Re: [flasher PATCH V2 1/4] Add crc32 verification of the flash image

[Thierry Reding <thierry.reding-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>]

[-- Attachment #1: Type: text/plain, Size: 1577 bytes --]

On Fri, Dec 06, 2013 at 10:35:33AM -0700, Stephen Warren wrote:
[...]
> diff --git a/tegra-uboot-flasher b/tegra-uboot-flasher
[...]
> @@ -125,6 +129,19 @@ def func_flash():
>      if args.debug:
>          print 'flash_img_size %d 0x%x' % (flash_img_size, flash_img_size)
>  
> +    imgf = file(flash_img, 'rb')
> +    imgd = imgf.read()
> +    imgf.close()
> +    flash_img_crc32 = binascii.crc32(imgd)
> +    if args.debug:
> +        print 'flash_img_crc32 %x' % flash_img_crc32
> +    flash_img_crc32_bs = (
> +        ((flash_img_crc32 & 0xff) << 24) |
> +        ((flash_img_crc32 & 0xff00) << 8) |
> +        ((flash_img_crc32 & 0xff0000) >> 8) |
> +        ((flash_img_crc32 & 0xff000000) >> 24)
> +    )

I would've thought that Python actually supported byteswapping with some
function, but it seems not (or at least not trivially). One could do
something like this:

	a = array.array('I', [crc])
	a.byteswap()
	crc = a[0]

But with the array module you apparently can't force 32-bit values. 'I'
will be at least 2, but 4 on 32-bit and 64-bit systems, 'L' will be 64
bits on 64-bit systems and 32 bits on 32-bit systems it seems. You could
check a.itemsize to determine the right type code. I suppose we won't be
running on anything but 32- or 64-bit systems for a while, but it still
said there's no support for sized types there.

It could possibly be done with the struct module as well, but that would
likely end up much more verbose too, so:

Reviewed-by: Thierry Reding <treding-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]