linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Mark Rutland <mark.rutland@arm.com>
To: Yang Yingliang <yangyingliang@huawei.com>
Cc: Catalin Marinas <Catalin.Marinas@arm.com>,
	Will Deacon <Will.Deacon@arm.com>,
	"bones@secretlab.ca" <bones@secretlab.ca>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"linux-arm-kernel@lists.infradead.org" 
	<linux-arm-kernel@lists.infradead.org>
Subject: Re: [RFC PATCH] arm64: Image: Allow the appending of a device tree binary
Date: Thu, 4 Jun 2015 11:04:50 +0100	[thread overview]
Message-ID: <20150604100450.GC599@leverpostej> (raw)
In-Reply-To: <1433410504-11628-1-git-send-email-yangyingliang@huawei.com>

Hi,

On Thu, Jun 04, 2015 at 10:35:04AM +0100, Yang Yingliang wrote:
> This patch provides the ability to boot using a device tree that is appended
> to the raw binary Image (e.g. cat Image <filename>.dtb > Image_w_dtb).
> 
> Both BE and LE conditions are tested.
> 
> Got references from e2a6a3aa ("ARM: zImage: Allow the appending of a device tree binary").

NAK for this as it stands. I really don't want to see appended DTB for
arm64.

We needed it for arm because old bootloaders didn't know how to pass a
DTB to the kernel. For arm64 DTB has always been a requirement, so
there's no legacy bootloader to support, and the rationale from arm does
not apply.

Adding appended DTB to arm64 encourages the broken assumption that
kernel and DTB are intimately coupled. It relies on arbitrary
limitations that are difficult and/or impossible to change in future in
a compatible fashion (e.g. the 2M carveout you make for the DT to live
in), and also makes it hard to replace either the kernel or DTB becuase
in the resulting Image there's no clear distinction between the two.
It's simply far too fragile.

Thanks,
Mark.

> 
> Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
> ---
>  arch/arm64/Kconfig              | 11 +++++++++++
>  arch/arm64/kernel/head.S        | 12 ++++++++++++
>  arch/arm64/kernel/vmlinux.lds.S | 17 +++++++++++++++++
>  3 files changed, 40 insertions(+)
> 
> diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
> index 7796af4..d402448 100644
> --- a/arch/arm64/Kconfig
> +++ b/arch/arm64/Kconfig
> @@ -660,6 +660,17 @@ endmenu
>  
>  menu "Boot options"
>  
> +config ARM64_APPENDED_DTB
> +	bool "Use appended device tree blob to Image"
> +	depends on OF
> +	help
> +	  With this option, the boot code will look for a device tree binary
> +	  (DTB) appended to Image
> +	  (e.g. cat Image <filename>.dtb > Image_w_dtb).
> +
> +	  With this option, the size of Image will be extended to be 2MB aligned.
> +	  The 2MB memory followed by Image is used for FDT blob.
> +
>  config CMDLINE
>  	string "Default kernel command string"
>  	default ""
> diff --git a/arch/arm64/kernel/head.S b/arch/arm64/kernel/head.S
> index 19f915e..ac00c8a 100644
> --- a/arch/arm64/kernel/head.S
> +++ b/arch/arm64/kernel/head.S
> @@ -274,6 +274,18 @@ ENDPROC(preserve_boot_args)
>   * The dtb must be 8-byte aligned and live in the first 512M of memory.
>   */
>  __vet_fdt:
> +#ifdef CONFIG_ARM64_APPENDED_DTB
> +	ldr	x21, =_edata
> +	add	x21, x21, x28
> +	ldr	w0, [x21]
> +#ifndef __AARCH64EB__
> +	ldr	w1, =0xedfe0dd0		// FDT magic word
> +#else
> +	ldr	w1, =0xd00dfeed
> +#endif
> +	cmp	w0, w1
> +	b.ne	1f			// no FDT found
> +#endif
>  	tst	x21, #0x7
>  	b.ne	1f
>  	cmp	x21, x24
> diff --git a/arch/arm64/kernel/vmlinux.lds.S b/arch/arm64/kernel/vmlinux.lds.S
> index a2c2986..c71bac3 100644
> --- a/arch/arm64/kernel/vmlinux.lds.S
> +++ b/arch/arm64/kernel/vmlinux.lds.S
> @@ -54,6 +54,18 @@ PECOFF_FILE_ALIGNMENT = 0x200;
>  #define PECOFF_EDATA_PADDING
>  #endif
>  
> +/*
> + * Set FDT pointer 2MB aligned to ensure
> + * it don't cross a 2-megabyte boundary.
> + */
> +FDT_FILE_ALIGNMENT = (1 << 21);
> +#ifdef CONFIG_ARM64_APPENDED_DTB
> +#define FDT_EDATA_PADDING	\
> +	.fdt_edata_padding : { BYTE(0); . = ALIGN(FDT_FILE_ALIGNMENT); }
> +#else
> +#define FDT_EDATA_PADDING
> +#endif
> +
>  #ifdef CONFIG_DEBUG_ALIGN_RODATA
>  #define ALIGN_DEBUG_RO			. = ALIGN(1<<SECTION_SHIFT);
>  #define ALIGN_DEBUG_RO_MIN(min)		ALIGN_DEBUG_RO
> @@ -149,8 +161,13 @@ SECTIONS
>  	_sdata = .;
>  	RW_DATA_SECTION(64, PAGE_SIZE, THREAD_SIZE)
>  	PECOFF_EDATA_PADDING
> +	FDT_EDATA_PADDING	/* 2MB aligned */
>  	_edata = .;
>  
> +#ifdef CONFIG_ARM64_APPENDED_DTB
> +	. += (1 << 21);		/* maximum 2MB for FDT blob */
> +#endif
> +
>  	BSS_SECTION(0, 0, 0)
>  
>  	. = ALIGN(PAGE_SIZE);
> -- 
> 1.8.0
> 
> 
> 
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> 

  reply	other threads:[~2015-06-04 10:05 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-06-04  9:35 [RFC PATCH] arm64: Image: Allow the appending of a device tree binary Yang Yingliang
2015-06-04 10:04 ` Mark Rutland [this message]
2015-06-04 10:07 ` Dave Martin

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=20150604100450.GC599@leverpostej \
    --to=mark.rutland@arm.com \
    --cc=Catalin.Marinas@arm.com \
    --cc=Will.Deacon@arm.com \
    --cc=bones@secretlab.ca \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=yangyingliang@huawei.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).