Message-ID: <3dba0c90-9673-eadd-4f82-353ae48567a7@linux.alibaba.com>
Date:   Mon, 24 Jul 2023 20:13:26 +0800
From:   Shuai Xue <xueshuai@...ux.alibaba.com>
To:     Jonathan Cameron <Jonathan.Cameron@...wei.com>,
        Bjorn Helgaas <helgaas@...nel.org>
Cc:     chengyou@...ux.alibaba.com, kaishen@...ux.alibaba.com,
        helgaas@...nel.org, yangyicong@...wei.com, will@...nel.org,
        baolin.wang@...ux.alibaba.com, robin.murphy@....com,
        linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        linux-pci@...r.kernel.org, rdunlap@...radead.org,
        mark.rutland@....com, zhuo.song@...ux.alibaba.com
Subject: Re: [PATCH v6 0/4] drivers/perf: add Synopsys DesignWare PCIe PMU
 driver support



On 2023/7/24 17:18, Jonathan Cameron wrote:
> Really a question for Bjorn I think, but here is my 2 cents...
> 
> The problem here is that we need to do that fundamental redesign of the
> way the PCI ports drivers work.  I'm not sure there is a path to merging
> this until that is done.  The bigger problem is that I'm not sure anyone
> is actively looking at that yet.  I'd like to look at this (as I have
> the same problem for some other drivers), but it is behind various
> other things on my todo list.
> 
> Bjorn might be persuaded on a temporary solution, but that would come
> with some maintenance problems, particularly when we try to do it
> 'right' in the future.  Maybe adding another service driver would be
> a stop gap as long as we know we won't keep doing so for ever. Not sure.

Thank you for your reply, and got your point, :)

+ Bjorn


>>>> The approach used here is to separately walk the PCI topology and
>>>> register the devices.  It can 'maybe' get away with that because no
>>>> interrupts and I assume resets have no nasty impacts on it because
>>>> the device is fairly simple.  In general that's not going to work.
>>>> CXL does a similar trick (which I don't much like, but too late
>>>> now), but we've also run into the problem of how to get interrupts
>>>> if not the main driver.
>>>
>>> Yes, this is a real problem.  I think the "walk all PCI devices
>>> looking for one we like" approach is terrible because it breaks a lot
>>> of driver model assumptions (no device ID to autoload module via udev,
>>> hotplug doesn't work, etc), but we don't have a good alternative right
>>> now.
>>>
>>> I think portdrv is slightly better because at least it claims the
>>> device in the usual way and gives a way for service drivers to
>>> register with it.  But I don't really like that either because it
>>> created a new weird /sys/bus/pci_express hierarchy full of these
>>> sub-devices that aren't really devices, and it doesn't solve the
>>> module load and hotplug issues.
>>>
>>> I would like to have portdrv be completely built into the PCI core and
>>> not claim Root Ports or Switch Ports.  Then those devices would be
>>> available via the usual driver model for driver loading and binding
>>> and for hotplug.
>>
>> Let me see if I understand this correctly as I can think of a few options
>> that perhaps are inline with what you are thinking.
>>
>> 1) All the portdrv stuff converted to normal PCI core helper functions
>>    that a driver bound to the struct pci_dev can use.
>> 2) Driver core itself provides a bunch of extra devices alongside the
>>    struct pci_dev one to which additional drivers can bind? - so kind
>>    of portdrv handling, but squashed into the PCI device topology?
>> 3) Have portdrv operated under the hood, so all the services etc that
>>    it provides don't require a driver to be bound at all.  Then
>>    allow usual VID/DID based driver binding.
>>
>> If 1 - we are going to run into class device restrictions and that will
>> just move where we have to handle the potential vendor specific parts.
>> We probably don't want that to be a hydra with all the functionality
>> and lookups etc driven from there, so do we end up with sub devices
>> of that new PCI port driver with a discover method based on either
>> vsec + VID or DVSEC with devices created under the main pci_dev.
>> That would have to include nastiness around interrupt discovery for
>> those sub devices. So ends up roughly like port_drv.
>>
>> I don't think 2 solves anything.
>>
>> For 3 - interrupts and ownership of facilities is going to be tricky
>> as initially those need to be owned by the PCI core (no device driver bound)
>> and then I guess handed off to the driver once it shows up?  Maybe that
>> driver should call a pci_claim_port() that gives it control of everything
>> and pci_release_port() that hands it all back to the core.  That seems
>> racey.
>
> Yes, 3 is the option I want to explore.  That's what we already do for
> things like ASPM.  Agreed, interrupts is a potential issue.  I think
> the architected parts of config space should be implicitly owned by
> the PCI core, with interfaces à la pci_disable_link_state() if drivers
> need them.
>
> Bjorn
> https://lore.kernel.org/lkml/ZGUAWxoEngmqFcLJ@bhelgaas/

@Bjorn Is there a path to merging this patch set until your explore is done?
And are you still actively looking at that yet?

I am not quite familiar with PCI core, but I would like to help work on that.

Thank you.

Best Regards,
Shuai

Powered by blists - more mailing lists