From: Julien Grall <julien.grall@linaro.org>
To: Ian Campbell <Ian.Campbell@citrix.com>
Cc: patches@linaro.org, Stefano.Stabellini@eu.citrix.com,
xen-devel@lists.xen.org
Subject: Re: [PATCH v2 2/2] xen/arm: If the DOM0 zImage as a DTB appended replace it
Date: Thu, 30 May 2013 15:47:30 +0100 [thread overview]
Message-ID: <51A76682.8090309@linaro.org> (raw)
In-Reply-To: <1369923870.18727.9.camel@zakaz.uk.xensource.com>
On 05/30/2013 03:24 PM, Ian Campbell wrote:
> On Thu, 2013-05-30 at 15:05 +0100, Julien Grall wrote:
>> If a DTB is appended to the DOM0 zImage, it's probably means that the linux
>> decompression stage will replace the DBT register value (r2 on arm32)
>> by the address of the appended DTB.
>
> I don't think we should do this, it certainly violates the principal of
> least surprise, since no other "bootloader" does anything like this
> AFAIK. Yes that means you occasionally trip over your DTB changes
> appearing not to take affect, but that's true on native and is one of
> the known and understood bad things about appended DTB.
Shall we load the DTB if there is one appended? If yes, I think the user
will be confused because Linux won't use the right device tree.
It's hard for the user to find the issue because there is no log.
> So I think this is a case of "don't do that then". At most we could
> issue a warning, I suppose. But really we shouldn't be making any
> assumptions (good or bad) about what happens to live in the memory just
> past the end of the kernel, it may or may not be an appended DTB.
I can add a warning and also fix the check the line:
if ( addr + end - start + sizeof(dtb_hdr) <= size )
must be replace by:
if ( end - start + sizeof(dtb_hdr) <= size )
>> In this case, to avoid issue with Linux, Xen needs to load the new device tree
>> just after the kernel.
>>
>> Signed-off-by: Julien Grall <julien.grall@linaro.org>
>> ---
>> xen/arch/arm/kernel.c | 26 ++++++++++++--------------
>> 1 file changed, 12 insertions(+), 14 deletions(-)
>>
>> diff --git a/xen/arch/arm/kernel.c b/xen/arch/arm/kernel.c
>> index f8c8850..1d6c927 100644
>> --- a/xen/arch/arm/kernel.c
>> +++ b/xen/arch/arm/kernel.c
>> @@ -123,20 +123,6 @@ static int kernel_try_zimage_prepare(struct kernel_info *info,
>> if ( (end - start) > size )
>> return -EINVAL;
>>
>> - /*
>> - * Check for an appended DTB.
>> - */
>> - if ( addr + end - start + sizeof(dtb_hdr) <= size )
>> - {
>> - copy_from_paddr(&dtb_hdr, addr + end - start,
>> - sizeof(dtb_hdr), DEV_SHARED);
>> - if (be32_to_cpu(dtb_hdr.magic) == DTB_MAGIC) {
>> - end += be32_to_cpu(dtb_hdr.total_size);
>> -
>> - if ( end > addr + size )
>> - return -EINVAL;
>> - }
>> - }
>>
>> info->zimage.kernel_addr = addr;
>>
>> @@ -154,6 +140,18 @@ static int kernel_try_zimage_prepare(struct kernel_info *info,
>> info->load = kernel_zimage_load;
>> info->type = KERNEL_ZIMAGE;
>>
>> + /*
>> + * If there is an appended DTB, ask XEN to replace the DTB
>> + * by the generate one.
>> + */
>> + if ( info->zimage.len + sizeof(dtb_hdr) <= size )
>> + {
>> + copy_from_paddr(&dtb_hdr, addr + end - start,
>> + sizeof(dtb_hdr), DEV_SHARED);
>> + if (be32_to_cpu(dtb_hdr.magic) == DTB_MAGIC)
>> + info->dtb_paddr = info->zimage.load_addr + info->zimage.len;
>> + }
>> +
>> return 0;
>> }
>>
>
>
next prev parent reply other threads:[~2013-05-30 14:47 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-05-30 14:05 [PATCH v2 0/2] Rework the way to compute dom0 DTB base address Julien Grall
2013-05-30 14:05 ` [PATCH v2 1/2] xen/arm: " Julien Grall
2013-05-31 11:45 ` Ian Campbell
2013-05-31 13:28 ` Julien Grall
2013-05-31 13:45 ` Julien Grall
2013-05-31 14:14 ` Ian Campbell
2013-05-30 14:05 ` [PATCH v2 2/2] xen/arm: If the DOM0 zImage as a DTB appended replace it Julien Grall
2013-05-30 14:24 ` Ian Campbell
2013-05-30 14:47 ` Julien Grall [this message]
2013-05-30 14:52 ` Ian Campbell
2013-05-30 15:01 ` Julien Grall
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=51A76682.8090309@linaro.org \
--to=julien.grall@linaro.org \
--cc=Ian.Campbell@citrix.com \
--cc=Stefano.Stabellini@eu.citrix.com \
--cc=patches@linaro.org \
--cc=xen-devel@lists.xen.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;
as well as URLs for NNTP newsgroup(s).