[U-Boot] [PATCH v3 6/9] spl: fit: add support for post-processing of images

[Simon Glass]

On 27 June 2016 at 07:19, Andreas Dannenberg <dannenberg at ti.com> wrote:
> From: Daniel Allred <d-allred at ti.com>
>
> The next stage boot loader image and the selected FDT can be post-
> processed by board/platform/device-specific code, which can include
> modifying the size and altering the starting source address before
> copying these binary blobs to their final destination. This might be
> desired to do things like strip headers or footers attached to the
> images before they were packaged into the FIT, or to perform operations
> such as decryption or authentication. Introduce new configuration
> option CONFIG_SPL_FIT_IMAGE_POST_PROCESS to allow controlling this
> feature. If enabled, a platform-specific post-process function must
> be provided.
>
> Signed-off-by: Daniel Allred <d-allred at ti.com>
> Signed-off-by: Andreas Dannenberg <dannenberg at ti.com>
> Reviewed-by: Tom Rini <trini at konsulko.com>
> ---
>  Kconfig              | 14 ++++++++++++++
>  common/spl/spl_fit.c | 21 ++++++++++++++++-----
>  include/image.h      | 17 +++++++++++++++++
>  3 files changed, 47 insertions(+), 5 deletions(-)

Reviewed-by: Simon Glass <sjg at chromium.org>

Nit below in case you do a new version.

>
> diff --git a/Kconfig b/Kconfig
> index 3ceff25..2afbaaf 100644
> --- a/Kconfig
> +++ b/Kconfig
> @@ -313,6 +313,20 @@ config SPL_LOAD_FIT
>           particular it can handle selecting from multiple device tree
>           and passing the correct one to U-Boot.
>
> +config SPL_FIT_IMAGE_POST_PROCESS
> +       bool "Enable post-processing of FIT artifacts after loading by the SPL"
> +       depends on SPL_LOAD_FIT && TI_SECURE_DEVICE
> +       help
> +         Allows doing any sort of manipulation to blobs after they got extracted
> +         from the U-Boot FIT image like stripping off headers or modifying the
> +         size of the blob, verification, authentication, decryption etc. in a
> +         platform or board specific way. In order to use this feature a platform
> +         or board-specific implementation of board_fit_image_post_process() must
> +         be provided. Also, anything done during this post-processing step would
> +         need to be comprehended in how the images were prepared before being
> +         injected into the FIT creation (i.e. the blobs would have been pre-
> +         processed before being added to the FIT image).
> +
>  config SYS_CLK_FREQ
>         depends on ARC || ARCH_SUNXI
>         int "CPU clock frequency"
> diff --git a/common/spl/spl_fit.c b/common/spl/spl_fit.c
> index 9874708..069e94d 100644
> --- a/common/spl/spl_fit.c
> +++ b/common/spl/spl_fit.c
> @@ -132,7 +132,7 @@ int spl_load_simple_fit(struct spl_load_info *info, ulong sector, void *fit)
>         int data_offset, data_size;
>         int base_offset, align_len = ARCH_DMA_MINALIGN - 1;
>         int src_sector;
> -       void *dst;
> +       void *dst, *src;
>
>         /*
>          * Figure out where the external images start. This is the base for the
> @@ -206,8 +206,13 @@ int spl_load_simple_fit(struct spl_load_info *info, ulong sector, void *fit)
>                 return -EIO;
>         debug("image: dst=%p, data_offset=%x, size=%x\n", dst, data_offset,
>               data_size);
> -       memcpy(dst, dst + get_aligned_image_overhead(info, data_offset),
> -              data_size);
> +       src = dst + get_aligned_image_overhead(info, data_offset);
> +
> +#ifdef CONFIG_SPL_FIT_IMAGE_POST_PROCESS
> +       board_fit_image_post_process((void **)&src, (size_t *)&data_size);
> +#endif
> +
> +       memcpy(dst, src, data_size);
>
>         /* Figure out which device tree the board wants to use */
>         fdt_len = spl_fit_select_fdt(fit, images, &fdt_offset);
> @@ -236,8 +241,14 @@ int spl_load_simple_fit(struct spl_load_info *info, ulong sector, void *fit)
>          */
>         debug("fdt: dst=%p, data_offset=%x, size=%x\n", dst, fdt_offset,
>               fdt_len);
> -       memcpy(load_ptr + data_size,
> -              dst + get_aligned_image_overhead(info, fdt_offset), fdt_len);
> +       src = dst + get_aligned_image_overhead(info, fdt_offset);
> +       dst = load_ptr + data_size;
> +
> +#ifdef CONFIG_SPL_FIT_IMAGE_POST_PROCESS
> +       board_fit_image_post_process((void **)&src, (size_t *)&fdt_len);
> +#endif
> +
> +       memcpy(dst, src, fdt_len);
>
>         return 0;
>  }
> diff --git a/include/image.h b/include/image.h
> index d788c26..93d39e1 100644
> --- a/include/image.h
> +++ b/include/image.h
> @@ -1173,4 +1173,21 @@ void android_print_contents(const struct andr_img_hdr *hdr);
>   */
>  int board_fit_config_name_match(const char *name);
>
> +#ifdef CONFIG_SPL_FIT_IMAGE_POST_PROCESS
> +/**
> + * board_fit_image_post_process() - Do any post-process on FIT binary data
> + *
> + * This is used to do any sort of image manipulation, verification, decryption
> + * etc. in a platform or board specific way. Obviously, anything done here would
> + * need to be comprehended in how the images were prepared before being injected
> + * into the FIT creation (i.e. the binary blobs would have been pre-processed
> + * before being added to the FIT image).
> + *
> + * @image: pointer to the image start pointer
> + * @size: pointer to the image size
> + * @return no return value (failure should be handled internally)
> + */
> +void board_fit_image_post_process(void **p_image, size_t *p_size);
> +#endif /* CONFIG_SPL_FIT_IMAGE_POST_PROCESS */

We don't need #ifdef in header files - it just makes the code harder
to read, and we'll still get a build error (with correct line number
info) if someone uses it when they should not.

> +
>  #endif /* __IMAGE_H__ */
> --
> 2.6.4
>

Regards,
Simon