Simon Glass [Fri, 14 Sep 2018 10:57:36 +0000 (04:57 -0600)]
binman: Allow writing a map file when something goes wrong
When we get a problem like overlapping regions it is sometimes hard to
figure what what is going on. At present we don't write the map file in
this case. However the file does provide useful information.
Catch any packing errors and write a map file (if enabled with -m) to aid
debugging.
Simon Glass [Fri, 14 Sep 2018 10:57:34 +0000 (04:57 -0600)]
binman: Correct fmap output on x86
Normally x86 platforms use the end-at-4gb option. This currently produces
an FMAP with positions which have a large offset. The use of end-at-4gb is
a useful convenience within binman, but we don't really want to export
a map with these offsets.
Fix this by subtracting the 'skip at start' parameter.
Also put the code which convers names to fmap format, for clarity.
Simon Glass [Fri, 14 Sep 2018 10:57:33 +0000 (04:57 -0600)]
binman: Record the parent section of each section
At present sections have no record of their parent so it is not possible
to traverse up the tree to the root and figure out the position of a
section within the image.
Change the constructor to record this information.
Simon Glass [Fri, 14 Sep 2018 10:57:32 +0000 (04:57 -0600)]
binman: Support x86 microcode in TPL
When TPL is used on x86 we may want to program the microcode (at least for
the first CPU) early in boot. Add support for this by refactoring the
existing code to be more generic.
Simon Glass [Fri, 14 Sep 2018 10:57:31 +0000 (04:57 -0600)]
binman: Support hashing entries
Sometimesi it us useful to be able to verify the content of entries with
a hash. Add an easy way to do this in binman. The hash information can be
retrieved from the device tree at run time.
Simon Glass [Fri, 14 Sep 2018 10:57:28 +0000 (04:57 -0600)]
binman: Support adding files
In some cases it is useful to add a group of files to the image and be
able to access them at run-time. Of course it is possible to generate
the binman config file with a set of blobs each with a filename. But for
convenience, add an entry type which can do this.
Add required support (for adding nodes and string properties) into the
state module.
Simon Glass [Fri, 14 Sep 2018 10:57:26 +0000 (04:57 -0600)]
binman: Support compressed entries
Add support for compressing blob entries. This can help reduce image sizes
for many types of data. It requires that the firmware be able to
decompress the data at run-time.
Simon Glass [Fri, 14 Sep 2018 10:57:25 +0000 (04:57 -0600)]
patman: Detect missing tools and report them
When tools are needed but not present, at present we just get an error
which can be confusing for the user. Try to be helpful by reporting the
tool as missing and suggesting a possible remedy.
Simon Glass [Fri, 14 Sep 2018 10:57:24 +0000 (04:57 -0600)]
binman: Support updating all device tree files
Binman currently supports updating the main device tree with things like
the position of each entry. Extend this support to SPL and TPL as well,
since they may need (a subset of) this information.
Also adjust DTB output files to have a .out extension since this seems
clearer than having a .dtb extension with 'out' in the name somwhere.
Also add a few missing comments and update the DT setup code to use
ReadFile and WriteFile().
Simon Glass [Fri, 14 Sep 2018 10:57:23 +0000 (04:57 -0600)]
binman: Allow control of whether a fake DT is used
We use a fake device tree in tests most of the time since tests don't
normally care about the actual data. For example, for U-Boot proper we use
U_BOOT_DTB_DATA which is just a four-character string. This makes testing
the image output against an expected value very easy.
However in some cases, such as when the test wants to check that the DT
output containing particular nodes, we do actually need the real DT. Add
support for this, along with a command-line option to select 'test mode'.
Simon Glass [Fri, 14 Sep 2018 10:57:22 +0000 (04:57 -0600)]
binman: Obtain the list of device trees from the config
We always have a device tree for U-Boot proper. But we may also have one
for SPL and TPL. Add a new Entry method to find out what DTs an entry
has, and use that list when updating DTs.
Simon Glass [Fri, 14 Sep 2018 10:57:21 +0000 (04:57 -0600)]
binman: Centralise device-tree updates within binman
At present we have a few calls to device-tree functions in binman and plan
to add more as we add new entry types which need to report their results.
It makes sense to put this code in a central place so that we can make
sure all device trees are updated. At present we only have U-Boot proper,
but plan to add SPL and TPL too.
Simon Glass [Fri, 14 Sep 2018 10:57:19 +0000 (04:57 -0600)]
binman: Move state information into a new module
At present the control module has state information in it, since it is the
primary user of this. But it is a bit odd to have entries and other
modules importing control to obtain this information.
It seems better to have a dedicated state module, which control can use as
well. Create a new module using code from control and update other modules
to use it.
Simon Glass [Fri, 14 Sep 2018 10:57:18 +0000 (04:57 -0600)]
binman: Add an entry method for getting the default filename
Various entry implementations provide a way to obtain the default filename
for an entry. But at present there is no base-class implementation for
this function. Add one so that the API is defined.
Simon Glass [Fri, 14 Sep 2018 10:57:13 +0000 (04:57 -0600)]
dtoc: Allow syncing of the device tree back to a file
At present we require the caller to manually update the device tree using
individual calls to libfdt functions. This is not ideal. It would be
better if we could make changes using the Python structure and then call a
Sync() function to write them back.
Add this feature to the Fdt class. Update binman and the tests to match.
Simon Glass [Fri, 14 Sep 2018 10:57:12 +0000 (04:57 -0600)]
binman: Support building a selection of images
Sometimes it is useful to build only a subset of the images provided by
the binman configuration. Add a -i option for this. It can be given
multiple times to build several images. If the option is not given, all
images are built.
Simon Glass [Fri, 14 Sep 2018 10:57:11 +0000 (04:57 -0600)]
binman: Tidy up the vblock entry
At present if there are two vblock entries an image their contents are
written to the same file in the output directory. This prevents checking
the contents of each separately.
Fix this by adding part of the entry path to the filename, and add some
missing comments.
Simon Glass [Fri, 14 Sep 2018 10:57:10 +0000 (04:57 -0600)]
binman: Add x86 support for starting TPL
Sometimes we want to include TPL for x86 platforms, such as when we want
to select between different SPL images (e.g. for Chrome OS verified boot).
Add support for this.
Simon Glass [Fri, 14 Sep 2018 10:57:09 +0000 (04:57 -0600)]
binman: Generate an error when text is not provided
When the value of a text entry is not provided an execption is generated
talking about a None type. This is confusing. Add a more explanatory error
and a test for this case.
Tom Rini [Wed, 26 Sep 2018 19:14:02 +0000 (15:14 -0400)]
Merge tag 'signed-efi-next' of git://github.com/agraf/u-boot
Patch queue for efi - 2018-09-26
A lot of goodness in this release. We're *very* close to running the
UEFI Shell and SCT natively. The only missing piece are HII protocols.
- FAT write support (needed for SCT)
- improved FAT directory support (needed for SCT)
- RTC support with QEMU -M virt
- Sandbox support (run UEFI binaries in Linux - yay)
- Proper UTF-16 support
- EFI_UNICODE_COLLATION_PROTOCOL support (for UEFI Shell)
- EFI_SIMPLE_TEXT_INPUT_EX_PROTOCOL support (for UEFI Shell)
- Fix window size determination
- Fix Tegra by explicitly unmapping RAM
- Clean up handle entanglement
- Lots of generic code cleanup
[trini: Fixup merge conflict in include/configs/qemu-arm.h] Signed-off-by: Tom Rini <trini@konsulko.com>
Alexander Graf [Wed, 26 Sep 2018 12:07:41 +0000 (14:07 +0200)]
sandbox: Fix set_working_fdt_addr users
When running sandbox with the new pointer sanitization we just recently
introduced, we're running into a case with FIT images where we end up
interpreting pointers as addresses.
What happened is that most callers of set_working_fdt_addr() simply
convert pointers into addresses without taking into account that they
might be 2 separate address spaces. Fix the callers up to map their
pointers into addresses.
This patch modifies PL bitstream loading sequence as per
latest Xilfpga which supports all variants of bitstream images
generated from vivado and from bootgen. With this new change in
Xilfpga, uboot doesn't need to validate and swap bitstream as it will
be taken care inside Xilfpga. ZynqMP PL driver now checks for supporting
PMUFW version before skipping the validation and swap sequence as there
can be old PMUFW which doesn't supports this feature. In this case, driver
uses old way of PL bitstream loading sequence.
Signed-off-by: Siva Durga Prasad Paladugu <siva.durga.paladugu@xilinx.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
arm64: zynqmp: Return pmufw version for zynqmp_pmufw_version()
Modify the zynqmp_pmufw_version() routine to return PMUFW version so
that it can be reused wherever required. Get PMUFW version from PMU
only once at bootup and later just return stored value.
Signed-off-by: Siva Durga Prasad Paladugu <siva.durga.paladugu@xilinx.com> Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Bin Meng [Mon, 17 Sep 2018 05:35:28 +0000 (22:35 -0700)]
Makefile: Use -fno-strict-aliasing globally
The -fstrict-aliasing option is implicitly enabled at levels -O2,
-O3, -Os by GCC. This option allows the compiler to assume the
strictest aliasing rules applicable to the language being compiled.
For example, the practice of reading from a different union member
than the one most recently written to (called "type-punning") is
common. In this case, "type-punning" only works if the memory is
accessed through the union type, but might not work by taking the
address, casting the resulting pointer and dereferencing the result,
which is an undefined behavior per the "strict aliasing rules".
GCC's -Wstrict-aliasing (included in -Wall) option does not catch
all cases, but does attempt to catch the more common pitfalls. So
there are cases that GCC does not report but the codes are violating
the "strict aliasing rules".
Given lots of codes that may be written to rely on "type-punning",
and Linux kernel disables it by -fno-strict-aliasing globally, since
U-Boot currently does this on nds32/riscv/x86 builds only, extend
this for all architecture builds.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Tom Rini <trini@konsulko.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Tuomas Tynkkynen [Wed, 12 Sep 2018 22:28:54 +0000 (01:28 +0300)]
ata: ahci: Loop over the actual number of ports, not the maximum
The loop in ahci_start_ports() is looping over the maximum number of
SCSI devices in the system, which can be larger than the amount of ports
a particular AHCI controller has. The extra looping isn't directly
harmful because the link_port_map bitmap won't have the bit set for a
nonexistent port, but it is wasteful. Replace the loop limit with the
port count of the AHCI controller instead.
Signed-off-by: Tuomas Tynkkynen <tuomas.tynkkynen@iki.fi> Reviewed-by: Simon Glass <sjg@chromium.org>
Chris Packham [Sat, 8 Sep 2018 09:39:04 +0000 (21:39 +1200)]
Add include/asm-generic/atomic.h
The arm, xtensa and mips version of atomic.h were already very similar
(the mips one was a copy of xtensa). Combine these implementations
together to produce a generic atomic.h that can be included by these
architectures (and any others that need it in future).
Signed-off-by: Chris Packham <judge.packham@gmail.com> Reviewed-by: Tom Rini <trini@konsulko.com>
Adam Ford [Wed, 5 Sep 2018 09:11:08 +0000 (04:11 -0500)]
mmc: omap_hsmmc: Fix pbias for omap3_logic to enable CD pin
The MMC card detect pin is connected to gpio127 on omap3_logic.
When setting up the pbias register for MMC, let's also enable
gpio_127 for the card detect. As part of the package deal,
gpio_126 and gpio_129 are also enabled.
Tom Rini [Fri, 31 Aug 2018 15:59:11 +0000 (11:59 -0400)]
dm: Update README.txt to clarify device tree usage
In the section about Device Trees add a paragraph at the end that
clarifies how we decide of a tree is valid or not. We say that all
bindings must either be in the specification (link provided) or in our
device-tree-bindings directory. We say that most of these come from the
Linux Kernel and as such some design decisions are made for us already,
but that in most cases we wish to retain compatibility.
Cc: Simon Glass <sjg@chromium.org> Cc: Bin Meng <bmeng.cn@gmail.com> Signed-off-by: Tom Rini <trini@konsulko.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sekhar Nori [Thu, 23 Aug 2018 11:41:31 +0000 (17:11 +0530)]
configs: am335x_evm: enable OF_BOARD_SETUP
Enable CONFIG_OF_BOARD_SETUP as it is needed for Beaglebone
black to overwrite the Ethernet phy address present in DT
in case the phy latches on to a different address.
Signed-off-by: Sekhar Nori <nsekhar@ti.com> Reviewed-by: Tom Rini <trini@konsulko.com>
Sekhar Nori [Thu, 23 Aug 2018 11:41:30 +0000 (17:11 +0530)]
board: ti: am335x: add support to fixup phy address
On beaglebone black, it can so happen that PHY address
is not latched correctly on reset and board boots with
PHY responding to a different address than that
programmed in device-tree. For example, see this report:
Workaround this by fixing up device-tree passed to kernel
by using the PHY address detected in hardware.
Beaglebone itself uses only one ethernet port and its DT
currently uses phy_id (obsoleted). But the function has
been written to handle multiple ports and phy_id as well
as phy-handle to make the function more generically useful.
Signed-off-by: Sekhar Nori <nsekhar@ti.com> Reviewed-by: Tom Rini <trini@konsulko.com>
Sekhar Nori [Thu, 23 Aug 2018 11:41:29 +0000 (17:11 +0530)]
drivers: net: cpsw: add support to update phy address
On some boards using TI CPSW, it may be possible that
PHY address was not latched correctly, and the actual
address that the phy responds on is different from that
set in device-tree. For example, see this problem report
on beaglebone black:
configs: sama5d27_som1_ek: Add defconfig for ENV/boot from uSD
This adds a defconfig for sama5d27_som1_ek board to get environment from
uSD. The defconfig is made from sama5d27_som1_ek_mmc_defconfig, with
'bootcmd' and 'bootargs' changed to kernel, device-tree and rootfs from
uSD. The environment is expected to be found in uSD's FAT partition.
configs: sama5d2_xplained: Add defconfig for ENV/boot from eMMC
This adds a defconfig for sama5d2_xplained board to get environment from
eMMC. The defconfig is made from sama5d2_xplained_mmc_defconfig, with
'bootcmd' and 'bootargs' changed to kernel, device-tree and rootfs from
eMMC. The environment is expected to be found in eMMC's FAT
partition.
Marek Vasut [Tue, 14 Aug 2018 09:27:02 +0000 (11:27 +0200)]
spl: Weed out CONFIG_SYS_TEXT_BASE usage
The SPL loaders assume that the CONFIG_SYS_TEXT_BASE memory location
is available and can be corrupted by loading ie. uImage or fitImage
headers there. Sometimes it could be beneficial to load the headers
elsewhere, ie. if CONFIG_SYS_TEXT_BASE is not yet writable while we
still want to parse the image headers in some local onchip memory to
ie. extract firmware from that image.
Add the possibility to override the location where the headers get
loaded by introducing new function, spl_get_load_buffer() which takes
two arguments -- offset from the CONFIG_SYS_TEXT_BASE and size of the
data that are to be loaded there -- and returns a valid buffer address
or hangs the system. The default behavior is the same as before, add
the offset to CONFIG_SYS_TEXT_BASE and return that address. User can
override the weak spl_get_load_buffer() function though.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Tom Rini <trini@konsulko.com> Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Create the handle of loaded images and the EFI_LOADED_IMAGE_PROTOCOL
inside efi_setup_loaded_image(). Do not use local variables.
Currently we expect the loaded image handle to point to the loaded image
protocol. Additionally we have appended private fields to the protocol.
With the patch the handle points to a loaded image object and the private
fields are added here. This matches how we handle the net and the gop
object.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
Alexander Graf [Sun, 23 Sep 2018 14:23:56 +0000 (16:23 +0200)]
efi_loader: Fix loaded_image handle passing from EL3
When running in EL3 mode on AArch64, we have to first drop to EL2
to execute a UEFI payload. When dropping down, the arguments to
the entry point have to stay identical to the ones for normal entry
though.
In commit ea54ad59286 ("efi_loader: pass handle of loaded image")
we incorrectly changed that logic and had the el3 entry path diverge.
Fix it up by syncing it back to what it's supposed to be.
Fixes: ea54ad59286 ("efi_loader: pass handle of loaded image") Signed-off-by: Alexander Graf <agraf@suse.de> Reviewed-by: Mark Kettenis <kettenis@openbsd.org>
Since GRUB patch d0c070179d4d ("arm/efi: Switch to arm64 linux loader",
2018-07-09) we do not need a workaround for GRUB on 32bit ARM anymore.
So let's eliminate function efi_exit_caches().
This will require Linux distributions to update grub-efi-arm to the GRUB
git HEAD (a tag containing the aforementioned GRUB patch is not available
yet).
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
CSI H can be used to position the cursor. The calling application may
specify a location that is beyond the limits of the screen. This may
lead to an illegal memory access.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
Currently we assign a lot of protocols to loaded images though
these protocols are not related to them. Instead they should be
installed on a separate handle. Via the device path it is the
parent to the devices like the network adapter.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
Not all terminals understand CSI [18t for querying the console size.
We should adhere to escape sequences documented in the console_codes
manpage and the ECMA-48 standard.
So here we follow a different approach. We position the cursor to the
bottom right and query its position. Before leaving the function we
restore the original cursor position.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
Alexander Graf [Sun, 16 Sep 2018 15:05:29 +0000 (17:05 +0200)]
efi_loader: Merge memory map entries
We currently do not combine memory entries that are adjacent and have
the same attributes. The problem with that is that our memory map can
easily grow multiple hundreds of entries in a simple UEFI Shell
environment.
So let's make sure we always combine all entries to make the memory
map as small as possible. That way every other piece of code that
loops through it should also gain some nice speed ups.
Alexander Graf [Fri, 22 Jun 2018 12:44:13 +0000 (14:44 +0200)]
sandbox: Allow to execute from RAM
With efi_loader, we may want to execute payloads from RAM. By default,
permissions on the RAM region don't allow us to execute from there though.
So let's change the default allocation scheme for RAM to also allow
execution from it. That way payloads that live in U-Boot RAM can be
directly executed.
Signed-off-by: Alexander Graf <agraf@suse.de> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Alexander Graf <agraf@suse.de>
Alexander Graf [Fri, 22 Jun 2018 12:44:12 +0000 (14:44 +0200)]
sandbox: Fix setjmp/longjmp
In sandbox, longjmp returns to itself in an endless loop because
os_longjmp() calls into longjmp() which is provided by U-Boot which
again calls os_longjmp().
Setjmp on the other hand must not return because otherwise the
return freees up stack elements that we need during longjmp().
The only straight forward fix that doesn't involve nasty hacks I
could find is to directly link against the system setjmp/longjmp
implementations. That means we just provide the compiler with
hints that the symbol will be available and actually fill them
out with versions from libc.
This approach should be reasonably platform agnostic
Signed-off-by: Alexander Graf <agraf@suse.de> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Alexander Graf <agraf@suse.de>
Simon Glass [Sat, 15 Sep 2018 06:50:56 +0000 (00:50 -0600)]
sandbox: Enhance map_to_sysmem() to handle foreign pointers
At present map_sysmem() maps an address into the sandbox RAM buffer,
return a pointer, while map_to_sysmem() goes the other way.
The mapping is currently just 1:1 since a case was not found where a more
flexible mapping was needed. PCI does have a separate and more complex
mapping, but uses its own mechanism.
However this arrange cannot handle one important case, which is where a
test declares a stack variable and passes a pointer to it into a U-Boot
function which uses map_to_sysmem() to turn it into a address. Since the
pointer is not inside emulated DRAM, this will fail.
Add a mapping feature which can handle any such pointer, mapping it to a
simple tag value which can be passed around in U-Boot as an address.
Signed-off-by: Simon Glass <sjg@chromium.org> Signed-off-by: Alexander Graf <agraf@suse.de>
Simon Glass [Sat, 15 Sep 2018 06:50:55 +0000 (00:50 -0600)]
sandbox: Add support for calling abort()
This function is useful to signal that the application needs to exit
immediate. It can be caught with a debugger (e.g. gdb). Add a stub for it
so that it can be called from within sandbox when an internal error
occurs.
Signed-off-by: Simon Glass <sjg@chromium.org> Signed-off-by: Alexander Graf <agraf@suse.de>
Simon Glass [Sat, 15 Sep 2018 06:50:54 +0000 (00:50 -0600)]
sandbox: Align RAM buffer to the machine page size
At present the sandbox RAM buffer is not aligned to any particular
address boundary. This makes the internal pointers somewhat random with
respect to the associated RAM buffer addresses.
Align the buffer to the page size of the machine to help with this. Note
that there is a header at the start of the allocated pointer. To avoid
returning a pointer which is not aligned to a page boundary, we waste
almost an entire page of memory for each allocation.
Signed-off-by: Simon Glass <sjg@chromium.org> Signed-off-by: Alexander Graf <agraf@suse.de>
projects / u-boot.git / log