Now the UCLASS_ETH device "node" field is owerwritten by some network drivers in
case of Ethernet PHYs which are linked to UCLASS_ETH device using
"phy-handle" DT property and when Ethernet PHY driver needs to read some
additional information from DT. In such cases following happens (in
general):
- network drivers
priv->phydev = phy_connect(priv->bus, priv->phyaddr, dev,
priv->interface);
<-- phydev is connected to dev which is UCLASS_ETH device
if (priv->phy_of_handle > 0)
dev_set_of_offset(priv->phydev->dev, priv->phy_of_handle);
<-- phydev->dev->node is overwritten by phy-handle DT node
As result, UCLASS_ETH device can't be used any more for DT accessing.
This patch adds additional ofnode node field to struct phy_device which can
be set explicitly by network drivers and used by PHY drivers, so
overwriting can be avoided. Also add helper function phy_get_ofnode()
which will check and return phy_device->node or dev_ofnode(phydev->dev) for
backward compatibility with existing drivers.
Joe Hershberger [Wed, 25 Jul 2018 17:59:22 +0000 (12:59 -0500)]
net: phy: Clean up includes of common.h
We want to be able to include some other system headers in phy.h but
that requires us to have included common.h in the top-level first.
Also, common.h includes config.h as the first thing it does, so don't
include it directly.
Series-to: u-boot
Series-cc: Grygorii Strashko <[email protected]>
Seried-cc: ti
Series-process-log: sort, uniq
Cover-letter:
Prepare for net: phy: prevent uclass_eth device "node" field overwriting
Prepare for [1] so that it doesn't break the build for a bunch of
boards. There are a number of reasons this series broke the build
but none of them depend on changes in the series, so fix up those
situations ahead of applying that series.
[1] https://patchwork.ozlabs.org/cover/940104/
END Signed-off-by: Joe Hershberger <[email protected]>
Joe Hershberger [Tue, 17 Jul 2018 20:02:30 +0000 (15:02 -0500)]
phy: Break include cycle
Because some phy wants to export some functions [1], export.h was
including the whole phy subsystem which pulls in lots of stuff that
causes some ordering and redefinition issues. Split out the only part
that is actually needed in export.h and include it there and in phy.h.
[1] commit 95279315076c ("board/ls2085rdb: Export functions for
standalone AQ FW load apps")
Joe Hershberger [Mon, 16 Jul 2018 20:33:51 +0000 (15:33 -0500)]
arm: Prevent redefinition error in fsl-layerscape
The include/phy.h will start including dm.h, which pulls in
linux/compat.h after the attempted redefinition in
arch/arm/include/asm/armv8/mmu.h, so move this include to allow
redefinition.
Murali Karicheri [Thu, 28 Jun 2018 19:26:34 +0000 (14:26 -0500)]
net: phy: dp83867: add workaround for incorrect RX_CTRL pin strap
The data manual for DP83867IR/CR, SNLS484E[1], revised march 2017,
advises that strapping RX_DV/RX_CTRL pin in mode 1 and 2 is not
supported (see note below Table 5 (4-Level Strap Pins)).
It further advises that if a board has this pin strapped in mode 1 and
mode 2, then bit[7] of Configuration Register 4 (address 0x0031) must
be cleared to 0. This is to ensure proper operation of PHY.
Since it is not possible to detect in software if RX_DV/RX_CTRL pin is
incorrectly strapped, add a device-tree property to advertise this and
allow corrective action in software.
[1] http://www.ti.com/lit/ds/snls484e/snls484e.pdf
Chris Packham [Mon, 9 Jul 2018 09:34:00 +0000 (21:34 +1200)]
net: mvgbe: convert to DM
Add driver model support to the mvgbe driver. As a temporary measure
both DM and non-DM uses are supported. Once all the users have been
converted the non-DM support can be dropped.
Andrew Thomas [Thu, 21 Jun 2018 23:21:01 +0000 (16:21 -0700)]
initialize net_mode.if_type
if_type is not correctly initialized
Failure to initialize if_type means that grub2/efinet sends
a bogus arp request. It therefore gets no response. On Raspberry Pi 3B+
this leads to a pause at:
lan78xx_eth Waiting for PHY auto negotiation to complete....... done
lan78xx_eth Waiting for PHY auto negotiation to complete....... done
Joe Hershberger [Wed, 4 Jul 2018 00:36:42 +0000 (19:36 -0500)]
net: Read bootfile from env on netboot_common()
Instead of depending on a env callback for bootfile, read it explicitly.
We do this because the bootfile can be specified on the command line and
if it is, we will overwrite the internal variable. If a netboot_common()
is called again with no bootfile parameter, we want to use the one in
the environment.
Joe Hershberger [Wed, 4 Jul 2018 00:36:39 +0000 (19:36 -0500)]
net: When checking prerequisites, consider boot_file_name
For net_boot_common, we allow the serverip to be specified as part of
the boot file name. For net commands that require serverip, include that
source as a valid specification of serverip.
Joe Hershberger [Wed, 4 Jul 2018 00:22:55 +0000 (19:22 -0500)]
net: Check subnet against the actual ip address in use for nfs
The check for sending to the gateway was not using the correct variable
for comparison, so it was reporting that packets are sent to the gateway
when they were not.
Joe Hershberger [Mon, 2 Jul 2018 19:47:54 +0000 (14:47 -0500)]
sandbox: eth-raw: Add a SIMPLE_BUS to enumerate host interfaces
Ask the OS for each of its interfaces and for each one, bind a U-Boot
device and then probe it. This will allocate the priv data structure
that is then populated.
Joe Hershberger [Mon, 2 Jul 2018 19:47:46 +0000 (14:47 -0500)]
net: Only call halt on a driver that has been init'ed
With driver model, we were not checking if the state of the device was
marked as active before calling the halt function. Check that the device
is probed and also marked as active. This avoids the case where we were
calling halt on the first device in net_init() and the driver would
operate on bogus data structures causing problems. In this case, the
priv was all 0, so halt() would close STDIN.
Joe Hershberger [Tue, 3 Jul 2018 01:06:49 +0000 (20:06 -0500)]
sandbox: Don't disable ctrlc() on sandbox if in raw mode
In raw mode, handle ctrl-c as normal. This allows normal ctrl-c behavior
such as aborting a command that is timing out without completely
terminating the sandbox executable.
In [1], Simon disabled this. His reason for it was that it interferes
with piping test scripts. Piping should be done in cooked mode, so this
change should still not interfere.
York Sun [Fri, 8 Jun 2018 23:37:27 +0000 (16:37 -0700)]
armv8: ls1046ardb: Enable IFC for SPL build
Commit a52ff334c5b1 ("armv8: ls1046ardb: SPL size reduction") reduced
image size for SPL. IFC was disabled. If PPA is loaded in SPL, MMU is
enabled as a result. Removing IFC skips IFC region in the MMU table,
causing later failure in RAM version U-boot when accessing CPLD
through IFC. Only disable IFC if PPA is not enabled for SPL.
Make the code structure more general so that more socs can be
added easily and also remove speed limitation restriction.
Add the ls1012a sata support as well.
One ls1012a, there is one SATA 3.0 advanced host controller interface
which is a high-performance SATA solution that delivers comprehensive
and fully-compliant generation 3 (1.5 Gb/s - 6.0 Gb/s) serial ATA
capabilities, in accordance with the serial ATA revision 3.0 of Serial
ATA International Organization.
Add sata node to support this feature.
York Sun [Tue, 26 Jun 2018 21:26:02 +0000 (14:26 -0700)]
armv8: layerscape: Enabled I-cache for SPL boot
Enable I-cache for SPL boot to boost performance. Earlier MMU was
enabled only for LS2080A and has since been dropped by commit f539c8a4a7a5 ("armv8: ls2080a: Drop early MMU for SPL build").
Ashish Kumar [Tue, 26 Jun 2018 09:18:30 +0000 (14:48 +0530)]
configs: Disable 4k erase sector size for spansion "s25fs512s" flash
4K erase size is used only in case of hydrid mode which is not
supported on any NXP platform with flash "s25fs512s".
Supported mode is uniform sector, with erase size 256kiB.
CONFIG_SPL_FSL_LS_PPA is needed only in case of falcon boot
Support for this is not present in PPA currently, so
removing the default option from defconfig
Tom Rini [Thu, 26 Jul 2018 15:17:24 +0000 (11:17 -0400)]
cmd/ubi.c: Fix format warning
On 64bit platforms we would otherwise see:
../cmd/ubi.c: In function 'ubi_volume_read':
../cmd/ubi.c:359:16: warning: format '%u' expects argument of type 'unsigned int', but argument 2 has type 'size_t {aka long unsigned int}' [-Wformat=]
Fixes: 68c7025d99db ("cmd: ubi: print load size after establishing volume size") Signed-off-by: Tom Rini <[email protected]>
The EFI subsystem accesses the real time clock and is enabled by default.
So we should drop any CONFIG_CMD_DATE dependency from the real time clock
drivers.
This unit test checks the CalculateCrc32 bootservice and checks the
headers of the system table, the boot services tablle, and the runtime
services table before and after ExitBootServices().
The system table contains a link to the list of configurations tables.
These include the device tree, SMBIOS table, and the ACPI table.
This array is currently statically linked. With the patch it is allocated
as EFI_RUNTIME_SERVICES_DATA. Due to the structure of the system table we
cannot work with a linked list here.
Both in the boot and the runtime services tables we have to specify the
UEFI spec revision. The same value is already used for the system
table. So let's use a common constant.
In the boot services table we have to provide the header signature.
fs_fat_write() is not able to write to subdirectories.
Currently if a filepath with a leading slash is passed, the slash is
treated as part of the filename to be created in the root directory.
Strip leading (back-)slashes.
Check that the remaining filename does not contain any illegal characters
(<>:"/\|?*). This way we will throw an error when trying to write to a
subdirectory.
Alexander Graf [Tue, 12 Jun 2018 05:48:37 +0000 (07:48 +0200)]
efi_loader: Rename sections to allow for implicit data
Some times gcc may generate data that is then used within code that may
be part of an efi runtime section. That data could be jump tables,
constants or strings.
In order to make sure we catch these, we need to ensure that gcc emits
them into a section that we can relocate together with all the other
efi runtime bits. This only works if the -ffunction-sections and
-fdata-sections flags are passed and the efi runtime functions are
in a section that starts with ".text".
Up to now we had all efi runtime bits in sections that did not
interfere with the normal section naming scheme, but this forces
us to do so. Hence we need to move the efi_loader text/data/rodata
sections before the global *(.text*) catch-all section.
With this patch in place, we should hopefully have an easier time
to extend the efi runtime functionality in the future.
Signed-off-by: Alexander Graf <[email protected]>
[agraf: Fix x86_64 breakage]
Simon Glass [Mon, 18 Jun 2018 14:08:28 +0000 (08:08 -0600)]
efi: Drop error return in efi_carve_out_dt_rsv()
This function currently returns an error code, but never uses it. There is
no function comment so it is not obvious why. Presuambly the error is not
important.
Update the function to explain its purpose and why it ignores the error.
Drop the useful error return value.
Simon Glass [Mon, 18 Jun 2018 14:08:20 +0000 (08:08 -0600)]
vsprintf: Handle NULL with %pU
At present a NULL pointer passed to printf for a %pU argument will cause
U-Boot to access memory at 0. Fix this by adding a check, and print
"(null)" instead.
Alexander Graf [Mon, 18 Jun 2018 15:23:15 +0000 (17:23 +0200)]
efi_loader: Expose U-Boot addresses in memory map for sandbox
We currently expose host addresses in the EFI memory map. That can be
bad if we ever want to use sandbox to boot strap a real kernel, because
then the kernel would fetch its memory table from our host virtual address
map. But to make that use case work, we would need to have full control
over the address space the EFI application sees.
So let's expose only U-Boot addresses to the guest until we get to the
point of allocation. EFI's allocation functions are fun - they can take
U-Boot addresses as input values for hints and return host addresses as
allocation results through the same uint64_t * parameter. So we need to
be extra careful on what to pass in when.
With this patch I am successfully able to run the efi selftest suite as
well as grub.efi on aarch64.
Simon Glass [Mon, 18 Jun 2018 15:23:12 +0000 (17:23 +0200)]
efi: sandbox: Adjust memory usage for sandbox
With sandbox the U-Boot code is not mapped into the sandbox memory range
so does not need to be excluded when allocating EFI memory. Update the EFI
memory init code to take account of that.
Signed-off-by: Simon Glass <[email protected]>
[agraf: Remove map_sysmem() call and header reference] Signed-off-by: Alexander Graf <[email protected]>
Alexander Graf [Mon, 18 Jun 2018 15:23:05 +0000 (17:23 +0200)]
efi_loader: Introduce ms abi vararg helpers
Varargs differ between sysv and ms abi. On x86_64 we have to follow the ms
abi though, so we also need to make sure we use x86_64 varargs helpers.
This patch introduces generic efi vararg helpers that adhere to the
respective EFI ABI. That way we can deal with them properly from efi
loader code and properly interpret variable arguments.
This fixes the InstallMultipleProtocolInterfaces tests in the efi selftests
on x86_64 for me.
When running on the sandbox the stack is not necessarily at a higher memory
address than the highest free memory.
There is no reason why the checking of the highest memory address should be
more restrictive for EFI_ALLOCATE_ANY_PAGES than for
EFI_ALLOCATE_MAX_ADDRESS.
Alexander Graf [Mon, 18 Jun 2018 15:23:00 +0000 (17:23 +0200)]
efi_loader: Allow SMBIOS tables in highmem
We try hard to make sure that SMBIOS tables live in the lower 32bit.
However, when we can not find any space at all there, we should not
error out but instead just fall back to map them in the full address
space instead.
This can for example happen on systems that do not have any RAM mapped
in the lower 32bits of address space. In that case having any SMBIOS
tables at all is better than having none.
Alexander Graf [Mon, 18 Jun 2018 15:22:58 +0000 (17:22 +0200)]
efi_loader: Use map_sysmem() in bootefi command
The bootefi command gets a few addresses as values passed in. In sandbox,
these values are in U-Boot address space, so we need to make sure we
explicitly call map_sysmem() on them to be able to access them.
Alexander Graf [Mon, 18 Jun 2018 15:22:57 +0000 (17:22 +0200)]
efi_loader: Use compiler constants for image loader
The EFI image loader tries to determine which target architecture we're
working with to only load PE binaries that match.
So far this has worked based on CONFIG defines, because the target CPU
was always indicated by a config define. With sandbox however, this is
not longer true as all sandbox targets only encompass a single CONFIG
option and so we need to use compiler defines to determine the CPU
architecture.
projects / J-u-boot.git / log