Fix mmap copying

Instead of attempting to map large files into memory at once, we map
chunks of at most `PHP_STREAM_MMAP_MAX` bytes, and repeat that until we
hit the point where `php_stream_seek()` fails (see bug 54902), and copy
the rest of the file by reading and writing small chunks.

We also fix the mapping behavior for zero bytes on Windows, which did
not error (as with `mmap()`), but would have mapped the remaining file.

Author: cmb69

NEWS

@@ -2,6 +2,10 @@ PHP                                                                        NEWS
 |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 ?? ??? 2020, PHP 7.4.12
 
+- Core:
+  . Fixed bug #80061 (Copying large files may have suboptimal performance).
+    (cmb)
+
 - MySQLnd:
   . Fixed bug #80115 (mysqlnd.debug doesn't recognize absolute paths with
     slashes). (cmb)

main/streams/php_stream_mmap.h

@@ -58,6 +58,8 @@ typedef struct {
 
 #define PHP_STREAM_MMAP_ALL 0
 
+#define PHP_STREAM_MMAP_MAX (512 * 1024 * 1024)
+
 #define php_stream_mmap_supported(stream)	(_php_stream_set_option((stream), PHP_STREAM_OPTION_MMAP_API, PHP_STREAM_MMAP_SUPPORTED, NULL) == 0 ? 1 : 0)
 
 /* Returns 1 if the stream in its current state can be memory mapped,

main/streams/plain_wrapper.c

@@ -826,6 +826,11 @@ static int php_stdiop_set_option(php_stream *stream, int option, int value, void
 							delta = (DWORD)range->offset - loffs;
 						}
 
+						/* MapViewOfFile()ing zero bytes would map to the end of the file; match *nix behavior instead */
+						if (range->length + delta == 0) {
+							return PHP_STREAM_OPTION_RETURN_ERR;
+						}
+
 						data->last_mapped_addr = MapViewOfFile(data->file_mapping, acc, 0, loffs, range->length + delta);
 
 						if (data->last_mapped_addr) {

main/streams/streams.c

@@ -1573,29 +1573,48 @@ PHPAPI int _php_stream_copy_to_stream_ex(php_stream *src, php_stream *dest, size
 
 	if (php_stream_mmap_possible(src)) {
 		char *p;
-		size_t mapped;
 
-		p = php_stream_mmap_range(src, php_stream_tell(src), maxlen, PHP_STREAM_MAP_MODE_SHARED_READONLY, &mapped);
+		do {
+			size_t chunk_size = (maxlen == 0 || maxlen > PHP_STREAM_MMAP_MAX) ? PHP_STREAM_MMAP_MAX : maxlen;
+			size_t mapped;
 
-		if (p) {
-			ssize_t didwrite = php_stream_write(dest, p, mapped);
-			if (didwrite < 0) {
-				*len = 0;
-				return FAILURE;
-			}
+			p = php_stream_mmap_range(src, php_stream_tell(src), chunk_size, PHP_STREAM_MAP_MODE_SHARED_READONLY, &mapped);
+
+			if (p) {
+				ssize_t didwrite;
+
+				if (php_stream_seek(src, mapped, SEEK_CUR) != 0) {
+					php_stream_mmap_unmap(src);
+					break;
+				}
+
+				didwrite = php_stream_write(dest, p, mapped);
+				if (didwrite < 0) {
+					*len = haveread;
+					return FAILURE;
+				}
 
-			php_stream_mmap_unmap_ex(src, mapped);
+				php_stream_mmap_unmap(src);
 
-			*len = didwrite;
+				*len = haveread += didwrite;
 
-			/* we've got at least 1 byte to read
-			 * less than 1 is an error
-			 * AND read bytes match written */
-			if (mapped > 0 && mapped == didwrite) {
-				return SUCCESS;
+				/* we've got at least 1 byte to read
+				 * less than 1 is an error
+				 * AND read bytes match written */
+				if (mapped == 0 || mapped != didwrite) {
+					return FAILURE;
+				}
+				if (mapped < chunk_size) {
+					return SUCCESS;
+				}
+				if (maxlen != 0) {
+					maxlen -= mapped;
+					if (maxlen == 0) {
+						return SUCCESS;
+					}
+				}
 			}
-			return FAILURE;
-		}
+		} while (p);
 	}
 
 	while(1) {