From 82336cb33e944ed313d12cf511ee6cf6f0a0deb3 Mon Sep 17 00:00:00 2001 From: Frank Praznik Date: Fri, 13 Sep 2024 10:41:26 -0400 Subject: [PATCH] wayland: Add support for the key repeat event (seat v10) The internal key repeat mechanism already disables itself if the key repeat interval is 0, and SDL tracks and handles the flagging of repeated keys itself, so just map the 'repeated' event to 'pressed'. --- src/video/wayland/SDL_waylandevents.c | 10 + src/video/wayland/SDL_waylandvideo.c | 4 +- wayland-protocols/wayland.xml | 292 +++++++++++++++++++------- 3 files changed, 233 insertions(+), 73 deletions(-) diff --git a/src/video/wayland/SDL_waylandevents.c b/src/video/wayland/SDL_waylandevents.c index e9325ac51d..4ee2c219c9 100644 --- a/src/video/wayland/SDL_waylandevents.c +++ b/src/video/wayland/SDL_waylandevents.c @@ -1972,6 +1972,16 @@ static void keyboard_handle_key(void *data, struct wl_keyboard *keyboard, Wayland_UpdateImplicitGrabSerial(seat, serial); + if (state == WL_KEYBOARD_KEY_STATE_REPEATED) { + // If this key shouldn't be repeated, just return. + if (seat->keyboard.xkb.keymap && !WAYLAND_xkb_keymap_key_repeats(seat->keyboard.xkb.keymap, key + 8)) { + return; + } + + // SDL automatically handles key tracking and repeat status, so just map 'repeated' to 'pressed'. + state = WL_KEYBOARD_KEY_STATE_PRESSED; + } + if (seat->keyboard.sdl_keymap != SDL_GetCurrentKeymap(true)) { SDL_SetKeymap(seat->keyboard.sdl_keymap, true); SDL_SetModState(seat->keyboard.pressed_modifiers | seat->keyboard.locked_modifiers); diff --git a/src/video/wayland/SDL_waylandvideo.c b/src/video/wayland/SDL_waylandvideo.c index 8514a903d6..8f51ae9748 100644 --- a/src/video/wayland/SDL_waylandvideo.c +++ b/src/video/wayland/SDL_waylandvideo.c @@ -80,7 +80,9 @@ #define SDL_WL_COMPOSITOR_VERSION 4 #endif -#if SDL_WAYLAND_CHECK_VERSION(1, 22, 0) +#if SDL_WAYLAND_CHECK_VERSION(1, 24, 0) +#define SDL_WL_SEAT_VERSION 10 +#elif SDL_WAYLAND_CHECK_VERSION(1, 22, 0) #define SDL_WL_SEAT_VERSION 9 #elif SDL_WAYLAND_CHECK_VERSION(1, 21, 0) #define SDL_WL_SEAT_VERSION 8 diff --git a/wayland-protocols/wayland.xml b/wayland-protocols/wayland.xml index 10e039d6ec..bee74a1008 100644 --- a/wayland-protocols/wayland.xml +++ b/wayland-protocols/wayland.xml @@ -46,7 +46,7 @@ compositor after the callback is fired and as such the client must not attempt to use it after that point. - The callback_data passed in the callback is the event serial. + The callback_data passed in the callback is undefined and should be ignored. @@ -212,7 +212,7 @@ - + The wl_shm_pool object encapsulates a piece of memory shared between the compositor and client. Through the wl_shm_pool @@ -262,17 +262,17 @@ created, but using the new size. This request can only be used to make the pool bigger. - This request only changes the amount of bytes that are mmapped - by the server and does not touch the file corresponding to the - file descriptor passed at creation time. It is the client's - responsibility to ensure that the file is at least as big as - the new pool size. + This request only changes the amount of bytes that are mmapped + by the server and does not touch the file corresponding to the + file descriptor passed at creation time. It is the client's + responsibility to ensure that the file is at least as big as + the new pool size. - + A singleton global object that provides support for shared memory. @@ -419,6 +419,21 @@ + + + + + + + + + + + + + + + @@ -442,6 +457,17 @@ + + + + + + Using this request a client can tell the server that it is not going to + use the shm object anymore. + + Objects created via this interface remain unaffected. + + @@ -453,9 +479,11 @@ client provides and updates the contents is defined by the buffer factory interface. - If the buffer uses a format that has an alpha channel, the alpha channel - is assumed to be premultiplied in the color channels unless otherwise - specified. + Color channels are assumed to be electrical rather than optical (in other + words, encoded with a transfer function) unless otherwise specified. If + the buffer uses a format that has an alpha channel, the alpha channel is + assumed to be premultiplied into the electrical color channel values + (after transfer function encoding) unless otherwise specified. Note, because wl_buffer objects are created from multiple independent factory interfaces, the wl_buffer interface is frozen at version 1. @@ -473,8 +501,10 @@ Sent when this wl_buffer is no longer used by the compositor. - The client is now free to reuse or destroy this buffer and its - backing storage. + + For more information on when release events may or may not be sent, + and what consequences it has, please see the description of + wl_surface.attach. If a client receives a release event before the frame callback requested in the same wl_surface.commit that attaches this @@ -847,6 +877,7 @@ + @@ -868,7 +899,7 @@ The icon surface is an optional (can be NULL) surface that provides an icon to be moved around with the cursor. Initially, the top-left corner of the icon surface is placed at the cursor - hotspot, but subsequent wl_surface.attach request can move the + hotspot, but subsequent wl_surface.offset requests can move the relative position. Attach requests must be confirmed with wl_surface.commit as usual. The icon surface is given the role of a drag-and-drop icon. If the icon surface already has another role, @@ -876,6 +907,10 @@ The input region is ignored for wl_surfaces with the role of a drag-and-drop icon. + + The given source may not be used in any further set_selection or + start_drag requests. Attempting to reuse a previously-used source + may send a used_source error. @@ -889,6 +924,10 @@ to the data from the source on behalf of the client. To unset the selection, set the source to NULL. + + The given source may not be used in any further set_selection or + start_drag requests. Attempting to reuse a previously-used source + may send a used_source error. @@ -1411,7 +1450,7 @@ + summary="surface was destroyed before its role object"/> @@ -1440,9 +1479,9 @@ When the bound wl_surface version is 5 or higher, passing any non-zero x or y is a protocol violation, and will result in an - 'invalid_offset' error being raised. The x and y arguments are ignored - and do not change the pending state. To achieve equivalent semantics, - use wl_surface.offset. + 'invalid_offset' error being raised. The x and y arguments are ignored + and do not change the pending state. To achieve equivalent semantics, + use wl_surface.offset. Surface contents are double-buffered state, see wl_surface.commit. @@ -1467,7 +1506,8 @@ the delivery of wl_buffer.release events becomes undefined. A well behaved client should not rely on wl_buffer.release events in this case. Alternatively, a client could create multiple wl_buffer objects - from the same backing storage or use wp_linux_buffer_release. + from the same backing storage or use a protocol extension providing + per-commit release notifications. Destroying the wl_buffer after wl_buffer.release does not change the surface contents. Destroying the wl_buffer before wl_buffer.release @@ -1479,6 +1519,13 @@ If wl_surface.attach is sent with a NULL wl_buffer, the following wl_surface.commit will remove the surface content. + + If a pending wl_buffer has been destroyed, the result is not specified. + Many compositors are known to remove the surface content on the following + wl_surface.commit, but this behaviour is not universal. Clients seeking to + maximise compatibility should not destroy pending buffers and should + ensure that they explicitly remove content from surfaces, even after + destroying buffers. @@ -1618,16 +1665,18 @@ Surface state (input, opaque, and damage regions, attached buffers, etc.) is double-buffered. Protocol requests modify the pending state, - as opposed to the current state in use by the compositor. A commit - request atomically applies all pending state, replacing the current - state. After commit, the new pending state is as documented for each - related request. + as opposed to the active state in use by the compositor. - On commit, a pending wl_buffer is applied first, and all other state - second. This means that all coordinates in double-buffered state are - relative to the new wl_buffer coming into use, except for - wl_surface.attach itself. If there is no pending wl_buffer, the - coordinates are relative to the current surface contents. + A commit request atomically creates a content update from the pending + state, even if the pending state has not been touched. The content + update is placed in a queue until it becomes active. After commit, the + new pending state is as documented for each related request. + + When the content update is applied, the wl_buffer is applied before all + other state. This means that all coordinates in double-buffered state + are relative to the newly attached wl_buffers, except for + wl_surface.attach itself. If there is no newly attached wl_buffer, the + coordinates are relative to the previous content update. All requests that need a commit to become effective are documented to affect double-buffered state. @@ -1666,10 +1715,12 @@ - This request sets an optional transformation on how the compositor - interprets the contents of the buffer attached to the surface. The - accepted values for the transform parameter are the values for - wl_output.transform. + This request sets the transformation that the client has already applied + to the content of the buffer. The accepted values for the transform + parameter are the values for wl_output.transform. + + The compositor applies the inverse of this transformation whenever it + uses the buffer contents. Buffer transform is double-buffered state, see wl_surface.commit. @@ -1725,11 +1776,11 @@ a buffer that is larger (by a factor of scale in each dimension) than the desired surface size. - If scale is not positive the invalid_scale protocol error is + If scale is not greater than 0 the invalid_scale protocol error is raised. + summary="scale for interpreting buffer contents"/> @@ -1784,6 +1835,9 @@ x and y, combined with the new surface size define in which directions the surface's size changes. + The exact semantics of wl_surface.offset are role-specific. Refer to + the documentation of specific roles for more information. + Surface location offset is double-buffered state, see wl_surface.commit. @@ -1802,10 +1856,15 @@ This event indicates the preferred buffer scale for this surface. It is sent whenever the compositor's preference changes. + Before receiving this event the preferred buffer scale for this surface + is 1. + It is intended that scaling aware clients use this event to scale their content and use wl_surface.set_buffer_scale to indicate the scale they have rendered with. This allows clients to supply a higher detail buffer. + + The compositor shall emit a scale value greater than 0. @@ -1815,16 +1874,19 @@ This event indicates the preferred buffer transform for this surface. It is sent whenever the compositor's preference changes. - It is intended that transform aware clients use this event to apply the - transform to their content and use wl_surface.set_buffer_transform to - indicate the transform they have rendered with. + Before receiving this event the preferred buffer transform for this + surface is normal. + + Applying this transformation to the surface buffer contents and using + wl_surface.set_buffer_transform might allow the compositor to use the + surface buffer more efficiently. - + A seat is a group of keyboards, pointer and touch devices. This object is published as a global during start up, or when such a @@ -1852,9 +1914,10 @@ - This is emitted whenever a seat gains or loses the pointer, - keyboard or touch capabilities. The argument is a capability - enum containing the complete set of capabilities this seat has. + This is sent on binding to the seat global or whenever a seat gains + or loses the pointer, keyboard or touch capabilities. + The argument is a capability enum containing the complete set of + capabilities this seat has. When the pointer capability is added, a client may create a wl_pointer object using the wl_seat.get_pointer request. This object @@ -1936,9 +1999,9 @@ The same seat names are used for all clients. Thus, the name can be shared across processes to refer to a specific wl_seat global. - The name event is sent after binding to the seat global. This event is - only sent once per seat object, and the name does not change over the - lifetime of the wl_seat global. + The name event is sent after binding to the seat global, and should be sent + before announcing capabilities. This event only sent once per seat object, + and the name does not change over the lifetime of the wl_seat global. Compositors may re-use the same seat name if the wl_seat global is destroyed and re-created later. @@ -1957,7 +2020,7 @@ - + The wl_pointer interface represents one or more input devices, such as mice, which control the pointer location and pointer_focus @@ -1992,9 +2055,9 @@ where (x, y) are the coordinates of the pointer location, in surface-local coordinates. - On surface.attach requests to the pointer surface, hotspot_x + On wl_surface.offset requests to the pointer surface, hotspot_x and hotspot_y are decremented by the x and y parameters - passed to the request. Attach must be confirmed by + passed to the request. The offset must be applied by wl_surface.commit as usual. The hotspot can also be updated by passing the currently set @@ -2248,7 +2311,7 @@ - + Discrete step information for scroll and other axes. @@ -2370,10 +2433,20 @@ - + The wl_keyboard interface represents one or more keyboards associated with a seat. + + Each wl_keyboard has the following logical state: + + - an active surface (possibly null), + - the keys currently logically down, + - the active modifiers, + - the active group. + + By default, the active surface is null, the keys currently logically down + are empty, the active modifiers and the active group are 0. @@ -2408,10 +2481,18 @@ The compositor must send the wl_keyboard.modifiers event after this event. + + In the wl_keyboard logical state, this event sets the active surface to + the surface argument and the keys currently logically down to the keys + in the keys argument. The compositor must not send this event if the + wl_keyboard already had an active surface immediately before this event. + + Clients should not use the list of pressed keys to emulate key-press + events. The order of keys in the list is unspecified. - + @@ -2422,8 +2503,10 @@ The leave notification is sent before the enter notification for the new focus. - After this event client must assume that all keys, including modifiers, - are lifted and also it must stop key repeating if there's some going on. + In the wl_keyboard logical state, this event resets all values to their + defaults. The compositor must not send this event if the active surface + of the wl_keyboard was not equal to the surface argument immediately + before this event. @@ -2432,9 +2515,18 @@ Describes the physical state of a key that produced the key event. + + Since version 10, the key can be in a "repeated" pseudo-state which + means the same as "pressed", but is used to signal repetition in the + key event. + + The key may only enter the repeated state after entering the pressed + state and before entering the released state. This event may be + generated multiple times while the key is down. + @@ -2448,6 +2540,20 @@ If this event produces a change in modifiers, then the resulting wl_keyboard.modifiers event must be sent after this event. + + In the wl_keyboard logical state, this event adds the key to the keys + currently logically down (if the state argument is pressed) or removes + the key from the keys currently logically down (if the state argument is + released). The compositor must not send this event if the wl_keyboard + did not have an active surface immediately before this event. The + compositor must not send this event if state is pressed (resp. released) + and the key was already logically down (resp. was not logically down) + immediately before this event. + + Since version 10, compositors may send key events with the "repeated" + key state when a wl_keyboard.repeat_info event with a rate argument of + 0 has been received. This allows the compositor to take over the + responsibility of key repetition. @@ -2459,6 +2565,17 @@ Notifies clients that the modifier and/or group state has changed, and it should update its local state. + + The compositor may send this event without a surface of the client + having keyboard focus, for example to tie modifier information to + pointer focus instead. If a modifier event with pressed modifiers is sent + without a prior enter event, the client can assume the modifier state is + valid until it receives the next wl_keyboard.modifiers event. In order to + reset the modifier state again, the compositor can send a + wl_keyboard.modifiers event with no pressed modifiers. + + In the wl_keyboard logical state, this event updates the modifiers and + group. @@ -2497,7 +2614,7 @@ - + The wl_touch interface represents a touchscreen associated with a seat. @@ -2566,6 +2683,8 @@ currently active on this client's surface. The client is responsible for finalizing the touch points, future touch points on this surface may reuse the touch point ID. + + No frame event is required after the cancel event. @@ -2665,10 +2784,9 @@ - - This describes the transform that a compositor will apply to a - surface to compensate for the rotation or mirroring of an - output device. + + This describes transformations that clients and compositors apply to + buffer contents. The flipped values correspond to an initial flip around a vertical axis followed by rotation. @@ -2700,6 +2818,10 @@ The geometry event will be followed by a done event (starting from version 2). + Clients should use wl_surface.preferred_buffer_transform instead of the + transform advertised by this event to find the preferred buffer + transform to use for a surface. + Note: wl_output only advertises partial information about the output position and identification. Some compositors, for instance those not implementing a desktop-style output layout or those exposing virtual @@ -2722,7 +2844,7 @@ + summary="additional transformation applied to buffer contents during presentation"/> @@ -2795,8 +2917,9 @@ This event contains scaling geometry information that is not in the geometry event. It may be sent after binding the output object or if the output scale changes - later. If it is not sent, the client should assume a - scale of 1. + later. The compositor will emit a non-zero, positive + value for scale. If it is not sent, the client should + assume a scale of 1. A scale larger than 1 means that the compositor will automatically scale surface buffers by this amount @@ -2804,12 +2927,9 @@ displays where applications rendering at the native resolution would be too small to be legible. - It is intended that scaling aware clients track the - current output of a surface, and if it is on a scaled - output it should use wl_surface.set_buffer_scale with - the scale of the output. That way the compositor can - avoid scaling the surface, and the client can supply - a higher detail image. + Clients should use wl_surface.preferred_buffer_scale + instead of this event to find the preferred buffer + scale to use for a surface. The scale event will be followed by a done event. @@ -3035,6 +3155,11 @@ If the parent wl_surface object is destroyed, the sub-surface is unmapped. + + A sub-surface never has the keyboard focus of any seat. + + The wl_surface.offset request is ignored: clients must use set_position + instead to move the sub-surface. @@ -3060,9 +3185,7 @@ surface area. Negative values are allowed. The scheduled coordinates will take effect whenever the state of the - parent surface is applied. When this happens depends on whether the - parent surface is in synchronized mode or not. See - wl_subsurface.set_sync and wl_subsurface.set_desync for details. + parent surface is applied. If more than one set_position request is invoked by the client before the commit of the parent surface, the position of a new request always @@ -3085,9 +3208,7 @@ The z-order is double-buffered. Requests are handled in order and applied immediately to a pending state. The final pending state is copied to the active state the next time the state of the parent - surface is applied. When this happens depends on whether the parent - surface is in synchronized mode or not. See wl_subsurface.set_sync and - wl_subsurface.set_desync for details. + surface is applied. A new sub-surface is initially added as the top-most in the stack of its siblings and parent. @@ -3148,4 +3269,31 @@ + + + This global fixes problems with other core-protocol interfaces that + cannot be fixed in these interfaces themselves. + + + + + + + + + This request destroys a wl_registry object. + + The client should no longer use the wl_registry after making this + request. + + The compositor will emit a wl_display.delete_id event with the object ID + of the registry and will no longer emit any events on the registry. The + client should re-use the object ID once it receives the + wl_display.delete_id event. + + + + +