Hacker News Reader

Comment by candiddevmike | Hacker News Reader

As someone with a passing interest in using TPM for crypto things, everytime I think deeply about the implementation details like this, I come back to needing some kind of recovery password/backup key setup that entirely negates the point of the TPM in the first place. They seem really neat, but I struggle to see the benefit they have for doing crypto when a tiny slip up means your users' keys are poof, gone. And the tiny slip up may not even be with your software, but some edge case in the OS/TPM stack.

View on HN

The TPM was never designed to be the only holder of a key that cannot be reset. The idea was that it prevents you from typing in a password or reseting an attestation signature in a database for 99% of boots, but if certain things in the boot process change (as determined by the firmware, the CPU, the OS, and the application using the TPM) it’s designed to lock you out so those things cannot change without anyone’s notice.

For that purpose they’re pretty good, though there are advantages to a more signature-oriented boot security option like Apple’s Secure Enclave. But that only works so well because Apple simply doesn’t permit altering parts of the macOS boot process. For Windows/Linux, you have a variety of hardware, firmware, and OS vendors all in the mix and agreeing on escrow of keys for all of them is hard.

Whether by design or accident, this is correct.

You backup a key or key creation mechanism or whatever elsewhere somewhere very safe.

Then almost never touch it, as the TPM authenticates.

The presumption is that the contents being secured are /so/ valuable that locking my device is preferable to any leak of them whatsoever.

This is military level security and just isn't appropriate for most consumers. Particularly around something so rarely exercised and utilized by users as the boot process. A simple warning with a long timeout would have sufficed.

Aside from that you have a hardware vendor, sourced into an integrated product from another vendor, sold to a user, with various third party software interacting with it. This was always going to result in questionable experiences for end users.

A warning doesn’t help at all. The main threat model for FDE is that someone steals your device and dumps the disk. If you don’t protect the boot process somehow, then you’re just storing the encryption key next to the data.

If you don’t care about that (which is not “military level security”, laptop thieves stealing creds is a thing), just don’t use FDE or use it with an on-boot password every time. No point in the theater.

> laptop thieves stealing creds is a thing

Two factor is a thing. FDE is such a 1990s idea.

Wow. That’s a new one. Where exactly do you think the authentication tokens you obtain using 2FA are stored?

The primary argument in favor of TPM's is the desire to assert against tampering to the boot system, and as a secondary effect it can be one of the solutions to reduce the need for users to type in passwords.

You can still use crypto without a TPM, including with full disk encryption, and for LUKS specifically you can use multiple passwords and mechanisms to unlock the system. Different solutions will give different benefits and drawbacks. Me and a friend wrote a remote password provider for Debian called Mandos which uses machines on the local network as a way for unattended boots. It does not address the issue of tampering with the bios/boot loader, but for the primary purpose of protecting against someone stealing the server/disks it serves the purpose of allowing us to use encrypted disk without drawbacks of typing in passwords, and the backup server, itself with encrypted disks, handles the risk of needing recovery passwords. At most one needs to have an additional backup key installed for the backup server.

TPM keys are great for things like SSH keys or Passkeys, which surprisingly works well even in Windows.

The private key is safe from any exfiltration, and usage only requires a short PIN instead of a long passphrase. The TPM ensures you're physically typing that PIN at the machine not a remote desktop window or other redirection that could be hacked.

Obviously, this is problematic/annoying for scripts and things that can't share the SSH session, because you need to PIN with every authentication. Also, for encryption, you want to use something where you can backup the private key before stashing it in the TPM. Windows allows you to do this with certificates that are exported for backup prior to encrypting the private key with an unexportable TPM key in Hello.

An easy solution to having to put your PIN in too often for SSH is to use the `ControlPersist` option in your SSH client config. This lets you only create a new SSH connection every 30s (or whatever you put), even if you’re doing separate operations. With a low timeout, there’s no realistic security risk (what’s the chance an attacker will only have control of your machine for 30s?).

I do this for GitHub in particular, because of tools that connect to the remote multiple times. Works with anything that uses the actual ssh executable under the hood.

Same with passkeys actually.

Passkeys get synced between your devices so they aren't any more fragile than passwords in a password manager.

Passkeys _may_ be synced, but that isn't guaranteed. For example a "device bound passkey" isn't synced.

There is a project under way to specify how to "sync" device-bound keys between authenticators: https://fidoalliance.org/specs/cx/cxp-v1.0-wd-20241003.html

Ideally this should have been hashed out before deploying passkeys everywhere, but I guess you can always register multiple passkeys for the sites that allow you to.

Iirc the original idea was that passkeys should be device specific. Of course that's impractical so now they're morphing to a long password that a human can't process.

In a few years someone will post "how about a long human retainable passphrase?" as a new and improved discovery.

They are still different to a password in that the service you are logging in to never gets the private key. So in the case the database gets compromised, if the service provider ensures no edits were made / restores a backup, there is no need to change your passkey since it was never exposed.

That was possible before passkeys.

https://www.rfc-editor.org/rfc/rfc9807.html

https://en.wikipedia.org/wiki/Password-authenticated_key_agr...

The big providers only want themselves to be able to backup passkeys. I do not want to handover my secrets to Apple/Microsoft/Google.

Apple Keychain syncing is end-to-end encrypted, Apple cannot see the contents of your synced keychain.

The benefit is that you don't enter the recovery password most of the time.

And when you do it should be rare and lead to a password reset.

Reply on news.ycombinator.com