On Mon, 11 Jan 2021 at 22:31, Ard Biesheuvel wrote:
>
> On Mon, 11 Jan 2021 at 22:05, Eric Biggers wrote:
> >
> > On Mon, Jan 11, 2021 at 05:52:30PM +0100, Ard Biesheuvel wrote:
> > > CRC-T10DIF is a very poor match for the crypto API:
> > > - every user in the kernel calls it via a library wrapp
On Mon, 11 Jan 2021 at 22:05, Eric Biggers wrote:
>
> On Mon, Jan 11, 2021 at 05:52:30PM +0100, Ard Biesheuvel wrote:
> > CRC-T10DIF is a very poor match for the crypto API:
> > - every user in the kernel calls it via a library wrapper around the
> > shash API, so all callers share a single inst
On Mon, Jan 11, 2021 at 05:52:30PM +0100, Ard Biesheuvel wrote:
> CRC-T10DIF is a very poor match for the crypto API:
> - every user in the kernel calls it via a library wrapper around the
> shash API, so all callers share a single instance of the transform
> - each architecture provides at most
On Mon, 11 Jan 2021 at 21:56, Peter Zijlstra wrote:
>
> On Mon, Jan 11, 2021 at 07:36:20PM +0100, Ard Biesheuvel wrote:
> > On Mon, 11 Jan 2021 at 18:27, Ard Biesheuvel wrote:
> > > On Mon, 11 Jan 2021 at 17:52, Ard Biesheuvel wrote:
>
> > > > Special request to Peter to take a look at patch #2,
On Mon, Jan 11, 2021 at 07:36:20PM +0100, Ard Biesheuvel wrote:
> On Mon, 11 Jan 2021 at 18:27, Ard Biesheuvel wrote:
> > On Mon, 11 Jan 2021 at 17:52, Ard Biesheuvel wrote:
> > > Special request to Peter to take a look at patch #2, and in particular,
> > > whether synchronize_rcu_tasks() is suf
On Mon, 11 Jan 2021 at 18:27, Ard Biesheuvel wrote:
>
> On Mon, 11 Jan 2021 at 17:52, Ard Biesheuvel wrote:
> >
> > CRC-T10DIF is a very poor match for the crypto API:
> > - every user in the kernel calls it via a library wrapper around the
> > shash API, so all callers share a single instance
On Mon, 11 Jan 2021 at 17:52, Ard Biesheuvel wrote:
>
> CRC-T10DIF is a very poor match for the crypto API:
> - every user in the kernel calls it via a library wrapper around the
> shash API, so all callers share a single instance of the transform
> - each architecture provides at most a single
CRC-T10DIF is a very poor match for the crypto API:
- every user in the kernel calls it via a library wrapper around the
shash API, so all callers share a single instance of the transform
- each architecture provides at most a single optimized implementation,
based on SIMD instructions for carr