If a chunk was larger than 4GiB, then chunk_data_sz would overflow and
blkcnt would not be calculated correctly. Upgrade it to a u64 and cast
its multiplicands as well. Also fix bytes_written while we're at it.
Signed-off-by: Sean Anderson <sean.anderson@seco.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
"TCG PC Client Platform Firmware Profile Specification"
requires to measure every attempt to load and execute
a OS Loader(a UEFI application) into PCR[4].
This commit adds the PE/COFF image measurement, extends PCR,
and appends measurement into Event Log.
Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Tested-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org>
Replace CONFIG_HASH_CALCULATE by CONFIG_HASH
Fix conversions between pointers and u64.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
The hash_calculate() symbol is provided by hash-checksum.c. It depends
on hash_progressive_lookup_algo(), provided when CONFIG_HASH=y.
The issue is that hash_calculate() is used by the efi_loader,
irregardless of CONFIG_FIT_SIGNATURE. As pointed out in
commit 87316da05f ("lib: introduce HASH_CALCULATE option"),
enabling hash_calculate() based on CONFIG_FIT_SIGNATURE is incorrect.
To resolve this, use CONFIG_HASH as the compile switch for
hash-checksum.c. This ensures that all dependencies are compiled, and
is the most natural Kconfig to use.
There is the issue of having to 'select HASH' in a couple of places
that already 'select SHA256'. This is a deeper problem with how hashes
are organized, and fixing it is beyonf the scope of this change.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Acked-by: Masahisa Kojima <masahisa.kojima@linaro.org>
When we think of Kconfig, we usually think of features that we like
to enable or not. Ideally, we wouldn't use Kconfig to fix a build
issue, although sometimes it might make sense. With Kconfig it's hard
to guarantee that the fix is universal. We can only say that it works
for the set of tested configurations. In the majority of cases, it's
preferable to let the linker figure things out for us.
The reverted commit attempted to fix a build issue by adding an
invisible Kconfig option. This is wrong in several ways:
It invents a new Kconfig variable when CONFIG_HASH already
exists for the same purpose.
Second, hash-checksum.c makes use of the hash_progressive_lookup_algo()
symbol, which is only provided with CONFIG_HASH, but this dependency
was not expressed in the reverted patch.
It feels like Kconfig is turning into a listing of all available
source files, and a buffet to 'select' which ones to compile. The
purpose of this revert is to enable the next change to make use of
CONFIG_HASH instead of adding to Kconfig.
This reverts commit 87316da05f.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Acked-by: Masahisa Kojima <masahisa.kojima@linaro.org>
Coverity reported 3 warnings on the current code.
CID 331856, 331855, 331854 on the latest scan.
Fix the rest of the warnings by initializing the variables before
passing them to tpm2_get_pcr_info().
In order to avoid future warnings and errors initialize them to 0 within
the function as well, since the values are always OR'ed after querying the
hardware.
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Compiling with -Wextra produces a number of (harmless) warnings for the EFI
TCG2 protocol. Let's fix those in case we ever enable the flag on the EFI
subsystem
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Before this patch efi_dp_from_name() only accommodated a maximum file path
length of 31 characters. This leads to boot failures due to file name
truncation.
Allow arbitrary path lengths.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
This is preparation for PE/COFF measurement support.
PE/COFF image hash calculation is same in both
UEFI Secure Boot image verification and measurement in
measured boot. PE/COFF image parsing functions are
gathered into efi_image_loader.c, and exposed even if
UEFI Secure Boot is not enabled.
This commit also adds the EFI_SIGNATURE_SUPPORT option
to decide if efi_signature.c shall be compiled.
Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org>
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Configuring a system with CONFIG_EFI_CAPSULE_AUTHENTICATE=y but without a
method to retrieve the public key data is a misconfiguration. We should get
a build failure for it. Therefore remove the weak efi_get_public_key_data()
implementation.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
When building with LTO, the system libc's `errno` variable used in
arch/sandbox/cpu/os.c conflicts with U-Boot's `errno` (defined in
lib/errno.c) with the following error:
.../ld: errno@@GLIBC_PRIVATE: TLS definition in /lib64/libc.so.6
section .tbss mismatches non-TLS reference in
/tmp/u-boot.EQlEXz.ltrans0.ltrans.o
To avoid this conflict use different asm label for this variable when
CONFIG_SANDBOX is enabled.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
When compiling with LTO, the compiler fails with an error saying that
`crc_table` causes a section type conflict with `efi_var_buf`.
This is because both are declared to be in the same section (via macro
`__efi_runtime_data`), but one is const while the other is not.
Put this variable into the section .rodata.efi_runtime, instead of
.data.efi_runtime, via macro __efi_runtime_rodata.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Heinrich Schuchardt <xypron.gpk@gmx.de>
Add $(CFLAGS_EFI) and remove $(CFLAGS_NON_EFI) for
efi_selftest_miniapp_exception.o.
The removal is needed when compiling with LTO - this object file needs
to be compiled without -flto.
The adding is for consistency with other miniapps.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
It seems that sometimes (happening on ARM64, for example with
turris_mox_defconfig) GCC, when linking with LTO, changes the symbol
names of some functions, for example lib/string.c's memcpy() function to
memcpy.isra.0.
This is a problem however when GCC for a code such as this:
struct some_struct *info = get_some_struct();
struct some struct tmpinfo;
tmpinfo = *info;
emits a call to memcpy() by builtin behaviour, to copy *info to tmpinfo.
This then results in the following linking error:
.../lz4.c:93: undefined reference to `memcpy'
.../uuid.c:206: more undefined references to `memcpy' follow
GCC's documentation says this about -nodefaultlibs option:
The compiler may generate calls to "memcmp", "memset", "memcpy" and
"memmove". These entries are usually resolved by entries in libc.
These entry points should be supplied through some other mechanism
when this option is specified.
Make these functions visible by using the __used macro to avoid this
error.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Simon Glass <sjg@chromium.org>
This commit does the same thing as Linux commit 33def8498fdd.
Use a more generic form for __section that requires quotes to avoid
complications with clang and gcc differences.
Remove the quote operator # from compiler_attributes.h __section macro.
Convert all unquoted __section(foo) uses to quoted __section("foo").
Also convert __attribute__((section("foo"))) uses to __section("foo")
even if the __attribute__ has multiple list entry forms.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Generally speaking BINMAN_FDT makes sense for OF_SEPARATE or OF_EMBED.
For the other OF_CONTROL methods, it's quite possible binman node is
not available as binman is invoked during the build phase instead of
runtime. Let's only turn it on for OF_SEPARATE or OF_EMBED by default.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Build error occurs when CONFIG_EFI_SECURE_BOOT or
CONFIG_EFI_CAPSULE_AUTHENTICATE is enabled,
because hash-checksum.c is not compiled.
Since hash_calculate() implemented in hash-checksum.c can be
commonly used aside from FIT image signature verification,
this commit itroduces HASH_CALCULATE option to decide
if hash-checksum.c shall be compiled.
Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
EFI_TCG2 depends not only on TPMv2 but also on the underlying algorithms.
So select the missing SHA1, SHA256, SHA384 and SHA512 we currently support
Reported-by: Michal Simek <michal.simek@xilinx.com>
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Add 'default y'.
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Up to now we are stopping the EFI subsystem if a TPMv2 exists but the
protocol fails to install. Now that we've switched the config to 'default
y' the sandbox TPM fails, since it doesn't support all the required
capabilities of the protocol.
Not installing the protocol is not catastrophic. If the protocol fails
to install the PCRs will never be extended to the expected values, so
some other entity later in the boot flow will eventually figure it out
and take the necessary actions.
While at it fix a corner case were the user can see an invalid error
message when the protocol failed to install. We do have a tcg2_uninit()
which we call when the protocol installation fails. There are cases though
that this might be called before the configuration table is installed (e.g
probing the TPM for capabilities failed). In that case the user will see
"Failed to delete final events config table". So stop printing it since it's
not an actual failure , simply because the config table was never installed
in the first place.
In order to stop printing it make efi_init_event_log() and create_final_event()
cleanup themselves and only call tcg2_uninit() when the protocol installation
fails.
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Instead of just failing, clean up the installed config table and
EventLog memory if logging an s-crtm event fails during the protocol
installation
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Eliminate label 'out:' by using return.
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
efi_init_event_log() calls tcg2_uninit() in case of failure.
We can skip that since the function is called on efi_tcg2_register()
which also cleans up if an error occurs
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
This is same issue as https://bugzilla.redhat.com/show_bug.cgi?id=1733817,
but that fix was wrongly partial reverted.
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_PROTOCOL.Read() with *buffer_size=0 and buffer=NULL to
obtain the needed size before doing the actual read.
So remove the check only for directory reading, file reading already
do the check by itself.
Fixes: db12f518edb0("efi_loader: implement non-blocking file services")
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Cc: Stefan Sørensen <stefan.sorensen@spectralink.com>
Tested-by: Peter Robinson <pbrobinson@gmail.com>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
The current capsule authentication code checks if the environment
variable capsule_authentication_enabled is set, for authenticating the
capsule. This is in addition to the check for the config symbol
CONFIG_EFI_CAPSULE_AUTHENTICATE. Remove the check for the environment
variable. The capsule will now be authenticated if the config symbol
is set.
Signed-off-by: Sughosh Ganu <sughosh.ganu@linaro.org>
Reviwed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Commit a479f103dc ("hash: Allow for SHA512 hardware implementations")
defined function definitions for hardware accelerated SHA384 and SHA512.
If CONFIG_SHA_HW_ACCEL=y, these functions are used.
We already have boards using CONFIG_SHA_HW_ACCEL=y but none implements the
new functions hw_sha384() and hw_sha512().
For implementing the EFI TCG2 protocol we need SHA384 and SHA512. The
missing hardware acceleration functions lead to build errors on boards like
peach-pi_defconfig.
Introduce a new Kconfig symbol CONFIG_SHA512_HW_ACCEL to control if the
functions hw_sha384() and hw_sha512() shall be used to implement the SHA384
and SHA512 algorithms.
Fixes: a479f103dc ("hash: Allow for SHA512 hardware implementations")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Building 32bit boards with the TCG2 protocol enabled leads to a build
warning due to a missing conversion.
lib/efi_loader/efi_tcg2.c:774:27:
error: cast to pointer from integer of different size
[-Werror=int-to-pointer-cast]
774 | ret = tcg2_create_digest((u8 *)data_to_hash, data_to_hash_len,
| ^
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Today of_address_to_resource() is called only in
ofnode_read_resource() for livetree support and
fdt_get_resource() is called when livetree is not supported.
The fdt_get_resource() doesn't do the address translation
so when it is required, but the address translation is done
by ofnode_read_resource() caller, for example in
drivers/firmware/scmi/smt.c::scmi_dt_get_smt_buffer() {
...
ret = ofnode_read_resource(args.node, 0, &resource);
if (ret)
return ret;
faddr = cpu_to_fdt32(resource.start);
paddr = ofnode_translate_address(args.node, &faddr);
...
The both behavior should be aligned and the address translation
must be called in fdt_get_resource() and removed for each caller.
Fixes: a44810123f ("dm: core: Add dev_read_resource() to read device resources")
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>
In case of failure at efi_get_variable_int("BootOrder"), we should
skip examining boot option variables and return an appropriate error
code which is the one the function returned.
Fixes: CID 331153 Code maintainability issues (UNUSED_VALUE)
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Indicated by Coverity Scan CID 331147
The ESRT was being refreshed in situations where the UpdateCapsule
procedure failed. In that scenario:
1) the ESRT refresh was superfluous.
2) a failed ESRT refresh return code overwrites the UpdateCapsule error
return code.
This commit ensures that the ESRT is only refreshed when the
UpdateCapsule performs successfully.
CC: Heinrich Schuchardt <xypron.glpk@gmx.de>
CC: Sughosh Ganu <sughosh.ganu@linaro.org>
CC: AKASHI Takahiro <takahiro.akashi@linaro.org>
CC: Tom Rini <trini@konsulko.com>
CC: Andre Przywara <andre.przywara@arm.com>
CC: nd@arm.com
Signed-off-by: Jose Marinho <jose.marinho@arm.com>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Bumping the digest list count, for all supported algorithms, can be
calculated outside of the individual switch statements. So let's do that
for every loop iteration instead and simplify the code a bit.
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
To avoid diverging function definitions we need to include efi_loader.h.
Fixes: fe179d7fb5 ("efi_loader: Add size checks to efi_create_indexed_name()")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Add 2 configs CONFIG_LMB_MEMORY_REGIONS and CONFIG_LMB_RESERVED_REGIONS
to change independently the max number of the regions in lmb
library.
When CONFIG_LMB_USE_MAX_REGIONS=y, move the lmb property arrays to
struct lmb and manage the array size with the element 'max' of struct
lmb_region; their are still allocated in stack.
When CONFIG_LMB_USE_MAX_REGIONS=n, keep the current location in
struct lmb_region to allow compiler optimization.
Increase CONFIG_LMB_RESERVED_REGIONS is useful to avoid lmb errors in
bootm when the number of reserved regions (not adjacent) is reached:
+ 1 region for relocated U-Boot
+ 1 region for initrd
+ 1 region for relocated linux device tree
+ reserved memory regions present in Linux device tree.
The current limit of 8 regions is reached with only 5 reserved regions
in DT.
see Linux kernel commit bf23c51f1f49 ("memblock: Move memblock arrays
to static storage in memblock.c and make their size a variable")
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Add a max parameter in lmb_region struct to handle test
in lmb_add_region without using the MAX_LMB_REGIONS
define.
This patch allows to modify these size independently for
memory of reserved regions in the next patches.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Remove the unused field size of struct lmb_region as it is initialized to 0
and never used after in lmb library.
See Linux kernel commit 4734b594c6ca ("memblock: Remove memblock_type.size
and add memblock.memory_size instead")
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Variable "final" will have SHA512 digest, but currently
the array size is not sufficient. Let's fix it.
Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
If realloc() fails, we should free the old buffer.
Fixes: c74cd8bd08: ("efi_loader: capsule: add capsule_on_disk support")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Remove function invocations using the EFI_CALL macro for those
functions that do not have an EFI_ENTRY call in their definition. Such
functions can use u-boot api's which rely on u-boot global data(gd)
pointer. The Arm and RiscV architectures maintain a separate gd
pointer, one for u-boot, and a separate gd for the efi application.
Calling a function through the EFI_CALL macro changes the gd pointer
to that used for the efi application, with u-boot gd being
unavailable. Any function then trying to dereference u-boot's gd will
result in an abort.
Fix this issue by removing the EFI_CALL macro for all of such
functions which do not begin by an EFI_ENTRY function call.
Signed-off-by: Sughosh Ganu <sughosh.ganu@linaro.org>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Most modern OpenSSL engines have methods overridden at the EVP level rather
than at RSA level, to make these engines work properly with mkimage, the RSA
signing code needs to switch to using EVP_* APIs as much as possible.
Signed-off-by: Donald Chan <hoiho@lab126.com>
[trini: Rebase on top of keyfile changes]
Signed-off-by: Tom Rini <trini@konsulko.com>
Keys can be derived from keydir, and the "key-name-hint" property of
the FIT. They can also be specified ad-literam via 'keyfile'. Update
the ECDSA signing path to use the appropriate one.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Keys can be derived from keydir, and the "key-name-hint" property of
the FIT. They can also be specified ad-literam via 'keyfile'. Update
the RSA signing path to use the appropriate one.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
mkimage supports rsa2048, and rsa4096 signatures. With newer silicon
now supporting hardware-accelerated ECDSA, it makes sense to expand
signing support to elliptic curves.
Implement host-side ECDSA signing and verification with libcrypto.
Device-side implementation of signature verification is beyond the
scope of this patch.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
fdt_add_bignum() is useful for algorithms other than just RSA. To
allow its use for ECDSA, move it to a common file under lib/.
The new file is suffixed with '-libcrypto' because it has a direct
dependency on openssl. This is due to the use of the "BIGNUM *" type.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
rsa-checksum.c sontains the hash_calculate() implementations. Despite
the "rsa-" file prefix, this function is useful for other algorithms.
To prevent confusion, move this file to lib/, and rename it to
hash-checksum.c, to give it a more "generic" feel.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
This introduces strlcat, which provides a safer interface than strncat. It
never copies more than its size bytes, including the terminating nul. In
addition, it never reads past dest[size - 1], even if dest is not
nul-terminated.
This also removes the stub for dwc3 now that we have a proper
implementation.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
strlcpy should always return the number of bytes copied. We were
accidentally missing the nul-terminator. We also always used to return a
non-zero value, even if we did not actually copy anything.
Fixes: 23cd138503 ("Integrate USB gadget layer and USB CDC driver layer")
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Similar to support for SHA1 and SHA256, allow the use of hardware hashing
engine by enabling the algorithm and setting CONFIG_SHA_HW_ACCEL /
CONFIG_SHA_PROG_HW_ACCEL.
Signed-off-by: Joel Stanley <joel@jms.id.au>
The optee_copy_fdt_nodes is only used to copy op-tee nodes
of U-Boot device tree (from gd->fdt_blob when OF_LIVE is not activated)
to external device tree but it is not compatible with OF_LIVE.
This patch migrates all used function fdt_ functions to read node on
old_blob to ofnode functions, compatible with OF_LIVE and remove this
parameter "old_blob".
The generated "device tree" is checked on stm32mp platform with OF_LIVE
activated.
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Although ptr arithmetics are allowed with extensions in gcc, they
are not allowed by the C spec. So switch to (void *)(uintptr_t) instead
Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
