On Thu, Jan 02, 2025 at 10:25:15AM -0500, Raymond Mao wrote:
[snip]
As I said, we need an kconfig here to decide whether a user should look for TPM log (and all other handoff information defined by the Firmware Handoff specification) from the bloblist or not. We don't have such kconfig now.
We do we need this knob? I don't think that we do. The case of bloblist not existing where we looked for it needs to work. And the case of the bloblist not having an entry needs to work (or if it *must* exist, that's a separate option to add, ie CONFIG_TPM_BLOBLIST_LOG_REQUIRED).
There's at least two sets of challenges here. One, being solved by vexpress64 right now, is that we didn't have CONFIG_BLOBLIST_PASSAGE as an actual option. And in that case, there's no U-Boot before full U-Boot and the bloblist exists for us. Two, U-Boot is what is creating the bloblist. The contentious parts are *when* it's created and *where* it resides prior to full U-Boot seeing it.
There isn't contention, so far as I am aware. The normal case is that U-Boot creates and uses the bloblist itself. Pre-U-Boot blobs (like TF-A, sadly) are not the normal case and should be discouraged in an open-source project. That doesn't mean we shouldn't support them, but it is the tail wagging the dog.
TBH, I am confused with this statement which means we should not use the bloblist library from the beginning to hand over data from the previous stage. If U-Boot bloblist only intends to consume the data created by itself, we have to introduce another library to do the handoff, like what was done in TF-A and OP-TEE - then finally we can have a standard handoff library that can be used in all projects and keep bloblist as it was.
Simon doesn't speak for the U-Boot project, he speaks for himself. I don't see why U-Boot shouldn't look for an use a bloblist for things that need to be passed from one stage of the boot process to another.