Ryujinx-SDL/src/video/kmsdrm/SDL_kmsdrmmouse.c

449 lines
15 KiB
C
Raw Normal View History

/*
Simple DirectMedia Layer
2020-01-17 05:49:25 +01:00
Copyright (C) 1997-2020 Sam Lantinga <slouken@libsdl.org>
Atomic KMSDRM backend by Manuel Alfayate Corchete <redwindwanderer@gmail.com>
This software is provided 'as-is', without any express or implied
warranty. In no event will the authors be held liable for any damages
arising from the use of this software.
Permission is granted to anyone to use this software for any purpose,
including commercial applications, and to alter it and redistribute it
freely, subject to the following restrictions:
1. The origin of this software must not be misrepresented; you must not
claim that you wrote the original software. If you use this software
in a product, an acknowledgment in the product documentation would be
appreciated but is not required.
2. Altered source versions must be plainly marked as such, and must not be
misrepresented as being the original software.
3. This notice may not be removed or altered from any source distribution.
*/
#include "../../SDL_internal.h"
#if SDL_VIDEO_DRIVER_KMSDRM
#include "SDL_kmsdrmvideo.h"
#include "SDL_kmsdrmmouse.h"
#include "SDL_kmsdrmdyn.h"
#include "SDL_assert.h"
#include "../../events/SDL_mouse_c.h"
#include "../../events/default_cursor.h"
static SDL_Cursor *KMSDRM_CreateDefaultCursor(void);
static SDL_Cursor *KMSDRM_CreateCursor(SDL_Surface * surface, int hot_x, int hot_y);
static int KMSDRM_ShowCursor(SDL_Cursor * cursor);
static void KMSDRM_MoveCursor(SDL_Cursor * cursor);
static void KMSDRM_FreeCursor(SDL_Cursor * cursor);
static void KMSDRM_WarpMouse(SDL_Window * window, int x, int y);
static int KMSDRM_WarpMouseGlobal(int x, int y);
/**********************************/
/* Atomic helper functions block. */
/**********************************/
int
drm_atomic_movecursor(KMSDRM_CursorData *curdata, uint16_t x, uint16_t y)
{
SDL_DisplayData *dispdata = (SDL_DisplayData *)SDL_GetDisplayDriverData(0);
if (!dispdata->cursor_plane) /* We can't move a non-existing cursor, but that's ok. */
return 0;
/* Do we have a set of changes already in the making? If not, allocate a new one. */
if (!dispdata->atomic_req)
dispdata->atomic_req = KMSDRM_drmModeAtomicAlloc();
add_plane_property(dispdata->atomic_req, dispdata->cursor_plane, "CRTC_X", x - curdata->hot_x);
add_plane_property(dispdata->atomic_req, dispdata->cursor_plane, "CRTC_Y", y - curdata->hot_y);
return 0;
}
/***************************************/
/* Atomic helper functions block ends. */
/***************************************/
/* Converts a pixel from straight-alpha [AA, RR, GG, BB], which the SDL cursor surface has,
to premultiplied-alpha [AA. AA*RR, AA*GG, AA*BB].
These multiplications have to be done with floats instead of uint32_t's, and the resulting values have
to be converted to be relative to the 0-255 interval, where 255 is 1.00 and anything between 0 and 255 is 0.xx. */
void alpha_premultiply_ARGB8888 (uint32_t *pixel) {
uint32_t A, R, G, B;
/* Component bytes extraction. */
A = (*pixel >> (3 << 3)) & 0xFF;
R = (*pixel >> (2 << 3)) & 0xFF;
G = (*pixel >> (1 << 3)) & 0xFF;
B = (*pixel >> (0 << 3)) & 0xFF;
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
/* Alpha pre-multiplication of each component. */
R = (float)A * ((float)R /255);
G = (float)A * ((float)G /255);
B = (float)A * ((float)B /255);
/* ARGB8888 pixel recomposition. */
(*pixel) = (((uint32_t)A << 24) | ((uint32_t)R << 16) | ((uint32_t)G << 8)) | ((uint32_t)B << 0);
}
static SDL_Cursor *
KMSDRM_CreateDefaultCursor(void)
{
return SDL_CreateCursor(default_cdata, default_cmask, DEFAULT_CWIDTH, DEFAULT_CHEIGHT, DEFAULT_CHOTX, DEFAULT_CHOTY);
}
/* Create a GBM cursor from a surface, which means creating a hardware cursor.
Most programs use software cursors, but protracker-clone for example uses
an optional hardware cursor. */
static SDL_Cursor *
KMSDRM_CreateCursor(SDL_Surface * surface, int hot_x, int hot_y)
{
SDL_VideoDevice *dev = SDL_GetVideoDevice();
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
SDL_VideoData *viddata = ((SDL_VideoData *)dev->driverdata);
KMSDRM_CursorData *curdata;
SDL_Cursor *cursor;
uint64_t usable_cursor_w, usable_cursor_h;
uint32_t bo_stride, pixel;
uint32_t *buffer = NULL;
size_t bufsize;
unsigned int i, j;
/* All code below assumes ARGB8888 format for the cursor surface, like other backends do.
Also, the GBM BO pixels have to be alpha-premultiplied, but the SDL surface we receive has
straight-alpha pixels, so we always have to convert. */
SDL_assert(surface->format->format == SDL_PIXELFORMAT_ARGB8888);
SDL_assert(surface->pitch == surface->w * 4);
if (!KMSDRM_gbm_device_is_format_supported(viddata->gbm_dev, GBM_FORMAT_ARGB8888,
GBM_BO_USE_CURSOR | GBM_BO_USE_WRITE))
{
SDL_SetError("Unsupported pixel format for cursor");
return NULL;
}
cursor = (SDL_Cursor *) SDL_calloc(1, sizeof(*cursor));
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (!cursor) {
SDL_OutOfMemory();
return NULL;
}
curdata = (KMSDRM_CursorData *) SDL_calloc(1, sizeof(*curdata));
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (!curdata) {
SDL_OutOfMemory();
SDL_free(cursor);
return NULL;
}
/* Find out what GBM cursor size is recommended by the driver. */
if (KMSDRM_drmGetCap(viddata->drm_fd, DRM_CAP_CURSOR_WIDTH, &usable_cursor_w) ||
KMSDRM_drmGetCap(viddata->drm_fd, DRM_CAP_CURSOR_HEIGHT, &usable_cursor_h))
{
SDL_SetError("Could not get the recommended GBM cursor size");
goto cleanup;
}
if (usable_cursor_w == 0 || usable_cursor_h == 0) {
SDL_SetError("Could not get an usable GBM cursor size");
goto cleanup;
}
/* hox_x and hot_y are the coordinates of the "tip of the cursor" from it's base. */
curdata->hot_x = hot_x;
curdata->hot_y = hot_y;
curdata->w = usable_cursor_w;
curdata->h = usable_cursor_h;
curdata->bo = KMSDRM_gbm_bo_create(viddata->gbm_dev, usable_cursor_w, usable_cursor_h,
GBM_FORMAT_ARGB8888, GBM_BO_USE_CURSOR | GBM_BO_USE_WRITE);
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (!curdata->bo) {
SDL_SetError("Could not create GBM cursor BO");
goto cleanup;
}
bo_stride = KMSDRM_gbm_bo_get_stride(curdata->bo);
bufsize = bo_stride * curdata->h;
/* Always use a temp buffer: it serves the purpose of storing the
alpha-premultiplied pixels (so we can copy them to the gbm BO
with a single gb_bo_write() call), and also copying from the
SDL surface, line by line, to a gbm BO with different pitch. */
buffer = (uint32_t*)SDL_malloc(bufsize);
if (!buffer) {
SDL_OutOfMemory();
goto cleanup;
}
if (SDL_MUSTLOCK(surface)) {
if (SDL_LockSurface(surface) < 0) {
/* Could not lock surface */
goto cleanup;
}
}
/* Clean the whole temporary buffer. */
SDL_memset(buffer, 0x00, bo_stride * curdata->h);
/* Copy from SDL surface to buffer, pre-multiplying by alpha each pixel as we go. */
for (i = 0; i < surface->h; i++) {
for (j = 0; j < surface->w; j++) {
pixel = ((uint32_t*)surface->pixels)[i * surface->w + j];
alpha_premultiply_ARGB8888 (&pixel);
SDL_memcpy(buffer + (i * curdata->w) + j, &pixel, 4);
}
}
if (SDL_MUSTLOCK(surface)) {
SDL_UnlockSurface(surface);
}
if (KMSDRM_gbm_bo_write(curdata->bo, buffer, bufsize)) {
SDL_SetError("Could not write to GBM cursor BO");
goto cleanup;
}
/* Free temporary buffer */
SDL_free(buffer);
buffer = NULL;
cursor->driverdata = curdata;
return cursor;
cleanup:
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (buffer) {
SDL_free(buffer);
}
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (cursor) {
SDL_free(cursor);
}
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (curdata) {
if (curdata->bo) {
KMSDRM_gbm_bo_destroy(curdata->bo);
}
SDL_free(curdata);
}
return NULL;
}
/* Show the specified cursor, or hide if cursor is NULL.
cur_cursor is the current cursor, and cursor is the new cursor.
A cursor is displayed on a display, so we have to add a pointer to dispdata
to the driverdata
*/
static int
KMSDRM_ShowCursor(SDL_Cursor * cursor)
{
SDL_VideoDevice *video_device = SDL_GetVideoDevice();
//SDL_VideoData *viddata = ((SDL_VideoData *)dev->driverdata);
SDL_Mouse *mouse;
KMSDRM_CursorData *curdata;
SDL_VideoDisplay *display = NULL;
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
SDL_DisplayData *dispdata = NULL;
KMSDRM_FBInfo *fb;
KMSDRM_PlaneInfo info = {0};
mouse = SDL_GetMouse();
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (!mouse) {
return SDL_SetError("No mouse.");
}
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (mouse->focus) {
display = SDL_GetDisplayForWindow(mouse->focus);
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (display) {
dispdata = (SDL_DisplayData*) display->driverdata;
}
}
/**********************************/
/* if cursor == NULL, HIDE cursor */
/**********************************/
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (!cursor) {
/* Hide CURRENT cursor, a cursor that is already on screen
2020-09-12 15:58:47 +02:00
and SDL is stored in mouse->cur_cursor. */
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (mouse->cur_cursor && mouse->cur_cursor->driverdata) {
if (dispdata && dispdata->cursor_plane) {
info.plane = dispdata->cursor_plane; /* The rest of the members are zeroed. */
drm_atomic_set_plane_props(&info);
if (drm_atomic_commit(display->device, SDL_TRUE))
return SDL_SetError("Failed atomic commit in KMSDRM_ShowCursor.");
}
return 0;
}
return SDL_SetError("Couldn't find cursor to hide.");
}
/************************************************/
/* If cursor != NULL, DO show cursor on display */
/************************************************/
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (!display) {
return SDL_SetError("Could not get display for mouse.");
}
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (!dispdata) {
return SDL_SetError("Could not get display driverdata.");
}
if (!dispdata->cursor_plane) {
return SDL_SetError("Hardware cursor plane not initialized.");
}
curdata = (KMSDRM_CursorData *) cursor->driverdata;
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (!curdata || !curdata->bo) {
return SDL_SetError("Cursor not initialized properly.");
}
curdata->crtc_id = dispdata->crtc->crtc->crtc_id;
curdata->plane = dispdata->cursor_plane;
curdata->video = video_device;
fb = KMSDRM_FBFromBO(curdata->video, curdata->bo);
info.plane = dispdata->cursor_plane;
info.crtc_id = curdata->crtc_id;
info.fb_id = fb->fb_id;
info.src_w = curdata->w;
info.src_h = curdata->h;
info.crtc_x = mouse->x - curdata->hot_x;
info.crtc_y = mouse->y - curdata->hot_y;
info.crtc_w = curdata->w;
info.crtc_h = curdata->h;
2020-09-12 15:58:47 +02:00
drm_atomic_set_plane_props(&info);
if (drm_atomic_commit(display->device, SDL_TRUE)) {
return SDL_SetError("Failed atomic commit in KMSDRM_ShowCursor.");
}
return 0;
}
/* Unset the cursor from the cursor plane, and ONLY WHEN THAT'S DONE,
DONE FOR REAL, and not only requested, destroy it by destroying the curso BO.
Destroying the cursor BO is an special an delicate situation,
because drm_atomic_set_plane_props() returns immediately, and we DON'T
want to get to gbm_bo_destroy() before the prop changes requested
in drm_atomic_set_plane_props() have effectively been done. So we
issue a BLOCKING atomic_commit here to avoid that situation.
REMEMBER you yan issue an atomic_commit whenever you want, and
the changes requested until that moment (for any planes, crtcs, etc.)
will be done. */
static void
KMSDRM_FreeCursor(SDL_Cursor * cursor)
{
KMSDRM_CursorData *curdata = NULL;
2020-09-12 15:58:47 +02:00
SDL_VideoDevice *video_device = SDL_GetVideoDevice();
KMSDRM_PlaneInfo info = {0};
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (cursor) {
curdata = (KMSDRM_CursorData *) cursor->driverdata;
2020-09-12 15:58:47 +02:00
if (video_device && curdata->bo && curdata->plane) {
info.plane = curdata->plane; /* The other members are zeroed. */
drm_atomic_set_plane_props(&info);
/* Wait until the cursor is unset from the cursor plane before destroying it's BO. */
2020-09-12 15:58:47 +02:00
if (drm_atomic_commit(video_device, SDL_TRUE)) {
SDL_SetError("Failed atomic commit in KMSDRM_FreeCursor.");
}
KMSDRM_gbm_bo_destroy(curdata->bo);
curdata->bo = NULL;
}
2020-09-12 15:58:47 +02:00
/* Even if the cursor is not ours, free it. */
SDL_free(cursor->driverdata);
SDL_free(cursor);
}
}
/* Warp the mouse to (x,y) */
static void
KMSDRM_WarpMouse(SDL_Window * window, int x, int y)
{
/* Only one global/fullscreen window is supported */
KMSDRM_WarpMouseGlobal(x, y);
}
/* Warp the mouse to (x,y) */
static int
KMSDRM_WarpMouseGlobal(int x, int y)
{
KMSDRM_CursorData *curdata;
SDL_Mouse *mouse = SDL_GetMouse();
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (mouse && mouse->cur_cursor && mouse->cur_cursor->driverdata) {
/* Update internal mouse position. */
SDL_SendMouseMotion(mouse->focus, mouse->mouseID, 0, x, y);
/* And now update the cursor graphic position on screen. */
curdata = (KMSDRM_CursorData *) mouse->cur_cursor->driverdata;
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (curdata->bo) {
if (drm_atomic_movecursor(curdata, x, y)) {
return SDL_SetError("drm_atomic_movecursor() failed.");
}
} else {
return SDL_SetError("Cursor not initialized properly.");
}
} else {
return SDL_SetError("No mouse or current cursor.");
}
return 0;
}
void
KMSDRM_InitMouse(_THIS)
{
/* FIXME: Using UDEV it should be possible to scan all mice
* but there's no point in doing so as there's no multimice support...yet!
*/
SDL_DisplayData *dispdata = (SDL_DisplayData *)SDL_GetDisplayDriverData(0);
SDL_Mouse *mouse = SDL_GetMouse();
mouse->CreateCursor = KMSDRM_CreateCursor;
mouse->ShowCursor = KMSDRM_ShowCursor;
mouse->MoveCursor = KMSDRM_MoveCursor;
mouse->FreeCursor = KMSDRM_FreeCursor;
mouse->WarpMouse = KMSDRM_WarpMouse;
mouse->WarpMouseGlobal = KMSDRM_WarpMouseGlobal;
/* Init cursor plane, if we haven't yet. */
if (!dispdata->cursor_plane) {
setup_plane(_this, &(dispdata->cursor_plane), DRM_PLANE_TYPE_CURSOR);
}
SDL_SetDefaultCursor(KMSDRM_CreateDefaultCursor());
}
void
KMSDRM_QuitMouse(_THIS)
{
/* Free the plane on which the cursor was being shown. */
SDL_DisplayData *dispdata = (SDL_DisplayData *)SDL_GetDisplayDriverData(0);
free_plane(&dispdata->cursor_plane);
}
/* This is called when a mouse motion event occurs */
static void
KMSDRM_MoveCursor(SDL_Cursor * cursor)
{
SDL_Mouse *mouse = SDL_GetMouse();
KMSDRM_CursorData *curdata;
/* We must NOT call SDL_SendMouseMotion() here or we will enter recursivity!
That's why we move the cursor graphic ONLY. */
Fixed bug 4966 - KMSDRM: Add dynamic modeset support Anthony Pesch * Remove triple buffering support. As far as I can tell, this goes against the libdrm API; the EGL implementations themselves control the buffering. Removing it isn't absolutely necessary as it seemingly works on the Pi at least, but I noticed this while doing my work and explained my reasoning in the commit. * Replace the crtc_ready logic which allocates an extra bo to perform the initial CRTC configuration (which is required before calling drmModePageFlip) with a call to drmModeSetCrtc after the front and back buffers are allocated, avoiding this allocation. * Standardized the SDL_*Data variable names and null checks to improve readability. Given that there were duplicate fields in each SDL_*Data structure, having generic names such as "data" at times was very confusing. * Removed unused fields from the SDL_*Data structures and moves all display related fields out of SDL_VideoData and into SDL_DisplayData. Not required since the code only supports a single display right now, but this was helpful in reading and understanding the code initially. * Implement KMSDRM_GetDisplayModes / KMSDRM_SetDisplayMode to provide dynamic modeset support. These changes have been tested on a Raspberry Pi 4 and a Dell XPS laptop with an HD 520. As an update, I went back over the triple buffer changes and left them in. I didn't entirely get the code originally, I had just seen it calling KMSDRM_gbm_surface_lock_front_buffer twice for a single swap and had removed it because I was paranoid of bugs stemming from it while working on the modeset changes. I've made a few small changes to the logic that had thrown me off originally and rebased the changes: * The condition wrapping the call to release buffer was incorrect. * The first call to KMSDRM_gbm_surface_lock_front_buffer has been removed. I don't understand why it existed. * Added additional comments describing what was going on in the code (as it does fix the buffer release pattern of the original code before it).
2020-02-09 20:44:22 +01:00
if (mouse && mouse->cur_cursor && mouse->cur_cursor->driverdata) {
curdata = (KMSDRM_CursorData *) mouse->cur_cursor->driverdata;
/* Some programs expect cursor movement even while they don't do SwapWindow() calls,
and since we ride on the atomic_commit() in SwapWindow() for cursor movement,
cursor won't move in these situations. We could do an atomic_commit() for each
cursor movement request, but it cripples the movement to 30FPS, so a future solution
is needed. SDLPoP "QUIT?" menu is an example of this situation. */
if (drm_atomic_movecursor(curdata, mouse->x, mouse->y)) {
SDL_SetError("drm_atomic_movecursor() failed.");
}
}
}
#endif /* SDL_VIDEO_DRIVER_KMSDRM */
/* vi: set ts=4 sw=4 expandtab: */