xhci: dbc: poll at different rate depending on data transfer activity

DbC driver starts polling for events immediately when DbC is enabled.
The current polling interval is 1ms, which keeps the CPU busy, impacting
power management even when there are no active data transfers.

Solve this by polling at a slower rate, with a 64ms interval as default
until a transfer request is queued, or if there are still are pending
unhandled transfers at event completion.

Tested-by: Uday M Bhat <uday.m.bhat@intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20240229141438.619372-9-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Add missing headers

Don't inherit headers "by chances" from asm/bug.h, asm/io.h,
etc... Include the needed headers explicitly.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231201150647.1307406-11-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Use sizeof(*pointer) instead of sizeof(type)

It is preferred to use sizeof(*pointer) instead of sizeof(type).
The type of the variable can change and one needs not change
the former (unlike the latter). No functional change intended.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231201150647.1307406-10-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Use sizeof_field() where it makes sense

Instead of doing custom calculations, use sizeof_field() macro.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231201150647.1307406-9-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Replace custom return value with proper Linux error code

Replace the custom return value with proper Linux error code.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231201150647.1307406-8-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Don't shadow error codes in store() functions

kstrtox() along with regmap API can return different error codes
based on the circumstances. Don't shadow them when returning to
the caller.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231201150647.1307406-7-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Check for errors first in xhci_dbc_stop()

The usual pattern is to check for errors and then continue if none.
Apply that pattern to xhci_dbc_stop() code.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231201150647.1307406-6-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Use ATTRIBUTE_GROUPS()

Embrace ATTRIBUTE_GROUPS() to avoid boiler plate code.
This should not introduce any functional changes.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231201150647.1307406-5-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Use sysfs_emit() to instead of scnprintf()

Follow the advice of the Documentation/filesystems/sysfs.rst and show()
should only use sysfs_emit() or sysfs_emit_at() when formatting the
value to be returned to user space.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231201150647.1307406-4-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Convert to use sysfs_streq()

It's standard approach to parse values from user space by using
sysfs_streq(). Make driver use it.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231201150647.1307406-3-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Drop duplicate checks for dma_free_coherent()

dma_free_coherent() is NULL-aware, not necessary to check for
the parameter twice. Drop duplicate conditionals in the caller.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20231201150647.1307406-2-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Provide sysfs option to configure dbc descriptors

When DbC is enabled the first port on the xHC host acts as a usb device.
xHC provides the descriptors automatically when the DbC device is
enumerated. Most of the values are hardcoded, but some fields such as
idProduct, idVendor, bcdDevice and bInterfaceProtocol can be modified.

Add sysfs entries that allow userspace to change these.
User can only change them before dbc is enabled, i.e. before writing
"enable" to dbc sysfs file as we don't want these values to change while
device is connected, or during enumeration.

Add documentation for these entries in
Documentation/ABI/testing/sysfs-bus-pci-drivers-xhci_hcd

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20230317154715.535523-9-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Fix memory leak in xhci_alloc_dbc()

If DbC is already in use, then the allocated memory for the xhci_dbc struct
doesn't get freed before returning NULL, which leads to a memleak.

Fixes: 534675942e90 ("xhci: dbc: refactor xhci_dbc_init()")
Cc: stable@vger.kernel.org
Signed-off-by: Rafael Mendonca <rafaelmendsr@gmail.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20220921123450.671459-3-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Don't call dbc_tty_init() on every dbc tty probe

The current workaround to call the dbc_tty_init() in probe is
not working in case we have several xhci devices with dbc enabled.

dbc_tty_init() should be called only once by a module init call when
module is loaded.

until dbgtty is its own module call dbc_tty_init() from xhci
module init call.

Same is true for unloading and dbc_tty_exit()

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20220216095153.1303105-5-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Rename xhci_dbc_init and xhci_dbc_exit

These names give the impression the functions are related to
module init calls, but are in fact creating and removing the dbc
fake device

Rename them to xhci_create_dbc_dev() and xhci_remove_dbc_dev().

We will need the _init and _exit names for actual dbc module init
and exit calls.

No functional changes

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20220216095153.1303105-4-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: create and remove dbc structure in dbgtty driver.

Turn the dbgtty closer to a device driver by allocating the dbc
structure in its own xhci_dbc_tty_probe() function, and freeing it
in xhci_dbc_tty_remove()

Remove xhci_do_dbc_exit() as its no longer needed.

allocate and create the dbc strcuture in xhci_dbc_tty_probe()

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20220216095153.1303105-3-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: refactor xhci_dbc_init()

Refactor xhci_dbc_init(), splitting it into logical
parts closer to the Linux device model.

- Create the fake dbc device, depends on xhci strucure
- Allocate a dbc structure, xhci agnostic
- Call xhci_dbc_tty_probe(), similar to actual probe.

Adjustments to xhci_dbc_exit and xhci_dbc_remove are also needed
as a result to the xhci_dbc_init() changes

Mostly non-functional changes, except for creating the dbc sysfs
entry earlier, together with the dbc structure.

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20220216095153.1303105-2-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

treewide: Use fallthrough pseudo-keyword

Replace the existing /* fall through */ comments and its variants with
the new pseudo-keyword macro fallthrough[1]. Also, remove unnecessary
fall-through markings when it is the case.

[1] https://www.kernel.org/doc/html/v5.7/process/deprecated.html?highlight=fallthrough#implicit-switch-case-fall-through

Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>

xhci: dbc: Make function xhci_dbc_ring_alloc() static

The sparse tool complains as follows:

drivers/usb/host/xhci-dbgcap.c:422:18: warning:
symbol 'xhci_dbc_ring_alloc' was not declared. Should it be static?

This function is not used outside ofxhci-dbgcap.c, so this commit
marks it static.

Fixes: ac286428c69f ("xhci: dbc: don't use generic xhci ring allocation functions for dbc.")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Link: https://lore.kernel.org/r/20200727171207.3101-1-weiyongjun1@huawei.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbgcap: remove dbc dependency on dbctty specific flag

dbc should not be aware of, or use any dbctty specific variables.
currenly dbc driver reads the port->registered flag to see if the
callbacks should be called.

Only makes these decisions based on dbc internal state instead.

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-27-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Add a operations structure to access driver functions

Don't call dbctty driver functions directly from dbc core code.
Introduce a new dbc_driver structure that contains function pointers
for disconnect and configure operations.

The driver (ttydbc) must provide these opeations when creating a dbc.

Name the structure dbc_driver instead of dbc_ops as we plan to
add more driver configureable values here, such as vid and pid.

Decouples dbc and dbctty.

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-26-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbctty: split dbc tty driver registration and unregistration functions.

Split the dbc tty driver registrations function into separate
init and probe parts.

The init part will register the tty driver, and should in the future be
called from module_init().
The probe part will become the normal probe function, but for now it is
called from the init part.

The unregister function is s likewise split into remove and exit parts.

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-25-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: simplify dbc requests allocation and queueing

Don't pass endpoint pointer, dbctty should not be aware of
struct dbc_ep, knowing the direction is enough.

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-23-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Pass dbc pointer to dbc start and stop functions.

xhci_dbc_start() and xhci_dbc_stop() functions only used xhci_hcd pointer
to get the dbc pointer.
Pass the dbc pointer instead of the xhci_hcd pointer as a parameter

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-22-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Pass dbc pointer to dbc memory init and cleanup functions

Dbc mem_init and mem_cleanup functions used xhci_hcd to get to the device
pointer. The device pointer can be accessed directly from dbc structure,
so pass a pointer to dbc as a parameter instead.

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-21-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: don't use generic xhci ring allocation functions for dbc.

The generic xhci ring allocations code needs struct xhci_hcd pointer, and
it allocates memory for the rings from dma pools created for the xhci
device.

In order to decouple xhci and DbC we have to create our own ring allocation
and free routines for DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-20-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Don't use generic xhci context allocation for dbc

The DbC context is different from the xhci device context.
It's a lot smaller as it only contains three 64 bytes sub-contexts;
the info, endpoint-out, and endpoint-in contexts. In total 192 bytes.
The context size (CSZ) field in HCCPARAMS1 xhci register does not alter
DbC context size like it does for xhci device contexts.

So don't use the geneic xhci context memory allocation, or the
dma pool that is intended for xhci device contexts.

In addition to saving memory this also helps decoupleing xhci and dbc code.

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-19-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Use dbc structure in the request completion instead of xhci_hcd

The dbc request completion callback doesn't need a xhci_hcd pointer.
The only user of the xhci_hcd pointer in dbgtty request callback was
the xhci_warn() function. Change it to dev_warn() instead.

While changing the callback function parameter to dbc in struct
xhci_requeset, move the struct xhci_request declaraion down a bit in the
header file to avoid compiler warinings

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-18-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Pass dbc pointer to get_in/out_ep() helper functions to get endpoints

Pass dbc pointer instead of struct xhci_hcd pointer to the get_in_ep() and
get_out_ep() helper functions.

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-17-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbgtty: Pass dbc pointer when registering a dbctty device

Pass dbc pointer to the xhci_dbc_tty_register_device() and
xhci_dbc_tty_unregister_device() functions instead of xhci_hcd pointer

These functions don't need a xhci_hcd pointer anymore, only use case was
the xhci_err() function, which is now changed to a dev_err() instead.

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-16-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Pass dbc pointer to dbc_handle_xfer_event() instead of xhci_hcd pointer

The event handling function only used xhci pointer to get the dbc pointer.
Pass the dbc pointer instead as a parameter

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-15-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Change to pass dbc pointer to xhci_do_dbc_stop()

Pass the dbc pointer instead of xhci_hcd pointer in order to
decouple xhci and dbc.

xhci_do_dbc_stop() only used xhci to get the dbc pointer.
Pass the dbc pointer instead as a parameter

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-14-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Pass dbc pointer to endpoint init and exit functions.

struct xhci_hcd pointer is not needed for dbc endpoint init and exit,
it was only used to get to the dbc structure.
Pass the dbc pointer as a parameter to these functions instead.

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-13-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Get the device pointer from dbc structure in dbc_ep_do_queue()

dbc_ep_do_queue() can now get the device pointer directly from dbc
structure instead of going through the xhci_hcd structure.

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-12-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Don't pass the xhci pointer as a parameter to xhci_dbc_init_context()

xhci_dbc_init_context() no longer needs the struct xhci_hcd pointer.
Pass the dbc pointer directly instead.

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-11-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Don't use xhci_write_64() as it takes xhci as a parameter

xhci_write_64() is essentially a wrapper for lo_hi_writeq(), but it
requires struct xhci_hcd * as a parameter.
Use lo_hi_writeq() directly instead

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-10-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Use dev_info() and similar instead of xhci_info()

To make this change possible we also need to change
dbc_handle_port_status() to take dbc pointer as parameter instead of
xhci_hcd pointer.

Note that xhci_info() used xhci_to_hcd(xhci)->self.controller as the
device while for dev_info we use xhci_to_hcd(xhci)->self.sysdev.

In many cases those are the same, but not for some device where
a dwc3 controller creates a xhci platform device. In th this case
self.controller may be the platform device while self.sysdev is
the actual device known to firmware (dwc3).

This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-9-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Add device pointer to dbc structure

Currently the dbc structure contains a pointer to struct xhci_hcd,
and dbc functions use that to dig up the underlying device pointer.

We are trying to decouple xhci and dbc code, and prepare for
code that use dbc such as dbctty into into real device drivers.
This is one step along the way.

Keep functionality the same and keep the xhci pointer, and
let the new device pointer point to the xhci device for now.

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-8-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Remove dbc_dma_free_coherent() wrapper

dbc_dma_free_coherent() takes struct xhci_hcd pointer as a parameter, but
does nothing more than calls dma_free_coherent().
Remove it and call dma_free_coherent() directly instead.

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-7-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Remove dbc_dma_alloc_coherent() wrapper

dbc_dma_alloc_coherent() takes struct xhci_hcd pointer as an parameter, but
does nothing more than calls dma_alloc_coherent().
Remove it and call dma_alloc_coherent() directly instead.

No functional changes
This change helps decoupling xhci and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-6-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Don't use generic xhci erst allocation and free functions

The generic erst allocation and free functions take struct xhci_hcd pointer
as a parameter. Create own erst helpers for DbC in order to decouple xhci
and DbC

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-5-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: dbc: Don't use generic xhci inc_deq() function for dbc

The generic inc_deq() helper takes struct xhci_hcd pointer as a parameter,
and is a lot more complex than needed for the DbC usecase.

In order to decouple xhci and DbC we have to create our own small
inc_evt_deq() helper, not relying on xhci.

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200723144530.9992-3-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: host: xhci-dbgcap: File headers are not good candidates for kerneldoc

Demote xhci-dbgcap's file header to a standard comment block.

Fixes the following W=1 kernel build warning(s):

drivers/usb/host/xhci-dbgcap.c:20: warning: Function parameter or member 'xhci' not described in 'dbc_dma_alloc_coherent'
drivers/usb/host/xhci-dbgcap.c:20: warning: Function parameter or member 'size' not described in 'dbc_dma_alloc_coherent'
drivers/usb/host/xhci-dbgcap.c:20: warning: Function parameter or member 'dma_handle' not described in 'dbc_dma_alloc_coherent'
drivers/usb/host/xhci-dbgcap.c:20: warning: Function parameter or member 'flags' not described in 'dbc_dma_alloc_coherent'

Cc: Mathias Nyman <mathias.nyman@intel.com>
Cc: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Link: https://lore.kernel.org/r/20200703174148.2749969-16-lee.jones@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: host: Remove call to memset after dma_alloc_coherent

In commit 518a2f1925c3
("dma-mapping: zero memory returned from dma_alloc_*"),
dma_alloc_coherent has already zeroed the memory.
So memset is not needed.

Signed-off-by: Fuqian Huang <huangfq.daxian@gmail.com>
Link: https://lore.kernel.org/r/20190715032010.7258-1-huangfq.daxian@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

drivers: Remove explicit invocations of mmiowb()

mmiowb() is now implied by spin_unlock() on architectures that require
it, so there is no reason to call it from driver code. This patch was
generated using coccinelle:

@mmiowb@
@@
- mmiowb();

and invoked as:

$ for d in drivers include/linux/qed sound; do \
spatch --include-headers --sp-file mmiowb.cocci --dir $d --in-place; done

NOTE: mmiowb() has only ever guaranteed ordering in conjunction with
spin_unlock(). However, pairing each mmiowb() removal in this patch with
the corresponding call to spin_unlock() is not at all trivial, so there
is a small chance that this change may regress any drivers incorrectly
relying on mmiowb() to order MMIO writes between CPUs using lock-free
synchronisation. If you've ended up bisecting to this commit, you can
reintroduce the mmiowb() calls using wmb() instead, which should restore
the old behaviour on all architectures other than some esoteric ia64
systems.

Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>

usb: xhci: dbc: Don't free all memory with spinlock held

The xhci debug capability (DbC) feature did its memory cleanup with
spinlock held. dma_free_coherent() warns if called with interrupts
disabled

move the memory cleanup outside the spinlock

Cc: stable <stable@vger.kernel.org>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: xhci: dbc: Fixing typo error.

Replaced "xhci_dbc_flush_reqests" with xhci_dbc_flush_requests".

Signed-off-by: Prabhat Chand Pandey <prabhat.chand.pandey@intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: xhci: dbc: remove redundant pointer dbc

Pointer dbc is being assigned but is never used hence it is
redundant and can be removed.

Cleans up clang warning:
variable 'dbc' set but not used [-Wunused-but-set-variable]

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: xhci: dbc: Don't decrement runtime PM counter if DBC is not started

pm_runtime_put_sync() gets called everytime in xhci_dbc_stop().

If dbc is not started, this makes the runtime PM counter incorrectly
becomes 0, and calls autosuspend function. Then we'll keep seeing this:
[54664.762220] xhci_hcd 0000:00:14.0: Root hub is not suspended

So only calls pm_runtime_put_sync() when dbc was started.

Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: xhci: dbc: Add SPDX identifiers to dbc files

Update the xhci dbc files with the correct SPDX license identifiers.

Fixes: dfba2174dc42 ("usb: xhci: Add DbC support in xHCI driver")
Cc: Philippe Ombredanne <pombredanne@nexb.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Philippe Ombredanne <pombredanne@nexB.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: xhci: dbc: Fix lockdep warning

The xHCI DbC implementation might enter a deadlock situation because
there is no sufficient protection against the shared data between
process and softirq contexts. This can lead to the following lockdep
warnings. This patch changes to use spin_{,un}lock_irq{save,restore}
to avoid potential deadlock.

[ 528.248084] ================================
[ 528.252914] WARNING: inconsistent lock state
[ 528.257756] 4.15.0-rc1+ #1630 Not tainted
[ 528.262305] --------------------------------
[ 528.267145] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage.
[ 528.273953] ksoftirqd/1/17 [HC0[0]:SC1[1]:HE0:SE0] takes:
[ 528.280075] (&(&port->port_lock)->rlock){+.?.}, at: [<ffffffff815396a8>] dbc_rx_push+0x38/0x1c0
[ 528.290043] {SOFTIRQ-ON-W} state was registered at:
[ 528.295570] _raw_spin_lock+0x2f/0x40
[ 528.299818] dbc_write_complete+0x27/0xa0
[ 528.304458] xhci_dbc_giveback+0xd1/0x200
[ 528.309098] xhci_dbc_flush_endpoint_requests+0x50/0x70
[ 528.315116] xhci_dbc_handle_events+0x696/0x7b0
[ 528.320349] process_one_work+0x1ee/0x6e0
[ 528.324988] worker_thread+0x4a/0x430
[ 528.329236] kthread+0x13e/0x170
[ 528.332992] ret_from_fork+0x24/0x30
[ 528.337141] irq event stamp: 2861
[ 528.340897] hardirqs last enabled at (2860): [<ffffffff810674ea>] tasklet_action+0x6a/0x250
[ 528.350460] hardirqs last disabled at (2861): [<ffffffff817dc1ef>] _raw_spin_lock_irq+0xf/0x40
[ 528.360219] softirqs last enabled at (2852): [<ffffffff817e0e8c>] __do_softirq+0x3dc/0x4f9
[ 528.369683] softirqs last disabled at (2857): [<ffffffff8106805b>] run_ksoftirqd+0x1b/0x60
[ 528.379048]
[ 528.379048] other info that might help us debug this:
[ 528.386443] Possible unsafe locking scenario:
[ 528.386443]
[ 528.393150] CPU0
[ 528.395917] ----
[ 528.398687] lock(&(&port->port_lock)->rlock);
[ 528.403821] <Interrupt>
[ 528.406786] lock(&(&port->port_lock)->rlock);
[ 528.412116]
[ 528.412116] *** DEADLOCK ***
[ 528.412116]
[ 528.418825] no locks held by ksoftirqd/1/17.
[ 528.423662]
[ 528.423662] stack backtrace:
[ 528.428598] CPU: 1 PID: 17 Comm: ksoftirqd/1 Not tainted 4.15.0-rc1+ #1630
[ 528.436387] Call Trace:
[ 528.439158] dump_stack+0x5e/0x8e
[ 528.442914] print_usage_bug+0x1fc/0x220
[ 528.447357] mark_lock+0x4db/0x5a0
[ 528.451210] __lock_acquire+0x726/0x1130
[ 528.455655] ? __lock_acquire+0x557/0x1130
[ 528.460296] lock_acquire+0xa2/0x200
[ 528.464347] ? dbc_rx_push+0x38/0x1c0
[ 528.468496] _raw_spin_lock_irq+0x35/0x40
[ 528.473038] ? dbc_rx_push+0x38/0x1c0
[ 528.477186] dbc_rx_push+0x38/0x1c0
[ 528.481139] tasklet_action+0x1d2/0x250
[ 528.485483] __do_softirq+0x1dc/0x4f9
[ 528.489630] run_ksoftirqd+0x1b/0x60
[ 528.493682] smpboot_thread_fn+0x179/0x270
[ 528.498324] kthread+0x13e/0x170
[ 528.501981] ? sort_range+0x20/0x20
[ 528.505933] ? kthread_delayed_work_timer_fn+0x80/0x80
[ 528.511755] ret_from_fork+0x24/0x30

Fixes: dfba2174dc42 ("usb: xhci: Add DbC support in xHCI driver")
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

USB: move many drivers to use DEVICE_ATTR_RW

Instead of "open coding" a DEVICE_ATTR() define, use the
DEVICE_ATTR_RW() macro instead, which does everything properly instead.

This does require a few static functions to be renamed to work properly,
but thanks to a script from Joe Perches, this was easily done.

Reported-by: Joe Perches <joe@perches.com>
Cc: Matthieu CASTET <castet.matthieu@free.fr>
Cc: Stanislaw Gruszka <stf_xl@wp.pl>
Cc: Peter Chen <Peter.Chen@nxp.com>
Cc: Mathias Nyman <mathias.nyman@intel.com>
Acked-by: Felipe Balbi <felipe.balbi@linux.intel.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Bin Liu <b-liu@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: fixup incorrect memset size parameter when clearing up DbC on exit.

Incorrect size was given to memset when zeroing the DbC endpoint
structures on exit. Use element size * ARRAY_SIZE to fix it

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Fixes: dfba2174dc42 ("usb: xhci: Add DbC support in xHCI driver")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: xhci: Add DbC support in xHCI driver

xHCI compatible USB host controllers(i.e. super-speed USB3 controllers)
can be implemented with the Debug Capability(DbC). It presents a debug
device which is fully compliant with the USB framework and provides the
equivalent of a very high performance full-duplex serial link. The debug
capability operation model and registers interface are defined in 7.6.8
of the xHCI specification, revision 1.1.

The DbC debug device shares a root port with the xHCI host. By default,
the debug capability is disabled and the root port is assigned to xHCI.
When the DbC is enabled, the root port will be assigned to the DbC debug
device, and the xHCI sees nothing on this port. This implementation uses
a sysfs node named <dbc> under the xHCI device to manage the enabling
and disabling of the debug capability.

When the debug capability is enabled, it will present a debug device
through the debug port. This debug device is fully compliant with the
USB3 framework, and it can be enumerated by a debug host on the other
end of the USB link. As soon as the debug device is configured, a TTY
serial device named /dev/ttyDBC0 will be created.

Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>