Command-line tool and Python library to calculate expected measurement of an AMD SEV/SEV-ES/SEV-SNP guest VM for confidential computing.
Install from pip:
pip install sev-snp-measure
This installs the sevsnpmeasure package and the sev-snp-measure
command-line script.
Clone the Github repo and run the script directly from the local directory:
git clone https://github.com/virtee/sev-snp-measure.git
cd sev-snp-measure
./sev-snp-measure.py --help
$ sev-snp-measure --help
usage: sev-snp-measure [-h] [--version] [-v] --mode {sev,seves,snp,snp:ovmf-hash,snp:svsm}
[--vcpus N] [--vcpu-type CPUTYPE] [--vcpu-sig VALUE] [--vcpu-family FAMILY]
[--vcpu-model MODEL] [--vcpu-stepping STEPPING] [--vmm-type VMMTYPE] --ovmf
PATH [--kernel PATH] [--initrd PATH] [--append CMDLINE]
[--guest-features VALUE] [--output-format {hex,base64}]
[--snp-ovmf-hash HASH] [--dump-vmsa] [--svsm PATH]
[--vars-size SIZE | --vars-file PATH]
Calculate AMD SEV/SEV-ES/SEV-SNP guest launch measurement
options:
-h, --help show this help message and exit
--version show program's version number and exit
-v, --verbose
--mode {sev,seves,snp,snp:ovmf-hash,snp:svsm}
Guest mode
--vcpus N Number of guest vcpus
--vcpu-type CPUTYPE Type of guest vcpu (EPYC, EPYC-v1, EPYC-v2, EPYC-IBPB, EPYC-v3, EPYC-v4,
EPYC-Rome, EPYC-Rome-v1, EPYC-Rome-v2, EPYC-Rome-v3, EPYC-Milan, EPYC-
Milan-v1, EPYC-Milan-v2, EPYC-Genoa, EPYC-Genoa-v1)
--vcpu-sig VALUE Guest vcpu signature value
--vcpu-family FAMILY Guest vcpu family
--vcpu-model MODEL Guest vcpu model
--vcpu-stepping STEPPING
Guest vcpu stepping
--vmm-type VMMTYPE Type of guest vmm (QEMU, ec2)
--ovmf PATH OVMF file to calculate hash from
--kernel PATH Kernel file to calculate hash from
--initrd PATH Initrd file to calculate hash from (use with --kernel)
--append CMDLINE Kernel command line to calculate hash from (use with --kernel)
--guest-features VALUE
Hex representation of the guest kernel features expected to be included
(defaults to 0x1); see README.md for possible values
--output-format {hex,base64}
Measurement output format
--snp-ovmf-hash HASH Precalculated hash of the OVMF binary (hex string)
--dump-vmsa Write measured VMSAs to vmsa<N>.bin (seves, snp, and snp:svsm modes only)
snp:svsm Mode:
AMD SEV-SNP with Coconut-SVSM. This mode additionally requires --svsm and either --vars-file
or --vars-size to be set.
--svsm PATH SVSM binary
--vars-size SIZE Size of the OVMF_VARS file in bytes (conflicts with --vars-file)
--vars-file PATH OVMF_VARS file (conflicts with --vars-size)
$ sev-snp-measure --mode snp --vcpus=1 --vcpu-type=EPYC-v4 --ovmf=OVMF.fd --kernel=vmlinuz --initrd=initrd.img --append="console=ttyS0 loglevel=7"
1c8bf2f320add50cb22ca824c17f3fa51a7a4296a4a3113698c2e31b50c2dcfa7e36dea3ebc3a9411061c30acffc6d5a
$ sev-snp-measure \
--mode snp:svsm \
--vmm-type=QEMU \
--vcpus=4 \
--vcpu-type=EPYC-v4 \
--ovmf=OVMF_CODE.fd \
--svsm=svsm.bin --vars-file=OVMF_VARS.fd
3447e476b226e317890a350003b56ee17becb48d1dc25dd6b5819a1192df3238f50cda0f0216bd5ae2a992ad7ab961c4
$ snp-create-id-block --help
usage: snp-create-id-block [-h] [--measurement VALUE] [--idkey PATH] [--authorkey PATH]
Calculate AMD SEV-SNP guest id block
optional arguments:
-h, --help show this help message and exit
--measurement VALUE Guest launch measurement in Base64 encoding
--idkey PATH id private key file
--authorkey PATH author private key file
After installing the sev-snp-measure package with pip, you can call it from
another Python application:
from sevsnpmeasure import guest,id_block
from sevsnpmeasure import vcpu_types
from sevsnpmeasure.sev_mode import SevMode
ld = guest.calc_launch_digest(SevMode.SEV_SNP, vcpus_num, vcpu_types.CPU_SIGS["EPYC-v4"],
ovmf_path, kernel_path, initrd_path, cmdline_str, guest_features)
print("Calculated measurement:", ld.hex())
block = id_block.snp_calc_id_block(ld,"id_key_file","author_key_file")
print("Calculated id block in base64", block)For SEV-ES and SEV-SNP, the initial CPU state (VMSA) includes the guest CPU signature in the edx register when you use the QEMU vmm. Therefore, starting the VM with a different type of guest CPU will modify the content of the VMSA, and therefore modify the calculated measurement.
You can choose the guest CPU type using --vcpu-type, or --vcpu-sig, or a
combination of --vcpu-family, --vcpu-model, and --vcpu-stepping. For
example, the following 3 invocations are identical:
sev-snp-measure --vcpu-type=EPYC-v4 ...sev-snp-measure --vcpu-sig=0x800f12 ...sev-snp-measure --vcpu-family=23 --vcpu-model=1 --vcpu-stepping=2 ...
Prior to Linux Kernel version 6.6, the default value was always calculated to 0x1, as the kernel only supported SNPActive. After the release of Linux Kernel 6.6, additional features were made available some of them enabled by default. Because of this, the new default value is 0x21 which is SNPActive + DebugSwap. Other possible combinations my be derived by generating a 64-bit hex value from the following chart:
| BIT FIELD | Description |
|---|---|
| 0 | SNPActive |
| 1 | vTOM |
| 2 | ReflectVC |
| 3 | RestrictedInjection |
| 4 | AlternateInjection |
| 5 | DebugSwap |
| 6 | PreventHostIBS |
| 7 | BTBIsolation |
| 8 | VmplSSS |
| 9 | SecureTSC |
| 10 | VmgexitParameter |
| 11 | Reserved, SBZ |
| 12 | IbsVirtualization |
| 13 | Reserved, SBZ |
| 14 | VmsaRegProt |
| 15 | SmtProtection |
| 63:16 | Reserved, SBZ |
The SEV-SNP digest gets generated in multiple steps that each have a digest as output. With that digest output, you can stop at any of these steps and continue generation of the full digest later. These are the steps:
- OVMF
- (optional) -kernel, -initrd, -append arguments
- Initial state of all vCPUs
In situations where only minor OVMF changes happen, you may not want to copy the full OVMF binary to the validation system. In these situations, you can cut digest calculation after the OVMF step and use its hash instead of the full binary.
To generate a hash, use the --mode snp:ovmf-hash parameter:
$ sev-snp-measure --mode snp:ovmf-hash --ovmf OVMF.fd
cab7e085874b3acfdbe2d96dcaa3125111f00c35c6fc9708464c2ae74bfdb048a198cb9a9ccae0b3e5e1a33f5f249819
On a different machine that only has access to an older but compatible OVMF binary, you can then ingest the hash again to generate a full measurement:
$ sev-snp-measure --mode snp --vcpus=1 --vcpu-type=EPYC-v4 --ovmf=OVMF.fd.old --ovmf-hash cab7e[...]
d52697c3e056fb8d698d19cc29adfbed5a8ec9170cb9eb63c2ac957d22b4eb647e25780162036d063a0cf418b8830acc
- sev-snp-measure-go: port of sev-snp-measure to Go
- libvirt tools: virt-dom-sev-validate, virt-dom-sev-vmsa-tool
- sev Rust crate and snpguest CLI tool
- snp-digest-rs
- AMD sev-tool, sev-guest, and sev-utils
- go-sev-guest
Run all unit tests:
pip install -r requirements.txt
make test
Check unit tests coverage:
pip install coverage
make coverage
# See HTML coverage report in htmlcov/
Check Python type hints:
pip install mypy
make typecheck
Check Python coding style:
pip install flake8
make lint
If you have any questions or issues you can create a new issue here
Pull requests are welcome!
Apache 2.0 license.