Stub out the handle_seccomp_options function and other unused
code and data if WITH_SECCOMP is not defined.
Currently the handle_seccomp_options would always be called with
a NULL string to parse if WITH_SECCOMP is not defined, so this just
eliminates the code that will never be called.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Have swtpm report PTM_CAP_SEND_COMMAND_HEADER flag for a TPM 2
indicating that it knows how to handle the TCG header prefix
for the data channel.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Add support for handling the TPM 2 command header prefix that looks like this:
struct tpm2_send_command_prefix {
uint32_t cmd;
uint8_t locality;
uint32_t size;
};
This patch converts the existing send function to use an array of struct iovec
using writev() where we can optinally prepend a response header and trailer.
We detect whether the command had the above type of header prefixed and if so
indicate this with an offset into the byte stream where the normal TPM command
starts, which follows right after this header. In case such a header was found
we prepend a header in the response and append a 4-byte trailer. For TPM 1.2
there are no changes and for raw TPM 2 commands there are no changes, either.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Have SWTPM_IO_Read() read at leat the 10 bytes from the (regular) TPM
header. This solves a particular problem with the TPM 2 header prefix for
which we will add support in the subsequent patch. In the prefix case the IBM
TSS 2 stack sends 4 bytes for the command, then 1 byte for the locality, and then
4 bytes for the length indicator followed by the command. If we just read once
we would only then get 4 bytes. Reading 10 bytes gets all of these plus the sub-
sequent TPM command or the whole TPM command in case this prefix header is
missing.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Now that we will always read all bytes available on a file descriptor
we don't need the MAINLOOP_READALL flag anymore.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Since we can only read the file descriptor for the passphrase once,
we cannot stop the TPM 1.2 and then restart it but we have to use
'swtpm_ioctl -i' to re-init it.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Also extend swtpm_setup to allow passing the file descriptor for
the key or password files or pipes.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Enable reading the password file from a file descriptor using
--key pwdfd=<fd>
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Get rid of stat() so that we can read from pipes as well.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
filelen is the length of the file as it was stat'ed. len is
the number of bytes read. They should be the same. So, use
len as a parameter to SHA512() rather than filelen.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Enable reading the state encryption or migration keys using a
file descriptor passed via the command line option parameter.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Consolidate the exist of the parse_key_options function.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Add missing kdf option documentation to help screen. The kdf
option parameter support has been there since 0.1.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
The CUSE TPM needs the prctl syscall to function correctly when
commands are passed via the thread pool.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Fix an illegal heap access while parsing the options by making
sure that we do not access the tok variable beyond its size
when comparing a character to '=' and later on when accessing
the value after the '='.
This bug was discovered by configuring as follows on Fedora 28:
CFLAGS="-fsanitize=address -g -ggdb" LIBS="-lasan" \
./configure --prefix=/usr
and running tests like this:
sudo bash -c "SWTPM_TEST_EXPENSIVE=1 make -j32 check"
The test case test_tpm2_ctrlchannel2 indicated the error.
It looks like gcc on Ubuntu Xenial on Travis did not detect this
error.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Extend the previously modified test case to also test on
Darwin now that we are able to convert a file descriptor
to a filename.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Enable support for the seccomp blacklist profile that is
enabled by default and can be disabled using the following option:
--seccomp action=none
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
According to the XDG spec,
https://specifications.freedesktop.org/basedir-spec/basedir-spec-latest.html:
"If $XDG_CONFIG_HOME is either not set or empty, a default equal to
$HOME/.config should be used."
This fixes setting up a TPM with libvirt running in a user session.
It works by checking if configuration files are readable in the
directory priority order (XDG_CONFIG_HOME, then HOME, then SYSCONFDIR).
When libvirt is running as a system instance, $HOME isn't set, so it
will fall back on @SYSCONFDIR@ (/etc usually)
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
To avoid failures with mktemp call on BSD type of systems replace
mktemp --tmpdir=<dir> with TMPDIR=<dir> mktemp, which seems more portable.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
OS X cannot resolve the path we get from genenv("_") when running a
test script on Travis (reason unknown). So we use _NSGetExecutablePath()
instead and display the path in the error message in case realpath()
fails.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
When the project is configured with --prefix=/usr/local several configuration
files are installed to /usr/local/etc but not read by the programs that
need them. This patch fixes this issue by replacing @SYSCONFDIR@ in the source
code of those files with the actual used path.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
For PKCS11 modules (such as SoftHSM) to sign a TPM EK or platform
certificate we use 'pkcs11:' prefixed URIs like those URIs for
TPM 1.2 starting with tpmkey:.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Implement terminate_proc, which first tries to gracefully terminate a
process and then, if synchronization is requested, waits for 1 second
for it to disappear and then kills it with SIGKILL if it didn't go away.
Use this function for stop_tpm and stop_tcsd, which will first try
to send a SIGTERM to a process assuming it will terminate while we
do something else and then, in case the process needs to be restarted,
make sure that it's gone before it is restarted.
This addresses an issue where previously SIGTERM was sent to a process
before it was restarted but the restarting failed due to the previous
process still holding a lock on the file because it wasn't able to react
to the SIGTERM fast enough.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Uninstall the signal handler before we close the signal handler's
notification file descriptor so that signals received after the file
descriptor is closed do not cause an error.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Make sure that the msg parameter passed into recvmsg wasn't modified
by the library and verify that the pointers and sizes it contains are
still valid after the call. This addresses an issue raised by Coverity.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Ceritficates for a TPM 2 have to be signed using SHA256 and those for
a TPM 1.2 have to be signed using SHA1 (following specs). We can use
either of these algorithms when using the GnuTLS TPM support since it
will create the hash and have the TPM 1.2 sign whatever it created.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
swtpm_setup.sh does not need to be owned by tss:tss and in the
Fedora package it's not even allowed. So remove the install hook
that was changing the ownership.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Implement the --tcsp-system-ps-file option to make a backup of
TCSD's system_ps_file for later use of the setup TPM with the
TCSD. We need this for a test case.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Have the CUSE TPM return the TPM start error, particularly
TPM_DECRYPT_ERROR in case a wrong key was passed, rather than the plain
TPM_FAIL error code. This was a difference with the 'other' swtpm
interfaces that already returned the original error code.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Gnutls supports TPM 1.2 keys in its API and we can use a TPM 1.2
signing key for signing the EK or platform certs.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
The crypto functions for AES encryption and decryption clash with crypto
function names of libtpms. Avoid this by prefixing them with SWTPM_ instead
of TPM_.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
This patch adds rules that are needed for running QEMU with libvirt 4.4
on F28. This allows one to test the libvirt management stack and QEMU with
swtpm and SELinux in enforcing mode.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Only when the CUSE TPM is built we need the glib-2.0 dependency
(which can be removed at some point probably).
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
glib2 did not clean up a proxy thread once the pool was torn down. The
reason was that an exclusive pool was used. Switch to a non-exclusive
pool, which will cause the last thread to terminate as well and we have
a clean valgrind log now.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Valgrind complains about uninitialized bytes being accessed by a writev()
call in libfuse. It points to a line in the error branch
'error_not_running', but the error does not seem to come from there.
Instead, if we initialize ptm_est to zeros the problem goes away.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
With the fuse_session now global, we can use it to exit the CUSE
TPM using fuse_session_exit().
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
To be able to get to the low level fuse_session data structure, which
allows us to terminate the program more gracefully than the exit() in
the ioctl processing code, we implement the low level cuse startup code.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
F27 needs an additional rule for starting the swtpm from libvirt.
Add the rule 'allow svirt_t swtpm_exec_t:file map'
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
gcc 4.2.1 does not accept the initializer '= { 0 }' and so we now switch
this to memset as well.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Valgrind complains that there are uninitialized bytes when ioctl() is called.
This is due to the fact that we are passing the structure to the ioctl()
without specifying how many bytes are valid in the passed buffer.
So, we initialize the buffers to 0 and the complaint is resolved.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
When the AES encryption and decryption function were extended to support
256bit keys the size of the user ivec on the stack was also increased to
32 bytes. For the user provided IV we now must only copy 16 bytes in case
an 128 bit key is used. Fix the length parameter to the memcpy().
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Install the private libswtpm_libtpms into $(libdir)/swtpm
rather than into the shared libaries directory.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
To make the RPM building easier for the --without-cuse case we have to omit
a couple of files from being built so that they are not left-overs when
packing the RPM without building swtpm-cuse.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
An out-of-tree build fails to build swtpm_svirt.pp due to missing
source files from the swtpm.pp policy. Add the swtpm.{te,if,fc} files
as dependencies of the swtpm_svirt.pp policy package.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Use .NOTPARALLEL for the SELinux policy build target to avoid running
the rule concurrently and deleting the tmp directory while another
build is running.
For some reason it is not possible to use $(MAKE) since otherwise the
building of the policy package fails.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
This patch is partially from Lukas Vrabec (lvrabec@redhat.com) and fixes
some nits in the SELinux policy.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Extend the SELinux policy with rules for svirt_tcg when
QEMU is not being started with KVM by libvirt.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Coverity scan shows an uninitialized access in case an error is
encountered and the cleanup path is taken before datum was initialized.
Fix the error by initializing the variable.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Coverity scan discovered a buffer overrun error in case the
datum->size = 0. Fix the error.
The function with the potential error is not called if datum->size == 0.
Also make the 'size' variable larger so there cannot be a overrun in
'size = datum->size + sizeof(buffer) - i'.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement command line support for choosing the cipher to use for
the TPM state encryption. Either aes-128-cbc or aes-256-cbc can be
used. The same cipher has to be passed on the swtpm command line
when using the TPM.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Use the flags in the header to check the size of the provided key(s) against
the one(s) from the user. There are the state and migration keys, each has
a different flag in the header if a 256bit key was used. We display an error
about the mismatching key size against expected size if the given key is of
the wrong size.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Extend the buffer the key is read into to hold an 256bit key.
Add support for aes-256-cbc. Also, aes-128-cbc is now a synonym for
aes-cbc.
Update the man pages to reflect the AES 128 support.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Enable 256 bit AES keys in the en- and decryption functions.
Set a flag in the header to at least indicate whether 128bit
or 256bit keys were used.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Remove the valid field from the symmtric key structure and
user userKeyLength > 0 instead, which is the same.
Also remove the tag and fill fields, which were originally used in
TPM 1.2.
Use function calls to test whether the file or migration key is
available where possible.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
To get rid of the hard coded AES key size of 128, we add a filed
userKeyLength that describes the length of the key being used.
We replace TPM_AES_BLOCK_SIZE with userKeyLength where possible.
Rename TPM_AES_BLOCK_SIZE to SWTPM_AES_BLOCK_SIZE.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Use pbkdf2 as the default kdf and sha512 for the existing
test case. Do away with file limit of 32 bytes. This may
break backwards compatibility for some but better to do this
before a release...
Switch the existing test cases to use kdf=sha512 on the command
line where necessary to that the state for these test cases
does not need to be recreated.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
On some systems /bin/bash does not exists but the bash is somewhere
else and can be invoked with /usr/bin/env bash.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
The building of rpm files fails due to some issues with the rules. To make
this work we explicily list the policy packages' dependencies and use them
in the rules. This now make 'make distcheck' work and lets us build an RPM.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
For sys_dependencies.h to be found during out-of-tree builds we
need to include -I$(top_srcdir)/include.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
OS/X also does not have a proc filesystem, so we use the '_'
environment variable to determine the path of the program.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
The daemon() call is deprecated on OS X. This patch works around it so
that we can compiled with -Wall -Werror and still can use the daemon()
call while avoiding compiler errors due to this function's deprecation
marking in stdlib.h.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Make swtpm_setup.sh work on FreeBSD.
A common fix for the problem related to echo stopping to write
data into a socket after the first occurrence of \x0a sees to
write it into a pipe and cat the data from there.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
A (draft) specification indicates the ECC key NVRAM indices to use.
Adapt the code to use them in case of EK ECC key.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
The OpenBSD stat tool uses -f%z to return the file size in bytes.
So we wrap stat in get_filesize and call it with different parameters
depending on the system.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
On OpenBSD 6.2 we have an older version of od that does not support the
-w parameter. Also we have to use the bash echo there since /bin/echo
does not support the -e parameter. We are using /bin/echo on the other
systems since there was (once) a bug when \x0a had to be converted to a
hex number.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
OpenBSD does not have a proc file system from which we could
determine the path that swtpm_setup was run from and where we
expect swtpm_setup.sh to be located in as well. Using getenv("_")
seems to work as a replacement in the OpenBSD case.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Replace usage of shuf with a get_random function that produces
a random number between an upper and lower bound.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Fix the number of bytes passed into the control channel for commands
and expected for responses. The pointer to the memory is ok.
This fixes uninitialized memory errors reported by valgrind.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement a common function for cleaning up before exit.
This should now always remove the pid file.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Introduce compile-time variable HAVE_TCSD if the TCSD could
be found. It influences whether TPM 1.2 related swtpm_setup
test cases can be run. If it is set, they can be run.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Get rid of the dependency on netstat so that we can at least support
swtpm_setup with TPM 2 on Cygwin and BSD later on. This is the first
step into this direction.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Check that the netstat tool, which may not be required to be installed,
is indeed installed and usable with a set of command line options.
If this tool is not installed it may end up causing swtpm_setup to
hang forever.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
The include file include/swtpm.h will be generated from swtpm.h.in and
reside under the $(top_builddir) rather than the $(top_srcdir).
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement --pcr-banks to allow a user to choose the set of active
PCR banks. We determine the PCR banks available and enable those
that the user chose and that are available.
The log will now print out the following:
Successfully activated PCR banks sha1,sha256 among sha1,sha256,sha384.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Remove the -s parameter to tpm2_nv_define since it was causing
a 0-size NVRAM location to be created for the platform cert.
Also use the nvindex parameter rather than TPM2_NV_INDEX_PlatformCert
for addressing the platform NVRAM index.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Adjust the flags for the TPM 2 NVRAM locations to adhere to the
specification:
TCG PC Client Platform: TPM Profile (PTP) Specification
Family "2.0"; Level 00; Rev 01.03 v22; May 22, 2017
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
To prepare for writing data into different locations for ECC keys,
assign a variable the location of the NVRAM to write RSA related data
into.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
We write the EK template into the NVRAM location when it is non-standard.
It's non-standard once the EK can be used for signing.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Add the --decryption option to enable key encipherment separately
from enabling signing for the EK. The key encipherment is not set
but needs to be set if --allow-signing is used and key encipherment
is also requested.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
In case of a TPM 2 we allow the creation of a signing key by passing
--allow signing. To also enable key encipherment, we add the --decryption
option to allow key encipherment and signing at the same time.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Use the standard EK and SRK handles per IWG spec
"TCG TPM v2.0 Provisioning Guidance"; Version 1.0, Rev 1.0, March 15 2017
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Create the tpm2_createprimary_rsa_params function that has common code
for creating a primary RSA key with parameters.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Don't require root privileges to run swtpm_setup with a TPM 2 target.
For TPM 1.2 we need the high privilges due to TrouSerS wanting to be
started as root (or 'tss'), but for TPM 2 we do not use any tools
to manufacture the initial state that would require high privileges.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
libtpms preview branch tpm2-preview.rev142 enables volatile state
marshalling and unmarshalling which in turn enables the suspending
and resumption of the TPM state. This patch enables the capabilty
bits and adds test cases for testing the TPM state suspending and
resumption.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Provide support for creating certificates for TPM2 ECC type of keys.
Extend the test cases and the man pages.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Support TPM 2 in swtpm_setup and swtpm_setup.sh.
Implement support for all command line options except for:
o --take-ownership and anything related to ownership passwords
o --lock-nvram
o --display
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
TPM2 allows the primary key to also be a signing key, so in case
--tpm2 is provided, --allow-signing can be provided as well in
case the primary can also be used for signing operations.
We use SHA256 for the signing algorithm when TPM 2 is being used.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Extend the swtpm_bios tool with a --tpm2 command line parameter
to support TPM 2 initialization.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Call the libtpms API for resetting the TPM Established flag rather
than sending a TPM command, which only works for TPM1.2.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement support for TPM2. Some of the capabilities are not supported yet in
this patch.
Extend the man pages with description for --tpm2.
Missing: configure should probe for needed API calls in libtpms
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
The subject alternative name need to have a special sequence of
bytes prepended to them for certtool to accept the data. Also TCG's
sample certificate does show the sequence. The byte sequence is of
the form: 0x30 <subsequent length> 0xa4 <subsequent length> <data>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
The previous patch's reversal was partially wrong. The authority
key id needs to be set explicitly from the key id of the signing
key of the issuing CA.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Remove the copying of the authority key id from the given issuer
cert to the created cert since this copies the wrong key id and
besides that it will be set automatically when the certificate is
created.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Create a .lock file in the directory with the TPM state and get
a lock on this file and hold on to the lock until swtpm ends.
This precludes other swtpm instances to step on the same state.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Whenever we encrypt the data we generate a new random IV and append a
tlv block with the IV to the byte stream. We mark the IV with different
tags depending on whether they are for the migration data or the (TPM)
data directly. All IVs are part of the HMAC and are added to it after
the data blob.
Adjust test cases that now return larger sizes of data. A constant
checksum over the data cannot be expected anymore, thus we have to remove
the verification of the checksum over the returned state (IV changes
every time).
The size of the blobs grow by 22 bytes, 6 for the tlv header, 16 bytes
for the IV (128 bit AES key).
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Get the IV from a tlv block in the data stream. If none is found, which
is the case when reading older state, we get a NULL pointer for the IV
and call the functions with the NULL pointer, which provides backwards
compatibility.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Enable a caller to pass an IV into the AES CBC encryption/decryption
function. If the caller passes NULL, we use the IV with all zeroes.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Refactor the functions calculating the HMAC so that we can later on
pass the IV for the AES CBC encryption as well.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Prepend tag-length-value (tlv) headers in front of all data being stored in
the byte stream following the header. This lets us uniquely identify plain
data (= TPM state), encrypted data (= encrytped TPM state), migration data
(which is wrapped plain or encrytped TPM state), and an HMAC block to
validate the plain data.
We keep support for version 1 for reading the data but convert them to
version 2 when writing them out. This way we loose backwards compatibility
(downgrading of swtpm is not possible), but it allows us to extend the state
in the future by adding addition blocks with tlv headers.
Version 1 of the encryption was prepending the hash on the plaintext data
then encrypting all of it. This method is not so good. In version 2 we now
use Encrypt-then-MAC (EtM) where we encrypt the data and then calculate an
HMAC on the encrypted data.
Files written by the swtpm didn't have a header before. Now they also get a
header. This means that the state written into files and the state retrieved
using the API (swtpm_ioctl --save) have the same format, but still differ
in so far as the API wraps the data in a tlv header for migration, which the
files written out as state would never get.
Adapt a couple of test cases show file sizes and hashes have changed now.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement functions for supporting tag-length-value headers
in the byte stream we store the TPM's data into.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Remove FreeBL support for swtpm since there will not be support
for FreeBL with TPM 2 in libtpms.
Since a lot of documentation shows --with-openssl, we leave that for
now.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Coverity found that the usage of strncpy may leave an unterminated
string. In this case it is ok, if the string is unterminated since
it would only be the part of a response and the client would have
to collect all the parts as indicated by the total length of the
string. So we use memcpy instead and leave a note in the code. So
far the strings would not nearly be 3k to get close to the maximum.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Rather than writing to files directly and having to validate the state in
those files using TPMLIB_ValidatetState(), we now use the new
TPMLIB_SetState() call to set the TPM's state blobs. The advantage of this
call is that it doesn't overwrite state files and ends up leaving state in
files that the TPM cannot use. Instead, it validates the state immediately
when the blob is set and returns an error in case the state cannot be
accepted.
We need to adapt one test case that now gets a failure earlier than before.
Before the TPM_INIT failed, now setting the encrypted blob fails because it
cannot be decrypted and thus cannot be accepted by the TPM.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
For TPM 1.2 the spec says that we must not set the subject, so we
do not set it but keep it around for TPM 2 certificates.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
To make the ASN.1 look like the one from the spec. we have to
use a lower-level GNUTLS API function to set it.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Put the code that creates the ASN.1 for the platform and
TPM manufacturer info into their own functions.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Use the swtpm_ioctl tool also to get TPM manufacturer, firmware
version and TPM model and pass it to the external tool creating
the certificate.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Simplify the two's complement implementation by converting the number
into a big endian and writing it out into a byte array that is prefixed
with a 0-byte. This covers all unsigned ints while the previous imple-
mentation would have been wrong once the number exceeded 255.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Calculate the two's completement of the spec_level and spec_revision
numbers so that ASN.1 properly stores them as unsigned integers.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Create ASN.1 for the Extended Key Usage field for the EK certificate
that has the oid 2.23.133.8.1 and for the platform certificate the
oid 2.23.133.8.2. Both are registered OIDs:
http://oid-info.com/get/2.23.133.8.1http://oid-info.com/get/2.23.133.8.2
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Non-privileged mode needs another rule and on Fedora 27 we need
a swtpm_exec_t related rule we did not need on Fedora 23.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
The extern cert program was called with --vmid <foo>--tpm-spec-family ...
and failed due to a missing whitespace. This patch fixes it.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Use the new --info parameter for swtpm_ioctl to get TPM specification
info from the swtpm and use this as a parameter for creating the EK
certificate.
Extend the man page.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement the --info option that takes a flag as an argument
and returns information about the TPM implementation. Only the
value '1' has been defined for now, which returns a line as
the following:
{"TPMSpecification":{"family":"1.2","level":2,"revision":116}}
Extend the man page.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
We quetry the swtpm for TPM specification info that goes into the
certificate for the EK.
Update the test cases that now see more capabilties being returned
by the swtpm.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Since the EK cert must have an Authority Key Id, we error out if
the signing certificate we are given does not have an Authority
Key Id. The typical reason why it wouldn't have it is that we are
using a self-signed (rootCA) certificate directly.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Log the output of the tool that is creating the certificate and
prefix every line with the name of the tool.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
To not loose stderr we have to run the exec command in a subshell
and suppress the output of the subshell otherwise the exec command
seems to close stderr on us.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Create the TPMSpecification SEQUENCE and add it to the subject
directory attributes of an EK cert.
The code generates the same ASN.1 for the Subject Directory Attributes
as the example in the EK spec has.
> openssl asn1parse -in ${cert} -strparse 603
0:d=0 hl=2 l= 30 cons: SEQUENCE
2:d=1 hl=2 l= 28 cons: SEQUENCE
4:d=2 hl=2 l= 5 prim: OBJECT :2.23.133.2.16
11:d=2 hl=2 l= 19 cons: SET
13:d=3 hl=2 l= 17 cons: SEQUENCE
15:d=4 hl=2 l= 3 prim: UTF8STRING :1.2
20:d=4 hl=2 l= 4 prim: INTEGER :41010000
26:d=4 hl=2 l= 4 prim: INTEGER :7B000000
Extend existing test case so they create the ASN.1 as well.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
TPM 1.2 certificates need a header when written to NVRAM. This patch
adds a --add-header option for this and prepends the header to.
Also see TCG PC Specific Inmplementation Specification section 7.4.4 and
7.4.5.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement a mode parameter for the tpmstate option so that the user
can choose what the file mode bits of the TPM's state files will be.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Replace the previously used swtpmcuse_svirt type of policy with
swtpm_svirt. This policy contains the minimum set of rules necessary
to start swtpm from within libvirt.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Address the issue that the pidfile needs to be removed at the end if
a file descriptor was passed in. So we have to look up the file
given the file descriptor and by using /proc/self/fd/.
We also want to make sure that the given file descriptor describes
a regular file.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Update the SELinux for swtpm so that swtpm can be launched from
libvirtd and QEMU can access swtpm's UnixIO socket and talk to it.
Use this as follows:
> cd src/selinux
> make clean all
> semodule -i *.pp
> restorecon /usr/bin/swtpm*
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement support for uid and gid parameters that allows us to set the file
ownership of a UnixIO socket.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement support for a mode parameter that allows us to set the file
mode bits on a UnixIO socket.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Create tpmlib_register_callbacks and call it to register the
libtpms callbacks a lot earlier. Before it was done in tpmlib_start,
which was invoked by CMD_INIT, which was after state blobs could
have been set already.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Validate the state blobs after they have been set (during
state migration) so that we can fall back to the migration
source in case they are not accepted by the TPM.
This patch requires the latest version of libtpms 0.6
from the master branch.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Avoid using the expect tool as much as possible. Use it only if
the user requested a special owner and/or SRK password.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Use the passwords as specified by the user also when only temporarily
taking ownership. Use the zero password in case nothing has been
provided by the user. This is a first step to loosen the dependency
on expect, which pulls in too many dependencies.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
In case the poll() is interrupted by a signal handler and we
get EINTR, we loop again assuming that the signal handler will
send us a notification to end.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
POLLERR is indicated in some error cases when using the chardev with the
vtpm_proxy with runc. In case the flag is set, we terminate the client
connection rather than endlessly looping.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
File descriptors 0, 1, and 2, when used for file descriptor passing,
as is the case in some older version of runc/docker-ce (17.06), do
not work. So we refuse accepting these values and log the error message.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Only pass through the vmid parameter to the tool creating the cert
if this parameter was set. The tool creating the cert may require
it or ignore it and is supposed to be able to track certificates
for which it minted certs using this parameter.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
The certificate chain could not be verified due to a wrong
API call for getting the authority key id where the key id
was gotten instead.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Determine the size of the unix_path parameter and then only copy
those number of bytes + terminating null into the destination.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Determine the size of the unix_path parameter and then only copy
those number of bytes + terminating null into the destination.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Check whether filename and input pointers are NULL pointers.
Per previous code this is currently not possibly but should
resolve some issues with Coverity.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Use void * for the command rather than tpm_header *, which we do
not need. This addresses a few Coverity complaints.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Move some of the debugging output into its own function and
differentiate output by TPM Version number.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Add options --overwrite and --not-overwrite to allow or prevent
overwriting of existing TPM state. If neiter of the options is
given and existing state is found, an error is returned.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement a command for setting and querying the buffer size the
TPM implementation (libtpms) is using. The setting of the
buffersize allows to reduce the size of the buffer to a size
that the interface can support so that these two sizes match
and the TPM will not produce larger responses than what the
interface can support.
Extend swtpm_ioctl with an option to set the buffersize.
Adapt the existing tests to reflect the newly supported command.
Implement a new test for getting/setting of the buffer size.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Changeset aa3999 introduced a segfault when calling
ctrlchannel_set_client_fd() with a NULL pointer. Like all the other
functions, we return with -1 in this case.
Since the segfault occurred on process shutdown no problems were
noticeable through bad test results or so.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Introduce another parameter to ctrlchannel_new and server_new
to pass the Unix socket's path. Implement ctrlchannel_free and
server_free to clean up and remove the Unix socket's path.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement ctrlchannel_set_client_fd() to set a new client
fd and return its current value.
Use this new function to set the file descriptor to -1 upon
closing its associate file descriptor in the main loop.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement server_set_fd() function to get the current file descriptor
and set a new one on the server structure.
Use this function to transfer the server file descriptor from the
server structure to the mainloop structure and have the main loop
close the file descriptor.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
On OpenBSD the socket's name would be cut short by one letter at the
end due to an off-by-one for the calculated sockaddr struct's length.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
The control channel of a UnixIO socket is not supported
on Cygwin, so do not present this flag. Return an error
if this command is run.
Adapt the test case.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Cast the exponent to unsigned long int to avoid a compiler
error reporting comparison of signed and unsigned integers.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Consider TPM_PORT environment at the time of parsing server options. This way we
can avoid the code duplication, hence removed obslote code.
Signed-off-by: Amarnath Valluri <amarnath.valluri@intel.com>
As objected by QEMU upstream developers to use two different sockets for
starting/using of swtpm, This commit adds support for passing unix domain
socket over control channel.
The summary of the changes include:
- Defined new control command CMD_SET_DATAFD, using this clients can send data
socket.
- set mlp.fd and mlp.flags outside of the mainloop
- updated the testcases
Signed-off-by: Amarnath Valluri <amarnath.valluri@intel.com>
In preparation for receiving control messages via UnixIO sockets,
switch to recvmsg() for receiving the first chunk of data.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Enable receiving control commands in small chnunks. Allow 500ms
for the whole control command to be received.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Add the missing PTM_CAP_GET_CONFIG to the returned flags.
Fix the order of the flags on the way.
Fix the test cases that test for the flags.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Remove the check for the FUSE_IOCTL_COMPAT flag to make a 32bit
CUSE swtpm work on 64bit host.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Fix possible illegal buffer access in case a character device
is used via ioctl. In this case we have to make sure we only
access the number of bytes in the ioctl structure itself.
In case of a socket, make sure we received enough bytes to
be able to access the payload.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Change the startup behavior with respect to requiring an INIT command
via control channel. We change this for the socket and chardev
interfaces so that the behavior now is the same as that of the CUSE
interface.
Introduce the --flag not-need-init command line option for the socket
and chardev interfaces to allow the old behavior using this option.
Adapt some of the test cases and swtpm_setup.sh.in that now need this
command line flag.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Allow running swtpm_setup directly as tss user using the following
command lines as an example for how to do it:
mkdir /tmp/test-tss
chown tss:tss /tmp/test-tss
chmod 770 /tmp/test-tss
su -c "swtpm_setup --tpm-state /tmp/test-tss" -s /bin/sh tss
This change prevents us from requiring invocation under the root
user.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Replace the show_help variable with change_user varaible with
opposite logic. The name of the variable is better for upcoming
changes.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
If the return code from the command processing function is 0,
get the error code from the command response and return that one
instead. The problem was that clients though resetting the
establishment bit worked in locality 0 for example, while it did
not.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Return the number of bytes received in case ctrlcmd does not
use an ioctl because in this case we need to know how many bytes
we received in the response.
The receiving of the state blob also needs to take into account
how many bytes were received in the initial response so we write
the proper amount of bytes from that response into the file.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
While parsing the options we modified the argv[] values and
the commas disappeared, which showed when checking for the
swtpm processes using 'ps'. So, work on a copy of the options
parameter.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
There was an offset bug in the code getting the state blobs
from a client.
Also fix an endianess bug in a control channel command.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement allow-set-locality for CUSE TPM and extend the existing
locality test case with commands that try to write the SetLocality
comamnd to the device.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Implement support for --locality allow-set-locality option that
is primarily useful in cases when the VTPM proxy access is enabled
by file descriptor passing.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Add options for locality handling so that commands in locality
4 for example are rejected per command line parameter. This is
useful when the vTPM is used with containers.
Also implement the custom TPM/TPM2_SetLocality command to allow
the Linux vTPM proxy driver to set the locality in which subsequent
TPM commands will be executed.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Fix the missing check of return code from handle_server_options
to check for negative value like all the other checks also do.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Use logprintf wherever possible. While the logging is not set up,
print error messages to stderr and prefix them with 'swtpm: '. After
the logging has been set up all error messages go into the log file.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Use a buffer for printing in TPM_PrintAll. Emit the buffer
upon a new line. This prepares for printing every line with
a prefix.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Add MAIN_LOOP_FLAG_READALL for non SOCK_STREAM client sockets. This allows
reading all socket data once.
Signed-off-by: Amarnath Valluri <amarnath.valluri@intel.com>
swtpm server is closing the client data connection after serving a single
request, which shall not happen. Hence added MAIN_LOOP_FLAG_KEEP_CONNECTION
mainloop flag to keep the client data connection alive, till client is alive.
Signed-off-by: Amarnath Valluri <amarnath.valluri@intel.com>
New option '--ctrl clientfd=<fd>' is added to the control socket parameters.
The passed 'fd' is used as control channel client-fd and treated as single
client mode.
Signed-off-by: Amarnath Valluri <amarnath.valluri@intel.com>
Initialize variables in ctrlchannel_return_state so that all error cases
are properly handled.
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
