HID: pidff: Clamp effect playback LOOP_COUNT value

Ensures the loop count will never exceed the logical_maximum.

Fixes implementation errors happening when applications use the max
value of int32/D

HID: pidff: Clamp effect playback LOOP_COUNT value

Ensures the loop count will never exceed the logical_maximum.

Fixes implementation errors happening when applications use the max
value of int32/DWORD as the effect iterations. This could be observed
when running software both native and in wine.

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Rename two functions to align them with naming convention

Driver uses "set" everywhere to indicate setting report values and
requesting HID_REQ_SET_REPORT

Signed-off-by: Tomasz Pakuła <

HID: pidff: Rename two functions to align them with naming convention

Driver uses "set" everywhere to indicate setting report values and
requesting HID_REQ_SET_REPORT

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Remove redundant call to pidff_find_special_keys

Probably left out as a mistake after Anssi created the helper macro

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Sig

HID: pidff: Remove redundant call to pidff_find_special_keys

Probably left out as a mistake after Anssi created the helper macro

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Support device error response from PID_BLOCK_LOAD

If an error happens on the device, the driver will no longer fall
into the trap of reading this status 60 times before it decides that
t

HID: pidff: Support device error response from PID_BLOCK_LOAD

If an error happens on the device, the driver will no longer fall
into the trap of reading this status 60 times before it decides that
this reply won't change to success/memory full.

Greatly reduces communication overhead during device error situation.

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Comment and code style update

Update comments to fully conform to the Linux comment styling.
Define Linux infinite effect duration (0) as FF_INFINITE

Chanage Oleg's name order

Signed-o

HID: pidff: Comment and code style update

Update comments to fully conform to the Linux comment styling.
Define Linux infinite effect duration (0) as FF_INFINITE

Chanage Oleg's name order

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Make sure to fetch pool before checking SIMULTANEOUS_MAX

As noted by Anssi some 20 years ago, pool report is sometimes messed up.
This worked fine on many devices but casued oops on VRS

HID: pidff: Make sure to fetch pool before checking SIMULTANEOUS_MAX

As noted by Anssi some 20 years ago, pool report is sometimes messed up.
This worked fine on many devices but casued oops on VRS DirectForce PRO.

Here, we're making sure pool report is refetched before trying to access
any of it's fields. While loop was replaced with a for loop + exit
conditions were moved aroud to decrease the possibility of creating an
infinite loop scenario.

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Factor out pool report fetch and remove excess declaration

We only want to refetch the pool report during device init. Reset
function is now called when uploading effects to an empty dev

HID: pidff: Factor out pool report fetch and remove excess declaration

We only want to refetch the pool report during device init. Reset
function is now called when uploading effects to an empty device so
extract pool fetch to separate function and call it from init before
autocenter check (autocenter check triggered reset during init).

Remove a superfluous pointer declaration and assigment as well.

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Use macros instead of hardcoded min/max values for shorts

Makes it obvious these magic values ARE in fact derived from min and
max values for s16 and u16

Signed-off-by: Tomasz Pakuła <t

HID: pidff: Use macros instead of hardcoded min/max values for shorts

Makes it obvious these magic values ARE in fact derived from min and
max values for s16 and u16

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Simplify pidff_rescale_signed

This function overrelies on ternary operators and makes it hard to parse
it mentally. New version makes it very easy to understand.

Signed-off-by: Tomasz P

HID: pidff: Simplify pidff_rescale_signed

This function overrelies on ternary operators and makes it hard to parse
it mentally. New version makes it very easy to understand.

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Move all hid-pidff definitions to a dedicated header

Do not clutter hid includes with stuff not needed outside of
the kernel.

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail

HID: pidff: Move all hid-pidff definitions to a dedicated header

Do not clutter hid includes with stuff not needed outside of
the kernel.

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Fix null pointer dereference in pidff_find_fields

This function triggered a null pointer dereference if used to search for
a report that isn't implemented on the device. This happened bo

HID: pidff: Fix null pointer dereference in pidff_find_fields

This function triggered a null pointer dereference if used to search for
a report that isn't implemented on the device. This happened both for
optional and required reports alike.

The same logic was applied to pidff_find_special_field and although
pidff_init_fields should return an error earlier if one of the required
reports is missing, future modifications could change this logic and
resurface this possible null pointer dereference again.

LKML bug report:
https://lore.kernel.org/all/CAL-gK7f5=R0nrrQdPtaZZr1fd-cdAMbDMuZ_NLA8vM0SX+nGSw@mail.gmail.com

Reported-by: Nolan Nicholson <nolananicholson@gmail.com>
Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Factor out code for setting gain

Makes it possible to easily set gain from inside hid-pidff.c

Changes in v7:
- Check if device gain field exists before setting device gain

Signed-off-b

HID: pidff: Factor out code for setting gain

Makes it possible to easily set gain from inside hid-pidff.c

Changes in v7:
- Check if device gain field exists before setting device gain

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Rescale time values to match field units

PID devices can use different exponents for time fields, while Linux
Force Feedback API only supports miliseconds.

Read the exponent of a given

HID: pidff: Rescale time values to match field units

PID devices can use different exponents for time fields, while Linux
Force Feedback API only supports miliseconds.

Read the exponent of a given time field and scale its value accordingly.

Changes in v7:
- Rescale all time fields, not only period

changes in v9:
- Properly assign fade_lenght, not attack_length to PID_FADE_TIME

Co-developed-by: Makarenko Oleg <oleg@makarenk.ooo>
Signed-off-by: Makarenko Oleg <oleg@makarenk.ooo>
Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Define values used in pidff_find_special_fields

Makes it clear where did these values came from

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kope

HID: pidff: Define values used in pidff_find_special_fields

Makes it clear where did these values came from

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Simplify pidff_upload_effect function

Merge a bit of code that reqeusts conditional effects upload.
Makes it clear, that effect handling should be identical for
SPRING, DAMPER, INERTIA a

HID: pidff: Simplify pidff_upload_effect function

Merge a bit of code that reqeusts conditional effects upload.
Makes it clear, that effect handling should be identical for
SPRING, DAMPER, INERTIA and FRICTION.

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Completely rework and fix pidff_reset function

Previously, it was assumed that DEVICE_CONTROL usage is always an array
but a lot of devices implements it as a bitmask variable. This led

HID: pidff: Completely rework and fix pidff_reset function

Previously, it was assumed that DEVICE_CONTROL usage is always an array
but a lot of devices implements it as a bitmask variable. This led to
the pidff_reset function not working and causing errors in such cases.

Selectors can come in three types. One selection of a set, N selections
and Any selection in form of bitmask as from USB Hid Usage Tables v1.5,
subsection 3.4.2.1

Added pidff_send_device_control which handles usage flag check which
decides whether DEVICE_CONTROL should be handled as "One selection of a
set" or "Any selection of a set".

Reset was triggered once, on device initialization. Now, it's triggered
every time when uploading an effect to an empty device (no currently
stored effects), tracked by pidff->effect_count variable.

Co-developed-by: Makarenko Oleg <oleg@makarenk.ooo>
Signed-off-by: Makarenko Oleg <oleg@makarenk.ooo>
Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Add PERIODIC_SINE_ONLY quirk

Some devices only support SINE periodic effect although they advertise
support for all PERIODIC effect in their HID descriptor. Some just do
nothing when try

HID: pidff: Add PERIODIC_SINE_ONLY quirk

Some devices only support SINE periodic effect although they advertise
support for all PERIODIC effect in their HID descriptor. Some just do
nothing when trying to play such an effect (upload goes fine), some express
undefined behavior like turning to one side.

This quirk forces all the periodic effects to be uploaded as SINE. This is
acceptable as all these effects are similar in nature and are mostly used as
rumble. SINE is the most popular with others seldom used (especially SAW_UP
and SAW_DOWN).

Fixes periodic effects for PXN and LITE STAR wheels

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Stop all effects before enabling actuators

Some PID compliant devices automatically play effects after boot (i.e.
autocenter spring) that prevent the rendering of other effects since
it

HID: pidff: Stop all effects before enabling actuators

Some PID compliant devices automatically play effects after boot (i.e.
autocenter spring) that prevent the rendering of other effects since
it is done outside the kernel driver.

This makes sure all the effects currently played are stopped after
resetting the device.
It brings compatibility to the Brunner CLS-P joystick and others

Reported-by: Jules Noirant <jules.noirant@orange.fr>
Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Add FIX_WHEEL_DIRECTION quirk

Most steering wheels simply ignore DIRECTION field, but some try to be
compliant with the PID standard and use it in force calculations. Games
often ignore

HID: pidff: Add FIX_WHEEL_DIRECTION quirk

Most steering wheels simply ignore DIRECTION field, but some try to be
compliant with the PID standard and use it in force calculations. Games
often ignore setting this field properly and/or there can be issues with
dinput8 -> wine -> SDL -> Linux API translation, and this value can be
incorrect. This can lead to partial/complete loss of Force Feedback or
even unexpected force reversal.

Sadly, this quirk can't be detected automatically without sending out
effects that would move an axis.

This fixes FFB on Moza Racing devices and others where effect direction
is not simply ignored.

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Add hid_pidff_init_with_quirks and export as GPL symbol

This lays out a way to provide an initial set of quirks to enable before
device initialization takes place. GPL symbol export need

HID: pidff: Add hid_pidff_init_with_quirks and export as GPL symbol

This lays out a way to provide an initial set of quirks to enable before
device initialization takes place. GPL symbol export needed for the
possibility of building HID drivers which use this function as modules.

Adding a wrapper function to ensure compatibility with the old behavior
of hid_pidff_init.

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Add PERMISSIVE_CONTROL quirk

With this quirk, a PID device isn't required to have a strict
logical_minimum of 1 for the the PID_DEVICE_CONTROL usage page.

Some devices come with weird v

HID: pidff: Add PERMISSIVE_CONTROL quirk

With this quirk, a PID device isn't required to have a strict
logical_minimum of 1 for the the PID_DEVICE_CONTROL usage page.

Some devices come with weird values in their device descriptors and
this quirk enables their initialization even if the logical minimum
of the DEVICE_CONTROL page is not 1.

Fixes initialization of VRS Direct Force Pro

Changes in v6:
- Change quirk name to better reflect it's intention

Co-developed-by: Makarenko Oleg <oleg@makarenk.ooo>
Signed-off-by: Makarenko Oleg <oleg@makarenk.ooo>
Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Add MISSING_PBO quirk and its detection

Some devices with only one axis are missing PARAMETER_BLOCK_OFFSET field
for conditional effects. They can only have one axis, so we're limiting
t

HID: pidff: Add MISSING_PBO quirk and its detection

Some devices with only one axis are missing PARAMETER_BLOCK_OFFSET field
for conditional effects. They can only have one axis, so we're limiting
the max_axis when setting the report for those effects.

Automatic detection ensures compatibility even if such device won't be
explicitly defined in the kernel.

Fixes initialization of VRS DirectForce PRO and possibly other devices.

Changes in v6:
- Fixed NULL pointer dereference. When PBO is missing, make sure not
to set it anyway

Co-developed-by: Makarenko Oleg <oleg@makarenk.ooo>
Signed-off-by: Makarenko Oleg <oleg@makarenk.ooo>
Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Add MISSING_DELAY quirk and its detection

A lot of devices do not include this field, and it's seldom used in force
feedback implementations. I tested about three dozen applications and

HID: pidff: Add MISSING_DELAY quirk and its detection

A lot of devices do not include this field, and it's seldom used in force
feedback implementations. I tested about three dozen applications and
none of them make use of the delay.

This fixes initialization of a lot of PID wheels like Cammus, VRS, FFBeast

This change has no effect on fully compliant devices

Co-developed-by: Makarenko Oleg <oleg@makarenk.ooo>
Signed-off-by: Makarenko Oleg <oleg@makarenk.ooo>
Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Clamp PERIODIC effect period to device's logical range

This ensures the effect can actually be played on the connected force
feedback device. Adds clamping functions used instead of resc

HID: pidff: Clamp PERIODIC effect period to device's logical range

This ensures the effect can actually be played on the connected force
feedback device. Adds clamping functions used instead of rescaling, as we
don't want to change the characteristics of the periodic effects.

Fixes edge cases found on Moza Racing and some other hardware where
the effects would not play if the period is outside the defined
logical range.

Changes in v6:
- Use in-kernel clamp macro instead of a custom solution

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...

HID: pidff: Do not send effect envelope if it's empty

Envelope struct is always initialized, but the envelope itself is
optional as described in USB PID Device class definition 1.0.

5.1.1.1 Type Sp

HID: pidff: Do not send effect envelope if it's empty

Envelope struct is always initialized, but the envelope itself is
optional as described in USB PID Device class definition 1.0.

5.1.1.1 Type Specific Block Offsets
...
4) Effects that do not use Condition Blocks use 1 Parameter Block and
an *optional* Envelope Block.

Sending out "empty" envelope breaks force feedback on some devices with
games that use SINE effect + offset to emulate constant force effect, as
well as generally breaking Constant/Periodic effects. One of the affected
brands is Moza Racing.

This change prevents the envelope from being sent if it contains all
0 values while keeping the old behavior of only sending it, if it differs
from the old one.

Changes in v6:
- Simplify the checks to make them clearer
- Fix possible null pointer dereference while calling
pidff_needs_set_envelope

Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Reviewed-by: Michał Kopeć <michal@nozomi.space>
Reviewed-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Paul Dino Jones <paul@spacefreak18.xyz>
Tested-by: Cristóferson Bueno <cbueno81@gmail.com>
Tested-by: Pablo Cisneros <patchkez@protonmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...