From: Theo Buehler <tb@theobuehler.org>
Subject: Re: openssl/1.1: missing endbr64 for padlock engine
To: Mark Kettenis <mark.kettenis@xs4all.nl>
Cc: ports@openbsd.org, kettenis@openbsd.org
Date: Sat, 24 Feb 2024 17:47:42 +0100

On Sat, Feb 24, 2024 at 05:39:23PM +0100, Mark Kettenis wrote:
> > Date: Sat, 24 Feb 2024 17:33:10 +0100
> > From: Theo Buehler <tb@theobuehler.org>
> > 
> > This adds missing endbr64 to the padlock engine in openssl/1.1:
> > 
> > ld: warning: padlock_capability: missing endbr64
> > ld: warning: padlock_key_bswap: missing endbr64
> > ld: warning: padlock_verify_context: missing endbr64
> > ld: warning: padlock_reload_key: missing endbr64
> > ld: warning: padlock_aes_block: missing endbr64
> > ld: warning: padlock_xstore: missing endbr64
> > ld: warning: padlock_sha1_oneshot: missing endbr64
> > ld: warning: padlock_sha1_blocks: missing endbr64
> > ld: warning: padlock_sha256_oneshot: missing endbr64
> > ld: warning: padlock_sha256_blocks: missing endbr64
> > ld: warning: padlock_sha512_blocks: missing endbr64
> > ld: warning: padlock_ecb_encrypt: missing endbr64
> > ld: warning: padlock_cbc_encrypt: missing endbr64
> > ld: warning: padlock_cfb_encrypt: missing endbr64
> > ld: warning: padlock_ofb_encrypt: missing endbr64
> > ld: warning: padlock_ctr32_encrypt: missing endbr64
> 
> Not opposed to fixing these, but I doubt we'll ever see a CPU that has
> both the VIA PadLock instructions and IBT ;).
> 
> (Sorry, didn't realize this until after I sent out that diff)

Me neither until deraadt pointed it out off-list.

> ok kettenis@ (but also fine if you decide to drop this)

Let's not do it then.

> Is this code still there in openssl 3.x?

Yes, it is there and it does not have endbr64 annotations unlike most
of the other code.