| Age | Commit message (Collapse) | Author |
|---|
|
Define constants for hptimer reg names and use them instead of magic
numbers in rockchip_stimer_init().
Signed-off-by: Jonas Karlman <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
The Rockchip RK3576 SoC uses a different DRAM base address, 0x40000000,
compared to prior SoCs.
Add default options that should work when 0x40000000 is used as DRAM
base address. Use same offsets as before, just below 64 MiB.
Signed-off-by: Jonas Karlman <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
New Rockchip SoCs will typically require use or an external TPL when
support for the SoC is added to U-Boot.
Move imply ROCKCHIP_EXTERNAL_TPL under SoC Kconfig symbol to remove a
future likelihood of a long "default y if" line.
Signed-off-by: Jonas Karlman <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
The ROCKCHIP_COMMON_STACK_ADDR Kconfig option was originally enabled
in the SoC specific Kconfig files to ease during the initial migration
to use common stack addresses.
All boards for the affected SoCs have been migrated to use common stack
addresses. Migrate to use an imply under the SoC symbol instead of
re-define the symbol in each SoC specific Kconfig file.
Signed-off-by: Jonas Karlman <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
A few Rockchip ARMv7 SoCs use 0x60000000 as DRAM base address instead of
the more common 0x0 DRAM base address used on AArch64 SoCs.
Add default options that should work for these ARMv7 SoCs. Same offsets
as before are used, just below 64 MiB. Hex values have also been padded
to improve alignment.
Signed-off-by: Jonas Karlman <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
The stack-pointer addresses used with ROCKCHIP_COMMON_STACK_ADDR expect
that DRAM is initialized by TPL or ROCKCHIP_EXTERNAL_TPL, that SPL has
access to full DRAM and SPL is loaded to/executed from start of DRAM.
Add depends on to ensure use of the ROCKCHIP_COMMON_STACK_ADDR symbol
does not cause problem for any board not using TPL and back-to-BROM
loading of SPL.
Signed-off-by: Jonas Karlman <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
The Kconfig symbol ROCKCHIP_COMMON_STACK_ADDR currently imply the
TPL_ROCKCHIP_COMMON_BOARD option when TPL=y. This is inconvenient for a
SoC with very limited SRAM to use a custom tpl.c together with the
common stack addresses.
Move any imply TPL_ROCKCHIP_COMMON_BOARD to under the SoC symbol, where
it belongs. Add the missing imply to RK3328 and PX30 use a SoC specific
tpl.c and only expect imply TPL_LIBGENERIC_SUPPORT.
Signed-off-by: Jonas Karlman <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Haikou carrierboard allows multiple adapter boards to be connected, for
now there exists the following adapter boards compatible with PX30
Ringneck:
- Haikou Video Demo on the Video Connector,
- Haikou LVDS 9904379 on the Video Connector,
So support DT overlays so we can use this mechanism instead of full DTB
containing both the carrierboard and the adapter.
Signed-off-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
In order to be able to use Device Tree Overlays, the fdtoverlay_addr_r
needs to be specified.
Follow what's been done for other Rockchip SoCs and leave 1MiB for the
base DTB before the address for the overlay.
Signed-off-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
I don't see a reason why this should only be enabled on a per-board
basis. The rng IP is inside the SoC and doesn't seem to rely on anything
external to it, therefore let's enable it on the SoC DTSI and remove the
now empty px30-evb-u-boot.dtsi.
Signed-off-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
The build system uses HAVE_VENDOR_COMMON_LIB to automatically include
board/$(VENDOR)/common/Makefile, therefore let's use that to implicitly
include board/theobroma-systems/common/Makefile and compile the common.c
file when building proper.
Signed-off-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Nothing is calling the function rk_board_init() and the io-domain driver
can handle the functions intended purpose based on information from DT.
Cleanup by removing the unused rk_board_init() function and re-sort
included headers.
Signed-off-by: Jonas Karlman <[email protected]>
Reviewed-by: Quentin Schulz <[email protected]>
Reviewed-by: Kever Yang <[email protected]>
|
|
Add sandbox64_lwip_defconfig to the list of tested boards.
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
Add sandbox64_lwip_defconfig based on sandbox64_defconfig with NET_LWIP
enabled.
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
Now that the DSA tests in test/dm/dsa.c are compatible with NET_LWIP,
remove the dependency of DM_DSA on NET.
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
Convert the tests to use the do_ping() interface which is now
common to NET and NET_LWIP. This allows running most network test with
SANDBOX and NET_LWIP. A few things to note though:
1. The ARP and IPv6 tests are enabled for NET only
2. The net_retry test is modified to use eth0 (eth@10002000) as the
active (but disabled) interface, and therefore we expect eth1
(eth@10003000) to be the fallback when "netretry" is "yes". This is in
replacement of eth7 (lan1) and eth0 (eth@10002000) respectively.
Indeed, it seems eth7 works with NET by chance and it certainly does not
work with NET_LWIP. I observed that even with NET,
sandbox_eth_disable_response(1, true) has no effect: remove it and
the test still passes. The interface ID is not correct to begin with; 1
corresponds to eth1 (eth@10003000) as shown by debug traces, it is not
eth7 (lan1). And using index 7 causes a SEGV. In fact, it is not the
call to sandbox_eth_disable_response() that prevents the stack from
processing the ICMP reply but the timeout caused by the call to
sandbox_eth_skip_timeout(). Here is what happens when trying to ping
using the eth7 (lan1) interface with NET:
do_ping(...)
net_loop(PING)
ping_start()
eth_rx()
sb_eth_recv()
time_test_add_offset(11000UL);
if (get_timer(0) - time_start > time_delta)
ping_timeout_handler() // ping error, as expected
And the same with NET_LWIP:
do_ping(...)
ping_loop(...)
sys_check_timeouts()
net_lwip_rx(...)
sb_eth_recv()
time_test_add_offset(11000UL);
netif->input(...) // the packet is processed succesfully
By choosing eth0 and sandbox_eth_disable_response(0, true), the incoming
packet is indeed discarded and things work as expected with both network
stacks.
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
Make the do_ping() function in cmd/net.c a global one by getting rid of
the static qualifier, and move the prototype declaration from net-lwip.h
to net-common.h. This makes the function available to other parts of
U-Boot when CONFIG_NET=y, as was already the case when
CONFIG_NET_LWIP=y.
This is a peparation step to make the sandbox tests use a common API
between NET and NET_LWIP.
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
The tests in test/dm/eth.c and test/dm/dsa.c use interfaces that have
no static IP addresses configured in the board's default environment
file. That will be a problem when NET_LWIP=y because the lwIP stack
refuses to send ICMP packets through an interface that doesn't have an
IP ("no route to host"). Therefore and in preparation for enabling the
sandbox tests with NET_LWIP, provide such addresses in
board/sandbox/sandbox.env.
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
Make the sandbox mock ethernet driver (drivers/net/sandbox.c) compatible
with NET_LWIP by not relying on any of the structures or functions
defined in net-legacy.h. This is done by providing local definitions of
the various protocol structures (Ethernet, ARP, IPv4, ICMP). Drop the
stub driver that was introduced specifically for NET_LWIP
(drivers/net/sandbox-lwip.c).
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
When the sandbox timer is available, use it. This allows skipping
time in the tests (sandbox_eth_skip_timeout()).
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
Use net_start_again() in do_ping() to determine if a failed ping should
be restarted on a different interface.
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
Implement net_start_again() when NET_LWIP=y in a very similar way to
NET. This will be used in a future commit to determine if a failed
ping needs to be tried again on a different interface.
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
The things that are done prior to executing a network command with
NET_LWIP are not consistent with what is done with NET. It impacts the
selection of the current device, and more precisely if the active device
is invalid NET would return an error while NET_LWIP would try to pick a
new device. This incorrect behavior was detected thanks to the eth_rotate
sandbox test (dm_test_eth_rotate()).
Fix it by re-using a sequence similar to what NET has in net_loop().
This piece of code is inserted in a function called net_lwip_eth_start()
renamed from net_lwip_eth_set_current().
Signed-off-by: Jerome Forissier <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
initr_net() initalizes the network devices by calling eth_initalize().
There is no good reason to disable this if no command line interface is
present.
Let initr_net() depend on CONFIG_NET || CONFIG_NET_LWIP.
Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Ilias Apalodimas <[email protected]>
|
|
CMD_PXE is already under CMD_NET in Kconfig that's why make no sense to
have another ifdef inside source code.
Signed-off-by: Michal Simek <[email protected]>
Reviewed-by: Quentin Schulz <[email protected]>
|
|
!A || (A && B) is equivalent to !A || B
Drop the middle expression from the statement
Signed-off-by: Bryan Brattlof <[email protected]>
Reviewed-by: Jerome Forissier <[email protected]>
|
|
In IPv6, the broadcast MAC address is not used. Instead, it should use
the multicast address (see RFC RFC2464).
Add IPV6_ALL_NODE_ETH_ADDR macro for clarity.
Signed-off-by: Sean Edmond <[email protected]>
Reviewed-by: Jerome Forissier <[email protected]>
|
|
The set_dfu_alt_info() function calls the ALLOC_CACHE_ALIGN_BUFFER()
macro to declare a `buf' variable pointer into an array allocated on the
stack. It then calls the memset() function to clear the useable portion
of the array using the idiomatic expression `sizeof(buf)'.
While this would indeed work fine for an array, in the present case we
end up clearing only the size of a pointer.
Fix this by specifying the explicit size `DFU_ALT_BUF_LEN' instead.
Fixes: 6b403ca4dcf4 ("fwu: DeveloperBox: add support for FWU")
Signed-off-by: Vincent Stehlé <[email protected]>
Cc: Masahisa Kojima <[email protected]>
Cc: Tom Rini <[email protected]>
Cc: Jassi Brar <[email protected]>
Reviewed-by: Mattijs Korpershoek <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mattijs Korpershoek <[email protected]>
|
|
f_acm calls usb_ep_disable(f_acm->ep_notify) unconditionally in
acm_start_ctrl(), even if the USB endpoint was never enabled before. This
causes crashes for some UDC drivers (e.g. ci_udc), because they dereference
data structures that are assigned only after having called usb_ep_enable().
The f_acm driver in U-Boot is similar to the Linux driver, where this issue
does not occur because usb_ep_disable() and usb_ep_enable() internally
track the enabled state. In Linux this change was made in commit
b0bac2581c19 ("usb: gadget: introduce 'enabled' flag in struct usb_ep") by
Robert Baldyga.
Fix the crashes for f_acm by making the same change in U-Boot. This makes
the API less bug-prone and avoids introducing crashes when adapting new
gadget drivers from Linux.
Signed-off-by: Stephan Gerhold <[email protected]>
Reviewed-by: Mattijs Korpershoek <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mattijs Korpershoek <[email protected]>
|
|
When using IOMUX, the "usbacm" console can be added/removed dynamically
from the stdout/stderr/stdin environment variables to allow temporarily
starting other USB gadgets (e.g. Fastboot).
However, right now acm_stdio_stop() does not completely undo
acm_stdio_start(): The USB gadget is unregistered, but as long as dev->priv
stays set acm_stdio_start() will never register the USB gadget again.
Clear dev->priv after we detach to make sure a start operation after a stop
operation registers the gadget again.
Fixes: fc2b399ac03b ("usb: gadget: Add CDC ACM function")
Signed-off-by: Stephan Gerhold <[email protected]>
Reviewed-by: Mattijs Korpershoek <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mattijs Korpershoek <[email protected]>
|
|
U-Boot has an older version of the Linux gadget API, where USB endpoints
returned by usb_ep_autoconfig() are not automatically claimed. As written
in the documentation comment:
"To prevent the endpoint from being returned by a later autoconfig call,
claim it by assigning ep->driver_data to some non-null value."
Right now f_acm doesn't do that, which means that e.g. ep_in and ep_notify
may end up being assigned the same endpoint. Surprisingly, the ACM console
is still somehow working, but this is not the expected behavior. It will
break with a later commit that disallows calling usb_ep_enable() multiple
times.
Fix this by assigning some data to ep->driver_data, similar to the other
gadget drivers.
Fixes: fc2b399ac03b ("usb: gadget: Add CDC ACM function")
Signed-off-by: Stephan Gerhold <[email protected]>
Reviewed-by: Mattijs Korpershoek <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mattijs Korpershoek <[email protected]>
|
|
The current code attempts to bind scmi_voltage_domain to regulator subnode
of the SCMI protocol node, so scmi_voltage_domain can then bind regulators
directly to subnodes of its node. This kind of behavior should not be in
core code, move it into scmi_voltage_domain driver code. Let the driver
descend into regulator node and bind regulators to its subnodes.
Fixes: 1f213ee4dbf2 ("firmware: scmi: voltage regulator")
Signed-off-by: Marek Vasut <[email protected]>
[Alice Guo: Fix scmi_regulator_bind]
Signed-off-by: Peng Fan <[email protected]>
|
|
This callback is used to send the 74 clock cycles after power up.
Signed-off-by: Mathieu Othacehe <[email protected]>
Reviewed-by: Peng Fan <[email protected]>
Signed-off-by: Peng Fan <[email protected]>
|
|
Add a new callback function *send_init_stream* which start a sequence of
at least 74 clock cycles.
The mmc core uses *mmc_send_init_stream* in order to invoke the callback
function. This will be used during power cycle where the specification
requires such a sequence after power up.
Signed-off-by: Jean-Jacques Hiblot <[email protected]>
Reviewed-by: Peng Fan <[email protected]>
Signed-off-by: Peng Fan <[email protected]>
|
|
%s/Relatvie/Relative'
Signed-off-by: Heinrich Schuchardt <[email protected]>
Acked-by: Peng Fan <[email protected]>
Signed-off-by: Peng Fan <[email protected]>
|
|
The default value of 0x800000 (8 MB) is somewhat limiting for us, as our
compressed kernel may grow up to around 4 MB. By choosing the commonly
used value of 0x2000000 (32 MB), we are definitely on the safe side.
This rather large amount should be fine, as we have 128 MB of RAM.
Signed-off-by: Ezra Buehler <[email protected]>
Reviewed-by: Stefan Roese <[email protected]>
|
|
Jiaxun Yang <[email protected]> says:
This is a huge series which promoted MIPS/Boston target into a
usable state, with fixes to drivers and general framework issues
I found in this process.
I also converted the target to OF_UPSTREAM.
This target is covered by QEMU, to test on QEMU:
```
make boston64r6el_defconfig
make
qemu-system-mips64el -M boston -cpu I6500 -bios ./u-boot.bin -nographic
```
Link: https://lore.kernel.org/r/[email protected]
|
|
Paul had left MIPS a couple of years ago, his email address is
no longer valid.
Replace it with his kenrel.org email, which has been used in
kernel and QEMU, in case we still want to reach him.
Signed-off-by: Jiaxun Yang <[email protected]>
Reviewed-by: Daniel Schwierzeck <[email protected]>
|
|
We can now boot with upstream devicetree.
Reviewed-by: Sumit Garg <[email protected]>
Signed-off-by: Jiaxun Yang <[email protected]>
Reviewed-by: Daniel Schwierzeck <[email protected]>
|
|
It is required to make OF_UPSTREAM work.
Reviewed-by: Sumit Garg <[email protected]>
Signed-off-by: Jiaxun Yang <[email protected]>
Reviewed-by: Daniel Schwierzeck <[email protected]>
|
|
In upstream devicetree, clk_boston is a child of syscon node
and there is no "regmap" property for clk_boston node.
Try to check parent device first to look for syscon.
Signed-off-by: Jiaxun Yang <[email protected]>
Reviewed-by: Daniel Schwierzeck <[email protected]>
|
|
Provide default environment variables on image loading address
to make the board useful.
Signed-off-by: Jiaxun Yang <[email protected]>
Reviewed-by: Daniel Schwierzeck <[email protected]>
|
|
This is a PC-like platform board.
Enable drivers for most on-board devices to make it useful.
Signed-off-by: Jiaxun Yang <[email protected]>
Reviewed-by: Daniel Schwierzeck <[email protected]>
|
|
Some drivers need this header.
Provide this dummy header as riscv did.
Signed-off-by: Jiaxun Yang <[email protected]>
Reviewed-by: Daniel Schwierzeck <[email protected]>
|
|
Remove hard dependencies to arch headers, get clock from clk
subsystem if arch clock function is not available, align
compatible strings with devicetree binding.
No functional change on existing platforms, just get it build
on other platforms.
Signed-off-by: Jiaxun Yang <[email protected]>
|
|
Ensure that we are using correct physical/virtual address for
DMA buffer write and hardware register settings.
The convention is: in ahci_ioports all pointers are virtual,
that will be converted to physical address when writing to
hardware registers or into sg/cmd_tbl.
Also fixed 64bit physical address support for dwc_ahsata, ensure
higher bits are written into registers/sg properly.
Use memalign for allocating aligned buffer in dwc_ahsata so we
don't have to do our own alignment in driver.
Signed-off-by: Jiaxun Yang <[email protected]>
|
|
For MIPS we are always looking gd->dram in virtual address so
PCI_MAP_SYSTEM_MEMORY should always be enabled.
If in future we ever want to make it physical we have to set
ARCH_MAP_SYSMEM.
Signed-off-by: Jiaxun Yang <[email protected]>
Reviewed-by: Daniel Schwierzeck <[email protected]>
|
|
Boston has a very limited memory range for PCI controllers, where
1MB can't easily fit into it.
Make alignment boundary of PCI memory resource allocation a Kconfig
option and default to 0x10000 for boston.
Signed-off-by: Jiaxun Yang <[email protected]>
Reviewed-by: Daniel Schwierzeck <[email protected]>
|
|
Probe size of ecam from devicetree properly and cap accessible
bus number accorading to ecam region size to ensure we don't go
beyond hardware address space.
Also disable all interrupts to ensure errors are handled silently.
Signed-off-by: Jiaxun Yang <[email protected]>
Reviewed-by: Daniel Schwierzeck <[email protected]>
|
|
Heinrich Schuchardt <[email protected]> says:
For QEMU we have developed supporting passsing through ACPI tables.
This functionality has been broken multipled times due to missing CI
builds.
* Two new defconfigs qemu_arm64_acpi and qemu-riscv64_smode_acpi.
* Assign the defconfigs to the respective maintainers.
Link: https://lore.kernel.org/r/[email protected]
|
