Evolve dm_pci_map_bar() to include an offset and length parameter. These
allow a portion of the memory to be mapped and range checks to be
applied.
Passing both the offset and length as zero results in the previous
behaviour and this is used to migrate the previous callers.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add tests for the functions dm_pci_bus_to_phys() and
dm_pci_phys_to_bus() which convert between PCI bus addresses and
physical addresses based on the ranges declared for the PCI controller.
The ranges of bus#1 are used for the tests, adding a translation to one
of the ranges to cover more cases.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
When converting between PCI bus and physical addresses, include a length
parameter that can be used to check that the entire range fits within
one of the PCI regions. This prevents an address being returned that
might be only partially valid for the range it is going to be used for.
Where the range check is not wanted, passing a length of 0 will have the
same behaviour as before this change.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
When parsing the `ranges` DT node, check that both extremes of the
regions are addressable without overflow. This assumption can then be
safely made when processing the regions.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The flags parameter of dm_pci_map_bar() is used for PCI region flags
rather than memory mapping flags. Fix the type to match that of the
region flags and stop using the regions flags as memory mapping flags.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Read the virtio PCI capability out of the device configuration space to
a struct rather than accessing fields directly from the configuration
space as they are needed. This both makes access to the fields easier
and avoids re-reading fields.
Re-reading fields could result in time-of-check to time-of-use problems,
should the value in the configuration space change. The range check of
the `bar` field and the later call to `dm_pci_read_bar32()` is an
example of where this could happen.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Ensure the virtio PCI capabilities are contained within the bounds of
the device's configuration space. The expected size of the capability is
passed when searching for the capability to enforce this check.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Check that the common config is at least as large as the struct it is
expected to contain. Only then is it safe to cast the pointer and be
safe from out-of-bounds accesses.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Make sure virtio notifications are written within their allocated
buffer.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The device config is optional, so check it was present and mapped before
trying to use the pointer. Bounds violations are an error, not just a
warning, so bail if the checks fail.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The length of the device config was erroneously being taken from the
notify capability. Correct this by finding the length in the device
capability.
Fixes: 550435edf8 ("virtio: pci: Support non-legacy PCI transport device")
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add a new config to control whether the driver for legacy virtio PCI
devices is included in the build. VIRTIO_PCI_LEGACY is included by
default when VIRTIO_PCI is selected, but it can also be independently
toggled.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Based on reading https://www.msys2.org/docs/ci/ and "Other Systems"
rework how we update MSYS2 to the current version. We run it once, to
perform nothing other than being the first run, then we run pacman
twice.
Signed-off-by: Tom Rini <trini@konsulko.com>
To quote the author:
U-Boot provides a verified-boot feature based around FIT, but there is
no standard way of implementing it for a board. At present the various
required pieces must be built up separately, to produce a working
implementation. In particular, there is no built-in support for selecting
A/B boot or recovery mode.
This series introduces VPL, a verified program loader phase for U-Boot.
Its purpose is to run the verified-boot process and decide which SPL
binary should be run. It is critical that this decision happens before
SPL runs, since SPL sets up SDRAM and we need to be able to update the
SDRAM-init code in the field.
Adding VPL into the boot flow provides a standard place to implement
verified boot. This series includes the phase itself, some useful Kconfig
options and a sandbox_vpl build for sandbox. No verfied-boot support is
provided in this series.
Most of the patches in this series are fixes and improvements to docs and
various Kconfig conditions for SPL.
Add an initial VPL build for sandbox. This includes the flow:
TPL (with of-platdata) -> VPL -> SPL -> U-Boot
To run it:
./tpl/u-boot-tpl -D
The -D is needed to get the default device tree, which includes the serial
console info.
Add a Makefile check for OF_HOSTFILE which is the option that enables
devicetree control on sandbox.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add support for VPL, a new phase of U-Boot. This runs after TPL. It is
responsible for selecting which SPL binary to run, based on a
verified-boot process.
Signed-off-by: Simon Glass <sjg@chromium.org>
The current logic checks several options to decide whether SPL/TPL need
the U-Boot devicetree to be built. In fact we can check OF_CONTROL, which
is enabled in all cases that matter.
Signed-off-by: Simon Glass <sjg@chromium.org>
This feature is not available in SPL unless common/ and lib/ are built.
Update the Kconfig to avoid build errors.
Signed-off-by: Simon Glass <sjg@chromium.org>
MPP55 is used as a reset connected to the L3 switch chip. This doesn't
matter for u-boot as it doesn't use the L3 switch but it is useful to
be able to toggle the switch in/out of reset for the OS.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>
This command is useful in U-boot scripts and it is being used by
OpenWrt bootscript for this board [1]. Otherwise shell scripting
commands are enabled by default in cmd/Kconfig.
[1] 852126680e/target/linux/mvebu/image/clearfog.bootscript (L7)
[2] e95afa5675/cmd/Kconfig (L1504)
Fixes: 0299c90f39 ("arm: mvebu: Add
SolidRun ClearFog Armada 38x initial support")
Signed-off-by: Josef Schlehofer <pepe.schlehofer@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>
The Zyxel NSA310s board has the network chip Marvell Alaska 88E1318S.
Use uclass mvgbe and the compatible driver M88E1310 driver to bring
up Ethernet.
- Use uclass mvgbe to bring up the network. And remove ad-hoc code.
- Remove CONFIG_RESET_PHY_R.
- Enable CONFIG_PHY_MARVELL to properly configure the network.
- Add phy mode RGMII to kirkwood-nsa310s.dts
- Miscellaneous changes: Move constants to .c file and remove header file
board/zyxel/nsa310s/nsa310s.h, add support for large USB and SATA HDDs,
use BIT macro, add/cleanup comments, and cosmetic changes.
Note that this patch is depended on the following patch:
https://patchwork.ozlabs.org/project/uboot/patch/20220412201820.10291-1-mibodhi@gmail.com/
Signed-off-by: Tony Dinh <mibodhi@gmail.com>
For most Kirkwood boards, the PHY page is already set to page 0
(in register 22) before phy_connect is invoked. But some board like
the Zyxel NSA310S (which uses the network chip MV88E1318S), the PHY page
is not set to page 0. There seems to be some bad data remained in
register 22 when the uclass MVGBE about to invoke phy_connect().
This patch enables the uclass MVGBE to always set the PHY page to 0
before phy_connect.
For reference, please see this discussion:
[RFC PATCH v2] arm: kirkwood: nsa310s: Use Marvell uclass mvgbe
and PHY driver for DM Ethernet.
https://lists.denx.de/pipermail/u-boot/2022-April/480946.html
This patch has been tested with the following Kirkwood boards:
NSA310S (88F6702, network chip MV88E1318S)
Sheevaplug (88F6281, network chip MV88E1318)
Pogo V4 (88F6192, network chip 88E1116R)
GF Home(88F6281, network chip 88E1116R)
Dreamplug (88F6281, network chip MV88E1318)
Dell Kace M300 (88F6282, network chip MV88E1318) - out of tree u-boot
Signed-off-by: Tony Dinh <mibodhi@gmail.com>
U-Boot for Turris Omnia is always compiled with MMC, SCSI and USB support,
so always enable macros for booting from these devices.
Signed-off-by: Pali Rohár <pali@kernel.org>
CONFIG_ETHPRIME defines primary ethernet device and env variable $ethact
stores currently active ethernet device.
So there is no point to set ethact= in default environment. Instead set
CONFIG_ETHPRIME properly.
Signed-off-by: Pali Rohár <pali@kernel.org>
Reviewed-by: Marek Behún <marek.behun@nic.cz>
PCIe Mini CEM 2.1 spec added support for USB3.0 mode on MiniPCIe cards.
USB3.0 and PCIe share same pins and only one function can be active at the
same time. PCIe Mini CEM 2.1 spec says that determining function is
platform specific and spec does not define any dedicated pin which could
say if card is USB3.0-based or PCIe-based.
Implement this platform specific decision (USB3.0 vs PCIe) for WWAN
MiniPCIe slot on Turris Omnia via U-Boot env variable "omnia_wwan_slot",
similarly like is implemented forced mode for MiniPCIe/mSATA slot via
"omnia_msata_slot" env variable. Value "usb3" for "omnia_wwan_slot" would
mean to set USB3.0 mode and value "pcie" original PCIe mode.
A385 SoC on Turris Omnia has configurable fifth SerDes line (exported to
MiniPCIe WWAN slot with SIM card) either to USB3.0 or PCIe functionality,
so implementation of this new PCIe Mini CEM 2.1 feature is simple, by just
configuring SerDes to USB 3.0 mode.
Other twos MiniPCIe slots on Turris Omnia do not have this new
functionality as their SerDes lines cannot be switched to USB3.0
functionality.
Note that A385 SoC does not have too many USB3.0 blocks, so activating
USB3.0 in MiniPCIe cause that one external USB3.0 USB-A port would loose
USB3.0 functionality and would be downgraded just to USB2.0.
By default this MiniPCIe WWAN slot is in PCIe mode, like before.
To set this MiniPCIe WWAN slot to USB3.0 mode, call U-Boot commands:
=> setenv omnia_wwan_slot usb3
=> saveenv
=> reset
Signed-off-by: Pali Rohár <pali@kernel.org>
Show error message when DT file does not contain sata or pcie node which
should be explicitly disabled. This can happen when U-Boot code for finding
those nodes is incomplete or when those DT nodes are in different
unexpected location. In any case it is needed to know if DT not was not
explicitly disabled as it could mean that combo slots where setup
incorrectly.
Signed-off-by: Pali Rohár <pali@kernel.org>
Move code for disabling sata and pcie DT nodes to own functions, so this
code can be called from other places in follow up patches.
Signed-off-by: Pali Rohár <pali@kernel.org>
Some PCIe-based MiniPCIe cards are broken and they do not ground PIN 43
which is required by PCIe mini CEM specs. Such broken cards are incorrectly
detected as mSATA cards because SATA specs requires that PIN 43 on mSATA
cards has to be disconnected.
PIN 43 on Turris Omnia is used only for MiniPCIe/mSATA card detection by
software in U-Boot SPL. Allow to override that U-Boot SPL detection by a
new "omnia_msata_slot" env variable (to value "pcie" or "sata") so broken
MiniPCIe cards can be used in combo mSATA/MiniPCIe slot too.
As configuration of PCIe vs SATA functionality is done in U-Boot SPL,
it is required to change env variable in permanent storage and reset the
board to take effect.
To force PCIe mode for broken MiniPCIe cards, call U-Boot commands:
=> setenv omnia_msata_slot pcie
=> saveenv
=> reset
Signed-off-by: Pali Rohár <pali@kernel.org>
By default use primary serdes map with PCIe function in combined
miniPCIe/mSATA slot. When SATA is detected change serdes map variable at
runtime.
Signed-off-by: Pali Rohár <pali@kernel.org>
BootROM maps SPI Flash to fixed address 0xD4000000 and this mapping is
active also when BootROM is executing binary kwbimage headers, which
includes also U-Boot SPL.
Therefore no initialization code is required to access SPI Flags from
U-Boot SPL. In proper U-Boot it is remapped to other location.
So in mvebu implementation of env_sf_get_env_addr() function returns
0xD4000000 when running in SPL and NULL when in proper U-Boot.
This change would allow to use U-Boot ENV in U-Boot SPL. Normally it is not
possible to read ENV because it is too big and U-Boot SPL does not have
such big malloc() pool to real all ENV variables.
Signed-off-by: Pali Rohár <pali@kernel.org>
Reviewed-by: Marek Behún <marek.behun@nic.cz>
In some cases it makes sense to use env_sf_init_addr() also in SPL mode.
Allow it for boards by providing custom implementation of weak function
env_sf_get_env_addr(). When this function returns NULL it signals that
address is invalid, like config option CONFIG_ENV_ADDR.
There is no change in default behavior or in config options.
Signed-off-by: Pali Rohár <pali@kernel.org>
UEFI:
* fix UEFI booting after integration of UEFI sub-system with driver model
* avoid random return values from devpath_is_partition()
* minor code clean ups
-----BEGIN PGP SIGNATURE-----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=IdkS
-----END PGP SIGNATURE-----
Merge tag 'efi-2022-07-rc2' of https://source.denx.de/u-boot/custodians/u-boot-efi
Pull request for efi-2022-07-rc2
UEFI:
* fix UEFI booting after integration of UEFI sub-system with driver model
* avoid random return values from devpath_is_partition()
* minor code clean ups
To quote the author:
I've been experimenting with ASAN on sandbox and turned up a few issues
that are fixed in this series.
Basic ASAN was easy to turn on, but integrating with dlmalloc was
messier and fairly intrusive. Even when I had it working, there was only
a small redzone between allocations which limits the usefulness.
I saw another series on the list by Sean Anderson to enable valgrind
which was finding a different set of issues, though there was one
overlap that Sean is fixing with
"[PATCH] IOMUX: Fix access past end of console_devices".
With these issues fixed, I was able to run the dm tests without any ASAN
issues. There are a couple of leaks reported at the end, but that's for
another day.
Use-after-free shouldn't be used, even in tests. It's bad practice and
makes the test brittle.
Signed-off-by: Andrew Scull <ascull@google.com>
Cc: Simon Glass <sjg@chromium.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
This tests calls regmap_read() which takes a uint pointer as an output
parameter. The test was passing a pointer to a u16 which resulted in an
overflow when the output was written. Fix this by following the
regmap_read() API and passing a uint pointer instead.
Signed-off-by: Andrew Scull <ascull@google.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Jean-Jacques Hiblot <jjhiblot@ti.com>
Cc: Pratyush Yadav <p.yadav@ti.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
