| Age | Commit message (Collapse) | Author |
|---|
|
This commit also supersedes PR#232 which was closed on Jul 1, 2021.
So that original codesign EKU codes cannot be bothered.
To enable the codesign check, ENABLE_CODESIGN_EKU can be set to 1.
To disable the codesign check, ENABLE_CODESIGN_EKU can be set to 0
or just omit this flag.
For example:
make xxxx ENABLE_CODESIGN_EKU=1 xxxx shim.efi
Signed-off-by: Dennis Tseng <dennis.tseng@suse.com>
|
|
Also modify the ModSign EKU check to use VerifyEKUsInPkcs7Signature() to
check the signer certificate instead of the certificate directly from the
key database.
This commit supersedes the PR#232 and PR#661 (Apply the EKU checks) so
that author's original codes can be quite independent of other modification.
To answer the question in PR#232, author also changed the conditional
statement to EFI_Status != EFI_SUCCESS right after VerifyEKUsInPkcs7Signature()
in Cryptlib/Pk/CryptPkcs7Verify.c
Signed-off-by: Dennis Tseng <dennis.tseng@suse.com>
Signed-off-by: Gary Lin <glin@suse.com>
|
|
In some cases the -Werror / -Wfoo / -Wno-foo / -Wno-error=foo bits
aren't propagating in a way that clang is happy with, so we get errors
about unused variables and the like in Cryptlib and OpenSSL.
We're never going to fix those nits in this tree, so this patch tries
even harder to make the compiler ignore them, or at least not end the
build with a benign error.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Cryptlib and OpenSSL both currently throw warnings with some compilers
using -Wunused-but-set-variable:
clang -std=gnu11 -ggdb -ffreestanding -fmacro-prefix-map=/home/pjones/devel/github.com/shim/main/= -fno-stack-protector -fno-strict-aliasing -fpic -fshort-wchar -nostdinc -m64 -mno-mmx -mno-sse -mno-red-zone -Os -Wall -Wextra -Wno-missing-field-initializers -Wno-unused-parameter -Werror -I/home/pjones/devel/github.com/shim/main/Cryptlib -I/home/pjones/devel/github.com/shim/main/Cryptlib/Include -I/home/pjones/devel/github.com/shim/main/gnu-efi/inc -I/home/pjones/devel/github.com/shim/main/gnu-efi/inc/x86_64 -I/home/pjones/devel/github.com/shim/main/gnu-efi/inc/protocol -isystem /home/pjones/devel/github.com/shim/main/include/system -isystem /usr/lib64/clang/16/include -DMDE_CPU_X64 -c -o Pk/CryptX509.o Pk/CryptX509.c
Pk/CryptX509.c:94:19: error: variable 'Index' set but not used [-Werror,-Wunused-but-set-variable]
UINTN Index;
^
clang -std=gnu11 -ggdb -ffreestanding -fmacro-prefix-map=/home/pjones/devel/github.com/shim/main/= -fno-stack-protector -fno-strict-aliasing -fpic -fshort-wchar -nostdinc -m64 -mno-mmx -mno-sse -mno-red-zone -Os -Wall -Wextra -Wno-missing-field-initializers -Wno-empty-body -Wno-implicit-fallthrough -Wno-unused-parameter -Werror -I/home/pjones/devel/github.com/shim/main/Cryptlib/OpenSSL -I/home/pjones/devel/github.com/shim/main/Cryptlib -I/home/pjones/devel/github.com/shim/main/Cryptlib/OpenSSL/Include/ -I/home/pjones/devel/github.com/shim/main/Cryptlib/OpenSSL/crypto -I/home/pjones/devel/github.com/shim/main/Cryptlib/Include -I/home/pjones/devel/github.com/shim/main/gnu-efi/inc -I/home/pjones/devel/github.com/shim/main/gnu-efi/inc/x86_64 -I/home/pjones/devel/github.com/shim/main/gnu-efi/inc/protocol -I/home/pjones/devel/github.com/shim/main/Cryptlib/OpenSSL/crypto/asn1 -I/home/pjones/devel/github.com/shim/main/Cryptlib/OpenSSL/crypto/evp -I/home/pjones/devel/github.com/shim/main/Cryptlib/OpenSSL/crypto/modes -I/home/pjones/devel/github.com/shim/main/Cryptlib/OpenSSL/crypto/include -isystem /home/pjones/devel/github.com/shim/main/include/system -isystem /usr/lib64/clang/16/include -DL_ENDIAN -D_CRT_SECURE_NO_DEPRECATE -D_CRT_NONSTDC_NO_DEPRECATE -DOPENSSL_SMALL_FOOTPRINT -DPEDANTIC -DMDE_CPU_X64 -c -o crypto/asn1/t_x509.o crypto/asn1/t_x509.c
crypto/asn1/t_x509.c:504:18: error: variable 'l' set but not used [-Werror,-Wunused-but-set-variable]
int ret = 0, l, i;
^
Since we normally build with -Werror, these cause builds to fail in
these cases. While the bad code should be addressed, it appears
generally safe, so we should solve it upstream.
This patch adds -Wno-unused-but-set-variable to the Cryptlib Makefile,
and removes the conditionalization on gcc in the OpenSSL Makefile, as
clang now has this argument, and since we don't support building with
clang for the final build, it's useful to have clang-based tools
working.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
There is one long-standing problem in CRT realloc wrapper, which will
cause the obvious buffer overflow issue when re-allocating one bigger
memory block:
void *realloc (void *ptr, size_t size)
{
//
// BUG: hardcode OldSize == size! We have no any knowledge about
// memory size of original pointer ptr.
//
return ReallocatePool ((UINTN) size, (UINTN) size, ptr);
}
This patch introduces one extra header to record the memory buffer size
information when allocating memory block from malloc routine, and re-wrap
the realloc() and free() routines to remove this BUG.
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Ting Ye <ting.ye@intel.com>
Cc: Jian J Wang <jian.j.wang@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qin Long <qin.long@intel.com>
Reviewed-by: Jian J Wang <jian.j.wang@intel.com>
Validated-by: Jian J Wang <jian.j.wang@intel.com>
Cherry picked from https://github.com/tianocore/edk2.git, commit
cf8197a39d07179027455421a182598bd6989999. Changes:
* `SIGNATURE_32` -> `EFI_SIGNATURE_32`
* Added definition of `MIN`
Fixes https://github.com/rhboot/shim/issues/538
Signed-off-by: Nicholas Bishop <nicholasbishop@google.com>
|
|
AuthenticodeVerify() calls OpenSSLs d2i_PKCS7() API to parse asn encoded
signed authenticode pkcs#7 data. when this successfully returns, a type
check is done by calling PKCS7_type_is_signed() and then
Pkcs7->d.sign->contents->type is used. It is possible to construct an asn1
blob that successfully decodes and have d2i_PKCS7() return a valid pointer
and have PKCS7_type_is_signed() also return success but have Pkcs7->d.sign
be a NULL pointer.
Looking at how PKCS7_verify() [inside of OpenSSL] implements checking for
pkcs7 structs it does the following:
- call PKCS7_type_is_signed()
- call PKCS7_get_detached()
Looking into how PKCS7_get_detatched() is implemented, it checks to see if
p7->d.sign is NULL or if p7->d.sign->contents->d.ptr is NULL.
As such, the fix is to do the same as OpenSSL after calling d2i_PKCS7().
- Add call to PKS7_get_detached() to existing error handling
Cc: Chao Zhang <chao.b.zhang@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Signed-off-by: Jian J Wang <jian.j.wang@intel.com>
Cherry-picked-from: https://github.com/tianocore/edk2/commit/26442d11e620a9e81c019a24a4ff38441c64ba10
|
|
EFI_BOOT_SERVICES includes CopyMem() and SetMem() functions which are
marked EFIAPI, and in the case of CopyMem() does not mark the source
argument as CONST.
This patch makes all our invocations work with that, so (once gnu-efi's
implementation is fixed to match) we can use the existing implementation
as the implementation in a mock EFI_BOOT_SERVICES.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Be much more explicit about exactly which va_* stuff comes from which
ABI in both shim and gnu-efi. This fixes the problem where we see:
| (null):0:(null)() v->name:"(null)" v->rtname:"(null)"
| (null):0:(null)() v->data_size:0 v->data:0x0
and similar messages where everything is NULL.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
gcc -Wextra, has a lot of good, useful checks, a few obnoxious checks,
and a few absolutely insane checks.
This enables -Wextra, but disables -Wmissing-field-initializers, because
it is irrational nonsense that just leads to worse code. It also
disables some specific things in the Cryptlib and Cryptlib/OpenSSL
trees:
Both:
-Wno-unused-parameter
- there are a fair number of functions that have to conform to some API
or another but have arguments that are unused, but haven't been
marked with UNUSED; we don't need to see warnings about them.
Cryptlib/OpenSSL:
-Wno-empty-body
- functions that exist merely to populate some API
-Wno-implicit-fallthrough
- these probably should get fixed someday, but I bet upstream will do
it and rebasing will solve it
-Wno-old-style-declaration
- this gripes if you write "const static" instead of "static const".
Again I expect rebasing will fix it at some point.
-Wno-unused-but-set-variable
- self explanatory, and again, I expect a rebase to solve it someday.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
There were a couple cases of "uninitialized variable" warnings in the
imported OpenSSL code; I used the current OpenSSL code as a guide for
picking the default values used here.
On my dev system there is one remaining build warning in OpenSSL's
crypto/asn1/x_pkey.c:X509_PKEY_new() function. Unfortunately it
involves some preprocessor crimes and the fix would be a bit ugly.
Fortunately it appears the warning here is harmless and can be
ignored.
As a point of reference, my build system is a current Arch install
with GCC v10.2.0 and GNU-EFI v 3.0.12.
Signed-off-by: Paul Moore <pmoore2@cisco.com>
|
|
This gets us the same working definition for VA_* va_* etc everywhere,
and it's the same definition edk2 is using.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Some time ago, commit e571428e212 ("Update to openssl to 1.0.2e")
changed the way we define the va_* (and VA_*) functions and macros.
Unfortunately, it only changed for some parts of the tree, and the
different parts of the tree need to both call each other and use the
same types in all cases. Additionally, they need to all be able to call
gnu-efi functions such as VPrint, which means they need the same va_list
type definitions everywhere.
This partially reverts that patch, adding EFIAPI back and unsetting
NO_BUILTIN_VA_FUNCS everywhere.
|
|
This re-structures our includes so we can be sure everything is always
including all the system headers in a uniform, predictable way.
Temporarily it also adds a bunch of junk at all the places we use
variadic functions to specifically pick either the MS (cdecl) or ELF
ABIs.
I'm not 100% sure that's all correct (see later patch) but it's enough
to allow this to build.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
|
|
This patch does some makefile cleanups, to fix the parts that are
actually just bad that the previous patch left in for clarity:
- removes -fno-builtin . This flag is implied by -ffreestanding , which
we use everywhere.
- gets rid of the two places where ARM has their own -O flags for no
real reason. Note that this will make those use -Os instead of -O2.
- export VERBOSE and DEBUG if they're set.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
This improves our static analysis targets by making them work better
with our make variables, and inhibits the use of ccache while building
those.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
These are all the NULL pointer dereferences (which all appear to be, at
worst, very difficult to hit) that gcc -fanalyzer finds in our OpenSSL
code.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Some of our makefile bits are a mess, as you may have noticed, making
changes to them difficult to review.
This patch attempts to make some parts of them vaguely less of a mess,
in order to facilitate review of follow-up changes. To so it:
- coalesces feature flags, optimizations, -W{no-,}, -W{no-}error,
include directives, and define/undefine directives into (mostly)
separate groups.
- exports them as appropriate so the sub-makes can use them
- Makes sure we have -Wextra -Werror everywhere, but adds -Wno-foo and
-Wno-error=foo directives at the appropriate places to keep the net
warnings the same.
- makes the arch defines in Cryptlib and Cryptlib/OpenSSL use the
appropriate ones, with no attempt to make them less stupid, without
changing the overall order.
- coalesces the various includes, with no attempt to make them less
stupid, without changing the overall order.
- One giant glaring whitespace fix in Cryptlib/OpenSSL/Makefile
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Signed-off-by: Peter Jones <pjones@redhat.com>
Upstream: pr#211
|
|
Upstream-commit-id: 6a73ca814af
|
|
Pass MDE_CPU_ARM, similar to how it is done for the other supported
architectures, otherwise the build fails in:
Cryptlib/Include/OpenSslSupport.h:55:2: error:
#error Unknown target architecture
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Upstream-commit-id: cb83c14628b
|
|
I don't think the x86 binaries clang builds will actually work unless
they just infer -maccumulate-outgoing-args from __attribute__((__ms_abi__),
but it's nice to have the analyzer working.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
system
Signed-off-by: Tamas K Lengyel <lengyelt@ainfosec.com>
|
|
We don't need the functions in CryptPem.c.
Signed-off-by: Gary Lin <glin@suse.com>
|
|
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
I'm still having some trouble with the offsetof() definition, so just
nerf it to what stddef.h would say anyway.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
I think this works around most of them.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Disable DES completely since it's already old and insecure.
This makes MokManager not support the DES based password hash but
probably no one is using it.
Signed-off-by: Gary Lin <glin@suse.com>
|
|
MD4 is known to be insecure and shim never uses it.
Signed-off-by: Gary Lin <glin@suse.com>
|
|
- Declare some functions in the proper headers
+ We missed them for a long time...
- Cast offsetof to UINTN
+ The original casting triggers the gcc warning since int can not
present the offset for the 64bit machines.
- Cast the "char" array to "CHAR8 *" to avoid the gcc warnings
- Implement atoi correctly
Signed-off-by: Gary Lin <glin@suse.com>
|
|
strcmp() and strcasecmp() are widely used in openssl. Implement those
two functions to eliminate the gcc warnings and the potential crash.
Signed-off-by: Gary Lin <glin@suse.com>
|
|
The changes in the openssl headers cause the inclusion of
CrtLibSupport.h eariler than the inclusion of stddef.h, so "offsetof"
was defined twice and this caused the followling build error:
In file included from Cryptlib/Include/openssl/buffer.h:23:0,
from Cryptlib/Include/openssl/x509.h:22,
from shim.c:56:
/usr/lib64/gcc/x86_64-suse-linux/6/include/stddef.h:417:0: error: "offsetof" redefined [-Werror]
#define offsetof(TYPE, MEMBER) __builtin_offsetof (TYPE, MEMBER)
In file included from Cryptlib/Include/limits.h:15:0,
from Cryptlib/Include/openssl/ossl_typ.h:13,
from Cryptlib/Include/openssl/x509.h:20,
from shim.c:56:
Cryptlib/Include/CrtLibSupport.h:192:0: note: this is the location of the previous definition
#define offsetof(type, member) ( (int) & ((type*)0) -> member )
We can lower the priority of the gcc include path or just remove the
path, but this might cause problem since the path was introduced on
purpose(*). Instead, including stddef.h first is more feasible.
(*) https://github.com/rhinstaller/shim/commit/d51739a416400ad348d8a1c7e3886abce11fff1b
Signed-off-by: Gary Lin <glin@suse.com>
|
|
OpenSSL changes quite a bit of the key validation, and most of the keys
I can find in the wild aren't marked as trusted by the new checker.
Intel noticed this too: https://github.com/vathpela/edk2/commit/f536d7c3ed
but instead of fixing the compatibility error, they switched their test
data to match the bug.
So that's pretty broken.
For now, I'm reverting OpenSSL 1.1.0e, because we need those certs in
the wild to work.
This reverts commit 513cbe2aea689bf968f171f894f3d4cdb43524d5.
This reverts commit e9cc33d6f2b7f35c6f5e349fd83fb9ae0bc66226.
This reverts commit 80d49f758ead0180bfe6161931838e0578248303.
This reverts commit 9bc647e2b23bcfd69a0077c0717fbc454c919a57.
This reverts commit ae75df6232ad30f3e8736e9449692d58a7439260.
This reverts commit e883479f35644d17db7efed710657c8543cfcb68.
This reverts commit 97469449fda5ba933a64280917e776487301a127.
This reverts commit e39692647f78e13d757ddbfdd36f440d5f526050.
This reverts commit 0f3dfc01e2d5e7df882c963dd8dc4a0dfbfc96ad.
This reverts commit 4da6ac819510c7cc4ba21d7a735d69b45daa5873.
This reverts commit d064bd7eef201f26cb926450a76260b5187ac689.
This reverts commit 9bc86cfd6f9387f0da9d5c0102b6aa5627e91c91.
This reverts commit ab9a05a10f16b33f7ee1e9da360c7801eebdb9d2.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
This lets you do:
mkdir build-x64 build-ia32
cd build-x64
make TOPDIR=.. -f ../Makefile
cd ../build-ia32
setarch i686 -B make ARCH=ia32 TOPDIR=.. -f ../Makefile
And not worry about generated sources and headers mixing and matching.
Signed-off-by: Peter Jones <pjones@redhat.com>
|
|
CryptPem only provides one function: RsaGetPrivateKeyFromPem(). Since we
don't need to retrieve any private key, it's safe to disable the
function.
Signed-off-by: Gary Lin <glin@suse.com>
|
|
Disable DES completely since it's already old and insecure.
This makes MokManager not support the DES based password hash but
probably no one is using it.
Signed-off-by: Gary Lin <glin@suse.com>
|
|
MD4 is known to be insecure and shim never uses it.
Signed-off-by: Gary Lin <glin@suse.com>
|
|
strcmp() and strcasecmp() are widely used in openssl. Implement those
two functions to eliminate the gcc warnings and the potential crash.
Signed-off-by: Gary Lin <glin@suse.com>
|
|
- Declare some functions in the proper headers
+ We missed them for a long time...
- Cast offsetof to UINTN
+ The original casting triggers the gcc warning since int can not
present the offset for the 64bit machines.
- Cast the "char" array to "CHAR8 *" to avoid the gcc warnings
- Implement atoi correctly
Signed-off-by: Gary Lin <glin@suse.com>
|
|
The changes in the openssl headers cause the inclusion of
CrtLibSupport.h eariler than the inclusion of stddef.h, so "offsetof"
was defined twice and this caused the followling build error:
In file included from Cryptlib/Include/openssl/buffer.h:23:0,
from Cryptlib/Include/openssl/x509.h:22,
from shim.c:56:
/usr/lib64/gcc/x86_64-suse-linux/6/include/stddef.h:417:0: error: "offsetof" redefined [-Werror]
#define offsetof(TYPE, MEMBER) __builtin_offsetof (TYPE, MEMBER)
In file included from Cryptlib/Include/limits.h:15:0,
from Cryptlib/Include/openssl/ossl_typ.h:13,
from Cryptlib/Include/openssl/x509.h:20,
from shim.c:56:
Cryptlib/Include/CrtLibSupport.h:192:0: note: this is the location of the previous definition
#define offsetof(type, member) ( (int) & ((type*)0) -> member )
We can lower the priority of the gcc include path or just remove the
path, but this might cause problem since the path was introduced on
purpose(*). Instead, including stddef.h first is more feasible.
(*) https://github.com/rhinstaller/shim/commit/d51739a416400ad348d8a1c7e3886abce11fff1b
Signed-off-by: Gary Lin <glin@suse.com>
|
|
- Delete the old openssl files and use the script to copy the new files
- Add "-DNO_SYSLOG" to CFLAGS and add crypto/include to the include path
Signed-off-by: Gary Lin <glin@suse.com>
|
|
- Update update.sh to copy the openssl 1.1.0 source files
- Refresh the supplemental patch to reflect the change
Signed-off-by: Gary Lin <glin@suse.com>
|
|
- Update to edk2 commit 7c410b3d4180087020c7734bf67cdc4ad9fdb136
CryptoPkg/BaseCryptLib: Adding NULL checking in time() wrapper.
- Update headers in Cryptlib/Include/openssl/ to 1.1.0e
+ Also copy the openssl internal headers
Signed-off-by: Gary Lin <glin@suse.com>
|
