| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Tue, 10 Aug 2021 11:52:34 -0600
From: Rob Herring <robh@...nel.org>
To: Mauro Carvalho Chehab <mchehab+huawei@...nel.org>
Cc: Linuxarm <linuxarm@...wei.com>, mauro.chehab@...wei.com,
Binghui Wang <wangbinghui@...ilicon.com>,
Gustavo Pimentel <gustavo.pimentel@...opsys.com>,
Jingoo Han <jingoohan1@...il.com>,
Xiaowei Song <songxiaowei@...ilicon.com>,
devicetree@...r.kernel.org,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
PCI <linux-pci@...r.kernel.org>, linux-phy@...ts.infradead.org
Subject: Re: [PATCH v3 0/4] DT schema changes for HiKey970 PCIe hardware to work
On Tue, Aug 10, 2021 at 11:13 AM Rob Herring <robh@...nel.org> wrote:
>
> On Tue, Aug 10, 2021 at 8:21 AM Mauro Carvalho Chehab
> <mchehab+huawei@...nel.org> wrote:
> >
> > Em Tue, 10 Aug 2021 07:44:50 -0600
> > Rob Herring <robh@...nel.org> escreveu:
> >
> > > On Tue, Aug 10, 2021 at 3:42 AM Mauro Carvalho Chehab
> > > <mchehab+huawei@...nel.org> wrote:
> > > >
> > > > Em Fri, 6 Aug 2021 10:23:35 -0600
> > > > Rob Herring <robh@...nel.org> escreveu:
> > > >
> > > > > On Thu, Aug 5, 2021 at 1:58 AM Mauro Carvalho Chehab
> > > > > <mchehab+huawei@...nel.org> wrote:
> > > > > >
> > > > > > Em Thu, 5 Aug 2021 09:46:12 +0200
> > > > > > Mauro Carvalho Chehab <mchehab+huawei@...nel.org> escreveu:
> > > > > >
> > > > > > > Em Wed, 4 Aug 2021 10:28:53 -0600
> > > > > > > Rob Herring <robh@...nel.org> escreveu:
> > > > > > >
> > > > > > > > On Wed, Aug 04, 2021 at 08:50:45AM +0200, Mauro Carvalho Chehab wrote:
> > > > > > > > > Em Tue, 3 Aug 2021 16:11:42 -0600
> > > > > > > > > Rob Herring <robh+dt@...nel.org> escreveu:
> > > > > > > > >
> > > > > > > > > > On Mon, Aug 2, 2021 at 10:39 PM Mauro Carvalho Chehab
> > > > > > > > > > <mchehab+huawei@...nel.org> wrote:
> > > > > > > > > > >
> > > > > > > > > > > Hi Rob,
> > > > > > > > > > >
> > > > > > > > > > > That's the third version of the DT bindings for Kirin 970 PCIE and its
> > > > > > > > > > > corresponding PHY.
> > > > > > > > > > >
> > > > > > > > > > > It is identical to v2, except by:
> > > > > > > > > > > - pcie@7,0 { // Lane 7: Ethernet
> > > > > > > > > > > + pcie@7,0 { // Lane 6: Ethernet
> > > > > > > > > >
> > > > > > > > > > Can you check whether you have DT node links in sysfs for the PCI
> > > > > > > > > > devices? If you don't, then something is wrong still in the topology
> > > > > > > > > > or the PCI core is failing to set the DT node pointer in struct
> > > > > > > > > > device. Though you don't rely on that currently, we want the topology
> > > > > > > > > > to match. It's possible this never worked on arm/arm64 as mainly
> > > > > > > > > > powerpc relied on this.
> > > > > > > > > >
> > > > > > > > > > I'd like some way to validate the DT matches the PCI topology. We
> > > > > > > > > > could have a tool that generates the DT structure based on the PCI
> > > > > > > > > > topology.
> > > > > > > > >
> > > > > > > > > The of_node node link is on those places:
> > > > > > > > >
> > > > > > > > > $ find /sys/devices/platform/soc/f4000000.pcie/ -name of_node
> > > > > > > > > /sys/devices/platform/soc/f4000000.pcie/of_node
> > > > > > > > > /sys/devices/platform/soc/f4000000.pcie/pci0000:00/0000:00:00.0/of_node
> > > > > > > > > /sys/devices/platform/soc/f4000000.pcie/pci0000:00/0000:00:00.0/pci_bus/0000:01/of_node
> > > > > > > > > /sys/devices/platform/soc/f4000000.pcie/pci0000:00/pci_bus/0000:00/of_node
> > > > > > > >
> > > > > > > > Looks like we're missing some...
> > > > > > > >
> > > > > > > > It's not immediately obvious to me what's wrong here. Only the root
> > > > > > > > bus is getting it's DT node set. The relevant code is pci_scan_device(),
> > > > > > > > pci_set_of_node() and pci_set_bus_of_node(). Give me a few days to try
> > > > > > > > to reproduce and debug it.
> > > > > > >
> > > > > > > I added a printk on both pci_set_*of_node() functions:
> > > > > > >
> > > > > > > [ 4.872991] (null): pci_set_bus_of_node: of_node: /soc/pcie@...00000
> > > > > > > [ 4.913806] (null): pci_set_of_node: of_node: /soc/pcie@...00000
> > > > > > > [ 4.978102] pci_bus 0000:01: pci_set_bus_of_node: of_node: /soc/pcie@...00000/pcie@0,0
> > > > > > > [ 4.990622] (null): pci_set_of_node: of_node: /soc/pcie@...00000/pcie@0,0
> > > > > > > [ 5.052383] pci_bus 0000:02: pci_set_bus_of_node: of_node: (null)
> > > > > > > [ 5.059263] (null): pci_set_of_node: of_node: (null)
> > > > > > > [ 5.085552] (null): pci_set_of_node: of_node: (null)
> > > > > > > [ 5.112073] (null): pci_set_of_node: of_node: (null)
> > > > > > > [ 5.138320] (null): pci_set_of_node: of_node: (null)
> > > > > > > [ 5.164673] (null): pci_set_of_node: of_node: (null)
> > > > > > > [ 5.233759] pci_bus 0000:03: pci_set_bus_of_node: of_node: (null)
> > > > > > > [ 5.240539] (null): pci_set_of_node: of_node: (null)
> > > > > > > [ 5.310545] pci_bus 0000:04: pci_set_bus_of_node: of_node: (null)
> > > > > > > [ 5.324719] pci_bus 0000:05: pci_set_bus_of_node: of_node: (null)
> > > > > > > [ 5.338914] pci_bus 0000:06: pci_set_bus_of_node: of_node: (null)
> > > > > > > [ 5.345516] (null): pci_set_of_node: of_node: (null)
> > > > > > > [ 5.415795] pci_bus 0000:07: pci_set_bus_of_node: of_node: (null)
> > > > > >
> > > > > > The enclosed patch makes the above a clearer:
> > > > > >
> > > > > > [ 4.800975] (null): pci_set_bus_of_node: of_node: /soc/pcie@...00000
> > > > > > [ 4.855983] pci 0000:00:00.0: pci_set_of_node: of_node: /soc/pcie@...00000
> > > > > > [ 4.879169] pci_bus 0000:01: pci_set_bus_of_node: of_node: /soc/pcie@...00000/pcie@0,0
> > > > > > [ 4.900602] pci 0000:01:00.0: pci_set_of_node: of_node: /soc/pcie@...00000/pcie@0,0
> > > > > > [ 4.953086] pci_bus 0000:02: pci_set_bus_of_node: of_node: (null)
> > > > >
> > > > > I believe the issue is we need another bridge node in the DT
> > > > > hierarchy. What we have is:
> > > > >
> > > > > Bus 0 is node /soc/pcie@...00000
> > > > > Bus 1 is device 0 on bus 0 is node /soc/pcie@...00000/pcie@0,0
> > > > > Bus 2 is device 0 on bus 1 in node ... whoops, there's no device 0
> > > > > under /soc/pcie@...00000/pcie@0,0
> > > > >
> > > > > So we need the hierarchy to be: /soc/pcie@...00000/pcie@...cie@...cie@{1,5,7}
> > > >
> > > > Adding a child pcie@0 produces the following output from my debug
> > > > patches:
> > >
> > > You removed your changes to the PCI code other than the debug print?
> >
> > Yes.
> >
> > > >
> > > > [ 4.984278] (null): pci_set_bus_of_node: of_node: /soc/pcie@...00000
> > > > [ 5.042992] pci 0000:00:00.0: pci_set_of_node: of_node: /soc/pcie@...00000
> > > > [ 5.083738] pci_bus 0000:01: pci_set_bus_of_node: of_node: /soc/pcie@...00000/pcie@0,0
> > > > [ 5.124377] pci 0000:01:00.0: pci_set_of_node: of_node: /soc/pcie@...00000/pcie@0,0
> > > > [ 5.168395] pci_bus 0000:02: pci_set_bus_of_node: of_node: /soc/pcie@...00000/pcie@0,0/pcie@0,0
> > > > [ 5.200719] pci 0000:02:01.0: pci_set_of_node: of_node: /soc/pcie@...00000/pcie@0,0/pcie@0,0
> > >
> > > This should not happen. The devfn doesn't match.
> > >
> > > > [ 5.247777] pci 0000:02:04.0: pci_set_of_node: of_node: /soc/pcie@...00000/pcie@0,0/pcie@0,0
> > > > [ 5.276768] pci 0000:02:05.0: pci_set_of_node: of_node: /soc/pcie@...00000/pcie@0,0/pcie@0,0
> > > > [ 5.305018] pci 0000:02:07.0: pci_set_of_node: of_node: /soc/pcie@...00000/pcie@0,0/pcie@0,0
> > > > [ 5.333093] pci 0000:02:09.0: pci_set_of_node: of_node: /soc/pcie@...00000/pcie@0,0/pcie@0,0
> > > > [ 5.395620] pci_bus 0000:03: pci_set_bus_of_node: of_node: (null)
> > > > [ 5.416333] pci 0000:03:00.0: pci_set_of_node: of_node: (null)
> > > > [ 5.451353] pci_bus 0000:04: pci_set_bus_of_node: of_node: (null)
> > > > [ 5.473970] pci_bus 0000:05: pci_set_bus_of_node: of_node: (null)
> > > > [ 5.487765] pci_bus 0000:06: pci_set_bus_of_node: of_node: (null)
> > > > [ 5.530219] pci 0000:06:00.0: pci_set_of_node: of_node: (null)
> > > > [ 5.560896] pci_bus 0000:07: pci_set_bus_of_node: of_node: (null)
> > > >
> > > > It produces the following sysfs nodes:
> > > >
> > > > $ find /sys/devices/platform/soc/f4000000.pcie/ -name of_node
> > > > /sys/devices/platform/soc/f4000000.pcie/of_node
> > > > /sys/devices/platform/soc/f4000000.pcie/pci0000:00/0000:00:00.0/of_node
> > > > /sys/devices/platform/soc/f4000000.pcie/pci0000:00/0000:00:00.0/0000:01:00.0/of_node
> > > > /sys/devices/platform/soc/f4000000.pcie/pci0000:00/0000:00:00.0/0000:01:00.0/pci_bus/0000:02/of_node
> > > > /sys/devices/platform/soc/f4000000.pcie/pci0000:00/0000:00:00.0/pci_bus/0000:01/of_node
> > > > /sys/devices/platform/soc/f4000000.pcie/pci0000:00/pci_bus/0000:00/of_node
> > > >
> > > >
> > > > I'm enclosing the DT schema I'm using.
> > > >
> > > >
> > > >
> > > > Thanks,
> > > > Mauro
> > > >
> > > > ---
> > > >
> > > > pcie@...00000 {
> > > > compatible = "hisilicon,kirin970-pcie";
> > > > reg = <0x0 0xf4000000 0x0 0x1000000>,
> > > > <0x0 0xfc180000 0x0 0x1000>,
> > > > <0x0 0xf5000000 0x0 0x2000>;
> > > > reg-names = "dbi", "apb", "config";
> > > > bus-range = <0x00 0xff>;
> > > > #address-cells = <3>;
> > > > #size-cells = <2>;
> > > > device_type = "pci";
> > > > phys = <&pcie_phy>;
> > > > ranges = <0x02000000 0x0 0x00000000
> > > > 0x0 0xf6000000
> > > > 0x0 0x02000000>;
> > > > num-lanes = <1>;
> > > > #interrupt-cells = <1>;
> > > > interrupts = <GIC_SPI 283 IRQ_TYPE_LEVEL_HIGH>;
> > > > interrupt-names = "msi";
> > > > interrupt-map-mask = <0 0 0 7>;
> > > > interrupt-map = <0x0 0 0 1
> > > > &gic GIC_SPI 282 IRQ_TYPE_LEVEL_HIGH>,
> > > > <0x0 0 0 2
> > > > &gic GIC_SPI 283 IRQ_TYPE_LEVEL_HIGH>,
> > > > <0x0 0 0 3
> > > > &gic GIC_SPI 284 IRQ_TYPE_LEVEL_HIGH>,
> > > > <0x0 0 0 4
> > > > &gic GIC_SPI 285 IRQ_TYPE_LEVEL_HIGH>;
> > > > reset-gpios = <&gpio7 0 0>;
> > > > hisilicon,clken-gpios = <&gpio27 3 0>, <&gpio17 0 0>,
> > > > <&gpio20 6 0>;
> > > > pcie@0,0 { // Lane 0: PCIe switch: Bus 1, Device 0
> > > > reg = <0x80 0 0 0 0>;
> > >
> > > s/0x80/0/
> > >
> > > > compatible = "pciclass,0604";
> > > > device_type = "pci";
> > > > #address-cells = <3>;
> > > > #size-cells = <2>;
> > > > ranges;
> > > > bus-range = <0x01 0xff>;
> > > > msi-parent = <&its_pcie>;
> > > >
> > > > pcie@0,0 { // Lane 0: upstream
> > > > reg = <0x010000 0 0 0 0>;
> > >
> > > While technically correct having the bus# in the address, that doesn't
> > > work for FDT since we don't know the bus assignment. So we should just
> > > use 0.
> >
> > Using 0 causes DTB compilation to produce a warning, due to the
> > bus-range.
What's the warning? 'bus-range' should be optional.
> > Without the bus-range, there will be runtime warnings,
> > as this will be assigned as bus 1.
>
> Okay, that might be something we need to fix.
Actually, I don't see how there's a problem. First, the only place we
read 'bus-range' is of_pci_parse_bus_range() and that's only called
for the host bridge. Any other occurrence of 'bus-range' should be
ignored. The only place we read 'reg' is of_pci_get_devfn() and we
mask just the devfn bits.
> [ 4.992572] pci_bus 0000:00: root bus resource [bus 00-01]
I don't see any way this can happen other than pcie@...00000 node
containing 'bus-range = <0 1>;'. It comes from
pci_host_bridge.windows.
Rob
Powered by blists - more mailing lists