Re: [PATCH v4 1/5] usb: host: export symbols for xhci-exynos to use xhci hooks

From: Jung Daehwan <dh10.jung@samsung.com>
To: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mathias Nyman <mathias.nyman@intel.com>,
	"open list:USB XHCI DRIVER" <linux-usb@vger.kernel.org>,
	open list <linux-kernel@vger.kernel.org>,
	Howard Yen <howardyen@google.com>,
	Jack Pham <jackp@codeaurora.org>, Puma Hsu <pumahsu@google.com>,
	"J . Avila" <elavila@google.com>,
	sc.suh@samsung.com,
	Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Subject: Re: [PATCH v4 1/5] usb: host: export symbols for xhci-exynos to use xhci hooks
Date: Thu, 28 Apr 2022 12:30:20 +0900	[thread overview]
Message-ID: <20220428033020.GD151827@ubuntu> (raw)
In-Reply-To: <Ymg8h6TiH4RpL93t@kroah.com>

[-- Attachment #1: Type: text/plain, Size: 3055 bytes --]

On Tue, Apr 26, 2022 at 08:40:07PM +0200, Greg Kroah-Hartman wrote:
> On Tue, Apr 26, 2022 at 12:31:33PM +0200, Greg Kroah-Hartman wrote:
> > On Tue, Apr 26, 2022 at 07:27:09PM +0900, Jung Daehwan wrote:
> > > On Tue, Apr 26, 2022 at 11:54:16AM +0200, Greg Kroah-Hartman wrote:
> > > > On Tue, Apr 26, 2022 at 06:18:44PM +0900, Daehwan Jung wrote:
> > > > > Export symbols for xhci hooks usage:
> > > > >     xhci_get_slot_ctx
> > > > >     xhci_get_endpoint_address
> > > > >     - Allow xhci hook to get ep_ctx from the xhci_container_ctx for
> > > > >       getting the ep_ctx information to know which ep is offloading and
> > > > >       comparing the context in remote subsystem memory if needed.
> > > > > 
> > > > >     xhci_ring_alloc
> > > > >     - Allow xhci hook to allocate vendor specific ring. Vendors could
> > > > >       alloc additional event ring.
> > > > > 
> > > > >     xhci_trb_virt_to_dma
> > > > >     - Used to retrieve the DMA address of vendor specific ring. Vendors
> > > > >       could get dequeue address of event ring.
> > > > > 
> > > > >     xhci_segment_free
> > > > >     xhci_link_segments
> > > > >     - Allow xhci hook to handle vendor specific segment. Vendors could
> > > > >       directly free or link segments of vendor specific ring.
> > > > > 
> > > > >     xhci_initialize_ring_info
> > > > >     - Allow xhci hook to initialize vendor specific ring.
> > > > > 
> > > > >     xhci_check_trb_in_td_math
> > > > >     - Allow xhci hook to Check TRB math for validation. Vendors could
> > > > >       check trb when allocating vendor specific ring.
> > > > > 
> > > > >     xhci_address_device
> > > > >     - Allow override to give configuration info to Co-processor.
> > > > > 
> > > > >     xhci_bus_suspend
> > > > >     xhci_bus_resume
> > > > >     - Allow override of suspend and resume for power scenario.
> > > > > 
> > > > >     xhci_remove_stream_mapping
> > > > >     - Allow to xhci hook to remove stream mapping. Vendors need to do it
> > > > >       when free-ing vendor specific ring if it's stream type.
> > > > > 
> > > > 
> > > > For the static functions that you are now exporting, they need to have
> > > > their functions declared in a .h file.  If not, you now get warnings
> > > > when you run sparse after applying this commit, right?
> > > > 
> > > > thanks,
> > > > 
> > > > greg k-h
> > > > 
> > > 
> > > I didn't get warnings in my environment. I tested to build default config +
> > > xhci_exynos by make command. Is there compile option to check it?
> > 
> > make "C=1"
> 
> And now 0-day reported it as well:
> 	https://protect2.fireeye.com/v1/url?k=eb7c3909-8af72c33-eb7db246-74fe4860008a-ea28471ebd7c5af8&q=1&e=c40c7ad3-ab59-48ea-92a0-19c7ed0d5fd9&u=https%3A%2F%2Flore.kernel.org%2Fr%2F202204262306.mzMIKFKO-lkp%40intel.com
> 
>
I agree they should be added in header instead manually including xhci-exynos.
I will modify it on next submission.

I also don't get warnings in my environment when I run sparse.
Let me check my environment again.

Best Regards,
Jung Daehwan

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]