On 24/08/26 07:47:33, Jan Kiszka wrote: > On 23.07.26 16:38, Jorge Ramirez-Ortiz via U-Boot wrote: > > OP-TEE secure storage (CFG_RPMB_FS) relies on an RPMB partition, but > > U-Boot's OP-TEE RPMB supplicant only speaks the legacy single-command > > interface, which is bound to eMMC. SoCs that are UFS-only and have no > > eMMC (for example the Qualcomm SA8775P) therefore cannot back OP-TEE > > secure storage from U-Boot today. This series adds that support. > > > > It introduces the transport-agnostic OP-TEE RPMB "subsystem" interface > > (PROBE_RESET / PROBE_NEXT / FRAMES), where the normal world enumerates > > the RPMB device and reports its kind, size and CID, then carries the > > signed frames. The legacy eMMC supplicant is preserved unchanged, only > > renamed to rpmb_emmc.c, with the new UFS backend added as a separate > > rpmb_ufs.c; the two are mutually exclusive via Kconfig (SUPPORT_UFS_RPMB > > depends on !SUPPORT_EMMC_RPMB) because the OP-TEE supplicant handles a > > single RPMB transport. The subsystem interface is UFS-only for now; eMMC > > can be migrated onto it later as the legacy path is retired. > > > > Just stumbled over this. Not sure if you saw [1] already. Our series > collide, and while yours lacks RPMB TA enumeration (to enable fTPM > e.g.), mine is based on the the in-tree emmc-only core. How to resolve > that best? Establish a common optee/rpmb.c for now so that the service > enumeration can be shared by both?
um, not sure what happened but my previous reply seems to have gone to some black hole (some mutt issue, might resurface later) No I didnt notice - I wasnt really tracking thought, just hoping that my series would be merged anytime now (I posted v4 last week). But yes I agree, a shared rpmb.c for enumeration seems like the right way to go. > > > On top of that it adds a UFS RPMB transport that moves JEDEC RPMB frames > > to and from the RPMB Well-Known LUN using SCSI SECURITY PROTOCOL IN/OUT. > > The per-region 16-byte CID is derived by BLAKE2b-hashing the exact > > device-id string the Linux kernel builds (ufshcd_create_device_id() > > plus a "-R<region>" suffix), so OP-TEE derives an RPMB key that matches > > the one Linux would use. > > > > The first patch is a standalone UFS descriptor fix the RPMB path depends > > on (UTF-16BE string decoding); the transport patches also include a > > power-on UNIT ATTENTION retry and a DMA-alignment bounce for the RPMB > > WLUN. > > > > Note: reading UFS descriptors reliably also requires the descriptor > > data-segment cache-invalidation fix, which has already been posted and > > merged separately, so this series is based on top of it. > > > > Tested on the Qualcomm IQ-9075-EVK (SA8775P): OP-TEE with CFG_RPMB_FS > > programs the RPMB key through U-Boot and reads/writes secure-storage > > objects, with the derived CID matching the Linux UFS device_id ABI. > > Looks like the QEMU model for UFS is still lacking RPMB support. Did you > happen to have a look if the eMMC bits could be reused for that? We are > doing most of the security stack integration tests via QEMU now, > including OTP key deployments - you would need a lot of new hardware > otherwise. I didn't look beyond confirming the support is still missing. I needed to do an end-to-end validation on real hardware before proposing the change anyway, and having done this kind of integration across several platforms — AMD, NXP, ST, now QC — I had a reasonable degree of confidence about how to derisk it...plus the thrill of bricking boards. But joke aside having the QEMU bits in place would be _really_ useful. if you want to merge your changes before mine I could rework my series on top. I am fine either way.
