Re: [RFC PATCH qemu] spapr/docs: Add a few words about x-vof

[Alexey Kardashevskiy <aik@ozlabs.ru>]

On 5/12/22 23:40, Daniel Henrique Barboza wrote:
> 
> 
> On 5/12/22 00:10, Alexey Kardashevskiy wrote:
>>
>>
>> On 5/12/22 06:42, Daniel Henrique Barboza wrote:
>>>
>>>
>>> On 5/6/22 02:51, Alexey Kardashevskiy wrote:
>>>> The alternative small firmware needs a few words of what it can and
>>>> absolutely cannot do; this adds those words.
>>>>
>>>> Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
>>>> ---
>>>>   docs/system/ppc/pseries.rst | 28 ++++++++++++++++++++++++++++
>>>>   1 file changed, 28 insertions(+)
>>>>
>>>> diff --git a/docs/system/ppc/pseries.rst b/docs/system/ppc/pseries.rst
>>>> index d9b65ad4e850..4c98a94f9add 100644
>>>> --- a/docs/system/ppc/pseries.rst
>>>> +++ b/docs/system/ppc/pseries.rst
>>>> @@ -32,14 +32,42 @@ Missing devices
>>>>   Firmware
>>>>   ========
>>>> +The pSeries platform in QEMU comes with 2 firmwares:
>>>> +
>>>>   `SLOF <https://github.com/aik/SLOF>`_ (Slimline Open Firmware) is an
>>>>   implementation of the `IEEE 1275-1994, Standard for Boot 
>>>> (Initialization
>>>>   Configuration) Firmware: Core Requirements and Practices
>>>>   <https://standards.ieee.org/standard/1275-1994.html>`_.
>>>> +SLOF performs bus scanning, PCI resource allocation, provides the 
>>>> client
>>>> +interface to boot from block devices and network.
>>>> +
>>>>   QEMU includes a prebuilt image of SLOF which is updated when a 
>>>> more recent
>>>>   version is required.
>>>> +VOF (Virtual Open Firmware) is a minimalistic firmware to work with
>>>> +``-machine pseries,x-vof=on``. When enabled, the firmware acts as a 
>>>> slim
>>>> +shim and QEMU implements parts of the IEEE 1275 Open Firmware 
>>>> interface.
>>>> +
>>>> +VOF does not have device drivers, does not do PCI resource 
>>>> allocation and
>>>> +relies on ``-kernel`` used with Linux kernels recent enough (v5.4+)
>>>> +to PCI resource assignment. It is ideal to use with petitboot.
>>>> +
>>>> +Booting via ``-kernel`` supports the following:
>>>> ++-------------------+-------------------+------------------+
>>>> +| kernel            | pseries,x-vof=off | pseries,x-vof=on |
>>>> ++===================+===================+==================+
>>>> +| vmlinux BE        |     ✓             |     ✓            |
>>>> ++-------------------+-------------------+------------------+
>>>> +| vmlinux LE        |     ✓             |     ✓            |
>>>> ++-------------------+-------------------+------------------+
>>>> +| zImage.pseries BE |     x             |     ✓¹           |
>>>> ++-------------------+-------------------+------------------+
>>>> +| zImage.pseries LE |     ✓             |     ✓            |
>>>> ++-------------------+-------------------+------------------+
>>>
>>> You need an empty line at the start and at the end of the table. 
>>> Otherwise it'll
>>> be rendered as regular text.
>>
>> How do you build htmls from these btw?
> 
> Had to do this yesterday because I changed machines recently. In a 
> Fedora 35
> system I did this:
> 
> 
> sudo dnf install python3-sphinx
> sudo dnf install python-sphinx_rtd_theme-doc
> pip install sphinx_rtd_theme


I only needed the last one, compiles htmls now. I was trying to figure 
out in Makefiles where is that "html" target and I just could not :)


> 
> (not sure if all steps are needed)
> 
> Then the generated docs will be under build/docs/manual .
> 
> 
>>
>>>
>>>> +Notes:
>>>
>>> I also don't believe you need the "Notes:" addendum here. It's clear 
>>> that you're
>>> making an observation about the zImage.pseries BE and x-vof=on case.
>>
>> But only this combination needs kernel-addr=0, other images do not 
>> need that with SLOF or VOF.
> 
> 
> I mentioned about the "Notes:" string. We can remove it and leave just the
> 
> +¹ must set kernel-addr=0
> 
> 
> Since it's clear that you're making a note about that item in the table.
> 
> 
> 
>>
>>
>>>
>>> Everything else LGTM. If no one else has any comment, and you're ok 
>>> with these
>>> changes I mentioned, I can amend it myself with my R-b.
>>
>> I'll probably repost after the other patch with kernel-addr is merged 
>> into your tree. Thanks,
> 
> I already picked it (just waiting some tests to finish). But feel free to
> send a v2 if you want to play around generating the docs to see how
> your patch looks like in the finished HTML.


Nah, the changes you made are fine so I am not planning on posting 
another version. Thanks for fixing it up.


> 
> 
> Thanks,
> 
> 
> Daniel
> 
>>
>>
>>>
>>>
>>>
>>> Thanks,
>>>
>>>
>>> Daniel
>>>
>>>
>>>> +¹ must set kernel-addr=0
>>>> +
>>>>   Build directions
>>>>   ================
>>

-- 
Alexey