xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Julien Grall <julien@xen.org>
To: Christopher Clark <christopher.w.clark@gmail.com>,
	Srinivas Bangalore <srini@yujala.com>
Cc: xen-devel <xen-devel@lists.xenproject.org>
Subject: Re: Porting Xen to Jetson Nano
Date: Thu, 23 Jul 2020 18:04:53 +0100	[thread overview]
Message-ID: <c04b542c-53b9-ee0d-981f-53ea4100f139@xen.org> (raw)
In-Reply-To: <CACMJ4GYQUXNGrqq_6wFLX4actMgTat-i5ThhS21Bjy3HO52bUQ@mail.gmail.com>

Hi,

On 23/07/2020 05:26, Christopher Clark wrote:
> On Wed, Jul 22, 2020 at 10:59 AM Srinivas Bangalore <srini@yujala.com> wrote:
>> Dear Xen experts,
>>
>> Would greatly appreciate some hints on how to move forward with this one…
> 
> Hi Srini,
> 
> I don't have any strong recommendations for you, but I do want to say
> that I'm very happy to see you taking this project on and I am hoping
> for your success. I have a newly-arrived Jetson Nano sitting on my
> desk here, purchased with the intention of getting Xen up and running
> on it, that I just haven't got to work on yet. I'm also familiar with
> Chris Patterson, Kyle Temkin and Ian Campbell's previous Tegra Jetson
> patches and it would be great to see some further progress made from
> those.

I agree that it would be good to have the support in upstream!

> 
> In my recent experience with the Raspberry Pi 4, one basic observation
> with ARM kernel bringup is that if your device tree isn't good, your
> dom0 kernel can be missing the configuration it needs to use the
> serial port correctly and you don't get any diagnostics from it after
> Xen attempts to launch it, so I would just patch the right serial port
> config directly into your Linux kernel (eg. hardcode specific things
> onto the kernel command line) so you're not messing about with that
> any more.
> 
> The other thing I would recommend is patching in some printks into the
> earliest part of the Xen parts of the Dom0 Linux kernel start code.
> Others who are more familar with Xen on ARM may have some better
> recommendations, but linux/arch/arm/xen/enlighten.c has a function
> xen_guest_init that looks like a good place to stuff some extra
> printks for some early proof-of-entry from your kernel, and that way
> you'll have some indication whether execution has actually commenced
> in there.

Linux provides earlyprintk facilities that can be used in Xen. To enable 
it, you need to have your kernel built with CONFIG_EARLY_PRINTK=y and 
CONFIG_XEN=y. This can then be enabled by passing earlyprintk=xenboot on 
your command line.

Note that Linux needs to detect you are using Xen before using 
earlyprintk. If you need earlier, then what I usually do is hacking 
xen_raw_console_write() (in drivers/tty/hvc/hvc_xen.c) and replace 'if 
(xen_domain())' with 'if (1)'.

> 
> I don't think you're going to get a great deal of enthusiasm on this
> list for Xen 4.8.5, unfortunately; most people around here work off
> Xen's staging branch, and I'd be surprised to hear of anyone having
> tried a 5.7 Linux kernel with Xen 4.8.5. I can understand why you
> might start there from the existing patch series though.

Right, 4.8.5 is now out of support and we improved Xen quite a lot since 
then. As a general recommendation, I would suggest to move the series to 
the latest staging once you get it working on 4.8.5.

However, I don't see any reason why 5.7 wouldn't boot on 4.8.5. I will 
have a look at your stack trace and answer there.

Best regards,

-- 
Julien Grall


  parent reply	other threads:[~2020-07-23 17:05 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-07-22 17:57 Porting Xen to Jetson Nano Srinivas Bangalore
2020-07-23  4:26 ` Christopher Clark
2020-07-23 15:42   ` Srinivas Bangalore
2020-07-23 17:04   ` Julien Grall [this message]
2020-07-23 17:19     ` Julien Grall
2020-07-23 18:04 ` Julien Grall
2020-07-24 15:01   ` Srinivas Bangalore
2020-07-24 17:25     ` Julien Grall
2020-07-27 22:09       ` Srinivas Bangalore
2020-07-28 16:56         ` Julien Grall
2020-07-28 17:18           ` Srinivas Bangalore
2020-07-28 18:01             ` Rich Persaud
2020-07-29 17:25       ` srini
2020-07-30  1:27         ` Stefano Stabellini
2020-07-30 18:58           ` srini
2020-07-31  1:16             ` Stefano Stabellini
  -- strict thread matches above, loose matches on Subject: below --
2020-07-20 10:50 Srinivas Bangalore

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=c04b542c-53b9-ee0d-981f-53ea4100f139@xen.org \
    --to=julien@xen.org \
    --cc=christopher.w.clark@gmail.com \
    --cc=srini@yujala.com \
    --cc=xen-devel@lists.xenproject.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).