As the PCIe specification recommend reading the Vendor ID register
to determine if a Function is present, read the Vendor ID of a
non-existent Function must not result in system error, so we'd better
make the first CFG read to Vendor ID instead of Header Type register
in the PCIe enumeration.
Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@nxp.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
This introduces two new APIs dm_pci_find_next_capability() and
dm_pci_find_next_ext_capability() to get PCI capability address
and PCI express extended capability address for a given PCI device
starting from a given offset.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The PCI controller can have DT subnodes describing extra properties
of particular PCI devices, ie. a PHY attached to an EHCI controller
on a PCI bus. This patch parses those DT subnodes and assigns a node
to the PCI device instance, so that the driver can extract details
from that node and ie. configure the PHY using the PHY subsystem.
Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Tom Rini <trini@konsulko.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
In int-ll64.h, we always use the following typedefs:
typedef unsigned int u32;
typedef unsigned long uintptr_t;
typedef unsigned long long u64;
This does not need to match to the compiler's <inttypes.h>.
Do not include it.
The use of PRI* makes the code super-ugly. You can simply use
"l" for printing uintptr_t, "ll" for u64, and no modifier for u32.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
This introduces two new APIs dm_pci_find_capability() and
dm_pci_find_ext_capability() to get PCI capability address and
PCI express extended capability address for a given PCI device.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The correct driver data comes from the matching 'id' instead of
'find_id' in pci_find_and_bind_driver().
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The flag to control whether to scan multi-function device during
enumeration should be cleared at the beginning of each iteration
if the device's function number equals to zero.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Currently only devfn is extracted in child_post_bind(). Now that
we have the live-tree version API to look up PCI vendor and device
id from the compatible string, let's extract and save them too.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
If U-Boot gets used as coreboot payload all pci resources got
assigned by coreboot. If a dts without any pci ranges gets used
the dm is not able to access pci device memory. To get things
working make use of a 1:1 mapping for bus <-> phy addresses.
This change makes it possible to get the e1000 U-Boot driver
working on a sandybridge device where U-Boot is used as coreboot
payload.
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
[bmeng: fixed 'u-boot' in the commit message]
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
If we use U-Boot as coreboot payload with a generic dts without
any ranges specified we fail in pci pre_probe and our pci bus
is not usable.
So convert decode_regions(..) into a void function and do the simple
error handling there.
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
[bmeng: fixed 'u-boot' in the commit message and checkpatch warning]
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Currently, qemu_arm_defconfig and qemu_arm64_defconfig only work with
the 'highmem=off' parameter passed to QEMU's virt machine. The reason is
that when 'highmem' is not disabled, QEMU appends 64-bit a memory
resource to the PCI controller's regions property in DT in addition to
the 32-bit PCI memory window in low memory. And the current DT parsing
code picks the last (thus the 64-bit one) memory resource, whose address
eventually gets silently truncated to 32 bits because
CONFIG_SYS_PCI_64BIT is not set, which obviously causes PCI to break.
Avoid this problem by ignoring memory regions whose addresses are above
the 32-bit boundary when CONFIG_SYS_PCI_64BIT is not set.
Signed-off-by: Tuomas Tynkkynen <tuomas.tynkkynen@iki.fi>
Reviewed-by: Simon Glass <sjg@chromium.org>
When U-Boot started using SPDX tags we were among the early adopters and
there weren't a lot of other examples to borrow from. So we picked the
area of the file that usually had a full license text and replaced it
with an appropriate SPDX-License-Identifier: entry. Since then, the
Linux Kernel has adopted SPDX tags and they place it as the very first
line in a file (except where shebangs are used, then it's second line)
and with slightly different comment styles than us.
In part due to community overlap, in part due to better tag visibility
and in part for other minor reasons, switch over to that style.
This commit changes all instances where we have a single declared
license in the tag as both the before and after are identical in tag
contents. There's also a few places where I found we did not have a tag
and have introduced one.
Signed-off-by: Tom Rini <trini@konsulko.com>
PCI enumeration may happen very early on an x86 board. The board
information pointer should have been checked in decode_regions()
as its space may not be allocated yet.
With this commit, Intel Galileo board boots again.
Fixes: 664758c ("pci: Fix decode regions for memory banks")
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Since memory banks may not be located behind each other we need to add
them separately.
Signed-off-by: Bernhard Messerklinger <bernhard.messerklinger@br-automation.com>
Reviewed-by: Hannes Schmelzer <hannes.schmelzer@br-automation.com>
This sort of pattern for implementing memory-mapped PCI config space
accesses appears in U-Boot twice already, and a third user is coming up.
So add helper functions to avoid code duplication, similar to how Linux
has pci_generic_config_write and pci_generic_config_read.
Signed-off-by: Tuomas Tynkkynen <tuomas.tynkkynen@iki.fi>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
This function returns the pointer to the value of a node property.
The current name ofnode_read_prop() is confusing. Follow the naming
of_get_property() from Linux.
The return type (const u32 *) is wrong. DT property values can be
strings as well as integers. This is why of_get_property/fdt_getprop
returns an opaque pointer.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Simon Glass <sjg@chromium.org>
The of_n_addr_cells() and of_n_size_cells() functions are useful for
getting the size of addresses in a node, but in a few places U-Boot needs
to obtain the actual property value for a node without walking up the
stack. Add functions for this and just the existing code to use it.
Add a comment to the existing ofnode functions which do not do the right
thing with a flat tree.
This fixes a problem reading PCI addresses.
Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Tested-on: Beaver, Jetson-TK1
Update the PCI uclass to support livetree. This mostly involves fixing
the address decoding from the device tree.
Signed-off-by: Simon Glass <sjg@chromium.org>
strdup uses malloc to allocate memory for str.
If we cannot bind to the generic driver we should release
the memory.
The problem was indicated by clang scan-build.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
In the description of function pci_match_one_id(), there are some
problems on arguments list and return value description, so correct
them.
Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@nxp.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
At present devices use a simple integer offset to record the device tree
node associated with the device. In preparation for supporting a live
device tree, which uses a node pointer instead, refactor existing code to
access this field through an inline function.
Signed-off-by: Simon Glass <sjg@chromium.org>
In pci_uclass_pre_probe an attempt is made to detect whether the parent
of a device is a PCI device and that the device is thus a bridge. This
was being done by checking whether the parent of the device is of the
UCLASS_ROOT class. This causes problems if the PCI controller is a child
of some other non-PCI node, for example a simple-bus node.
For example, if the device tree contains something like the following
then pci_uclass_pre_probe would incorrectly believe that the PCI
controller is a bridge, with a PCI parent:
/ {
some_child {
compatible = "simple-bus";
#address-cells = <1>;
#size-cells = <1>;
ranges = <>;
pci_controller: pci@10000000 {
compatible = "my-pci-controller";
device_type = "pci";
reg = <0x10000000 0x2000000>;
};
};
};
Avoid this incorrect detection of bridges by instead checking whether
the parent devices class is UCLASS_PCI and treating a device as a bridge
when this is true, making use of device_is_on_pci_bus to perform this
test.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Quite a few places have a bind() method which just calls dm_scan_fdt_dev().
We may as well call dm_scan_fdt_dev() directly. Update the code to do this.
Signed-off-by: Simon Glass <sjg@chromium.org>
The terminal condition in the area where a PCI device is scanned is wrong,
and 1f.7 isn't scanned.
Signed-off-by: Yoshinori Sato <ysato@users.sourceforge.jp>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Two comments are missing a parameter and there is an extra blank line. Also
two of the region access macros are misnamed. Correct these problems.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
It is common to read a config register value, clear and set some bits, then
write back the updated value. Add functions to do this in one step, for
convenience.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
With CONFIG_DM_PCI enabled, PCI buses are not enumerated at boot, as they
are without that config option enabled. No command exists to enumerate the
PCI buses. Hence, unless some board-specific code causes PCI enumeration,
PCI-based Ethernet devices are not detected, and network access is not
available.
This patch implements "pci enum" in the CONFIG_DM_PCI case, thus giving a
mechanism whereby PCI can be enumerated.
do_pci()'s handling of case 'e' is moved into a single location before the
dev variable is assigned, in order to skip calculation of dev. The enum
sub-command doesn't need the dev value, and skipping its calculation
avoids an irrelevant error being printed.
Using a command to initialize PCI like this has a disadvantage relative to
enumerating PCI at boot. In particular, Ethernet devices are not probed
during PCI enumeration, but only when used. This defers setting variables
such as ethact, ethaddr, etc. until the first network-related command is
executed. Hopefully this will not cause further issues. Perhaps in the
long term, we need a "net start/enum" command too?
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add a driver-model version of the pci_write_bar32 function so that this is
supported in the new API.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
This function should not be used by driver-model code, so move it to the
compatibility portion.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The current comments are confusing. We don't actually bind a generic device
when the device tree has no information. We try to scan available PCI
drivers. Update the comments to reflect this.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
At present the PCI address map functions use the old API. Add new functions
for this so that drivers can be converted.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add a driver-model function for reading the PCI BAR from a device.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Add a function which scans the driver model device information rather
than scanning the PCI bus again.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Bin Meng <bmeng.cn@gmail.com>
Add a function which scans the driver model device information rather
than scanning the PCI bus again.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Bin Meng <bmeng.cn@gmail.com>
At present we are using legacy functions even in the auto-configuration code
used by driver model. Add a new pci_auto.c version which uses the correct
API.
Create a new pci_internal.h header to hold functions that are used within
the PCI subsystem, but are not exported to other drivers.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Bin Meng <bmeng.cn@gmail.com>
Most driver model PCI functions have a dm_ prefix. At some point, when the
old code is converted to driver model and the old functions are removed, we
will drop that prefix.
For consistency, we should use the dm_ prefix for all driver model
functions. Update pci_bus_find_bdf() accordingly.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Most driver model PCI functions have a dm_ prefix. At some point, when the
old code is converted to driver model and the old functions are removed, we
will drop that prefix.
For consistency, we should use the dm_ prefix for all driver model
functions. Update pci_get_bdf() accordingly.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
This function looks up the controller and returns a pointer to each region
type.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Stephen Warren <swarren@nvidia.com>
Tested-by: Stephen Warren <swarren@nvidia.com>
A PCI bus may be a bridge device where the controller is the bridge's
parent. Add a function to return the controller device, given a PCI device.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Stephen Warren <swarren@nvidia.com>
Tested-by: Stephen Warren <swarren@nvidia.com>
Provide a few functions to support using 32-bit access to emulate 8- and
16-bit access.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Stephen Warren <swarren@nvidia.com>
Tested-by: Stephen Warren <swarren@nvidia.com>
At present we add a new resource entry for every range entry. But some range
entries refer to configuration regions. To make this work, avoid adding two
regions of the same type. The later ranges will overwrite the earlier
(configuration) ones.
There does not seem to be a way to distinguish the configuration ranges
other than by ordering (as per the device tree binding).
We could perhaps instead just store one region of each type in a simple
array. Once we are sure that we don't need to support multiple regions, we
could change this. It would be easier to do it when all drivers are
converted to use driver model for PCI.
Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Stephen Warren <swarren@nvidia.com>
SDRAM doesn't always start at 0. Adjust the region mapping so that it works
on platforms where SDRAM is somewhere else.
This needs testing on other platforms.
Signed-off-by: Simon Glass <sjg@chromium.org>
Tested-by: Stephen Warren <swarren@nvidia.com>
To support graphics card behind a PCI bridge, the bridge control
register (offset 0x3e) in the configuration space must turn on
VGA address forwarding.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
Currently pci_last_busno() only checks the last bridge device
under the first UCLASS_PCI device. This is not the case when
there are multiple bridge devices.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
The current code returns 0 even if it failed to find or bind a driver. The
caller then has to check the returned device to see if it is NULL. It is
better to return an error code in this case so that it is clear what
happened.
Adjust the code to return -EPERM, indicating that the device was not bound
because it is not needed for pre-relocation use. Add comments so that the
return value is clear.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
One debug() statement is missing a newline. The other has a repeated word.
Fix these.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
When the auto-configuration process fails for a device (generally due to
lack of memory) we should return the error correctly so that we don't
continue to try memory allocations which will fail.
Adjust the code to check for errors and abort if something goes wrong.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
