bicarusandGitHub
8b2f38307b
## Link to GitHub Issue or related Pull Request, if one exists #0 ## Description of change Significantly speeds up API screen capture and D3D9 screenshots saving. Two reasons for doing this: 1. We now have a 4K game (GITADORA) and existing capture code was taking multiple seconds. 2. Renewed user interest on streaming as we have a couple more companion apps in active development. **API screen capture (streaming), 1280x720:** 14.3ms -> 6.3ms per frame. Back buffer copies go to pooled `D3DPOOL_SYSTEMMEM` surfaces via `GetRenderTargetData` instead of allocating a lockable render target every frame, and TooJpeg is replaced with libjpeg-turbo (encode 9.8ms -> 3.0ms). MSAA remains unsupported **Screenshots for GITADORA arena model, across 4 screens with one of them 4K**: 4068ms -> 124ms. `D3DXSaveSurfaceToFileA` is replaced with fpng (encode 4043ms -> 76ms) and the screens encode in parallel. Dropping D3DX also removes the `d3dx9_43.dll` ... `d3dx9_24.dll` probing loop, so screenshots no longer fail outright on machines with no D3DX9 runtime installed. Screenshot surfaces are read on the present thread, so no D3D call reaches another thread for screenshots. This fixes a hang in DDR X2 introduced earlier in the branch: its device has no internal locking, and reading the surface on a pool thread while the present thread sat inside `GetRenderTargetData` left the game's own render thread deadlocked. ## Testing - **GITADORA** (arena model, D3D9Ex, 4K main plus three subscreens, windowed) with `-screenshotsub`: three sets of four screenshots, images verified correct. Completion order differs between sets, so the screens really are encoding in parallel. - **LovePlus** (KLP, plain D3D9, 768x1360): covers the inline path used by games whose image processing must not leave the present thread. - **API screen capture** through a companion app: live video correct throughout. - **Print Screen** bound as the screenshot key: the clipboard copy succeeded on every shot. - Quitting the game after capturing leaves no `IDirect3DDevice9` reference count warning, so the pooled readback surfaces are released along with the device.
111 lines
3.2 KiB
C
111 lines
3.2 KiB
C
/*
|
|
* jcomapi.c
|
|
*
|
|
* This file was part of the Independent JPEG Group's software:
|
|
* Copyright (C) 1994-1997, Thomas G. Lane.
|
|
* libjpeg-turbo Modifications:
|
|
* Copyright (C) 2024, D. R. Commander.
|
|
* For conditions of distribution and use, see the accompanying README.ijg
|
|
* file.
|
|
*
|
|
* This file contains application interface routines that are used for both
|
|
* compression and decompression.
|
|
*/
|
|
|
|
#define JPEG_INTERNALS
|
|
#include "jinclude.h"
|
|
#include "jpeglib.h"
|
|
|
|
|
|
/*
|
|
* Abort processing of a JPEG compression or decompression operation,
|
|
* but don't destroy the object itself.
|
|
*
|
|
* For this, we merely clean up all the nonpermanent memory pools.
|
|
* Note that temp files (virtual arrays) are not allowed to belong to
|
|
* the permanent pool, so we will be able to close all temp files here.
|
|
* Closing a data source or destination, if necessary, is the application's
|
|
* responsibility.
|
|
*/
|
|
|
|
GLOBAL(void)
|
|
jpeg_abort(j_common_ptr cinfo)
|
|
{
|
|
int pool;
|
|
|
|
/* Do nothing if called on a not-initialized or destroyed JPEG object. */
|
|
if (cinfo->mem == NULL)
|
|
return;
|
|
|
|
/* Releasing pools in reverse order might help avoid fragmentation
|
|
* with some (brain-damaged) malloc libraries.
|
|
*/
|
|
for (pool = JPOOL_NUMPOOLS - 1; pool > JPOOL_PERMANENT; pool--) {
|
|
(*cinfo->mem->free_pool) (cinfo, pool);
|
|
}
|
|
|
|
/* Reset overall state for possible reuse of object */
|
|
if (cinfo->is_decompressor) {
|
|
cinfo->global_state = DSTATE_START;
|
|
/* Try to keep application from accessing now-deleted marker list.
|
|
* A bit kludgy to do it here, but this is the most central place.
|
|
*/
|
|
((j_decompress_ptr)cinfo)->marker_list = NULL;
|
|
((j_decompress_ptr)cinfo)->master->marker_list_end = NULL;
|
|
} else {
|
|
cinfo->global_state = CSTATE_START;
|
|
}
|
|
}
|
|
|
|
|
|
/*
|
|
* Destruction of a JPEG object.
|
|
*
|
|
* Everything gets deallocated except the master jpeg_compress_struct itself
|
|
* and the error manager struct. Both of these are supplied by the application
|
|
* and must be freed, if necessary, by the application. (Often they are on
|
|
* the stack and so don't need to be freed anyway.)
|
|
* Closing a data source or destination, if necessary, is the application's
|
|
* responsibility.
|
|
*/
|
|
|
|
GLOBAL(void)
|
|
jpeg_destroy(j_common_ptr cinfo)
|
|
{
|
|
/* We need only tell the memory manager to release everything. */
|
|
/* NB: mem pointer is NULL if memory mgr failed to initialize. */
|
|
if (cinfo->mem != NULL)
|
|
(*cinfo->mem->self_destruct) (cinfo);
|
|
cinfo->mem = NULL; /* be safe if jpeg_destroy is called twice */
|
|
cinfo->global_state = 0; /* mark it destroyed */
|
|
}
|
|
|
|
|
|
/*
|
|
* Convenience routines for allocating quantization and Huffman tables.
|
|
* (Would jutils.c be a more reasonable place to put these?)
|
|
*/
|
|
|
|
GLOBAL(JQUANT_TBL *)
|
|
jpeg_alloc_quant_table(j_common_ptr cinfo)
|
|
{
|
|
JQUANT_TBL *tbl;
|
|
|
|
tbl = (JQUANT_TBL *)
|
|
(*cinfo->mem->alloc_small) (cinfo, JPOOL_PERMANENT, sizeof(JQUANT_TBL));
|
|
tbl->sent_table = FALSE; /* make sure this is false in any new table */
|
|
return tbl;
|
|
}
|
|
|
|
|
|
GLOBAL(JHUFF_TBL *)
|
|
jpeg_alloc_huff_table(j_common_ptr cinfo)
|
|
{
|
|
JHUFF_TBL *tbl;
|
|
|
|
tbl = (JHUFF_TBL *)
|
|
(*cinfo->mem->alloc_small) (cinfo, JPOOL_PERMANENT, sizeof(JHUFF_TBL));
|
|
tbl->sent_table = FALSE; /* make sure this is false in any new table */
|
|
return tbl;
|
|
}
|