[RESEND PATCH 1/2] dt-ops: Add helper API to dump fdt blob

[Dave Young]

On 04/16/18 at 08:39pm, Bhupesh Sharma wrote:
> Hi Dave,
> 
> On Mon, Apr 16, 2018 at 1:43 PM, Dave Young <dyoung at redhat.com> wrote:
> > Hi Bhupesh,
> >
> > On 04/03/18 at 07:28pm, Bhupesh Sharma wrote:
> >> Hi James,
> >>
> >> Sorry for the delay, I had a long weekend last week.
> >>
> >> On Tue, Mar 27, 2018 at 7:01 PM, James Morse <james.morse at arm.com> wrote:
> >> > Hi Akashi, Bhupesh,
> >> >
> >> > On 27/03/18 10:04, AKASHI Takahiro wrote:
> >> >> On Mon, Mar 26, 2018 at 02:29:31PM +0530, Bhupesh Sharma wrote:
> >> >>> On Tue, Mar 20, 2018 at 12:36 AM, Bhupesh Sharma <bhsharma at redhat.com> wrote:
> >> >>>> On Mon, Mar 19, 2018 at 8:15 PM, AKASHI Takahiro
> >> >>>> <takahiro.akashi at linaro.org> wrote:
> >> >>>>> On Mon, Mar 19, 2018 at 04:05:38PM +0530, Bhupesh Sharma wrote:
> >> >>>>>> At several occasions it would be useful to dump the fdt
> >> >>>>>> blob being passed to the second (kexec/kdump) kernel
> >> >>>>>> when '-d' flag is specified while invoking kexec/kdump.
> >> >>>>>
> >> >>>>> Why not just save binary data to a file and interpret it
> >> >>>>> with dtc command?
> >> >
> >> > I'd prefer this too. It would also let us debug any issue where kexec-tools
> >> > produces an invalid DTB. It also lets us test booting the kernel from firmware
> >> > with that DTB.
> >>
> >> I captured the use case where it is not possible to do so. I have seen
> >> primary kernel crash before we can get to the command prompt to save
> >> the dtb blob. Since the arm64 crashkernel still seems to have issues
> >> itself while booting on acpi enabled machines (see
> >> <https://www.spinics.net/lists/arm-kernel/msg616632.html>), so we are
> >> trying to debug a problem which has two undefined variables :)
> >>
> >> >>>> Well, there are a couple of reasons for that which can be understood
> >> >>>> from a system which is in a production environment (for e.g. I have
> >> >>>> worked on one arm64 system in the past which used a yocto based
> >> >>>> distribution in which kexec -p was launched with a -d option as a part
> >> >>>> of initial ramfs scriptware):
> >> >
> >> > and panics before you get an interactive prompt or persistent storage? I think
> >> > this would be a pretty niche use-case. You could always base64-dump the dtb to
> >> > stdout from your script.
> >>
> >> That is pretty basic case on several new arm64 development boards
> >> (e.g. qualcomm, huawei etc) where we are debugging issues in primary
> >> kernel boot (and we are not even able to reach the command prompt).
> >>
> >> If the crashkernel crashes even before the primary kernel does because
> >> of the issues in the way DTB is passed to the crashkernel (which can
> >> include wrong DTB fields), we better have mechanisms to track the same
> >> rather than adding debug prints to the kernels.
> >
> > In this case, since userspace always need to run 'kexec -l', there
> > should be chance to save dtb and dump them in scripts if needed.
> > If this is doable in scripts I also tend not to add the code in kexec-tools.
> 
> Perhaps you missed my latest email which explained the use case better
> (please see [1]).
> Please note that I am talking about the 'kexec -p' or the kdump use
> case rather than the 'kexec -l' or the warm reboot to the second
> kernel.

It is a typo, I meant about 'kexec -p' as well

> 
> In several cases its useful if we are not able to reach the command
> prompt to just modify the scriptware to launch 'kexec -p' when the
> ramfs starts up so that we can catch crashes which happen before the
> primary kernel reaches the command prompt. For example, on freescale
> boards I found this scriptware mechanism quite useful on yocto based
> distributions, as you can be porting a new upstream kernel on the
> board and thus is not very stable while booting and the crashkernel
> also has issues which causes it to crash. In such cases although we
> see a panic message from the crashkernel it is very difficult to
> deduce that it was because of some wrongly fixed up dtb property.

Since kexec -p need to be automated in scripts, add something to dump
dtb is also doable.

> 
> Getting a debug log which contains the dtb dump is very useful in
> above cases for debugging.
> 
> Also saving the dtb requires 'dtc' tool to be installed which is an
> additional dependency.

I think it should be not a big problem,  dtc is designed to do this.

> 
> Since we have debug facility/messages already available when we
> execute 'kexec -p' with '-d' flag we can use the dtb dumps from the
> tool to debug crashes than happen in the crashkernel also because of
> wrong dtb fields fixed up by 'kexec-tools'. I was loathe to keep this
> patch locally and apply to the upstream 'kexec-tools' when debugging
> these issues as I think it makes sense to improve our debugging
> capabilities if we know there are issues around the same.

Hmm, do you have the example kexec-tools fix-up the dtb content?

> 
> Ok, let me resend this patch with a cover letter this time to explain
> the use case better and also to capture the dtb logs.
> 
> [1] https://marc.info/?l=kexec&m=152382169120505&w=2

Thanks
Dave