff00000875c9ac>] dwc3_suspend_common+0x50/0x60
[ 33.331080] [<ffff00000875ca04>] dwc3_runtime_suspend+0x48/0x6c
[ 33.336810] [<ffff0000085b82f4>] pm_generic_runtime_suspend+0x28/0x38
[ 33.342627] [<ffff0000085bace0>] __rpm_callback+0x150/0x254
[ 33.349222] [<ffff0000085bae08>] rpm_callback+0x24/0x78
[ 33.354604] [<ffff0000085b9fd8>] rpm_suspend+0xe0/0x4e4
[ 33.359813] [<ffff0000085bb784>] pm_runtime_work+0xdc/0xf0
[ 33.365028] [<ffff0000080d7b30>] process_one_work+0x12c/0x28c
[ 33.370576] [<ffff0000080d7ce8>] worker_thread+0x58/0x3b8
[ 33.376393] [<ffff0000080dd4a8>] kthread+0x100/0x12c
[ 33.381776] [<ffff0000080836c0>] ret_from_fork+0x10/0x50
Fix this by disabling it as the first thing in phy_exit().
Fixes: e78f3d15e1 ("phy: qcom-qmp: new qmp phy driver for qcom-chipsets")
Signed-off-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Signed-off-by: Manu Gautam <mgautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Fixing the clk enable failure path in qcom_qmp_phy_init()
and cleanup the reset control deassertion failure path in
qcom_qmp_phy_com_init().
Fixes: e78f3d15e1 ("phy: qcom-qmp: new qmp phy driver for qcom-chipsets")
Cc: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Add definitions required to enable QMP phy support for IPQ8074.
Signed-off-by: smuthayy <smuthayy@codeaurora.org>
Signed-off-by: Varadarajan Narayanan <varada@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Presently, the phy pipe clock's name is assumed to be either
usb3_phy_pipe_clk_src or pcie_XX_pipe_clk_src (where XX is the
phy lane's number). However, this will not work if an SoC has
more than one instance of the phy. Hence, instead of assuming
the name of the clock, fetch it from the DT.
Reviewed-by: Vivek Gautam <vivek.gautam@codeaurora.org>
Signed-off-by: Varadarajan Narayanan <varada@codeaurora.org>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Evan Green