Enable raw & on disk capsule updates and provide configs required
for updating MTD devices. Also resync after savedefconfig.
Signed-off-by: Wadim Egorov <w.egorov@phytec.de>
Siddharth Vadapalli <s-vadapalli@ti.com> says:
Hello,
This series adds support for USB DFU boot on TI's AM62A SoC which has
two instances of DWC3 USB Controllers namely USB0 and USB1. The USB0
instance of the USB Controller supports USB DFU boot:
ROM => tiboot3.bin => tispl.bin => u-boot.img
USB DFU Boot requires the USB Controller to be configured for Gadget
mode of operation. Since the USB0 instance of the DWC3 USB Controller
supports both Host and Gadget modes of operation via the Type-C interface
on the AM62A7-SK board, the device-tree specifies the "dr_mode" as "OTG".
However, there is currently no support for dynamically switching the "mode"
from Host to Gadget and vice-versa with the help of a state-machine.
The OTG mode is treated as a separate mode in itself rather than being
treated as an intermediate stage before assuming the Host/Gadget mode.
Due to this, USB DFU boot via the Type-C interface doesn't work as the
USB Controller hasn't been appropriately configured for Device/Gadget
mode of operation. One option is to change the device-tree to specify
"dr_mode" as "peripheral" and force the controller to assume the Device
role. This will imply that the U-Boot device-tree for AM62A diverges
from its Linux counterpart. Therefore, with the intent of keeping the
device-tree uniform across Linux and U-Boot, and at the same time, in
order to enable USB DFU boot in "OTG" mode with the DWC3 Controller,
the first patch in this series sets the "mode" on the basis of the
caller function, rather than using the "dr_mode" property in the
device-tree. There are only two callers of "dwc3_generic_probe()",
each of which clearly specify the expected mode of configuration.
This will enable both Host and Device mode of operation based on the
command executed by the user, thereby truly supporting "OTG"
functionality when the USB Controller supports it.
The second patch in this series adds USB DFU environment for AM62A,
enabling USB DFU Boot and USB DFU flash on AM62A.
In addition to the patches in this series, the following device-tree
changes will be required to test USB DFU on AM62A (bootph-all property
to be added to ensure that USB Controller is present at all stages
for DFU Boot):
https://gist.github.com/Siddharth-Vadapalli-at-TI/53ba02cb0ff4a09c47e920d08247065f
The above device-tree changes will be made to the Linux device-tree,
which shall ensure that the same shall be a part of U-Boot device-tree
eventually.
The USB DFU config fragments for AM62x have been used for enabling
USB DFU boot on AM62a as follows:
R5 => am62ax_evm_r5_defconfig + am62x_r5_usbdfu.config
A53 => am62ax_evm_a53_defconfig + am62x_a53_usbdfu.config
Logs validating USB DFU boot with this series:
https://gist.github.com/Siddharth-Vadapalli-at-TI/daa71da1b0e478a51afea42605fb2d2c
Link: https://lore.kernel.org/r/20241126120322.1760862-1-s-vadapalli@ti.com
Include the TI K3 DFU environment to support DFU Boot and DFU Flash.
Signed-off-by: Siddharth Vadapalli <s-vadapalli@ti.com>
Reviewed-by: Roger Quadros <rogerq@kernel.org>
There are only two callers of "dwc3_generic_probe()", namely:
1. dwc3_generic_peripheral_probe()
2. dwc3_generic_host_probe()
Currently, the "mode" is set based on the device-tree node of the
platform device. Also, the DWC3 core doesn't support updating the "mode"
dynamically at runtime if it is set to "OTG", i.e. "OTG" is treated as a
separate mode in itself, rather than being treated as a mode which should
eventually lead to "host"/"peripheral".
Given that the callers of "dwc3_generic_probe()" clarify the expected
"mode" of the USB Controller, use that "mode" instead of the one
specified in the device-tree. This shall allow the USB Controller to
function both as a "Host" and as a "Peripheral" when the "mode" is "otg"
in the device-tree, based on the caller of "dwc3_generic_probe()".
Signed-off-by: Siddharth Vadapalli <s-vadapalli@ti.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Roger Quadros <rogerq@kernel.org>
Enables ESM configs for j721s2 and disables them for AM68 as AM68
includes J721s2 configs by default.
Signed-off-by: Manorit Chawdhry <m-chawdhry@ti.com>
Initialize the 3 instances of SOC ESM & PMIC ESM.
This is needed for watchdog functionality.
Signed-off-by: Keerthy <j-keerthy@ti.com>
Signed-off-by: Udit Kumar <u-kumar1@ti.com>
Enable Quality of Service (QoS) blocks for Display SubSystem (DSS), by
servicing the DSS - DDR traffic from the Real-Time (RT) queue. This is
done by setting the DSS DMA orderID to greater than 7.
DDR intensive software applications can overwhelm the DSS's access to
the DDR because of their higher frequency DDR accesses. This can cause
flickering in display with certain applications running parallely if
the DSS traffic is being serviced through non-RT queue.
Signed-off-by: Jayesh Choudhary <j-choudhary@ti.com>
Reviewed-by: Devarsh Thakkar <devarsht@ti.com>
Enable Quality of Service (QoS) blocks for Display SubSystem (DSS), by
servicing the DSS - DDR traffic from the Real-Time (RT) queue. This is
done by setting the DSS DMA orderID to greater than 7.
The C7x and VPAC can overwhelm the DSS's access to the DDR because of
their higher frequency DDR accesses. This can cause flickering in
display with certain edgeAI models running parallely if the DSS traffic
is being serviced through non-RT queue.
Signed-off-by: Jayesh Choudhary <j-choudhary@ti.com>
Reviewed-by: Devarsh Thakkar <devarsht@ti.com>
This reverts commit 1fdf53ace1, reversing
changes made to e5aef1bbf1.
I had missed that this caused too much size growth on rcar3_salvator-x.
Signed-off-by: Tom Rini <trini@konsulko.com>
VBE uses a crc8 checksum to verify that the nvdata is valid, so make
sure it is available if VBE is enabled.
Signed-off-by: Simon Glass <sjg@chromium.org>
With VBE we want to use FIT in all phases of the boot. Add Kconfig
options to support this.
Disable the options for sandbox_vpl for now.
Signed-off-by: Simon Glass <sjg@chromium.org>
The variable 'ret' is defined twice, which is not intended. This may
have been a local merge error.
Signed-off-by: Simon Glass <sjg@chromium.org>
Fixes: 2eefeb6d893 ("spl: Report a loader failure")
U-Boot uses ulong for addresses. It is confusing to use uintptr_t in a
few places, since it makes people wonder if the types are compatible.
Change the few occurences in SPL to use ulong
Signed-off-by: Simon Glass <sjg@chromium.org>
The current check looks only at SPL, but TPL or VPL might have a
different setting. Update the condition.
Signed-off-by: Simon Glass <sjg@chromium.org>
The current check looks only at SPL, but TPL or VPL might have a
different setting. Update the condition.
Signed-off-by: Simon Glass <sjg@chromium.org>
If a loader returns an error code it is silently ignored. Show a message
to at least provide some feedback to the user.
Signed-off-by: Simon Glass <sjg@chromium.org>
Some code has crept in which ignores this parameter. Fix this and add a
little debugging.
Signed-off-by: Simon Glass <sjg@chromium.org>
Fixes: b1307f884a ("fit: Support compression for non-kernel components (e.g. FDT)")
The name of the bootdev device is not that important, particular in SPL.
Save a little code space by using a simpler name.
Signed-off-by: Simon Glass <sjg@chromium.org>
When the best-match feature fails to find something, use the provided
config name as a fallback. The allows SPL to select a suitable config
when best-match is enabled.
Signed-off-by: Simon Glass <sjg@chromium.org>
The for() loop ends up being in the code even if the log_debug() does
nothing. Add a condition to fix this.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>
Sometimes clocks provided to a consumer might not have .set_rate
operation (like gate or mux clocks), but have CLK_SET_PARENT_RATE flag
set. In that case it's usually possible to find a parent up the tree
which is capable of setting the rate (div, pll, etc). Implement a simple
lookup procedure for such cases, to traverse the clock tree until
.set_rate capable parent is found, and use that parent to actually
change the rate. The search will stop once the first .set_rate capable
clock is found, which is usually enough to handle most cases.
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
The boot_fdt_add_mem_rsv_regions() function can be called twice, e.g.
first time during the board init (as a part of LMB init), and then when
booting the OS with 'booti' command:
lmb_add_region_flags
lmb_reserve_flags
boot_fdt_reserve_region
boot_fdt_add_mem_rsv_regions
^
|
+-----------------------+
| (1) | (2)
lmb_reserve_common image_setup_linux
lmb_init ...
initr_lmb do_booti
board_init_r 'booti'
That consequently leads to the attempt of reserving the same memory
areas (described in the 'reserved-memory' dts node) in LMB. The
lmb_add_region_flags() returns -EEXIST error code in such cases, but
boot_fdt_reserve_region() handles all negative error codes as a failure
to reserve fdt memory region, printing corresponding error messages,
which are essentially harmless, but misleading. For example, this is the
output of 'booti' command on E850-96 board:
=> booti $loadaddr - $fdtaddr
...
ERROR: reserving fdt memory region failed
(addr=bab00000 size=5500000 flags=2)
ERROR: reserving fdt memory region failed
(addr=f0000000 size=200000 flags=4)
...
Starting kernel ...
The mentioned false positive error messages are observed starting with
commit 1d9aa4a283 ("lmb: Fix the allocation of overlapping memory
areas with !LMB_NONE"), which removes the check for the already added
memory regions in lmb_add_region_flags(), making it return -1 for
!LMB_NONE cases. Another commit 827dee587b ("fdt: lmb: add reserved
regions as no-overwrite") changes flags used for reserving memory in
boot_fdt_add_mem_rsv_regions() from LMB_NONE to LMB_NOOVERWRITE. So
together with the patch mentioned earlier, it makes
lmb_add_region_flags() return -1 when called from
boot_fdt_reserve_region().
Since then, the different patch was implemented, returning -EEXIST error
code in described cases, which is:
lmb: Return -EEXIST in lmb_add_region_flags() if region already added
Handle -EEXIST error code as a normal (successful) case in
lmb_reserve_flags() and don't print any messages.
Fixes: 1d9aa4a283 ("lmb: Fix the allocation of overlapping memory areas with !LMB_NONE")
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Tested-by: Michal Simek <michal.simek@amd.com>
An attempt to add the already added LMB region using
lmb_add_region_flags() ends up in lmb_addrs_overlap() check, which
eventually leads to either returning 0 if 'flags' is LMB_NONE, or -1
otherwise. It makes it impossible for the user of this function to catch
the case when the region is already added and differentiate it from
regular errors. That in turn may lead to incorrect error handling in the
caller code, like reporting misleading errors or interrupting the normal
code path where it could be treated as the normal case. An example is
boot_fdt_reserve_region() function, which might be called twice (e.g.
during board startup in initr_lmb(), and then during 'booti' command
booting the OS), thus trying to reserve exactly the same memory regions
described in the device tree twice, which produces an error message on
second call.
Return -EEXIST error code in case when the added region exists and it's
not LMB_NONE; for LMB_NONE return 0, to conform to unit tests
(specifically test_alloc_addr() in test/lib/lmb.c) and the preferred
behavior described in commit 1d9aa4a283 ("lmb: Fix the allocation of
overlapping memory areas with !LMB_NONE"). The change of
lmb_add_region_flags() return values is described in the table below:
Return case Pre-1d9 1d9 New
-----------------------------------------------------------
Added successfully 0 0 0
Failed to add -1 -1 -1
Already added, flags == LMB_NONE 0 0 0
Already added, flags != LMB_NONE 0 -1 -EEXIST
Rework all affected functions and their documentation. Also fix the
corresponding unit test which checks reserving the same region with the
same flags to account for the changed return value.
No functional change is intended (by this patch itself).
Fixes: 1d9aa4a283 ("lmb: Fix the allocation of overlapping memory areas with !LMB_NONE")
Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Tested-by: Michal Simek <michal.simek@amd.com>
Since RPI works well using EFI and has no size limitations with regards
to U-Boot, add the needed structures to support capsule updates.
While at it update the most commonly used defconfigs and include capsule
support and U-Boot commands needed by EFI
Tested-by: Sughosh Ganu <sughosh.ganu@linaro.org>
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
