OMAP5432 did go into production with AVS class0 registers which were
mutually exclusive from AVS Class 1.5 registers.
Most OMAP5-uEVM boards use the pre-production Class1.5 which has
production efuse registers set to 0. However on production devices,
these are set to valid data.
scale_vcore logic is already smart enough to detect this and use the
"Nominal voltage" on devices that do not have efuse registers populated.
On a test production device populated as follows:
MPU OPP_NOM:
=> md.l 0x04A0021C4 1
4a0021c4: 03a003e9 ....
(0x3e9 = 1.01v) vs nom voltage of 1.06v
MPU OPP_HIGH:
=> md.l 0x04A0021C8 1
4a0021c8: 03400485 ..@.
MM OPP_NOM:
=> md.l 0x04A0021A4 1
4a0021a4: 038003d4 ....
(0x3d4 = 980mV) vs nom voltage of 1.025v
MM OPP_OD:
=> md.l 0x04A0021A8 1
4a0021a8: 03600403 ..`.
CORE OPP_NOM:
=> md.l 0x04A0021D8 1
4a0021d8: 000003cf ....
(0x3cf = 975mV) vs nom voltage of 1.040v
Since the efuse values are'nt currently used, we do not regress on
existing pre-production samples (they continue to use nominal voltage).
But on boards that do have production samples populated, we can leverage
the optimal voltages necessary for proper operation.
Tested on:
a) 720-2644-001 OMAP5UEVM with production sample.
b) 750-2628-222(A) UEVM5432G-02 with pre-production sample.
Data based on OMAP5432 Technical reference Manual SWPU282AF (May
2012-Revised Aug 2016)
NOTE: All collaterals on OMAP5432 silicon itself seems to have been
removed from ti.com, though EVM details are still available:
http://www.ti.com/tool/OMAP5432-EVM
Signed-off-by: Nishanth Menon <nm@ti.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Refactor OMAP3/4/5 code so that we have only one get_device_type()
function for all platforms.
Details:
- Add ctrl variable for AM33xx and OMAP3 platforms (like it's done for
OMAP4/5), so we can obtain status register in common way
- For now ctrl structure for AM33xx/OMAP3 contains only status register
address
- Run hw_data_init() in order to assign ctrl to proper structure
- Remove DEVICE_MASK and DEVICE_GP definitions as they are not used
(DEVICE_TYPE_MASK and GP_DEVICE are used instead)
- Guard structs in omap_common.h with #ifdefs, because otherwise
including omap_common.h on non-omap4/5 board files breaks compilation
Buildman script was run for all OMAP boards. Result output:
arm: (for 38/616 boards)
all +352.5
bss -1.4
data +3.5
rodata +300.0
spl/u-boot-spl:all +284.7
spl/u-boot-spl:data +2.2
spl/u-boot-spl:rodata +252.0
spl/u-boot-spl:text +30.5
text +50.4
(no errors to report)
Tested on AM57x EVM and BeagleBoard xM.
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
[trini: Rework the guards as to not break TI81xx]
Signed-off-by: Tom Rini <trini@konsulko.com>
While in theory this value could be used in places outside of "omap5"
(such as OMAP4), we only make use of it today in OMAP5, so place the
Kconfig entry there. Given that Kconfig lets us provide a default, we
drop CONFIG_DEFAULT_OMAP_RESET_TIME_MAX_USEC entirely. The contents of
doc/README.omap-reset-time make a good help entry, so adjust them
slightly and delete the file. Move the comment about range to where we
use the value now, and have Kconfig enforce the upper bound.
Signed-off-by: Tom Rini <trini@konsulko.com>
GPIO_TO_PIN(bank, bank_gpio) returns the GPIO index
from the GPIO bank number and bank's GPIO offset number.
Signed-off-by: Roger Quadros <rogerq@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
When using early malloc the allocated memory can overflow into the SRAM
scratch space, move NON_SECURE_SRAM_IMG_END down a bit to allow more
dynamic allocation at the expense of a slightly smaller maximum image
size.
Signed-off-by: Andrew F. Davis <afd@ti.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Update the board pinmux for AM572x-IDK board using latest PMT[1] and the
board files named am572x_gp_evm_A3a_sr2p0 that were autogenerated on
19th October, 2016 by "Ahmad Rashed<a-rashed@ti.com>".
[1] https://dev.ti.com/pinmux/app.html#/default/
Signed-off-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Add the pmic_data for LP873x PMIC which is used to power
up dra71x-evm.
Note: As per the DM[1] DRA71x supports only OP_NOM. So, updating
the efuse registers only to use OPP_NOM irrespective of any
CONFIG_DRA7_<VOLT>_OPP_{NOM,od,high} is defined.
[1] http://www.ti.com/product/DRA718/technicaldocuments
Signed-off-by: Keerthy <j-keerthy@ti.com>
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Redefine the macros used to define the voltage values and the
efuse register offsets based on OPP for all the voltage domains.
This is done using Kconfig macros that can be set in a defconfig
or selected during a config step. This allows a voltage domain
to be configured/set to a corresponding voltage value depending
on the OPP selection choice.
The Kconfig choices have been added for MPU, DSPEVE, IVA and GPU
voltage domains, with the MPU domain restricted to OPP_NOM. The
OPP_OD and OPP_HIGH options will be added when the support for
configuring the MPU clock frequency is added. The clock
configuration for other voltage domains is out of scope in
u-boot code.
The CORE voltage domain does not have separate voltage values
and efuse register offset at different OPPs, while the MPU
voltage domain only has different efuse register offsets for
different OPPs, but uses the same voltage value. Any different
choices of OPPs for voltage domains on common ganged-rails
is automatically taken care to select the corresponding
highest OPP voltage value.
Signed-off-by: Suman Anna <s-anna@ti.com>
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
CONFIG_AM57XX is just an unnecessary macro that is redundant given So,
remove the same instead of spreading through out the u-boot source
code and getting in the way to maintain common code for DRA7x family.
Acked-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
On all TI platforms the ROM defines a "downloaded image" area at or near
the start of SRAM which is followed by a reserved area. As it is at
best bad form and at worst possibly harmful in corner cases to write in
this reserved area, we stop doing that by adding in the define
NON_SECURE_SRAM_IMG_END to say where the end of the downloaded image
area is and make SRAM_SCRATCH_SPACE_ADDR be one kilobyte before this.
At current we define the end of scratch space at 0x228 bytes past the
start of scratch space this this gives us a lot of room to grow. As
these scratch uses are non-optional today, all targets are modified to
respect this boundary.
Tested on OMAP4 Pandaboard, OMAP3 Beagle xM
Cc: Albert Aribaud <albert.u.boot@aribaud.net>
Cc: Nagendra T S <nagendra@mistralsolutions.com>
Cc: Vaibhav Hiremath <hvaibhav@ti.com>
Cc: Lokesh Vutla <lokeshvutla@ti.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Igor Grinberg <grinberg@compulab.co.il>
Cc: Nikita Kiryanov <nikita@compulab.co.il>
Cc: Paul Kocialkowski <contact@paulk.fr>
Cc: Enric Balletbo i Serra <eballetbo@gmail.com>
Cc: Adam Ford <aford173@gmail.com>
Cc: Steve Sakoman <sakoman@gmail.com>
Cc: Stefan Roese <sr@denx.de>
Cc: Thomas Weber <weber@corscience.de>
Cc: Hannes Schmelzer <oe5hpm@oevsv.at>
Cc: Thomas Chou <thomas@wytron.com.tw>
Cc: Masahiro Yamada <yamada.masahiro@socionext.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Joe Hershberger <joe.hershberger@ni.com>
Cc: Sam Protsenko <semen.protsenko@linaro.org>
Cc: Heiko Schocher <hs@denx.de>
Cc: Samuel Egli <samuel.egli@siemens.com>
Cc: Michal Simek <michal.simek@xilinx.com>
Cc: Wolfgang Denk <wd@denx.de>
Cc: Mateusz Kulikowski <mateusz.kulikowski@gmail.com>
Cc: Ben Whitten <ben.whitten@gmail.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Bin Meng <bmeng.cn@gmail.com>
Cc: Sekhar Nori <nsekhar@ti.com>
Cc: Mugunthan V N <mugunthanvnm@ti.com>
Cc: "B, Ravi" <ravibabu@ti.com>
Cc: "Matwey V. Kornilov" <matwey.kornilov@gmail.com>
Cc: Ladislav Michl <ladis@linux-mips.org>
Cc: Ash Charles <ashcharles@gmail.com>
Cc: "Kipisz, Steven" <s-kipisz2@ti.com>
Cc: Daniel Allred <d-allred@ti.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Tested-by: Lokesh Vutla <lokeshvutla@ti.com>
Acked-by: Lokesh Vutla <lokeshvutla@ti.com>
Tested-by: Ladislav Michl <ladis@linux-mips.org>
The register offset of i2c_sysc offset is not correct as per
omap5[1]/dra7[2] TRM, correct the offsets as per the
documentation.
[1] - http://www.ti.com/lit/pdf/swpu249
[2] - http://www.ti.com/lit/pdf/spruhz6
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
The pmic registers for variants of am57xx boards are different
hence we need to assign them carefully based on the board type.
Add a function to assign omap_vcores after the board detection.
Signed-off-by: Keerthy <j-keerthy@ti.com>
Define specific macros for the voltage values for all voltage
domains for all applicable OPPs - OPP_NOM, OPP_OD and OPP_HIGH.
No separate macros are defined for VD_MPU and VD_CORE at OPP_OD
and OPP_HIGH as these use the same values as OPP_NOM.
The current macros will be used as common macros that can be
redefined appropriately based on a selected OPP configuration
at build time.
Signed-off-by: Suman Anna <s-anna@ti.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
The voltage values for each voltage domain at an OPP is identical
across all the SoCs in the DRA7 family. The current code defines
one set of macros for DRA75x/DRA74x SoCs and another set for DRA72x
macros. Consolidate both these sets into a single set.
This is done so as to minimize the number of macros used when voltage
values will be added for other OPPs as well.
Signed-off-by: Suman Anna <s-anna@ti.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Define a set of common macros for the efuse register offsets
(different for each OPP) that are used to get the AVS Class 0
voltage values and ABB configuration values. Assign these
common macros to the register offsets for OPP_NOM by default
for all voltage domains. These common macros can then be
redefined properly to point to the OPP specific efuse register
offset based on the desired OPP to program a specific voltage
domain.
Signed-off-by: Suman Anna <s-anna@ti.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
The current OPP_NOM voltage values defined for the MPU and CORE
voltage domains are based on the initial DRA75x_74x_SR1.1_DM data
manual. As per this DM, the PMIC boot voltage can be set to either
1.10V or 1.15V for VD_MPU, and either 1.06V or 1.15V for VD_CORE.
While the current values are correct, the latter set of values
are the values that are common across all DRA75x, DRA72x SoCs and
for all current Silicon revisions. So, update both the MPU and CORE
OPP_NOM voltages to 1.15V.
The macros are also slightly reorganized so that both the MPU and
CORE voltage domain values are defined together.
Signed-off-by: Suman Anna <s-anna@ti.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Add platforms specific phy mode configuration bits to be used
to configure phy mode in control module.
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
ABB should be initialized for all required domains voltage domain
for DRA7: IVA, GPU, EVE in addition to the existing MPU domain. If
we do not do this, kernel configuring just the frequency using the
default boot loader configured voltage can fail on many corner lot
units and has been hard to debug. This specifically is a concern with
DRA7 generation of SoCs since other than VDD_MPU, all other domains
are only permitted to setup the voltages to required OPP only at boot.
Reported-by: Richard Woodruff <r-woodruff2@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Since we setup the voltage and frequency for the MM domain, we *must*
setup the ABB configuration needed for the domain as well. If we do not
do this, kernel configuring just the frequency using the default boot
loader configured voltage can fail on many corner lot units.
Reported-by: Richard Woodruff <r-woodruff2@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
do_set_iodelay can now be used from board files based on needs of the
platforms variation they have.
Signed-off-by: Nishanth Menon <nm@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Since many platforms may need different pad configuration required
depending on variation of the platform with minor deltas, it is
easier to maintain a sub step based approach to allow for pin mux
and iodelay configuration which may depend on the platform variations
and need to be done in IO isolation.
While we retain the older __recalibrate_iodelay function which provides
a ready sequencing, __recalibrate_iodelay_start and
__recalibrate_iodelay_end may be alternatively used now and the callers
will be responsible for the correct sequencing of operations.
Signed-off-by: Nishanth Menon <nm@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Add support for detection of SR2.0 version of DRA72x family of
processors.
Signed-off-by: Ravi Babu <ravibabu@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
There is no distinction between essential and non-essential mux configuration,
so it doesn't make sense to have an "essential" prefix.
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
Many TI EVMs have capability to store relevant board information
such as DDR description in EEPROM. Further many pad configuration
variations can occur as part of revision changes in the platform.
In-order to support these at runtime, we for a board detection hook
which is available for override from board files that may desire to do
so.
NOTE: All TI EVMs are capable of detecting board information based on
early clocks that are configured. However, in case of additional needs
this can be achieved within the override logic from within the board
file.
Signed-off-by: Steve Kipisz <s-kipisz2@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Centralize gpi2c_init into omap_common from the sys_proto header so
that the information can be reused across SoCs.
Signed-off-by: Steve Kipisz <s-kipisz2@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Early clock initialization is currently done in two stages for OMAP4/5
SoCs. The first stage is the initialization of console clocks and
then we initialize basic clocks for functionality necessary for SoC
initialization and basic board functionality.
By splitting up prcm_init and centralizing this clock initialization,
we setup the code for follow on patches that can do board specific
initialization such as board detection which will depend on these
basic clocks.
As part of this change, since the early clock initialization
is centralized, we no longer need to expose the console clock
initialization.
NOTE: we change the sequence slightly by initializing console clocks
timer after the io settings are complete, but this is not expected
to have any functioanlity impact since we setup the basic IO drive
strength initialization as part of do_io_settings.
Signed-off-by: Steve Kipisz <s-kipisz2@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
introduce BIT() definition, used in at91_udc gadget
driver.
Signed-off-by: Heiko Schocher <hs@denx.de>
[remove all other occurrences of BIT(x) definition]
Signed-off-by: Andreas Bießmann <andreas.devel@googlemail.com>
Acked-by: Stefan Roese <sr@denx.de>
Acked-by: Anatolij Gustschin <agust@denx.de>
Implemented board_usb_init(), board_usb_cleanup() and
usb_gadget_handle_interrupts() in omap5 board file that
can be invoked by various gadget drivers.
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Implemented board_usb_init(), board_usb_cleanup() and
usb_gadget_handle_interrupts() in beagle_x15 board file that
can be invoked by various gadget drivers.
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Enabled clocks for the second dwc3 controller and second USB PHY present in
DRA7.
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
DP83865 ethernet phy used on DRA74x-evm is quirky and the datasheet
provided IODELAY values for standard RGMII phys do not work.
Silicon Revision(SR) 2.0 provides an alternative bit configuration
that allows us to do a "gross adjustment" to launch the data off a
different internal clock edge. Manual IO Delay overrides are still
necessary to fine tune the clock-to-data delays. This is a necessary
workaround for the quirky ethernet Phy we have on the platform.
NOTE: SMA registers are spare "kitchen sink" registers that does
contain bits for other workaround as necessary as well. Hence the
control for the same is introduced in a generic SoC specific, board
generic location.
Signed-off-by: Nishanth Menon <nm@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Add support for detection of ES2.0 version of DRA7 family of
processors. ES2.0 is an incremental revision with various fixes
including the following:
- reset logic fixes
- few assymetric aging logic fixes
- MMC clock rate fixes
- Ethernet speed fixes
- edma fixes for mcasp
[ravibabu@ti.com: posted internal for an older bootloader]
Signed-off-by: Ravi Babu <ravibabu@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Acked-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Enable TI_EDMA3 and SPL_DMA support, so as to reduce boot time. With
DMA enabled there is almost 3x improvement in read performance. This
helps in reducing boot time in qspiboot mode
Also add EDMA3 base address for DRA7XX and AM57XX.
Signed-off-by: Vignesh R <vigneshr@ti.com>
Reviewed-by: Jagan Teki <jteki@openedev.com>
This cleans up the SPL boot devices for omap platforms and introduces support
for missing boot devices.
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
This introduces OMAP3 support for the common omap boot code, as well as a
major cleanup of the common omap boot code.
First, the omap_boot_parameters structure becomes platform-specific, since its
definition differs a bit across omap platforms. The offsets are removed as well
since it is U-Boot's coding style to use structures for mapping such kind of
data (in the sense that it is similar to registers). It is correct to assume
that romcode structure encoding is the same as U-Boot, given the description
of these structures in the TRMs.
The original address provided by the bootrom is passed to the U-Boot binary
instead of a duplicate of the structure stored in global data. This allows to
have only the relevant (boot device and mode) information stored in global data.
It is also expected that the address where the bootrom stores that information
is not overridden by the U-Boot SPL or U-Boot.
The save_omap_boot_params is expected to handle all special cases where the data
provided by the bootrom cannot be used as-is, so that spl_boot_device and
spl_boot_mode only return the data from global data.
All of this is only relevant when the U-Boot SPL is used. In cases it is not,
save_boot_params should fallback to its weak (or board-specific) definition.
save_omap_boot_params should not be called in that context either.
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
Now all manual mode configurations are done as part of
IO delay recalibration sequence, remove the hack done for
CPSW.
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
In addition to the regular mux configuration, certain pins of DRA7
require to have "manual mode" also programmed, when predefined
delay characteristics cannot be used for the interface.
struct iodelay_cfg_entry is introduced for populating
manual mode IO timings.
For configuring manual mode, along with the normal pad
configuration do the following steps:
- Select MODESELECT field of each assocaited PAD.
CTRL_CORE_PAD_XXX[8]:MODESELECT = 1(Enable MANUAL_MODE macro along with mux)
- Populate A_DELAY, G_DELAY values that are specified in DATA MANUAL.
And pass the offset of the CFG_XXX register in iodelay_cfg_entry.
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
On DRA7, in addition to the regular muxing of pins, an additional
hardware module called IODelay which is also expected to be
configured. This "IODelay" module has it's own register space that is
independent of the control module.
It is advocated strongly in TI's official documentation considering
the existing design of the DRA7 family of processors during mux or
IODelay recalibration, there is a potential for a significant glitch
which may cause functional impairment to certain hardware. It is
hence recommended to do muxing as part of IOdelay recalibration.
IODELAY recalibration sequence:
- Complete AVS voltage change on VDD_CORE_L
- Unlock IODLAY config registers.
- Perform IO delay calibration with predefined values.
- Isolate all the IOs
- Update the delay mechanism for each IO with new calibrated values.
- Configure PAD configuration registers
- De-isolate all the IOs.
- Relock IODELAY config registers.
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
In addition to the regular mux configuration, certain pins of DRA7
require to have "virtual mode" also programmed.
This allows for predefined delay characteristics to be used by the SoC
to meet timing characterstics needed for the interface.
Provide easy to use macro to do the same.
For configuring virtual mode, along with normal pad configuration add
the following two steps:
- Select MODESELECT field of each assocaited PAD.
CTRL_CORE_PAD_XXX[8]:MODESELECT = 1
- DELAYMODE filed should be configured with value given in DATA Manual.
CTRL_CORE_PAD_XXX[7:4]:DELAYMODE =[0-15] (as given in DATA manual).
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Adopting the pinctrl register definitions from Linux kernel
to be consistent.
Old definitions will be removed once all the pinctrl data
is adapted to new definitions.
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
do_set_mux32() is redefined in dra7xx and beagle_x15 boards.
IO delay recalibration sequence also needs this.
Making it generic to avoid duplication.
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Implemented board_usb_init(), board_usb_cleanup() and
usb_gadget_handle_interrupts() in dra7xx board file that
can be invoked by various gadget drivers.
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
This patch enables the workaround for ARM errata 798870 for OMAP5 /
DRA7 which says "If back-to-back speculative cache line fills (fill
A and fill B) are issued from the L1 data cache of a CPU to the
L2 cache, the second request (fill B) is then cancelled, and the
second request would have detected a hazard against a recent write or
eviction (write B) to the same cache line as fill B then the L2 logic
might deadlock."
An l2auxctlr accessor implementation for OMAP5 and DRA7 is introduced
here as well.
Signed-off-by: Praveen Rao <prao@ti.com>
Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Tested-by: Matt Porter <mporter@konsulko.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
On the DRA72x (J6Eco) EVM one PMIC SMPS is powering three SoC
core rails. This concept of using one SMPS to supply multiple
core domains (in various, although limited combinations, per
primary device use case) has now become common and is used by
many customer J6/J6Eco designs; it is supported by a number of
corresponding PMIC OTP versions.
This patch implements correct operation of the core voltages
scaling routine by ensuring that each SMPS that is supplying
more than one domain shall be written only once, and with the
highest voltage of those fused in the SoC (or of those defined
in the corresponding header if fuse read is disabled or fails)
for the power rails belonging to the group.
The patch also replaces some PMIC-related magic numbers with
the appropriate definitions. The default OPP_NOM voltages for
the DRA7xx SoCs are updated as well, per the latest DMs.
Signed-off-by: Lubomir Popov <l-popov@ti.com>
just add a few ifdefs around because this
device is very similar to dra7xxx.
Signed-off-by: Felipe Balbi <balbi@ti.com>
Reviewed-by: Tom Rini <trini@ti.com>