Re: [PATCH v4 6/8] fsverity: improve performance by using multibuffer hashing
From: Eric Biggers <ebiggers@kernel.org>
Date: 2024-06-06 05:28:03
Also in:
dm-devel, linux-crypto, netdev
On Thu, Jun 06, 2024 at 10:00:05AM +0800, Herbert Xu wrote:
On Wed, Jun 05, 2024 at 12:14:10PM -0700, Eric Biggers wrote:quoted
This would at most apply to AH, not to ESP. Is AH commonly used these days?No AH is completely useless. However, this applies perfectly to ESP, in conjunction with authenc. Obviously we would need to add request linking to authenc (AEAD) as well so that it can pass it along to sha. BTW, does any of this interleaving apply to AES? If so we should explore adding request linking to skcipher as well.
With AES, interleaving would only help with non-parallelizable modes such as CBC encryption. Anyone who cares about IPsec performance should of course be using AES-GCM, which is parallelizable. Especially since my other patch https://lore.kernel.org/linux-crypto/20240602222221.176625-2-ebiggers@kernel.org/ (local) is making AES-GCM twice as fast... With hashing we unfortunately don't have the luxury of there being widely used and accepted parallelizable algorithms. In particular, all the SHAs are serialized. So that's why interleaving makes sense there. In any case, it seems that what you're asking for at this point is far beyond the scope of this patchset. - Eric _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel