linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Bjorn Andersson <bjorn.andersson@linaro.org>
To: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: ohad@wizery.com, mcoquelin.stm32@gmail.com,
	alexandre.torgue@st.com, loic.pallardy@st.com,
	arnaud.pouliquen@st.com, linux-remoteproc@vger.kernel.org,
	linux-stm32@st-md-mailman.stormreply.com,
	linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 10/12] remoteproc: stm32: Introduce new parse fw ops for synchronisation
Date: Wed, 13 May 2020 22:13:36 -0700	[thread overview]
Message-ID: <20200514051336.GA396285@builder.lan> (raw)
In-Reply-To: <20200424202505.29562-11-mathieu.poirier@linaro.org>

On Fri 24 Apr 13:25 PDT 2020, Mathieu Poirier wrote:

> Introduce new parse firmware rproc_ops functions to be used when
> synchonising with an MCU.
> 
> Mainly based on the work published by Arnaud Pouliquen [1].
> 
> [1]. https://patchwork.kernel.org/project/linux-remoteproc/list/?series=239877
> 
> Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
> Reviewed-by: Loic Pallardy <loic.pallardy@st.com>
> ---
>  drivers/remoteproc/stm32_rproc.c | 51 +++++++++++++++++++++++++++++++-
>  1 file changed, 50 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/remoteproc/stm32_rproc.c b/drivers/remoteproc/stm32_rproc.c
> index 86d23c35d805..b8ae8aed5585 100644
> --- a/drivers/remoteproc/stm32_rproc.c
> +++ b/drivers/remoteproc/stm32_rproc.c
> @@ -215,7 +215,34 @@ static int stm32_rproc_elf_load_rsc_table(struct rproc *rproc,
>  	return 0;
>  }
>  
> -static int stm32_rproc_parse_fw(struct rproc *rproc, const struct firmware *fw)
> +static int stm32_rproc_sync_elf_load_rsc_table(struct rproc *rproc,
> +					       const struct firmware *fw)
> +{
> +	struct resource_table *table = NULL;
> +	struct stm32_rproc *ddata = rproc->priv;
> +
> +	if (ddata->rsc_va) {

Does it really make sense to try to sync with a remote that doesn't have
a resource table?

> +		table = (struct resource_table *)ddata->rsc_va;
> +		/* Assuming that the resource table fits in 1kB is fair */
> +		rproc->cached_table = kmemdup(table, RSC_TBL_SIZE, GFP_KERNEL);

It's unfortunate that we need to create a clone of the resource table
that we found in ram, and then return the original memory when the core
ask for the loaded table...

I wonder if we somehow can avoid this in the core (i.e. skip overwriting
table_ptr with the cached_table during stop)

> +		if (!rproc->cached_table)
> +			return -ENOMEM;
> +
> +		rproc->table_ptr = rproc->cached_table;
> +		rproc->table_sz = RSC_TBL_SIZE;
> +		return 0;
> +	}
> +
> +	rproc->cached_table = NULL;
> +	rproc->table_ptr = NULL;
> +	rproc->table_sz = 0;
> +
> +	dev_warn(&rproc->dev, "no resource table found for this firmware\n");
> +	return 0;
> +}
> +
> +static int stm32_rproc_parse_memory_regions(struct rproc *rproc,
> +					    const struct firmware *fw)
>  {
>  	struct device *dev = rproc->dev.parent;
>  	struct device_node *np = dev->of_node;
> @@ -268,9 +295,30 @@ static int stm32_rproc_parse_fw(struct rproc *rproc, const struct firmware *fw)
>  		index++;
>  	}
>  
> +	return 0;
> +}
> +
> +static int stm32_rproc_parse_fw(struct rproc *rproc, const struct firmware *fw)
> +{
> +	int ret = stm32_rproc_parse_memory_regions(rproc, fw);
> +
> +	if (ret)
> +		return ret;
> +
>  	return stm32_rproc_elf_load_rsc_table(rproc, fw);
>  }
>  
> +static int stm32_rproc_sync_parse_fw(struct rproc *rproc,
> +				     const struct firmware *fw)

Rather than having a function parse_fw that is passed no fw and return
some state that was setup in probe, how about just do this during probe?

Regards,
Bjorn

> +{
> +	int ret = stm32_rproc_parse_memory_regions(rproc, fw);
> +
> +	if (ret)
> +		return ret;
> +
> +	return stm32_rproc_sync_elf_load_rsc_table(rproc, fw);
> +}
> +
>  static irqreturn_t stm32_rproc_wdg(int irq, void *data)
>  {
>  	struct platform_device *pdev = data;
> @@ -544,6 +592,7 @@ static struct rproc_ops st_rproc_ops = {
>  static __maybe_unused struct rproc_ops st_rproc_sync_ops = {
>  	.start		= stm32_rproc_sync_start,
>  	.stop		= stm32_rproc_stop,
> +	.parse_fw       = stm32_rproc_sync_parse_fw,
>  };
>  
>  static const struct of_device_id stm32_rproc_match[] = {
> -- 
> 2.20.1
> 

  reply	other threads:[~2020-05-14  5:15 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-24 20:24 [PATCH v2 00/12] remoteproc: stm32: Add support for synchronising with M4 Mathieu Poirier
2020-04-24 20:24 ` [PATCH v2 01/12] remoteproc: stm32: Decouple rproc from memory translation Mathieu Poirier
2020-05-14  4:57   ` Bjorn Andersson
2020-04-24 20:24 ` [PATCH v2 02/12] remoteproc: stm32: Request IRQ with platform device Mathieu Poirier
2020-05-14  4:57   ` Bjorn Andersson
2020-04-24 20:24 ` [PATCH v2 03/12] remoteproc: stm32: Decouple rproc from DT parsing Mathieu Poirier
2020-04-29 13:37   ` Arnaud POULIQUEN
2020-04-30 20:58     ` Mathieu Poirier
2020-05-14  4:59   ` Bjorn Andersson
2020-04-24 20:24 ` [PATCH v2 04/12] remoteproc: stm32: Remove memory translation " Mathieu Poirier
2020-05-14  5:03   ` Bjorn Andersson
2020-04-24 20:24 ` [PATCH v2 05/12] remoteproc: stm32: Parse syscon that will manage M4 synchronisation Mathieu Poirier
2020-05-14  5:03   ` Bjorn Andersson
2020-04-24 20:24 ` [PATCH v2 06/12] remoteproc: stm32: Get coprocessor state Mathieu Poirier
2020-04-29 13:38   ` Arnaud POULIQUEN
2020-05-01 17:40     ` Mathieu Poirier
2020-04-24 20:25 ` [PATCH v2 07/12] remoteproc: stm32: Get loaded resource table for synchronisation Mathieu Poirier
2020-04-24 20:25 ` [PATCH v2 08/12] remoteproc: stm32: Introduce new start ops " Mathieu Poirier
2020-04-29 14:50   ` Arnaud POULIQUEN
2020-04-24 20:25 ` [PATCH v2 09/12] remoteproc: stm32: Update M4 state in stm32_rproc_stop() Mathieu Poirier
2020-04-29 14:52   ` Arnaud POULIQUEN
2020-04-24 20:25 ` [PATCH v2 10/12] remoteproc: stm32: Introduce new parse fw ops for synchronisation Mathieu Poirier
2020-05-14  5:13   ` Bjorn Andersson [this message]
2020-04-24 20:25 ` [PATCH v2 11/12] remoteproc: stm32: Introduce new loaded rsc " Mathieu Poirier
2020-05-14  5:15   ` Bjorn Andersson
2020-04-24 20:25 ` [PATCH v2 12/12] remoteproc: stm32: Set synchronisation state machine if needed Mathieu Poirier
2020-04-29 14:47   ` Arnaud POULIQUEN
2020-05-01 17:54     ` Mathieu Poirier
2020-04-29 15:08 ` [PATCH v2 00/12] remoteproc: stm32: Add support for synchronising with M4 Arnaud POULIQUEN
2020-05-01 17:59   ` Mathieu Poirier

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=20200514051336.GA396285@builder.lan \
    --to=bjorn.andersson@linaro.org \
    --cc=alexandre.torgue@st.com \
    --cc=arnaud.pouliquen@st.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-remoteproc@vger.kernel.org \
    --cc=linux-stm32@st-md-mailman.stormreply.com \
    --cc=loic.pallardy@st.com \
    --cc=mathieu.poirier@linaro.org \
    --cc=mcoquelin.stm32@gmail.com \
    --cc=ohad@wizery.com \
    /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).