!939 Optimize fopen function

Merge pull request !939 from Wang Yaofeng/fopen_optimization

musl_src.gni

@@ -2175,6 +2175,13 @@ musl_src_porting_file = [
   "src/stdio/fmemopen.c",
   "src/stdio/freopen.c",
   "src/stdio/stdin.c",
+  "src/stdio/__fmodeflags.c",
+  "src/stdio/fopen.c",
+  "src/stdio/ofl.c",
+  "src/stdio/fclose.c",
+  "src/stdio/__toread.c",
+  "src/stdio/__towrite.c",
+  "src/stdio/stderr.c",
   "src/internal/stdio_impl.h",
   "src/internal/vdso.c",
   "src/time/clock_gettime.c",

porting/linux/user/src/internal/stdio_impl.h

@@ -17,6 +17,7 @@
 #define F_ERR 32
 #define F_SVB 64
 #define F_APP 128
+#define F_NOBUF 256
 
 struct _IO_FILE {
 	unsigned flags;
@@ -31,6 +32,14 @@ struct _IO_FILE {
 	off_t (*seek)(FILE *, off_t, int);
 	unsigned char *buf;
 	size_t buf_size;
+	/* when allocating buffer dynamically, base == buf - UNGET,
+	 * free base when calling fclose.
+	 * otherwise, base == NULL, cases:
+	 * 1. in stdout, stdin, stdout, base is static array.
+	 * 2. call setvbuf to set buffer or non-buffer.
+	 * 3. call fmemopen, base == NULL && buf_size != 0.
+	 */
+	unsigned char *base;
 	FILE *prev, *next;
 	int fd;
 	int pipe_pid;
@@ -67,6 +76,7 @@ hidden ssize_t __flush_buffer(FILE *f);
 
 hidden int __toread(FILE *);
 hidden int __towrite(FILE *);
+hidden int __falloc_buf(FILE *);
 
 hidden void __stdio_exit(void);
 hidden void __stdio_exit_needed(void);
@@ -84,11 +94,14 @@ hidden size_t __fwritex(const unsigned char *, size_t, FILE *);
 hidden int __putc_unlocked(int, FILE *);
 
 hidden FILE *__fdopen(int, const char *);
-hidden int __fmodeflags(const char *);
+hidden FILE *__fdopenx(int, int);
+hidden int __fmodeflags(const char *, int *);
 
 hidden FILE *__ofl_add(FILE *f);
 hidden FILE **__ofl_lock(void);
 hidden void __ofl_unlock(void);
+hidden void __ofl_free(FILE *f);
+hidden FILE *__ofl_alloc();
 
 struct __pthread;
 hidden void __register_locked_file(FILE *, struct __pthread *);

porting/linux/user/src/stdio/__fdopen.c

@@ -21,7 +21,7 @@
 #include <string.h>
 #include "libc.h"
 
-static size_t get_bufsize(int fd)
+static size_t __get_bufsize(int fd)
 {
 	struct stat st;
 	size_t buf_size = 0;
@@ -37,50 +37,79 @@ static size_t get_bufsize(int fd)
 	return buf_size;
 }
 
-FILE *__fdopen(int fd, const char *mode)
+int __falloc_buf(FILE *f)
 {
-	FILE *f;
-	struct winsize wsz;
-	size_t buf_size = 0;
-
-	/* Check for valid initial mode character */
-	if (!strchr("rwa", *mode)) {
-		errno = EINVAL;
+	/* return if already allocated, or F_NOBUF set */
+	if (f->buf != NULL || f->buf_size != 0 || f->flags & F_NOBUF) {
 		return 0;
 	}
 
+	/* Default,  base and buf are NULL,and buf_size = 0 */
+	size_t buf_size = 0;
+
 	/* get buffer size via file stat */
-	buf_size = get_bufsize(fd);
+	buf_size = __get_bufsize(f->fd);
 
-	/* Allocate FILE+buffer or fail */
-	if (!(f = malloc(sizeof *f + UNGET + buf_size))) {
-		return 0;
+	/* alloc R/W buffer */
+	f->base = (unsigned char *)malloc(UNGET + buf_size * sizeof(unsigned char));
+	if (!f->base) {
+		errno = -ENOMEM;
+		return errno;
 	}
 
-	/* Zero-fill only the struct, not the buffer */
-	memset(f, 0, sizeof *f);
+	/* reserve UNGET buffer */
+	f->buf = f->base + UNGET;
+	f->buf_size = buf_size;
+
+	return 0;
+}
 
-	/* Impose mode restrictions */
-	if (!strchr(mode, '+')) {
-		f->flags = (*mode == 'r') ? F_NOWR : F_NORD;
+FILE *__fdopen(int fd, const char *mode)
+{
+	FILE *f = NULL;
+	int file_flags = 0;
+	int mode_flags = 0;
+
+	/* Compute the flags to pass to open() */
+	mode_flags = __fmodeflags(mode, &file_flags);
+	if (mode_flags < 0) {
+		return NULL;
 	}
 
-	/* Apply close-on-exec flag */
-	if (strchr(mode, 'e')) {
+	if (mode_flags & O_CLOEXEC) {
 		__syscall(SYS_fcntl, fd, F_SETFD, FD_CLOEXEC);
 	}
 
-	/* Set append mode on fd if opened for append */
-	if (*mode == 'a') {
+	if (mode_flags & O_APPEND) {
 		int flags = __syscall(SYS_fcntl, fd, F_GETFL);
 		if (!(flags & O_APPEND))
 			__syscall(SYS_fcntl, fd, F_SETFL, flags | O_APPEND);
-		f->flags |= F_APP;
 	}
 
+	f = __fdopenx(fd, file_flags);
+	if (f) {
+		return f;
+	}
+
+	return NULL;
+}
+weak_alias(__fdopen, fdopen);
+
+FILE *__fdopenx(int fd, int flags)
+{
+	FILE *f = 0;
+	struct winsize wsz;
+
+	/* Allocate FILE or fail */
+	if (!(f = __ofl_alloc())) {
+		return NULL;
+	}
+
+	/* Zero-fill only the struct, not the buffer */
+	memset(f, 0, sizeof *f);
+
+	f->flags = flags;
 	f->fd = fd;
-	f->buf = (unsigned char *)f + sizeof *f + UNGET;
-	f->buf_size = buf_size;
 
 	/* Activate line buffered mode for terminals */
 	f->lbf = EOF;
@@ -103,4 +132,3 @@ FILE *__fdopen(int fd, const char *mode)
 	return __ofl_add(f);
 }
 
-weak_alias(__fdopen, fdopen);

porting/linux/user/src/stdio/__fmodeflags.c

+#include "stdio_impl.h"
+#include <fcntl.h>
+#include <string.h>
+#include <errno.h>
+
+int __fmodeflags(const char *mode, int *flags)
+{
+	int mode_flags = 0;
+	int options = 0;
+
+	switch (*mode) {
+		case 'r':
+			mode_flags = O_RDONLY;
+			*flags = F_NOWR;
+			break;
+		case 'w':
+			mode_flags = O_WRONLY;
+			options = O_TRUNC | O_CREAT;
+			*flags = F_NORD;
+			break;
+		case 'a':
+			mode_flags = O_WRONLY;
+			options = O_APPEND | O_CREAT;
+			*flags = F_NORD | F_APP;
+			break;
+		default:
+			errno = EINVAL;
+			return -EINVAL;
+	}
+
+	mode++;
+	while (*mode != '\0') {
+		switch (*mode) {
+			case '+':
+				mode_flags = O_RDWR;
+				*flags &= ~(F_NORD | F_NOWR);
+				break;
+			case 'x':
+				/* need O_CREAT check */
+				options |= O_EXCL;
+				break;
+			case 'e':
+				options |= O_CLOEXEC;
+				break;
+			case 'b':
+				break;
+			default:
+				/* only accept "+xeb" */
+				break;
+		}
+		mode++;
+	}
+
+	return mode_flags | options;
+}

porting/linux/user/src/stdio/__toread.c

+#include <stdio_impl.h>
+
+int __toread(FILE *f)
+{
+	f->mode |= f->mode-1;
+	if (f->wpos != f->wbase) f->write(f, 0, 0);
+	f->wpos = f->wbase = f->wend = 0;
+	if (f->flags & F_NORD) {
+		f->flags |= F_ERR;
+		return EOF;
+	}
+
+	/* Alloc file buffer if needed */
+	if (__falloc_buf(f) < 0) {
+		f->flags |= F_ERR;
+		return EOF;
+	}
+
+	f->rpos = f->rend = f->buf + f->buf_size;
+
+	return (f->flags & F_EOF) ? EOF : 0;
+}
+
+hidden void __toread_needs_stdio_exit()
+{
+	__stdio_exit_needed();
+}

porting/linux/user/src/stdio/__towrite.c

+#include "stdio_impl.h"
+
+int __towrite(FILE *f)
+{
+	f->mode |= f->mode-1;
+	if (f->flags & F_NOWR) {
+		f->flags |= F_ERR;
+		return EOF;
+	}
+	/* Clear read buffer (easier than summoning nasal demons) */
+	f->rpos = f->rend = 0;
+
+	/* Alloc file buffer if needed */
+	if (__falloc_buf(f) < 0) {
+		f->flags |= F_ERR;
+		return EOF;
+	}
+
+	/* Activate write through the buffer. */
+	f->wpos = f->wbase = f->buf;
+	f->wend = f->buf + f->buf_size;
+	return 0;
+}
+
+hidden void __towrite_needs_stdio_exit()
+{
+	__stdio_exit_needed();
+}

porting/linux/user/src/stdio/fclose.c

+#include "stdio_impl.h"
+#include <stdlib.h>
+
+static void dummy(FILE *f) { }
+weak_alias(dummy, __unlist_locked_file);
+
+int fclose(FILE *f)
+{
+	int r;
+	
+	FLOCK(f);
+	r = fflush(f);
+	r |= f->close(f);
+	FUNLOCK(f);
+
+	/* Past this point, f is closed and any further explict access
+	 * to it is undefined. However, it still exists as an entry in
+	 * the open file list and possibly in the thread's locked files
+	 * list, if it was closed while explicitly locked. Functions
+	 * which process these lists must tolerate dead FILE objects
+	 * (which necessarily have inactive buffer pointers) without
+	 * producing any side effects. */
+
+	if (f->flags & F_PERM) return r;
+
+	__unlist_locked_file(f);
+
+	free(f->getln_buf);
+	/* release base instead of buf which may be modified by setvbuf
+	 * or iniitalize by local variable */
+	free(f->base);
+	__ofl_free(f);
+
+	return r;
+}

porting/linux/user/src/stdio/fopen.c

+#include "stdio_impl.h"
+#include <fcntl.h>
+#include <string.h>
+#include <errno.h>
+
+FILE *fopen(const char *restrict filename, const char *restrict mode)
+{
+	FILE *f = NULL;
+	int fd = -1;
+	int file_flags = 0;
+	int mode_flags = 0;
+
+	/* Compute the flags to pass to open() */
+	mode_flags = __fmodeflags(mode, &file_flags);
+	if (mode_flags < 0) {
+		return NULL;
+	}
+
+	fd = sys_open(filename, mode_flags, 0666);
+	if (fd < 0) return 0;
+	if (mode_flags & O_CLOEXEC)
+		__syscall(SYS_fcntl, fd, F_SETFD, FD_CLOEXEC);
+
+	f = __fdopenx(fd, file_flags);
+	if (f) {
+		return f;
+	}
+
+	__syscall(SYS_close, fd);
+	return 0;
+}
+
+weak_alias(fopen, fopen64);

porting/linux/user/src/stdio/fread.c

@@ -9,6 +9,7 @@ int __fill_buffer(FILE *f)
 	if (r != 0) {
 		return r;
 	}
+
 	int k = f->readx(f, f->buf, f->buf_size);
 	if (k <= 0) {
 		f->flags |= (k == 0) ? F_EOF : F_ERR;
@@ -32,6 +33,12 @@ size_t fread(void *restrict destv, size_t size, size_t nmemb, FILE *restrict f)
 
 	FLOCK(f);
 
+	/* allocate file buffer if needed */
+	if (__falloc_buf(f) < 0) {
+		f->flags |= F_ERR;
+		goto exit;
+	}
+
 	f->mode |= f->mode-1;
 
 	while (l > 0) {
@@ -47,7 +54,9 @@ size_t fread(void *restrict destv, size_t size, size_t nmemb, FILE *restrict f)
 		if (l == 0) {
 			goto exit;
 		}
-		/* if user buffer is longer than file buffer, read directly */
+		/* if user buffer is longer than file buffer,
+		 * maybe buffer size is 0, non-buffer mode,
+		 * read directly */
 		if (l > f->buf_size) {
 			break;
 		}

porting/linux/user/src/stdio/freopen.c

@@ -12,7 +12,8 @@
 
 FILE *freopen(const char *restrict filename, const char *restrict mode, FILE *restrict f)
 {
-	int fl = __fmodeflags(mode);
+	int file_flags = 0;
+	int fl = __fmodeflags(mode, &file_flags);
 	FILE *f2;
 
 	FLOCK(f);

porting/linux/user/src/stdio/ofl.c

+#include "stdio_impl.h"
+#include "lock.h"
+#include "fork_impl.h"
+#include <stdlib.h>
+
+#define DEFAULT_ALLOC_FILE (8)
+
+static FILE *ofl_head = NULL;
+static FILE *ofl_free = NULL;
+
+static volatile int ofl_lock[1];
+volatile int *const __stdio_ofl_lockptr = ofl_lock;
+
+FILE **__ofl_lock()
+{
+	LOCK(ofl_lock);
+	return &ofl_head;
+}
+
+void __ofl_unlock()
+{
+	UNLOCK(ofl_lock);
+}
+
+FILE *__ofl_alloc()
+{
+	unsigned char *fsb = NULL;
+	size_t cnt = 0;
+	FILE *f = NULL;
+
+	LOCK(ofl_lock);
+	if (ofl_free) {
+		f = ofl_free;
+		ofl_free = ofl_free->next;
+		UNLOCK(ofl_lock);
+
+		return f;
+	}
+	UNLOCK(ofl_lock);
+
+	/* alloc new FILEs(8) */
+	fsb = (unsigned char *)malloc(DEFAULT_ALLOC_FILE * sizeof(FILE));
+	if (!fsb) {
+		return NULL;
+	}
+
+	LOCK(ofl_lock);
+	ofl_free = (FILE*)fsb;
+	ofl_free->prev = NULL;
+	f = ofl_free;
+
+	for (cnt = 1; cnt < DEFAULT_ALLOC_FILE; cnt++) {
+		FILE *tmp = (FILE*)(fsb + cnt * sizeof(FILE));
+		tmp->next = NULL;
+		f->next = tmp;
+		tmp->prev = f;
+		f = f->next;
+	}
+
+	/* reset and move to next free FILE */
+	f = ofl_free;
+	ofl_free = ofl_free->next;
+	if (ofl_free) {
+		ofl_free->prev = NULL;
+	}
+
+	UNLOCK(ofl_lock);
+	return f;
+}
+
+void __ofl_free(FILE *f)
+{
+	LOCK(ofl_lock);
+	if (!f) {
+		return;
+	}
+
+	/* remove from head list */
+	if (f->prev) {
+		f->prev->next = f->next;
+	}
+	if (f->next) {
+		f->next->prev = f->prev;
+	}
+	if (f == ofl_head) {
+		ofl_head = f->next;
+	}
+
+	/* push to free list */
+	f->next = ofl_free;
+	if (ofl_free) {
+		ofl_free->prev = f;
+	}
+	ofl_free = f;
+
+	UNLOCK(ofl_lock);
+}

porting/linux/user/src/stdio/setvbuf.c

+#include "stdio_impl.h"
+
+/* The behavior of this function is undefined except when it is the first
+ * operation on the stream, so the presence or absence of locking is not
+ * observable in a program whose behavior is defined. Thus no locking is
+ * performed here. No allocation of buffers is performed, but a buffer
+ * provided by the caller is used as long as it is suitably sized. */
+
+int setvbuf(FILE *restrict f, char *restrict buf, int type, size_t size)
+{
+	f->lbf = EOF;
+
+	if (type == _IONBF) {
+		f->buf_size = 0;
+		f->flags |= F_NOBUF;
+	} else if (type == _IOLBF || type == _IOFBF) {
+		if (buf && size >= UNGET) {
+			f->buf = (void *)(buf + UNGET);
+			f->buf_size = size - UNGET;
+		}
+		if (type == _IOLBF && f->buf_size)
+			f->lbf = '\n';
+		f->flags &= ~F_NOBUF;
+	} else {
+		return -1;
+	}
+
+	f->flags |= F_SVB;
+
+	return 0;
+}

porting/linux/user/src/stdio/stderr.c

+#include "stdio_impl.h"
+
+#undef stderr
+
+static unsigned char buf[UNGET];
+hidden FILE __stderr_FILE = {
+	.buf = buf+UNGET,
+	.buf_size = 0,
+	.fd = 2,
+	.flags = F_PERM | F_NORD | F_NOBUF,
+	.lbf = -1,
+	.write = __stdio_write,
+	.seek = __stdio_seek,
+	.close = __stdio_close,
+	.lock = -1,
+};
+FILE *const stderr = &__stderr_FILE;
+FILE *volatile __stderr_used = &__stderr_FILE;

porting/linux/user/src/stdio/vfprintf.c

@@ -674,6 +674,13 @@ int vfprintf(FILE *restrict f, const char *restrict fmt, va_list ap)
 	FLOCK(f);
 	olderr = f->flags & F_ERR;
 	if (f->mode < 1) f->flags &= ~F_ERR;
+
+	/* allocate file buffer if need */
+	if (__falloc_buf(f) < 0) {
+		f->flags |= F_ERR;
+		ret = -1;
+	}
+
 	if (!f->buf_size) {
 		saved_buf = f->buf;
 		f->buf = internal_buf;