Re: [PATCH v2 13/19] remoteproc: core: Append resource only if spare resource present

[Lee Jones]

On Wed, 31 Aug 2016, Loic Pallardy wrote:

> This patch renames rproc_add_resource_table_entry in __add_rsc_tbl_entry
> to have shorter function name and adds spare resource support.
> To guarantee remoteproc won't overwrite firmware data when copying
> back modified resource table, __add_rsc_tbl_entry verifies first that
> resource table owns a spare resource and uses spare bytes to create
> a new resource entry. Spare resource is updated according to changes.
> 
> Signed-off-by: Loic Pallardy <loic.pallardy@xxxxxx>
> ---
>  drivers/remoteproc/remoteproc_core.c | 54 ++++++++++++++++++++----------------
>  1 file changed, 30 insertions(+), 24 deletions(-)
> 
> diff --git a/drivers/remoteproc/remoteproc_core.c b/drivers/remoteproc/remoteproc_core.c
> index aff1a00..25a429b 100644
> --- a/drivers/remoteproc/remoteproc_core.c
> +++ b/drivers/remoteproc/remoteproc_core.c
> @@ -1107,39 +1107,34 @@ static int __update_rsc_tbl_entry(struct rproc *rproc,
>  	return !updated;
>  }
>  
> -static struct resource_table*
> -rproc_add_resource_table_entry(struct rproc *rproc,
> +static int __add_rsc_tbl_entry(struct rproc *rproc,

Once again, I prefer plain English over cryptic abbreviations.  Makes
things much more difficult for developers who are new to, or are just
dipping into RemoteProc code.

>  			       struct rproc_request_resource *request,
> -			       struct resource_table *old_table, int *tablesz)
> +			       struct resource_table *table, int tablesz)
>  {
> -	struct resource_table *table;
>  	struct fw_rsc_hdr h;
> +	struct fw_rsc_spare spare;
>  	void *new_rsc_loc;
>  	void *fw_header_loc;
>  	void *start_of_rscs;
>  	int new_rsc_offset;
> -	int size = *tablesz;
> -	int i;
> +	int new_spare_offset;
> +	int i, spare_index = 0;
>  
>  	h.type = request->type;
>  
> -	new_rsc_offset = size;
> +	/* check available spare size */

In keeping with the existing comments, please use correct grammar.

Capital letters to start and for names etc.

Much more professional IMO.

> +	spare.len = __get_rsc_tbl_spare_size(rproc, table, tablesz, &spare_index);
> +	if (spare.len < (sizeof(h) + request->size + 4)) /* new offset entry */

Not sure that comment makes the code any clearer?

All you're doing here is checking if we have enough space, right?

I think the 4 is a 'magic' number.  I'd either provide a comment (like
I did below), or define it.

> +		return -EPERM;

What does this have to do with permissions?

> -	/*
> -	 * Allocate another contiguous chunk of memory, large enough to
> -	 * contain the new, expanded resource table.
> -	 *
> -	 * The +4 is for the extra offset[] element in the top level header
> -	 */
> -	size += sizeof(struct fw_rsc_hdr) + request->size + 4;
> -	table = devm_kmemdup(&rproc->dev, old_table, size, GFP_KERNEL);
> -	if (!table)
> -		return ERR_PTR(-ENOMEM);
> +	new_rsc_offset = table->offset[spare_index];
>  
>  	/* Shunt table by 4 Bytes to account for the extra offset[] element */
>  	start_of_rscs = (void *)table + table->offset[0];
>  	memmove(start_of_rscs + 4,
>  		start_of_rscs, new_rsc_offset - table->offset[0]);
> +
> +	spare.len -= 4;

This probably deserves a comment too.

/*
 * The spare area is finite.  Since we are increasing the size of the
 * header and shunting the tables, we need to reduce the size of the
 * available 'spare' area by the shunt size.
 */

>  	new_rsc_offset += 4;
>  
>  	/* Update existing resource entry's offsets */
> @@ -1153,13 +1148,27 @@ rproc_add_resource_table_entry(struct rproc *rproc,
>  	/* Copy new firmware header into table */
>  	fw_header_loc = (void *)table + new_rsc_offset;
>  	memcpy(fw_header_loc, &h, sizeof(h));
> +	spare.len -= sizeof(h);
>  
>  	/* Copy new resource entry into table */
>  	new_rsc_loc = (void *)fw_header_loc + sizeof(h);
>  	memcpy(new_rsc_loc, request->resource, request->size);
> +	spare.len -= request->size;
>  
> -	*tablesz = size;
> -	return table;
> +	/* create new rsc spare resource at the end of remaining spare */

Same comment about using nice grammar in comments.

> +	new_spare_offset = new_rsc_offset + sizeof(h) + request->size;
> +	h.type = RSC_SPARE;
> +
> +	fw_header_loc = (void *)table + new_spare_offset;
> +	memcpy(fw_header_loc, &h, sizeof(h));
> +
> +	new_rsc_loc = (void *)fw_header_loc + sizeof(h);
> +	memcpy(new_rsc_loc, &spare, sizeof(spare));
> +
> +	/* update spare offset */
> +	table->offset[spare_index] = new_spare_offset;
> +
> +	return 0;
>  }
>  
>  static struct resource_table*
> @@ -1203,12 +1212,9 @@ rproc_apply_resource_overrides(struct rproc *rproc,
>  			continue;
>  
>  		/* Didn't find matching resource entry -- creating a new one. */
> -		table = rproc_add_resource_table_entry(rproc, resource,
> -						       table, &size);
> -		if (IS_ERR(table))
> +		updated = __add_rsc_tbl_entry(rproc, resource, table, size);
> +		if (updated)
>  			goto out;
> -
> -		*orig_table = table;
>  	}
>  
>  	rproc_dump_resource_table(rproc, table, size);

-- 
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org â Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog