From: Iuliana Prodan <iuliana.prodan@nxp.com>
To: Corentin Labbe <clabbe.montjoie@gmail.com>,
"alexandre.torgue@st.com" <alexandre.torgue@st.com>,
"davem@davemloft.net" <davem@davemloft.net>,
"herbert@gondor.apana.org.au" <herbert@gondor.apana.org.au>,
"mcoquelin.stm32@gmail.com" <mcoquelin.stm32@gmail.com>,
"mripard@kernel.org" <mripard@kernel.org>,
"wens@csie.org" <wens@csie.org>,
Horia Geanta <horia.geanta@nxp.com>,
Aymen Sghaier <aymen.sghaier@nxp.com>
Cc: "linux-arm-kernel@lists.infradead.org"
<linux-arm-kernel@lists.infradead.org>,
"linux-crypto@vger.kernel.org" <linux-crypto@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"linux-stm32@st-md-mailman.stormreply.com"
<linux-stm32@st-md-mailman.stormreply.com>,
"linux-sunxi@googlegroups.com" <linux-sunxi@googlegroups.com>
Subject: Re: [PATCH RFC 09/10] crypto: engine: permit to batch requests
Date: Thu, 16 Jan 2020 11:34:34 +0000 [thread overview]
Message-ID: <VI1PR04MB4445182A8A827C3EA303A2978C360@VI1PR04MB4445.eurprd04.prod.outlook.com> (raw)
In-Reply-To: 20200114135936.32422-10-clabbe.montjoie@gmail.com
On 1/14/2020 4:00 PM, Corentin Labbe wrote:
> Now everything is ready, this patch permits to choose the number of
> request to batch.
>
> Signed-off-by: Corentin Labbe <clabbe.montjoie@gmail.com>
> ---
> crypto/crypto_engine.c | 32 +++++++++++++++++++++++++++-----
> include/crypto/engine.h | 2 ++
> 2 files changed, 29 insertions(+), 5 deletions(-)
>
> diff --git a/crypto/crypto_engine.c b/crypto/crypto_engine.c
> index e23a398ba330..e9cd9ec9a732 100644
> --- a/crypto/crypto_engine.c
> +++ b/crypto/crypto_engine.c
> @@ -114,6 +114,7 @@ static void crypto_pump_requests(struct crypto_engine *engine,
> }
>
> engine->ct = 0;
> +retry:
> /* Get the fist request from the engine queue to handle */
> backlog = crypto_get_backlog(&engine->queue);
> async_req = crypto_dequeue_request(&engine->queue);
> @@ -151,7 +152,10 @@ static void crypto_pump_requests(struct crypto_engine *engine,
> }
> engine->cur_reqs[engine->ct].prepared = true;
> }
> - engine->ct++;
> + if (++engine->ct < engine->rmax && engine->queue.qlen > 0) {
> + spin_lock_irqsave(&engine->queue_lock, flags);
> + goto retry;
> + }
> if (!enginectx->op.do_one_request) {
> dev_err(engine->dev, "failed to do request\n");
> ret = -EINVAL;
> @@ -393,15 +397,18 @@ int crypto_engine_stop(struct crypto_engine *engine)
> EXPORT_SYMBOL_GPL(crypto_engine_stop);
>
> /**
> - * crypto_engine_alloc_init - allocate crypto hardware engine structure and
> + * crypto_engine_alloc_init2 - allocate crypto hardware engine structure and
> * initialize it.
> * @dev: the device attached with one hardware engine
> * @rt: whether this queue is set to run as a realtime task
> + * @rmax: The number of request that the engine can batch in one
> + * @qlen: The size of the crypto queue
> *
> * This must be called from context that can sleep.
> * Return: the crypto engine structure on success, else NULL.
> */
> -struct crypto_engine *crypto_engine_alloc_init(struct device *dev, bool rt)
> +struct crypto_engine *crypto_engine_alloc_init2(struct device *dev, bool rt,
> + int rmax, int qlen)
I think one _alloc_init function is enough, which will set the size of
crypto-engine queue (not hardcoded like it is now) and the number of
requests that the engine can execute in parallel.
{
> struct sched_param param = { .sched_priority = MAX_RT_PRIO / 2 };
> struct crypto_engine *engine;
> @@ -421,12 +428,12 @@ struct crypto_engine *crypto_engine_alloc_init(struct device *dev, bool rt)
> engine->priv_data = dev;
> snprintf(engine->name, sizeof(engine->name),
> "%s-engine", dev_name(dev));
> - engine->rmax = 1;
> + engine->rmax = rmax;
> engine->cur_reqs = devm_kzalloc(dev, sizeof(struct cur_req) * engine->rmax, GFP_KERNEL);
> if (!engine->cur_reqs)
> return NULL;
>
> - crypto_init_queue(&engine->queue, CRYPTO_ENGINE_MAX_QLEN);
> + crypto_init_queue(&engine->queue, qlen);
> spin_lock_init(&engine->queue_lock);
>
> engine->kworker = kthread_create_worker(0, "%s", engine->name);
> @@ -443,6 +450,21 @@ struct crypto_engine *crypto_engine_alloc_init(struct device *dev, bool rt)
>
> return engine;
> }
> +EXPORT_SYMBOL_GPL(crypto_engine_alloc_init2);
> +
> +/**
> + * crypto_engine_alloc_init - allocate crypto hardware engine structure and
> + * initialize it.
> + * @dev: the device attached with one hardware engine
> + * @rt: whether this queue is set to run as a realtime task
> + *
> + * This must be called from context that can sleep.
> + * Return: the crypto engine structure on success, else NULL.
> + */
> +struct crypto_engine *crypto_engine_alloc_init(struct device *dev, bool rt)
> +{
> + return crypto_engine_alloc_init2(dev, rt, 1, CRYPTO_ENGINE_MAX_QLEN);
> +}
> EXPORT_SYMBOL_GPL(crypto_engine_alloc_init);
>
> /**
> diff --git a/include/crypto/engine.h b/include/crypto/engine.h
> index 55d3dbc2498c..fe0dfea8bf07 100644
> --- a/include/crypto/engine.h
> +++ b/include/crypto/engine.h
> @@ -115,6 +115,8 @@ void crypto_finalize_skcipher_request(struct crypto_engine *engine,
> int crypto_engine_start(struct crypto_engine *engine);
> int crypto_engine_stop(struct crypto_engine *engine);
> struct crypto_engine *crypto_engine_alloc_init(struct device *dev, bool rt);
> +struct crypto_engine *crypto_engine_alloc_init2(struct device *dev, bool rt,
> + int rmax, int qlen);
> int crypto_engine_exit(struct crypto_engine *engine);
>
> #endif /* _CRYPTO_ENGINE_H */
>
next prev parent reply other threads:[~2020-01-16 11:34 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-01-14 13:59 [PATCH RFC 00/10] crypto: engine: permit to batch requests Corentin Labbe
2020-01-14 13:59 ` [PATCH RFC 01/10] crypto: sun8i-ce: move iv data to request context Corentin Labbe
2020-01-14 13:59 ` [PATCH RFC 02/10] crypto: sun8i-ce: increase task list size Corentin Labbe
2020-01-14 13:59 ` [PATCH RFC 03/10] crypto: sun8i-ce: split into prepare/run/unprepare Corentin Labbe
2020-01-14 13:59 ` [PATCH RFC 04/10] crypto: sun8i-ce: introduce the slot number Corentin Labbe
2020-01-14 13:59 ` [PATCH RFC 05/10] crypto: engine: transform cur_req in an array Corentin Labbe
2020-01-16 11:34 ` Iuliana Prodan
2020-01-14 13:59 ` [PATCH RFC 06/10] crypto: engine: introduce ct Corentin Labbe
2020-01-16 11:34 ` Iuliana Prodan
2020-01-16 13:21 ` Corentin Labbe
2020-01-14 13:59 ` [PATCH RFC 07/10] crypto: sun8i-ce: handle slot > 0 Corentin Labbe
2020-01-14 13:59 ` [PATCH RFC 08/10] crypto: engine: add slot parameter Corentin Labbe
2020-01-14 13:59 ` [PATCH RFC 09/10] crypto: engine: permit to batch requests Corentin Labbe
2020-01-16 11:34 ` Iuliana Prodan [this message]
2020-01-17 16:13 ` Iuliana Prodan
2020-01-14 13:59 ` [PATCH RFC 10/10] crypto: sun8i-ce: use the new batch mechanism Corentin Labbe
2020-01-16 11:33 ` [PATCH RFC 00/10] crypto: engine: permit to batch requests Iuliana Prodan
2020-01-16 13:16 ` Corentin Labbe
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=VI1PR04MB4445182A8A827C3EA303A2978C360@VI1PR04MB4445.eurprd04.prod.outlook.com \
--to=iuliana.prodan@nxp.com \
--cc=alexandre.torgue@st.com \
--cc=aymen.sghaier@nxp.com \
--cc=clabbe.montjoie@gmail.com \
--cc=davem@davemloft.net \
--cc=herbert@gondor.apana.org.au \
--cc=horia.geanta@nxp.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-crypto@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-stm32@st-md-mailman.stormreply.com \
--cc=linux-sunxi@googlegroups.com \
--cc=mcoquelin.stm32@gmail.com \
--cc=mripard@kernel.org \
--cc=wens@csie.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).