Re: [PATCH v2] PCI: dwc: fix enumeration end when reaching root subordinate

From: Kishon Vijay Abraham I <kishon@ti.com>
To: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>,
	Fabio Estevam <festevam@gmail.com>
Cc: Koen Vandeputte <koen.vandeputte@ncentric.com>,
	<linux-pci@vger.kernel.org>,
	Binghui Wang <wangbinghui@hisilicon.com>,
	Bjorn Helgaas <bhelgaas@google.com>,
	Jesper Nilsson <jesper.nilsson@axis.com>,
	Jianguo Sun <sunjianguo1@huawei.com>,
	Jingoo Han <jingoohan1@gmail.com>,
	Lucas Stach <l.stach@pengutronix.de>,
	Mika Westerberg <mika.westerberg@linux.intel.com>,
	Minghuan Lian <minghuan.Lian@freescale.com>,
	Mingkai Hu <mingkai.hu@freescale.com>,
	Murali Karicheri <m-karicheri2@ti.com>,
	Pratyush Anand <pratyush.anand@gmail.com>,
	Richard Zhu <hongxing.zhu@nxp.com>,
	Roy Zang <tie-fei.zang@freescale.com>,
	Shawn Guo <shawn.guo@linaro.org>,
	Stanimir Varbanov <svarbanov@mm-sol.com>,
	Thomas Petazzoni <thomas.petazzoni@free-electrons.com>,
	Xiaowei Song <songxiaowei@hisilicon.com>,
	Zhou Wang <wangzhou1@hisilicon.com>,
	Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Subject: Re: [PATCH v2] PCI: dwc: fix enumeration end when reaching root subordinate
Date: Mon, 5 Mar 2018 18:38:48 +0530	[thread overview]
Message-ID: <13b24b99-9de7-3306-a8dc-509c06af2f63@ti.com> (raw)
In-Reply-To: <20180305122440.GA29780@e107981-ln.cambridge.arm.com>

Hi Lorenzo,

On Monday 05 March 2018 05:54 PM, Lorenzo Pieralisi wrote:
> On Mon, Mar 05, 2018 at 08:55:43AM -0300, Fabio Estevam wrote:
>> Hi Lorenzo,
>>
>> On Mon, Mar 5, 2018 at 6:49 AM, Lorenzo Pieralisi
>> <lorenzo.pieralisi@arm.com> wrote:
>>
>>> It is a balance of urgency and making sure it is extensively tested -
>>> I'd prefer it to go via usual release cycle (and -next) and then it will
>>> trickle into stable kernels, let me know if that's not OK.
>>>
>>> I would understand your point if dwc maintainers were more proactive
>>> in testing their respective controllers - all of them should be affected
>>> by this fix but I have just heard from a few of them.
>>
>> We got this patch tested by: Koen, myself, Sebastian and the folks at
>> Pengutronix.
>>
>> Looks like a decent amount of testing IMHO.
> 
> IIUC you all tested the same dwc host bridge variant (ie imx6) - I want
> to understand if it works across dwc variants because this patch affects
> them all.
> 
>> In this case I would prefer that we could fix the regression into
>> 4.16-rc cycle rather than waiting until 4.17.
> 
> I will decide what to do shortly - I would really appreciate if other
> dwc host bridge maintainers (that are CC'ed) can share the testing effort.

For some reason I don't see the issues mentioned in this patch in dra7xx. The
root bus has a subordinate bus number as 01 but I'm able to read the
configuration space of the devices behind the bridge with bus number 2. I'll
have to take a closer look at what exactly happens.

root@dra7xx-evm:~# lspci -v
00:00.0 PCI bridge: Texas Instruments Multicore DSP+ARM KeyStone II SOC (rev
01) (prog-if 00 [Normal decode])
        Flags: bus master, fast devsel, latency 0, IRQ 255
        Memory at 20100000 (64-bit, non-prefetchable) [size=1M]
        Bus: primary=00, secondary=01, subordinate=01, sec-latency=0
        Capabilities: [40] Power Management version 3
        Capabilities: [50] MSI: Enable- Count=1/1 Maskable- 64bit+
        Capabilities: [70] Express Root Port (Slot-), MSI 00
        Capabilities: [100] Advanced Error Reporting
lspci: Unable to load libkmod resources: error -12

01:00.0 PCI bridge: Pericom Semiconductor Device 2304 (rev 05) (prog-if 00
[Normal decode])
        Flags: fast devsel
        Bus: primary=01, secondary=02, subordinate=04, sec-latency=0
        Capabilities: [40] Power Management version 3
        Capabilities: [5c] Vital Product Data
        Capabilities: [64] Vendor Specific Information: Len=34 <?>
        Capabilities: [b0] Subsystem: Device 0000:0000
        Capabilities: [c0] Express Upstream Port, MSI 00
        Capabilities: [100] Advanced Error Reporting
        Capabilities: [140] Virtual Channel
        Capabilities: [20c] Power Budgeting <?>
        Capabilities: [230] Latency Tolerance Reporting

02:01.0 PCI bridge: Pericom Semiconductor Device 2304 (rev 05) (prog-if 00
[Normal decode])
        Flags: fast devsel
        Bus: primary=02, secondary=03, subordinate=03, sec-latency=0
        Capabilities: [40] Power Management version 3
        Capabilities: [4c] MSI: Enable- Count=1/1 Maskable- 64bit+
        Capabilities: [64] Vendor Specific Information: Len=34 <?>
        Capabilities: [b0] Subsystem: Device 0000:0000
        Capabilities: [c0] Express Downstream Port (Slot-), MSI 00
        Capabilities: [100] Advanced Error Reporting
        Capabilities: [140] Virtual Channel
        Capabilities: [20c] Power Budgeting <?>
        Capabilities: [220] Access Control Services

02:02.0 PCI bridge: Pericom Semiconductor Device 2304 (rev 05) (prog-if 00
[Normal decode])
        Flags: fast devsel
        Bus: primary=02, secondary=04, subordinate=04, sec-latency=0
        Capabilities: [40] Power Management version 3
        Capabilities: [4c] MSI: Enable- Count=1/1 Maskable- 64bit+
        Capabilities: [64] Vendor Specific Information: Len=34 <?>
        Capabilities: [b0] Subsystem: Device 0000:0000
        Capabilities: [c0] Express Downstream Port (Slot-), MSI 00
        Capabilities: [100] Advanced Error Reporting
        Capabilities: [140] Virtual Channel
        Capabilities: [20c] Power Budgeting <?>
        Capabilities: [220] Access Control Services

Thanks
Kishon