The set_parent implementations do not make sense when OF_PLATDATA is
enabled. We guard these against OF_PLATDATA and don't populate the
set_parent-op when this is the case.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Implement the setting parent for gmac clock, and add internal
pll div set for mac clk.
Signed-off-by: David Wu <david.wu@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The RK3288 CRU-node assigns rates to a number of clocks that are not
implemented in the RK3288 clock-driver (but which have been
sufficiently initialised from rkclk_init()): for these clocks, we
implement the gmac clock set parent, but simply ignore the
others' set_rate() operation and return 0 to signal success.
Signed-off-by: David Wu <david.wu@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Assuming mac_clk is fed by an external clock, set clk_rmii_src
clock select control register from IO for rgmii interface.
Signed-off-by: David Wu <david.wu@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The rk3328 soc has two gmac controllers, one is gmac2io,
the other is gmac2phy. We use the gmac2io rgmii interface
for 1000M phy here.
Signed-off-by: David Wu <david.wu@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The RK3399 CRU-node assigns rates to a number of clocks that are not
implemented in the RK3399 clock-driver (but which have been
sufficiently initialised from rkclk_init()): for these clocks, we
simply ignore the set_rate() operation and return 0 to signal success.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Tested-by: David Wu <david.wu@rock-chips.com>
Series-changes: 2
- Fixed David's email address.
This implements the (newly added) set_parent() operation for the
RK3399 with a focus on allowing the RGMII clock parent to be
configured via the assigned-clock-parents property of the GMAC node.
This implementation supports only the GMAC (in fact only the RGMII
clock parent) and allows to set this clock's parent either to the
internal SCLK_GMAC or to an external clock input (identifiable by it
providing a 'clock-output-name' of "gmac_clkin").
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Tested-by: David Wu <david.wu@rock-chips.com>
Series-changes: 2
- Fixed David's email address.
Bind rockchip reset to clock-controller with rockchip_reset_bind().
Signed-off-by: Elaine Zhang <zhangqing@rock-chips.com>
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Add rk3128 clock driver and cru structure definition.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The RK3399 clk driver still has a left-over use of extract_bits, which
can be replaced by using bitfield_extract from include/bitfield.h.
This rewrites the invocation to use the shared function.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Using priv for new sysreset driver binding.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Init the CPU and its buses to speed up the boot time.
Move rkclk_init() to a place after rk3399_configure_cpu has defined
at the same time, or else there will be a warning.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
U-Boot widely uses error() as a bit noisier variant of printf().
This macro causes name conflict with the following line in
include/linux/compiler-gcc.h:
# define __compiletime_error(message) __attribute__((error(message)))
This prevents us from using __compiletime_error(), and makes it
difficult to fully sync BUILD_BUG macros with Linux. (Notice
Linux's BUILD_BUG_ON_MSG is implemented by using compiletime_assert().)
Let's convert error() into now treewide-available pr_err().
Done with the help of Coccinelle, excluing tools/ directory.
The semantic patch I used is as follows:
// <smpl>
@@@@
-error
+pr_err
(...)
// </smpl>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
[trini: Re-run Coccinelle]
Signed-off-by: Tom Rini <trini@konsulko.com>
The pd_bus hclk/pclk source is pd_bus aclk, not the PLL.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Fix typo RK322X/RK3036 in rk322x clock driver.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The clk_saradc is dividing from the 24M, clk_saradc=24MHz/(saradc_div_con+1).
SARADC integer divider control register is 8-bits width.
Signed-off-by: David Wu <david.wu@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The clk_saradc is dividing from the 24M, clk_saradc=24MHz/(saradc_div_con+1).
SARADC integer divider control register is 8-bits width.
Signed-off-by: David Wu <david.wu@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The clk_saradc is dividing from the 24M, clk_saradc=24MHz/(saradc_div_con+1).
SARADC integer divider control register is 10-bits width.
Signed-off-by: David Wu <david.wu@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The clk_saradc is dividing from the 24M, clk_saradc=24MHz/(saradc_div_con+1).
SARADC integer divider control register is 8-bits width.
Signed-off-by: David Wu <david.wu@rock-chips.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The clk_saradc is dividing from the 24M, clk_saradc=24MHz/(saradc_div_con+1).
SARADC integer divider control register is 10-bits width.
Signed-off-by: David Wu <david.wu@rock-chips.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Update the clock driver for the RK3399 to support a live device tree.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The generic ehci-driver (ehci-generic.c) will try to enable the clocks
listed in the DTSI. If this fails (e.g. due to clk_enable not being
implemented in a driver and -ENOSYS being returned by the clk-uclass),
the driver will bail our and print an error message.
This implements a minimal clk_enable for the RK3399 and supports the
clocks mandatory for the EHCI controllers; as these are enabled by
default we simply return success.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Update the clock driver for the RK3368 to support a live device tree.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Version-changes: 2
- use the dev_read_addr_ptr function in clk_rk3368.c
When using 32-bit addresses dtoc works correctly. For 64-bit addresses it
does not since it ignores the #address-cells and #size-cells properties.
Update the tool to use fdt64_t as the element type for reg properties when
either the address or size is larger than one cell. Use the correct value
so that C code can obtain the information from the device tree easily.
Alos create a new type, fdt_val_t, which is defined to either fdt32_t or
fdt64_t depending on the word size of the machine. This type corresponds
to fdt_addr_t and fdt_size_t. Unfortunately we cannot just use those types
since they are defined to phys_addr_t and phys_size_t which use
'unsigned long' in the 32-bit case, rather than 'unsigned int'.
Add tests for the four combinations of address and size values (32/32,
64/64, 32/64, 64/32). Also update existing uses for rk3399 and rk3368
which now need to use the new fdt_val_t type.
Signed-off-by: Simon Glass <sjg@chromium.org>
Suggested-by: Heiko Stuebner <heiko@sntech.de>
Reported-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Tested-by: Kever Yang <kever.yang@rock-chips.com>
Use DIV_ROUND_UP instead RATE_TO_DIV for all Rockchip SoC
clock driver.
Add or fix the div-field overflow check at the same time.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
dwmmc controller has default internal divider by 2,
and we always provide double of the clock rate request by
dwmmc controller. Sync code for all Rockchip SoC with:
4055b46 rockchip: clk: rk3288: fix mmc clock setting
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Acked-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
[fixup for 'missing DIV_ROUND_UP' conflict for clk_rk3288.c:]
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
As SPI support may be useful in the boot-flow, this adds support for
configuring the SPI controller's clocks in the RK3368 clock driver.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
With the clock support in rk3368_clk_set_rate() conditionalized on
various feature definitions, 'priv' can remain unused (e.g. in the
SPL build when only MMC is enabled).
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
To enable the GMAC on the RK3368, we need to set up the clocking
appropriately to generate a tx_clk for the MAC.
This adds an implementation that implements the use of the <&ext_gmac>
clock (i.e. an external 125MHz clock for RGMII provided by the PHY).
This is the clock setup used by the boards currently supported by
U-Boot (i.e. Geekbox, Sheep and RK3368-uQ7).
This includes the change from commit
- rockchip: clk: rk3368: define GMAC_MUX_SEL_EXTCLK
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
As part of the DRAM initialisation process (running as part of the TPL
stage) on the RK3368, we need to set up the DRAM PLL.
This implements support for configuring the PLL to for 1200, 1332 or
1600 MHz (i.e. for DDR3-1200, DDR3-1333, DDR3-1600 operating modes).
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The original clock support for MMC/SD cards on the RK3368 suffered
from a tendency to select a divider less-or-equal to the the one
giving the requested clock-rate: this can lead to higher-than-expected
(or rather: higher than supported) clock rates for the MMC/SD
communiction.
This change rewrites the MMC/SD clock generation to:
* always generate a clock less-than-or-equal to the requested clock
* support reparenting among the CPLL, GPLL and OSC24M parents to
generate the highest clock that does not exceed the requested rate
In addition to this, the Linux DTS uses HCLK_MMC/HCLK_SDMMC instead of
SCLK_MMC/SCLK_SDMMC: to match this (and to ensure that clock setup
always works), we adjust the driver appropriately.
This includes the changes from:
- rockchip: clk: rk3368: convert MMC_PLL_SEL_* definitions to shifted-value form
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
To implement a TPL stage (incl. its DRAM controller setup) for the
RK3368, we'll want to configure the DPLL (DRAM PLL).
This commit implements setting the DPLL (CLK_DDR) and provides PLL
configuration details for the common DRAM operating speeds found on
RK3368 boards.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The RK3368 has a somewhat temperamental BootROM (which I learned the
hard way) when it comes to reconfiguring the CPLL and GPLL (in fact,
experiments show that changing the GPLL broke things for me, while
changing the CPLL seems to be more benign). These should not be
modified by the SPL stage, if we intend to return to the BootROM for
chain booting the next stage.
This commit changes the clock initialisation to not change CPLL/GPLL
before returning to the BootROM (i.e. in TPL). As it's safe to change
these settings if we no longer intend to return to U-Boot, we'll run
the full PLL setup a little later (i.e. in SPL).
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
With the RK3368's limited TPL size, we'll want to use OF_PLATFDATA for
the SPL stage. This implements support for OF_PLATDATA in the clock
driver for the RK3368.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The RK3368 TRM recommends to configure the bandwith adjustment (CON2)
for PLLs to NF/2. This implements this for all reconfigurations of
PLLs and removes the 'has_bwadj' flag (as the RK3368 always has the
bandwidth-adjustment feature according to its manual).
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The clk driver for the RK3368 picked the wrong data structure's size
for its auto-alloc size: the size was calculated on the structure
representing the CRU hardware block instead of the priv structure.
As the CRU's register file is much larger than the driver's priv,
this did not cause any pain (except wasting memory).
Fix this by using the correct data structure's size.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
According to rk3036 TRM, pll_con1[12] should be set to '1' for the pll
integer mode, while the '0' means the frac mode.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
The variables gpll_init_cfg and apll_init_cfg are unused in this file,
remove them.
Cc: Simon Glass <sjg@chromium.org>
Cc: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Detect with a previous boot loader has already set up the clocks and set
them up again so that U-Boot gets what it expects.
Signed-off-by: Simon Glass <sjg@chromium.org>
Shifted masks are the standard approach with rockchip since it allows
use of the mask without shifting it each time. Update the definitions and
the driver to match.
Signed-off-by: Simon Glass <sjg@chromium.org>
- hclk/pclk_div range should use '<=' instead of '<'
- use GPLL for pd_bus clock source
- pd_bus HCLK/PCLK clock rate should not bigger than ACLK
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Embeded the shift in mask MACRO definition in cru header file
and clock driver.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Add driver to setup the various PLLs and peripheral
clocks on the RK3368.
Signed-off-by: Andy Yan <andy.yan@rock-chips.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
These support the flat device tree. We want to use the dev_read_..()
prefix for functions that support both flat tree and live tree. So rename
the existing functions to avoid confusion.
In the end we will have:
1. dev_read_addr...() - works on devices, supports flat/live tree
2. devfdt_get_addr...() - current functions, flat tree only
3. of_get_address() etc. - new functions, live tree only
All drivers will be written to use 1. That function will in turn call
either 2 or 3 depending on whether the flat or live tree is in use.
Note this involves changing some dead code - the imx_lpi2c.c file.
Signed-off-by: Simon Glass <sjg@chromium.org>
This allows requests (via the DTS) for PCLK_HDMI_CTRL/PCLK_VIO_GRF,
which are clock gates in the HDMI output path for the RK3399.
As these are enabled by default (i.e. after reset), we don't implement
any logic to actively open/close these clock gates and simply assume
that their reset-default has not been changed.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The (non-secure) efuse node in the DTS requests PCLK_EFUSE1024NS.
To allow us to add a efuse-driver (and more importantly, to allow
probes of such a driver to succeed), we need need to accept requests
for PCLK_EFUSE1024NS and return a non-error result.
As PCLK_EFUSE1024NS is enabled by default (i.e. after reset), we don't
implement any logic to manage this clock gate and simply assume that
the reset-default has not been changed.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Tested-by: Klaus Goger <klaus.goger@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The clocking of the designware MMC controller in the upstream
(i.e. Linux) RK3399 has changed/does not match what the current DTS in
U-Boot uses: the first clock entry now is HCLK_SDMMC instead of
SCLK_SDMMC.
With the simple clock driver used for the RK3399, this needs a change
in the selector understood by the various case statements in the driver
to ensure that the driver still loads successfully.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Acked-by: Simon Glass <sjg@chromium.org>
For the RK3399, i2c_set_rate (and by extension: our spi_set_rate,
which had been mindlessly following the template of the i2c_set_rate
implementation) miscalculates the rate returned due to a off-by-one
error resulting from the following sequence of events:
1. calculates 'src_div := src_freq / target_freq'
2. stores 'src_div - 1' into the register (the actual divider applied
in hardware is biased by adding 1)
3. returns the result of the DIV_RATE(src_freq, src_div) macro, which
expects the (decremented) divider from the hardware-register and
implictly adds 1 (i.e. 'DIV_RATE(freq, div) := freq / (div + 1)')
This can be observed with the SPI driver, which sets a rate of 99MHz
based on the GPLL frequency of 594MHz: the hardware generates a clock
of 99MHz (src_div is 6, the bitfield in the register correctly reads 5),
but reports a frequency of 84MHz (594 / 7) on return.
To fix, we have two options:
* either we bias (i.e. "DIV_RATE(GPLL, src_div - 1)"), which doesn't
make for a particularily nice read
* we simply call the i2c/spi_get_rate function (introducing additional
overhead for the additional register-read), which reads the divider
from the register and then passes it through the DIV_RATE macro
Given that this code is not time-critical, the more readable solution
(i.e. calling the appropriate get_rate function) is implemented in this
change.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Tested-by: Klaus Goger <klaus.goger@theobroma-systems.com>
Acked-by: Simon Glass <sjg@chromium.org>
Philipp Tomsich
David Wu
Elaine Zhang
Kever Yang
Masahiro Yamada
Simon Glass
Tom Rini
Andy Yan