mail archive of the barebox mailing list
 help / color / mirror / Atom feed
From: Christian Mauderer <christian.mauderer@embedded-brains.de>
To: Ahmad Fatoum <a.fatoum@pengutronix.de>, barebox@lists.infradead.org
Subject: Re: [PATCH v2] FIT: Parse `load` and `entry` addresses.
Date: Tue, 14 Jul 2020 09:56:10 +0200	[thread overview]
Message-ID: <64a9b074-bd8b-67ef-a7eb-2de72c500bed@embedded-brains.de> (raw)
In-Reply-To: <c70e7be2-9457-c7d9-80e5-1683fffa6d05@pengutronix.de>

Hello Ahmad,

On 14/07/2020 09:50, Ahmad Fatoum wrote:
> Hello Christian,
> 
> On 7/14/20 9:17 AM, Christian Mauderer wrote:
>> According to the U-Boot documentation for the FIT file format, the load
>> and entry have to be allways defined for a "kernel" or "standalone".
>> But Barebox ignored the parameters. That changes with this patch.
>>
>> For backward compatibility the default address is still used for images
>> without `load` or `entry`.
>>
>> Signed-off-by: Christian Mauderer <christian.mauderer@embedded-brains.de>
>> ---
>>  arch/arm/mach-layerscape/ppa.c |  3 +-
>>  common/bootm.c                 |  7 +++--
>>  common/image-fit.c             | 57 ++++++++++++++++++++++++++++++++--
>>  include/image-fit.h            |  3 +-
>>  4 files changed, 63 insertions(+), 7 deletions(-)
>>
>> diff --git a/arch/arm/mach-layerscape/ppa.c b/arch/arm/mach-layerscape/ppa.c
>> index 477e89478..3eb7ab641 100644
>> --- a/arch/arm/mach-layerscape/ppa.c
>> +++ b/arch/arm/mach-layerscape/ppa.c
>> @@ -82,7 +82,8 @@ static int ppa_init(void *ppa, size_t ppa_size, void *sec_firmware_addr)
>>  	}
>>  
>>  
>> -	ret = fit_open_image(fit, conf, "firmware", &buf, &firmware_size);
>> +	ret = fit_open_image(fit, conf, "firmware", &buf, &firmware_size,
>> +			     NULL, NULL);
>>  	if (ret) {
>>  		pr_err("Cannot open firmware image in ppa FIT image: %s\n",
>>  		       strerror(-ret));
>> diff --git a/common/bootm.c b/common/bootm.c
>> index 366f31455..94600cae3 100644
>> --- a/common/bootm.c
>> +++ b/common/bootm.c
>> @@ -222,7 +222,7 @@ int bootm_load_initrd(struct image_data *data, unsigned long load_address)
>>  		unsigned long initrd_size;
>>  
>>  		ret = fit_open_image(data->os_fit, data->fit_config, "ramdisk",
>> -				     &initrd, &initrd_size);
>> +				     &initrd, &initrd_size, NULL, NULL);
>>  
>>  		data->initrd_res = request_sdram_region("initrd",
>>  				load_address,
>> @@ -348,7 +348,7 @@ void *bootm_get_devicetree(struct image_data *data)
>>  		unsigned long of_size;
>>  
>>  		ret = fit_open_image(data->os_fit, data->fit_config, "fdt",
>> -				     &of_tree, &of_size);
>> +				     &of_tree, &of_size, NULL, NULL);
>>  		if (ret)
>>  			return ERR_PTR(ret);
>>  
>> @@ -622,7 +622,8 @@ int bootm_boot(struct bootm_data *bootm_data)
>>  		}
>>  
>>  		ret = fit_open_image(data->os_fit, data->fit_config, "kernel",
>> -				     &data->fit_kernel, &data->fit_kernel_size);
>> +				     &data->fit_kernel, &data->fit_kernel_size,
>> +				     &data->os_address, &data->os_entry);
>>  		if (ret)
>>  			goto err_out;
>>  	}
>> diff --git a/common/image-fit.c b/common/image-fit.c
>> index 2681d62a9..310013121 100644
>> --- a/common/image-fit.c
>> +++ b/common/image-fit.c
>> @@ -517,12 +517,30 @@ int fit_has_image(struct fit_handle *handle, void *configuration,
>>  	return 1;
>>  }
>>  
>> +static int fit_get_address(struct device_node *image, const char *property,
>> +			   unsigned long *addr)
>> +{
>> +	const __be32 *cell;
>> +	int len = 0;
>> +
>> +	cell = of_get_property(image, property, &len);
>> +	if (!cell)
>> +		return -EINVAL;
>> +	if (len > sizeof(*addr))
>> +		return -ENOTSUPP;
>> +	
>> +	*addr = (unsigned long)of_read_number(cell, len / sizeof(*cell));
>> +	return 0;
>> +}
>> +
>>  /**
>>   * fit_open_image - Open an image in a FIT image
>>   * @handle: The FIT image handle
>>   * @name: The name of the image to open
>>   * @outdata: The returned image
>>   * @outsize: Size of the returned image
>> + * @load: The load address given by the image
>> + * @entry: The entry address given by the image
>>   *
>>   * Open an image in a FIT image. The returned image is freed during fit_close().
>>   * @configuration holds the cookie returned from fit_open_configuration() if
>> @@ -532,14 +550,21 @@ int fit_has_image(struct fit_handle *handle, void *configuration,
>>   * then only the hash is checked (because opening the configuration already
>>   * checks the RSA signature of all involved nodes).
>>   *
>> + * The load address and entry point of the image description in the FIT will be
>> + * parsed if they exist and if the @load and @entry parameters are not NULL.
>> + * Otherwise @load and @entry won't be changed.
>> + *
>>   * Return: 0 for success, negative error code otherwise
>>   */
>>  int fit_open_image(struct fit_handle *handle, void *configuration,
>>  		   const char *name, const void **outdata,
>> -		   unsigned long *outsize)
>> +		   unsigned long *outsize, unsigned long *load,
>> +		   unsigned long *entry)
>>  {
>>  	struct device_node *image;
>>  	const char *unit, *type = NULL, *desc= "(no description)";
>> +	char load_str[sizeof("; load: 0x12345678")] = "";
> 
> Not big enough for sizeof(unsigned long) == 8

Yes, sorry. I was too focused on my 32 bit target.

> 
>> +	char entry_str[sizeof("; entry (relative to load): 0x12345678")] = "";
> 
> Same
> 
>>  	const void *data;
>>  	int data_len;
>>  	int ret = 0;
>> @@ -559,7 +584,35 @@ int fit_open_image(struct fit_handle *handle, void *configuration,
>>  		return -ENOENT;
>>  
>>  	of_property_read_string(image, "description", &desc);
>> -	pr_info("image '%s': '%s'\n", unit, desc);
> 
> 
>> +
>> +	if (load) {
>> +		ret = fit_get_address(image, "load", load);
>> +		if (ret < 0) {
>> +			pr_info("Couldn't get load address in %s. Use default.\n", image->full_name);
>> +			ret = 0;
> 
> You don't have to reset ret in between. I'd just skip this.
> 

You are right. I missed that ret is set in any case later. I somehow
thought it can be returned directly in some case. If that would have
been the case not resetting would have broken compatibility with FIT
images without load and entry.

>> +		}
>> +	}
>> +
>> +	if (load && entry) {
>> +		ret = fit_get_address(image, "entry", entry);
>> +		if (ret < 0) {
>> +			pr_info("Couldn't get entry point in %s. Use default.\n", image->full_name);
>> +			ret = 0;
>> +		} else {
>> +			/* Barebox uses an entry relative to load but the FIT
>> +			 * images assume an absolute entry. */
>> +			*entry -= *load;
>> +		}
>> +	}
>> +
>> +	if (load) {
>> +		snprintf(load_str, sizeof(load_str), "; load: 0x%08lx", *load);
>> +	}
> 
> No { } for single expression blocks.

Sorry. Used to another coding style.

> 
>> +	if (load) {
>> +		snprintf(entry_str, sizeof(entry_str),
>> +			 "; entry (relative to load): 0x%08lx", *entry);
>> +	}
> 
> This should be guarded by if (entry), but I'd prefer you drop this. See below
> 
>> +	pr_info("image '%s': '%s'%s%s\n", unit, desc, load_str, entry_str);
> 
> Could you use pr_info without a trailing \n and then use pr_cont to complete
> the line?

I missed the pr_cont. My first try has been multiple pr_info which did
add a lot of "FIT: " into the prints. I'll change it.

> 
>>  
>>  	of_property_read_string(image, "type", &type);
>>  	if (!type) {
>> diff --git a/include/image-fit.h b/include/image-fit.h
>> index 27c9e8351..038732d0d 100644
>> --- a/include/image-fit.h
>> +++ b/include/image-fit.h
>> @@ -31,7 +31,8 @@ int fit_has_image(struct fit_handle *handle, void *configuration,
>>  		  const char *name);
>>  int fit_open_image(struct fit_handle *handle, void *configuration,
>>  		   const char *name, const void **outdata,
>> -		   unsigned long *outsize);
>> +		   unsigned long *outsize, unsigned long *load,
>> +		   unsigned long *entry);
>>  
>>  void fit_close(struct fit_handle *handle);
>>  
>>
> 

-- 
--------------------------------------------
embedded brains GmbH
Herr Christian Mauderer
Dornierstr. 4
D-82178 Puchheim
Germany
email: christian.mauderer@embedded-brains.de
Phone: +49-89-18 94 741 - 18
Fax:   +49-89-18 94 741 - 08
PGP: Public key available on request.

Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

      reply	other threads:[~2020-07-14  7:56 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-07-14  7:17 Christian Mauderer
2020-07-14  7:50 ` Ahmad Fatoum
2020-07-14  7:56   ` Christian Mauderer [this message]

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=64a9b074-bd8b-67ef-a7eb-2de72c500bed@embedded-brains.de \
    --to=christian.mauderer@embedded-brains.de \
    --cc=a.fatoum@pengutronix.de \
    --cc=barebox@lists.infradead.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