On 30/06/2020 10:59, Herbert Xu wrote:
On Tue, Jun 30, 2020 at 10:58:28AM +0300, Tero Kristo wrote:
Openssl uses init->update->final sequencing, and it would be nice to have
this use crypto accelerator. Otherwise we basically support cipher
acceleration from openssl, but not hashing.
You're r
On Tue, Jun 30, 2020 at 10:58:28AM +0300, Tero Kristo wrote:
>
> Openssl uses init->update->final sequencing, and it would be nice to have
> this use crypto accelerator. Otherwise we basically support cipher
> acceleration from openssl, but not hashing.
You're referring to algif_hash, right? If so
On 30/06/2020 10:54, Herbert Xu wrote:
On Tue, Jun 30, 2020 at 10:52:57AM +0300, Tero Kristo wrote:
How about making the buffer size configurable during probe time via module
param, and/or runtime via sysfs parameter? This would allow system designers
to tune it based on expected usecases.
Wh
On Tue, Jun 30, 2020 at 10:52:57AM +0300, Tero Kristo wrote:
>
> How about making the buffer size configurable during probe time via module
> param, and/or runtime via sysfs parameter? This would allow system designers
> to tune it based on expected usecases.
Why do you even need to support this?
On 30/06/2020 10:46, Herbert Xu wrote:
On Tue, Jun 30, 2020 at 10:20:06AM +0300, Tero Kristo wrote:
Only up-to block size? This would limit the buffer to 64-128 bytes.
The exported hash state is not supposed to be used to support
hardware that does not support partial hashing.
These states c
On Tue, Jun 30, 2020 at 10:20:06AM +0300, Tero Kristo wrote:
>
> Only up-to block size? This would limit the buffer to 64-128 bytes.
The exported hash state is not supposed to be used to support
hardware that does not support partial hashing.
These states could potentially be placed on the stack
On 30/06/2020 07:49, Herbert Xu wrote:
On Fri, Jun 26, 2020 at 12:15:42PM +0300, Tero Kristo wrote:
I have been experimenting with an alternate approach, where I have a small
buffer within the context, this would be more like the way other drivers do
this. If the buffer is closed before running
On Fri, Jun 26, 2020 at 12:15:42PM +0300, Tero Kristo wrote:
>
> I have been experimenting with an alternate approach, where I have a small
> buffer within the context, this would be more like the way other drivers do
> this. If the buffer is closed before running out of space, I can push this
> to
On 26/06/2020 07:31, Herbert Xu wrote:
On Mon, Jun 15, 2020 at 10:14:48AM +0300, Tero Kristo wrote:
+static int sa_sha_update(struct ahash_request *req)
+{
+ struct sa_sha_req_ctx *rctx = ahash_request_ctx(req);
+ struct scatterlist *sg;
+ void *buf;
+ int pages;
+
On Mon, Jun 15, 2020 at 10:14:48AM +0300, Tero Kristo wrote:
>
> +static int sa_sha_update(struct ahash_request *req)
> +{
> + struct sa_sha_req_ctx *rctx = ahash_request_ctx(req);
> + struct scatterlist *sg;
> + void *buf;
> + int pages;
> + struct page *pg;
> +
> + if (!re
10 matches
Mail list logo
