Hi Aidan, > This is v3 of the wolfTPM TPM 2.0 stack integration for U-Boot. > > wolfTPM (https://github.com/wolfSSL/wolfTPM) is a portable, GPLv2 TPM 2.0
I go to the github link and it says GPL-3.0 license so which is it? IANAL but I'm not sure if the GPLv3 is compatible with GPLv2 in the context of U-Boot because it would make the whole output a GPLv3 application > library that provides a full TPM 2.0 command set, an SPI/MMIO HAL, and > firmware-update support for Infineon SLB9672/SLB9673 hardware. This > series wires it into U-Boot as an optional backend behind the existing > 'tpm2' command, alongside support for QEMU+swtpm, sandbox emulation, and > real Raspberry Pi 4 + Infineon SLB9672 hardware. So that provides the technical detail, but what value does it provide to U-Boot? We already have a relatively complete TPM stack, do people do firmware updates in the early boot process, why else would someone choose to use this over the existing implementation? One part of what I want to know in a cover letter is what value this provides to the project, what are the pros/cons etc, like what maintenance looks like. Peter > Branch (full 14-commit history including the subtree squash + merge): > https://github.com/aidangarske/u-boot wolftpm-v2-patches > > Note on the subtree commits omitted from this email thread: > The branch above contains 14 commits; the email series is 12 patches. > Two commits are deliberately not sent to the list: > > * "Squashed 'lib/wolftpm/' content from commit 664db130d57" > - the parentless squash commit produced by `git subtree add`. > Its diff is ~3.4MB / ~90k lines and would be rejected by the > mailing list on size. > * "Merge commit 'd42fd7b146...' as 'lib/wolftpm'" > - the corresponding subtree merge commit. Merges have no patch > form and are routinely omitted by `git format-patch`. > > Please pull from the branch above (or wolfssl/wolfTPM @ 664db130d57) > to inspect the imported wolfTPM source. Subsequent updates will go > via tools/update-subtree.sh, matching how mbedTLS, lwIP, and > dts/upstream are maintained in tree. > > Changes since v2: > - Replaced the lib/wolftpm git submodule with a git subtree import > (squash + merge), matching the convention used for mbedTLS, lwIP, > and dts/upstream. tools/update-subtree.sh is updated to know > about the wolftpm subtree (path lib/wolftpm, upstream > https://github.com/wolfssl/wolfTPM.git). > [feedback: Ilias Apalodimas] > - Reverted the changes to include/linux/byteorder/generic.h. The > redefinition workaround for cpu_to_beXX / beXX_to_cpu now lives > on the wolfTPM side: include/configs/user_settings.h pulls in > <asm/byteorder.h> up front so U-Boot's macros are defined before > wolfTPM's #ifndef-guarded fallbacks in tpm2_packet.h. > [feedback: Ilias Apalodimas] > > Testing: > - QEMU arm64 + swtpm Python test framework > (./test/py/test.py --bd qemu_arm64 -k "test_wolftpm and not ut_cmd"): > 19 passed, 2 skipped (matching doc/usage/cmd/wolftpm.rst). > - Manual QEMU arm64 + swtpm walkthrough per > doc/usage/cmd/wolftpm.rst section "Building and Running wolfTPM > with U-Boot using QEMU": tpm2 help/info/autostart/startup/ > get_capability/pcr_read/pcr_print/caps all return expected output. > - Raspberry Pi 4 + Infineon SLB9672 (real hardware): all wolfTPM > cmd tests pass, including firmware update path. > > v2 thread: > https://lore.kernel.org/u-boot/?q=PATCH+v2+tpm+wolfTPM > > Aidan Garske (12): > tpm: export tpm_show_device, tpm_set_device, and get_tpm > include/hash: add SHA384 hash wrapper declaration for wolfTPM > spi: add BCM2835/BCM2711 hardware SPI controller driver > dts: add TPM device tree nodes for RPi4, QEMU, and sandbox > tpm: add wolfTPM build rules and Kconfig > tpm: add wolfTPM headers and SHA384 glue code > tpm: add wolfTPM driver helpers and Kconfig options > cmd: refactor tpm2 command into frontend/backend architecture > tpm: add sandbox TPM SPI emulator > test: add wolfTPM C unit tests and Python integration tests > doc: add wolfTPM documentation > configs: enable wolfTPM in rpi_4_defconfig > > -- > 2.47.3 >
