Do not use data from the loaded image object after deleting it.
Fixes: 126a43f15b ("efi_loader: unload applications upon Exit()")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
load_options passed from do_efibootmgr() to do_bootefi_exec() may contain
invalid data from the stack which will lead to an invalid free().
Fixes: 0ad64007fe ("efi_loader: set load options in boot manager")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Up to now the escape key was not correctly detected in UEFI applications.
We had to hit it twice for a single escape to be recognized.
Use a 10 ms delay to detect if we are dealing with the escape key or an
escape sequence.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
IS_ENABLED() contains parentheses. But we should still put extra
parentheses around it in an if statement for readability.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
When opening an OP-TEE session we need to check the internal return
value of OP-TEE call arguments as well the return code of the
function itself.
The code was also ignoring to close the OP-TEE session in case the
shared memory registration failed.
Fixes: f042e47e8f ("efi_loader: Implement EFI variable handling via OP-TEE")
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Currently the size of the buffer to keep UEFI variables in memory is fixed
at 16384 bytes. This size has proven to be too small for some use cases.
Make the size of the memory buffer for UEFI variables customizable.
Reported-by: Paulo Alcantara (SUSE) <pc@cjr.nz>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Enable EFI_SET_TIME on the sandbox and QEMU ARM to ensure that we compile
and test the relevant code.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
If GetNextVariableName() is called with a non-existing combination of
VariableName and VendorGuid, return EFI_INVALID_PARAMETER.
If GetNextVariableName() is called with a string that is not zero
terminated, return EFI_INVALID_PARAMETER.
Reformat a line over 80 characters.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Since U-boot EFI implementation is getting richer it makes sense to
add support for EFI_TCG2_PROTOCOL taking advantage of any hardware TPM
available on the device.
This is the initial implementation of the protocol which only adds
support for GetCapability(). It's limited in the newer and safer
TPMv2 devices.
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Acked-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
When copying a string with must allocate a byte for the terminating '\0' in
the target buffer.
Fixes: fbe4c7df00 ("efi_loader: enable file SetInfo()")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Add missing newline to log messages in efi_rng_register() otherwise
something like below would be shown
Scanning disk virtio-blk#31...
Found 2 disks
Missing RNG device for EFI_RNG_PROTOCOLNo EFI system partition
Signed-off-by: Paulo Alcantara (SUSE) <pc@cjr.nz>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
When calling ExitBootServices during out unit tests we should not detach
devices as we need console output for runtime tests.
Fixes: 529441ca89 ("efi_loader: Disable devices before handing over control")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
When deleting a non-volatile variable it was deleted from memory but the
deletion was not persisted to the file system.
SetVariable() may be called with attributes == 0 to delete a variable. To
determine if the deletion shall be persisted we have to consider the
non-volatile flag in the attributes of the deleted variable and not the
value passed in the call parameter.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
This function will be used from several places in UEFI subsystem
to generate some specific form of utf-16 variable name.
For example, L"Capsule0001"
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Move function to separate module.
Use char * as argument instead of u16 *.
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
If this option, CONFIG_EFI_SETUP_EARLY, is enabled, the initialisation
of UEFI subsystem will be done as part of U-Boot initialisation.
Please note that this option won't be enabled explicitly by users,
instead, should be enabled implicitly by other configuration options.
Specifically, this feature will be utilised in implementing capsule-on-disk
feature.
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
If ChildHandle indicates the sole child of the driver, disconnect the
driver.
This fixes the test results for UEFI SCT 2.6 A
sub-tests 5.1.3.12.43, 5.1.3.12.44, 5.1.3.12.45.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Adjust the SetTime() and GetTime() runtime services to correctly convert
the daylight saving time information when communicating with the RTC.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
U-Boot Driver Model is supposed to remove devices with either
DM_REMOVE_ACTIVE_DMA or DM_REMOVE_OS_PREPARE flags set, before exiting.
Our bootm command does that by explicitly calling calling
"dm_remove_devices_flags(DM_REMOVE_ACTIVE_ALL);" and we also disable any
USB devices.
The EFI equivalent is doing none of those at the moment. As a result
probing an fTPM driver now renders it unusable in Linux. During our
(*probe) callback we open a session with OP-TEE, which is supposed to
close with our (*remove) callback. Since the (*remove) is never called,
once we boot into Linux and try to probe the device again we are getting
a busy error response. Moreover all uclass (*preremove) functions won't
run.
So let's fix this by mimicking what bootm does and disconnect devices
when efi_exit_boot_services() is called.
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Order header files according to https://www.denx.de/wiki/U-Boot/CodingStyle
Suggested-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
The sandbox is built with the SDL2 library with invokes the X11 library
which in turn calls getc(). But getc() in glibc is defined as
int getc(FILE *)
This does not match our definition.
int getc(void)
The sandbox crashes when called with parameter -l.
Rename our library symbol getc() to getchar().
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
With DM enabled the ethernet code will receive a packet, call
the push method that's set by the EFI network implementation
and then free the packet. Unfortunately the push methods only
sets a flag that the packet needs to be handled, but the code
that provides the packet to an EFI application runs after the
packet has already been freed.
To rectify this issue, adjust the push method to accept the packet
and store it in a temporary buffer. The EFI application then gets
the data copied from that buffer. This way the packet is cached
until is is needed.
The DM Ethernet stack tries to receive 32 packets at once, thus
we better allocate as many buffers as the stack.
Signed-off-by: Patrick Wildt <patrick@blueri.se>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
CONFIG_EFI_GRUB_ARM32 is only needed for architectures with caches that are
not managed via CP15 (or for some outdated buggy versions of GRUB). It
makes more sense to disable the setting per architecture than per defconfig.
Move QEMU's CONFIG_EFI_GRUB_ARM32_WORKAROUND=n from defconfig to Kconfig.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
strsep() changes the address that its first argument points to.
We cannot use the changed address as argument of free().
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
U-Boot offers a EFI_LOAD_FILE2_PROTOCOL which the Linux EFI stub can use to
load an initial RAM disk. Update the function comments of the
implementation.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Having an EFI_RNG_PROTOCOL without a backing RNG device leads to failure
to boot Linux 5.8.
Only install the EFI_RNG_PROTOCOL if we have a RNG device.
Reported-by: Scott K Logan <logans@cottsay.net>
Cc: Neil Armstrong <narmstrong@baylibre.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Refine text for overlap_only_ram description to
match to what exactly flag does and aling description
with other functions.
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
On RISC-V the global data pointer is stored in register gp. When a UEFI
binary calls the EFI API we have to restore it.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
efi_var_mem_notify_exit_boot_services() is invoked when ExitBootServices()
is called by the UEFI payload.
efi_var_mem_notify_exit_boot_services() should not be defined as
__efi_runtime as it is invoking EFI_ENTRY() and EFI_EXIT() which themselves
are not __efi_runtime.
Fixes: f1f990a8c9 ("efi_loader: memory buffer for variables")
Fixes: e01aed47d6 ("efi_loader: Enable run-time variable support for tee based variables")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Currently if the bootefi command fails due to missing authentication, the
user gets no feedback.
Write a log message 'Image not authenticated' if LoadImage() fails due to
missing authentication.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Bootxxxx variables are provided by the user and therefore cannot be
trusted. We have to validate them before usage.
A device path provided by a Bootxxxx variable must have an end node within
the indicated device path length.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
We need to check that device paths provided via UEFI variables are not
malformed.
Provide function efi_dp_check_length() to check if a device path has an
end node within a given number of bytes.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
If ResetSystem() is not implemented at runtime, it should return instead
of hanging in an endless loop. This allows the operating system to reset
the system by other means as Linux does. It also matches what EDK II
suggests in comments for functions ResetShutdown() and ResetWarm() in
OvmfPkg/Library/ResetSystemLib/ResetSystemLib.c.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Under the current implementation, all the signatures, if any, in
a signed image must be verified before loading it.
Meanwhile, UEFI specification v2.8b section 32.5.3.3 says,
Multiple signatures are allowed to exist in the binary’s certificate
table (as per PE/COFF Section “Attribute Certificate Table”). Only
one hash or signature is required to be present in db in order to pass
validation, so long as neither the SHA-256 hash of the binary nor any
present signature is reflected in dbx.
This patch makes the semantics of signature verification compliant with
the specification mentioned above.
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Reported-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Under the new file-based variable implementation, the secure state
is always and falsely set to 0 (hence, the secure boot gets disabled)
after the reboot even if PK (and other signature database) has already
been enrolled in the previous boot.
This is because the secure state is set up *before* loading non-volatile
variables' values from saved data.
This patch fixes the order of variable initialization and secure state
initialization.
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Fixes: 5f7dcf079d ("efi_loader: UEFI variable persistence")
In this commit, efi_signature_verify(with_sigdb) will be re-implemented
using pcks7_verify_one() in order to support certificates chain, where
the signer's certificate will be signed by an intermediate CA (certificate
authority) and the latter's certificate will also be signed by another CA
and so on.
What we need to do here is to search for certificates in a signature,
build up a chain of certificates and verify one by one. pkcs7_verify_one()
handles most of these steps except the last one.
pkcs7_verify_one() returns, if succeeded, the last certificate to verify,
which can be either a self-signed one or one that should be signed by one
of certificates in "db". Re-worked efi_signature_verify() will take care
of this step.
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
This is a bug fix; Setting an authenticated variable may fail due to
a memory corruption in the authentication.
A temporary buffer will, if needed, be allocated to parse a variable's
authentication data, and some portion of buffer, specifically signer's
certificates, will be referenced by efi_signature_verify().
So the buffer should be kept valid until the authentication process
is finished.
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Tested-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Up to now we used the value of the bootargs environment variable as load
options in the boot manager. This is not correct. The data has to be taken
from the Boot#### variable.
Let the boot manager copy the optional data of the EFI_LOAD_OPTION as load
options to the loaded image protocol.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
The bootefi bootmgr command has to set the load options for a loaded image
from the value of BootXXXX variable. If the boot manager is not used, the
value is set from the environment variable bootargs (or efi_selftest).
Factor out a common function efi_set_load_options().
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Some amount of code was moved/derived from efi_variable.c regarding
UEFI secure boot, in particluar in the commit 012c56ac76 ("efi_loader:
restructure code for TEE variables").
So add the orignal author's copyright notice.
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
The Kconfig symbol CONFIG_STACK_SIZE is used both by ARM and Microblaze
with the same meaning. Move it to menu 'General setup' so that we can use
it for all architectures.
Use the value of CONFIG_STACK_SIZE instead of a hard coded 16 MiB value for
reserving memory in the UEFI sub-system.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
We recently added functions for storing/restoring variables
from a file to a memory backed buffer marked as __efi_runtime_data
commit f1f990a8c9 ("efi_loader: memory buffer for variables")
commit 5f7dcf079d ("efi_loader: UEFI variable persistence")
Using the same idea we now can support GetVariable() and GetNextVariable()
on the OP-TEE based variables as well.
So let's re-arrange the code a bit and move the commmon code for
accessing variables out of efi_variable.c. Create common functions for
reading variables from memory that both implementations can use on
run-time. Then just use those functions in the run-time variants of the
OP-TEE based EFI variable implementation and initialize the memory
buffer on ExitBootServices()
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
The current code does not trim the output buffer correctly.
In fact it doesn't trim the buffer at all, since it calculates a wrong
value for it, which isn't even applied.
So let's remove the unused temporary size variable and trim the buffer
correctly.
Since we are editing efi_get_next_variable_name_int(), fix an indentation
error along the way.
Fixes: f042e47e8f ("efi_loader: Implement EFI variable handling via OP-TEE")
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Remove superfluous conversion to (u8 *) for memcpy argument.
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Use logging instead of printf() for messages occurring when scanning block
devices during the initialization of the UEFI sub-system.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
When reading a directory, EFI_BUFFER_TOO_SMALL should be returned when
the supplied buffer is too small, so a use-case is to call
efi_file_read with *buffer_size=0 and buffer=NULL to obtain the needed
size before doing the actual read.
So move the buffer!=NULL check to after the buffer size has been checked.
This fix allows the Redhat shim fallback to run and e.g. Fedora 32 now
boots out of the box.
Signed-off-by: Stefan Sørensen <stefan.sorensen@spectralink.com>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Currently we only check for the return value of tee_invoke_func().
Although OP-TEE and StMM will correctly set param[1].u.value.a and we'll
eventually return an error, the correct thing to do is check for the
OP_TEE return code as well.
So let's check for that and move tee_shm_free() and tee_close_session()
before exiting with an error to make sure we always clear the registered
memory.
Fixes: f042e47e8f ("efi_loader: Implement EFI variable handling via OP-TEE")
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Use EFI_DEVICE_ERROR for TEE communication problems.
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
