Fix T77456: Broken vertex paint undo on high-poly objects.

This is a critical fix that should also be backported to 2.83.1

Fairly stupid bug in fact, code detecting changes across undo steps was
assuming that each BHEAD (a block of data in blendfiles) would not be larger
than one memory chunk... Which is the case in alsmost every situation,
besides some super-heavy geometries, and other similar things (images
would also be affected e.g.).
This commit is contained in:
Bastien Montagne 2020-06-06 16:14:11 +02:00 committed by Jeroen Bakker
parent 864f8cbb86
commit e4ac4769f1
Notes: blender-bot 2023-02-14 02:13:08 +01:00
Referenced by issue #77348, Blender LTS: Maintenance Task 2.83
1 changed files with 7 additions and 3 deletions

View File

@ -1222,6 +1222,10 @@ static int fd_read_from_memfile(FileData *filedata,
static MemFileChunk *chunk = NULL;
size_t chunkoffset, readsize, totread;
if (r_is_memchunck_identical != NULL) {
*r_is_memchunck_identical = true;
}
if (size == 0) {
return 0;
}
@ -1272,13 +1276,13 @@ static int fd_read_from_memfile(FileData *filedata,
filedata->file_offset += readsize;
seek += readsize;
if (r_is_memchunck_identical != NULL) {
/* `is_identical` of current chunk represent whether it changed compared to previous undo
/* `is_identical` of current chunk represents whether it changed compared to previous undo
* step. this is fine in redo case (filedata->undo_direction > 0), but not in undo case,
* where we need an extra flag defined when saving the next (future) step after the one we
* want to restore, as we are supposed to 'come from' that future undo step, and not the
* one before current one. */
*r_is_memchunck_identical = filedata->undo_direction > 0 ? chunk->is_identical :
chunk->is_identical_future;
*r_is_memchunck_identical &= filedata->undo_direction > 0 ? chunk->is_identical :
chunk->is_identical_future;
}
} while (totread < size);