In preparation to adding more flags to configuration data, use pointer
to struct qcom_pcie_cfg directly inside struct qcom_pcie, rather than
duplicating all its fields. This would save us from the boilerplate code
that just copies flag values from one struct to another one.

Link: https://lore.kernel.org/r/20220223101435.447839-3-dmitry.baryshkov@linaro.orgSigned-off-by: NDmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Reviewed-by: NBjorn Andersson <bjorn.andersson@linaro.org>
Acked-by: NStanimir Varbanov <svarbanov@mm-sol.com>

find_first{,_zero}_bit is a more effective analogue of 'next' version if
start == 0. This patch replaces 'next' with 'first' where things look
trivial.
Signed-off-by: NYury Norov <yury.norov@gmail.com>
Tested-by: NWolfram Sang <wsa+renesas@sang-engineering.com>

Get "syscon" pcie_mode and pcie_id offset from the argument of
"ti,syscon-pcie-id" and "ti,syscon-pcie-mode" phandle respectively.
Previously a subnode to "syscon" node was added which has the
exact memory mapped address of pcie_mode and pcie_id but now the
offset of pcie_mode and pcie_id within "syscon" is now being passed
as argument to "ti,syscon-pcie-id" and "ti,syscon-pcie-mode" phandle.

If the offset is not provided in "ti,syscon-pcie-id"/"ti,syscon-pcie-mode",
the full memory mapped address of pcie_ctrl is used in order to maintain
old DT compatibility.

Similar change for J721E is as discussed in [1]

[1] -> http://lore.kernel.org/r/CAL_JsqKiUcO76bo1GoepWM1TusJWoty_BRy2hFSgtEVMqtrvvQ@mail.gmail.com

Link: https://lore.kernel.org/r/20211126083119.16570-3-kishon@ti.comSigned-off-by: NKishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>

The current Layerscape PCIe driver directly uses the physical layer
LTSSM code to check the link-up state, which treats the > L0 states
as link-up. This is not correct, since there is not explicit map
between link-up state and LTSSM. So this patch changes to use the
DWC common link-up check function.

Link: https://lore.kernel.org/r/20211224094000.8513-1-Zhiqiang.Hou@nxp.comSigned-off-by: NHou Zhiqiang <Zhiqiang.Hou@nxp.com>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Reviewed-by: NRob Herring <robh@kernel.org>

The only usage of pci_ep_ops is to assign its address to the ops field
in the dw_pcie_ep struct which is a pointer to const struct dw_pcie_ep_ops.
Make it const to allow the compiler to put it in read-only memory.

Link: https://lore.kernel.org/r/20211204220316.88655-1-rikard.falkeborn@gmail.comSigned-off-by: NRikard Falkeborn <rikard.falkeborn@gmail.com>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Reviewed-by: NKrzysztof Wilczyński <kw@linux.com>

The sparse checker complains about converting pointers between address
spaces.  We correctly stored an __iomem pointer in struct spear13xx_pcie,
but discarded the __iomem when extracting app_base, causing one warning.
Then we passed the non-__iomem pointer to writel(), which expects an
__iomem pointer, causing another warning.

Add the appropriate annotations.

The sparse warnings look like this:

  $ make C=2 drivers/pci/controller/
  drivers/pci/controller/dwc/pcie-spear13xx.c:72:54: warning: incorrect type in initializer (different address spaces)
  drivers/pci/controller/dwc/pcie-spear13xx.c:72:54:    expected struct pcie_app_reg *app_reg
  drivers/pci/controller/dwc/pcie-spear13xx.c:72:54:    got void [noderef] __iomem *app_base
  drivers/pci/controller/dwc/pcie-spear13xx.c:78:26: warning: incorrect type in argument 2 (different address spaces)
  drivers/pci/controller/dwc/pcie-spear13xx.c:78:26:    expected void volatile [noderef] __iomem *addr
  drivers/pci/controller/dwc/pcie-spear13xx.c:78:26:    got unsigned int *

Link: https://lore.kernel.org/r/20211223213749.1314142-3-helgaas@kernel.orgSigned-off-by: NBjorn Helgaas <bhelgaas@google.com>
Cc: Pratyush Anand <pratyush.anand@gmail.com>

The sparse checker complains about converting pointers between address
spaces.  The pci_config_window.priv pointer is a generic void *, but
hisi_pcie_map_bus() needs a void __iomem *.

This isn't a problem in other drivers because they store the __iomem
pointer in a driver struct.  Add a trivial struct hisi_pcie to avoid the
warning.

The sparse warning looks like this:

  $ make C=2 drivers/pci/controller/
  drivers/pci/controller/dwc/pcie-hisi.c:61:37: warning: incorrect type in initializer (different address spaces)
  drivers/pci/controller/dwc/pcie-hisi.c:61:37:    expected void [noderef] __iomem *reg_base
  drivers/pci/controller/dwc/pcie-hisi.c:61:37:    got void *priv

Link: https://lore.kernel.org/r/20211223213749.1314142-2-helgaas@kernel.orgSigned-off-by: NBjorn Helgaas <bhelgaas@google.com>
Cc: Zhou Wang <wangzhou1@hisilicon.com>

Rename struct uniphier_pcie_priv to uniphier_pcie to match the convention
of <driver>_pcie. No functional change intended.

[bhelgaas: drop extraneous uniphier_pcie_irq_ack() from patch]
Link: https://lore.kernel.org/r/20211223011054.1227810-21-helgaas@kernel.orgSigned-off-by: NFan Fei <ffclaire1224@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Cc: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>

Rename tegra_pcie_dw to tegra194_pcie to match the convention of
<driver>_pcie. No functional change intended.

[bhelgaas: rename functions similarly]
Link: https://lore.kernel.org/r/20211223011054.1227810-20-helgaas@kernel.orgSigned-off-by: NFan Fei <ffclaire1224@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: Jonathan Hunter <jonathanh@nvidia.com>

Rename struct intel_pcie_port to intel_pcie to match the convention of
<driver>_pcie. No functional change intended.

Link: https://lore.kernel.org/r/20211223011054.1227810-11-helgaas@kernel.orgSigned-off-by: NFan Fei <ffclaire1224@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Cc: Rahul Tanwar <rtanwar@maxlinear.com>

The kirin driver only needs the device data, not the whole struct
of_device_id.  Use of_device_get_match_data() instead of of_match_device().
No functional change intended.

[bhelgaas: commit log]
Link: https://lore.kernel.org/r/20211223011054.1227810-8-helgaas@kernel.orgSigned-off-by: NFan Fei <ffclaire1224@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Cc: Xiaowei Song <songxiaowei@hisilicon.com>
Cc: Binghui Wang <wangbinghui@hisilicon.com>

The keystone driver only needs the device data, not the whole struct
of_device_id.  Use of_device_get_match_data() instead of of_match_device().
No functional change intended.

[bhelgaas: commit log]
Link: https://lore.kernel.org/r/20211223011054.1227810-7-helgaas@kernel.orgSigned-off-by: NFan Fei <ffclaire1224@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Cc: Kishon Vijay Abraham I <kishon@ti.com>

The dra7xx driver only needs the device data, not the whole struct
of_device_id.  Use of_device_get_match_data() instead of of_match_device().
No functional change intended.

[bhelgaas: commit log]
Link: https://lore.kernel.org/r/20211223011054.1227810-6-helgaas@kernel.orgSigned-off-by: NFan Fei <ffclaire1224@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Cc: Kishon Vijay Abraham I <kishon@ti.com>

The designware-plat driver only needs the device data, not the whole struct
of_device_id.  Use of_device_get_match_data() instead of of_match_device().
No functional change intended.

[bhelgaas: commit log]
Link: https://lore.kernel.org/r/20211223011054.1227810-5-helgaas@kernel.orgSigned-off-by: NFan Fei <ffclaire1224@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Cc: Jingoo Han <jingoohan1@gmail.com>
Cc: Gustavo Pimentel <gustavo.pimentel@synopsys.com>

The artpec6 driver only needs the device data, not the whole struct
of_device_id.  Use of_device_get_match_data() instead of of_match_device().
No functional change intended.

[bhelgaas: commit log]
Link: https://lore.kernel.org/r/20211223011054.1227810-3-helgaas@kernel.orgSigned-off-by: NFan Fei <ffclaire1224@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NJesper Nilsson <jesper.nilsson@axis.com>

i.MX8MM PCIe works mostly like the i.MX8MQ one, but has a different PHY
and allows to output the internal PHY reference clock via the refclk pad.
Add the i.MX8MM PCIe support based on the standalone PHY driver.

Link: https://lore.kernel.org/r/1640312885-31142-2-git-send-email-hongxing.zhu@nxp.comTested-by: NMarcel Ziswiler <marcel.ziswiler@toradex.com>
Tested-by: NTim Harvey <tharvey@gateworks.com>
Signed-off-by: NRichard Zhu <hongxing.zhu@nxp.com>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Reviewed-by: NTim Harvey <tharvey@gateworks.com>

There is no need to call the dev_err() function directly to print a
custom message when handling an error from either the platform_get_irq()
or platform_get_irq_byname() functions as both are going to display an
appropriate error message in case of a failure.

This change is as per suggestions from Coccinelle, e.g.,
  drivers/pci/controller/dwc/pcie-qcom-ep.c:556:2-9: line 556 is redundant because platform_get_irq() already prints an error

Related:
  https://lore.kernel.org/all/20210310131913.2802385-1-kw@linux.com/
  https://lore.kernel.org/all/20200802142601.1635926-1-kw@linux.com/

Link: https://lore.kernel.org/r/20211027112931.37182-1-kw@linux.comSigned-off-by: NKrzysztof Wilczyński <kw@linux.com>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>

cpu_to_be16() returns __be16 value but the driver uses u16 and that's
incorrect. Fix it by using __be16 as the data type of bdf_be variable.

The issue was spotted by the below sparse warning:

sparse warnings: (new ones prefixed by >>)
>> drivers/pci/controller/dwc/pcie-qcom.c:1305:30: sparse: sparse: incorrect type in initializer (different base types) @@     expected unsigned short [usertype] bdf_be @@     got restricted __be16 [usertype] @@
   drivers/pci/controller/dwc/pcie-qcom.c:1305:30: sparse:     expected unsigned short [usertype] bdf_be
   drivers/pci/controller/dwc/pcie-qcom.c:1305:30: sparse:     got restricted __be16 [usertype]

Link: https://lore.kernel.org/r/20211130080924.266116-1-manivannan.sadhasivam@linaro.orgReported-by: Nkernel test robot <lkp@intel.com>
Signed-off-by: NManivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Reviewed-by: NKrzysztof Wilczyński <kw@linux.com>

cgroup.h (therefore swap.h, therefore half of the universe)
includes bpf.h which in turn includes module.h and slab.h.
Since we're about to get rid of that dependency we need
to clean things up.

v2: drop the cpu.h include from cacheinfo.h, it's not necessary
and it makes riscv sensitive to ordering of include files.
Signed-off-by: NJakub Kicinski <kuba@kernel.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Acked-by: NKrzysztof Wilczyński <kw@linux.com>
Acked-by: NPeter Chen <peter.chen@kernel.org>
Acked-by: NSeongJae Park <sj@kernel.org>
Acked-by: NJani Nikula <jani.nikula@intel.com>
Acked-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/all/20211120035253.72074-1-kuba@kernel.org/  # v1
Link: https://lore.kernel.org/all/20211120165528.197359-1-kuba@kernel.org/ # cacheinfo discussion
Link: https://lore.kernel.org/bpf/20211202203400.1208663-1-kuba@kernel.org

On imx6 and perhaps others when pcie probes you get a:
imx6q-pcie 33800000.pcie: invalid resource

This occurs because the atu is not specified in the DT and as such it
should not be remapped.

Link: https://lore.kernel.org/r/20211101180243.23761-1-tharvey@gateworks.com
Fixes: 281f1f99 ("PCI: dwc: Detect number of iATU windows")
Signed-off-by: NTim Harvey <tharvey@gateworks.com>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Reviewed-by: NRob Herring <robh@kernel.org>
Acked-by: NRichard Zhu <hongxing.zhu@nxp.com>
Cc: Richard Zhu <hongxing.zhu@nxp.com>

If 'of_device_get_match_data()' fails, previous 'pm_runtime_get_sync()/
pm_runtime_enable()' should be undone.

To fix it, the easiest is to move this block of code before the memory
allocations and the pm_runtime_xxx calls.

Link: https://lore.kernel.org/r/4d03c636193f64907c8dacb17fa71ed05fd5f60c.1636220582.git.christophe.jaillet@wanadoo.fr
Fixes: b89ff410 ("PCI: qcom: Replace ops with struct pcie_cfg in pcie match data")
Signed-off-by: NChristophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Reviewed-by: NStephen Boyd <swboyd@chromium.org>

Include PCI_ERROR_RESPONSE along with 0xffffffff in the comment about
identifying config read errors. This makes checks for config read errors
easier to find. Comment change only.

Link: https://lore.kernel.org/r/6ae6b071d92052dc511407513e2a7c0035aff9e7.1637243717.git.naveennaidu479@gmail.comSigned-off-by: NNaveen Naidu <naveennaidu479@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>

If config pci_ops.read() methods return failure, the PCI_OP_READ() and
PCI_USER_READ_CONFIG() wrappers use PCI_SET_ERROR_RESPONSE() to set the
data value, so there's no need to set it in the pci_ops.read() methods
themselves.

Drop the unnecessary data value fabrication when pci_ops.read() fails.

Link: https://lore.kernel.org/r/f87e22bc09a471d2cf15ad05dfd6432f57739aed.1637243717.git.naveennaidu479@gmail.comSigned-off-by: NNaveen Naidu <naveennaidu479@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>

If config pci_ops.read() methods return failure, the PCI_OP_READ() and
PCI_USER_READ_CONFIG() wrappers use PCI_SET_ERROR_RESPONSE() to set the
data value, so there's no need to set it in the pci_ops.read() methods
themselves.

Drop the unnecessary data value fabrication when pci_ops.read() fails.

Link: https://lore.kernel.org/r/7da7ea760abc5f85cad6e9b0d3e59eebd93f50d3.1637243717.git.naveennaidu479@gmail.comSigned-off-by: NNaveen Naidu <naveennaidu479@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>

If config pci_ops.read() methods return failure, the PCI_OP_READ() and
PCI_USER_READ_CONFIG() wrappers use PCI_SET_ERROR_RESPONSE() to set the
data value, so there's no need to set it in the pci_ops.read() methods
themselves.

Drop the unnecessary data value fabrication when pci_ops.read() fails.

Link: https://lore.kernel.org/r/765c2f27d85fcee4eb137023dc0a8f864eee9e36.1637243717.git.naveennaidu479@gmail.comSigned-off-by: NNaveen Naidu <naveennaidu479@gmail.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>

Now that everything is in place at the poweroff sequence, this driver can
use module_platform_driver(), which allows it to be removed.

Link: https://lore.kernel.org/r/53b40494252444a9b830827922c4e3a301b8f863.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>

The logic under .remove ops is missing a call to dw_pcie_host_deinit(). Add
it, in order to allow the DWC core to be properly cleaned up.

Link: https://lore.kernel.org/r/838621e1c84ebaac153ccd9c36ea5e1254c61ead.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>

The logic at kirin_pcie_gpio_request() enables some clkreq GPIO lines.
Disable them during power-off.

Link: https://lore.kernel.org/r/f403e590843de1a581cade2d534d34715706f54e.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>

Instead of having two copies of the same logic, place the power-off logic
in a separate function.

No functional changes.

Link: https://lore.kernel.org/r/64f6e8da3e5fff38b6c8fcb208ace46efe6555bb.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>

In order to prepare for module unload, add a power_off method for HiKey
960.

Link: https://lore.kernel.org/r/b095818b0d7fadae4cae200f481caf7a66e61fb4.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>

There's nothing preventing this driver from being loaded as a module.
Change its config from bool to tristate.

Link: https://lore.kernel.org/r/b5e7cfe9df09b492750bd6db0f0c911eaae8c2d4.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>

This driver misses the MODULE_* macros. Add them.

Link: https://lore.kernel.org/r/f7a951d0c2009f5765214fc2e83e24cf41585023.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>

Now that everything is in place, add a compatible for Kirin 970.

Link: https://lore.kernel.org/r/ac8c730c0300b90d96bdaaf387d458d8949241a9.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>

On HiKey970, there's a PEX 8606 PCI bridge on its PHY with 6 lanes. Only 4
lanes are connected:

  lane 0 - connected to Kirin 970 (upstream)
  lane 4 - M.2 slot
  lane 5 - mini PCIe slot
  lane 6 - on-board Ethernet controller

Each lane has its own PERST# GPIO pin and needs a clock request.

Add support to parse a DT schema containing the above data.

HiKey 970 requires a little more waiting time for the PCI bridge - which is
outside the SoC - to finish the PERST# reset, and then initialize the eye
diagram.

Increase the waiting time for the PERST# signals accordingly.

[bhelgaas: squash refcount fix from Wan Jiabing <wanjiabing@vivo.com>:
https://lore.kernel.org/r/20211103062518.25695-1-wanjiabing@vivo.com
and drop "parent" refcount per
https://lore.kernel.org/all/20211103143059.GA683503@bhelgaas/]
Link: https://lore.kernel.org/r/bb391a0e0f0863b66e645048315fab1a4f63f277.1634812676.git.mchehab+huawei@kernel.org
Link: https://lore.kernel.org/all/9a365cffe5af9ec5a1f79638968c3a2efa979b65.1634622716.git.mchehab+huawei@kernel.org/Signed-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>
Cc: Kishon Vijay Abraham I <kishon@ti.com>

The PHY layer need to access APB registers too, for Kirin 970.  So place
them into a named regmap.

Link: https://lore.kernel.org/r/daf0e4bda5a69a5ac8484e70f09351a959805c8c.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>

The pcie-kirin driver contains both PHY and generic PCI driver.

The best would be, instead, to support a PCI PHY driver, making the driver
more generic.

However, it is too late to remove the Kirin 960 PHY, as a change like that
would make the DT schema incompatible with past versions.

So, add support for an external PHY driver without removing the existing
Kirin 960 PHY from it.

Link: https://lore.kernel.org/r/f38361df2e9d0dc5a38ff942b631f7fef64cdc12.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Reviewed-by: NKishon Vijay Abraham I <kishon@ti.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>

The pcie-kirin PCIe driver contains internally a PHY interface for
Kirin 960.

As the next patches will add support for using an external PHY driver,
reorganize the driver in a way that the PHY part will be self-contained.

This could be moved to a separate PHY driver, but a change like that would
mean a non-backward-compatible DT schema change.

Link: https://lore.kernel.org/r/ad2f4aa6bbb71d5c9af0139704672f75f12644fc.1634812676.git.mchehab+huawei@kernel.orgSigned-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Acked-by: NXiaowei Song <songxiaowei@hisilicon.com>
Cc: Kishon Vijay Abraham I <kishon@ti.com>

The condition register PCI_RCV_INTX is used in irq_mask() and irq_unmask()
callbacks. Accesses to register can occur at the same time without a lock.
Add a lock into each callback to prevent the issue.

And INTX mask and unmask fields in PCL_RCV_INTX register should only be
set/reset for each bit. Clearing by PCL_RCV_INTX_ALL_MASK should be
removed.

INTX status fields in PCL_RCV_INTX register only indicates each INTX
interrupt status, so the handler can't clear by writing 1 to the field.
The status is expected to be cleared by the interrupt origin.
The ack function has no meaning, so should remove it.
Suggested-by: NPali Rohár <pali@kernel.org>
Link: https://lore.kernel.org/r/1631924579-24567-1-git-send-email-hayashi.kunihiko@socionext.com
Fixes: 7e6d5cd8 ("PCI: uniphier: Add UniPhier PCIe host controller support")
Signed-off-by: NKunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Acked-by: NPali Rohár <pali@kernel.org>
Acked-by: NMarc Zyngier <maz@kernel.org>

The SC8180x platform comes with 4 PCIe controllers, typically used for
things such as NVME storage or connecting a SDX55 5G modem. Add a
compatible for this, that just reuses the 1.9.0 ops.

Link: https://lore.kernel.org/linux-arm-msm/20210725040038.3966348-4-bjorn.andersson@linaro.org/
Link: https://lore.kernel.org/r/20210823154958.305677-2-bjorn.andersson@linaro.orgSigned-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>
[lorenzo.pieralisi@arm.com: updated match data structure]
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Reviewed-by: NRob Herring <robh@kernel.org>

On the SC7280, the clock source for gcc_pcie_1_pipe_clk_src must be the
TCXO while gdsc is enabled. After PHY init successful clock source should
switch to pipe clock for gcc_pcie_1_pipe_clk_src.

Link: https://lore.kernel.org/r/1633628923-25047-6-git-send-email-pmaliset@codeaurora.orgSigned-off-by: NPrasad Malisetty <pmaliset@codeaurora.org>
Signed-off-by: NLorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Reviewed-by: NStephen Boyd <swboyd@chromium.org>