/*
- * Copyright (C) 1991, NeXT Computer, Inc. All Rights Reserverd.
- *
- * @APPLE_LICENSE_HEADER_START@
- *
- * The contents of this file constitute Original Code as defined in and
- * are subject to the Apple Public Source License Version 1.1 (the
- * "License"). You may not use this file except in compliance with the
- * License. Please obtain a copy of the License at
- * http://www.apple.com/publicsource and read it before using this file.
- *
- * This Original Code and all software distributed under the License are
- * distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER
- * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
- * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT. Please see the
- * License for the specific language governing rights and limitations
- * under the License.
- *
- * @APPLE_LICENSE_HEADER_END@
- *
+ * Copyright (C) 1991, NeXT Computer, Inc. All Rights Reserverd.
*
* File: fsx.c
* Author: Avadis Tevanian, Jr.
*
* Small changes to work under Linux -- davej.
*
+ * Checks for mmap last-page zero fill.
*/
-#include <sys/types.h>
-#include <sys/stat.h>
-#include <sys/param.h>
+#include "global.h"
+
#include <limits.h>
#include <time.h>
#include <strings.h>
#include <sys/file.h>
#include <sys/mman.h>
-#include <limits.h>
+#ifdef HAVE_ERR_H
#include <err.h>
+#endif
#include <signal.h>
#include <stdio.h>
+#include <stddef.h>
#include <stdlib.h>
#include <string.h>
-#include <unistd.h>
#include <stdarg.h>
#include <errno.h>
+#ifdef AIO
+#include <libaio.h>
+#endif
#ifndef MAP_FILE
# define MAP_FILE 0
int logcount = 0; /* total ops */
/*
- * Define operations
+ * The operation matrix is complex due to conditional execution of different
+ * features. Hence when we come to deciding what operation to run, we need to
+ * be careful in how we select the different operations. The active operations
+ * are mapped to numbers as follows:
+ *
+ * lite !lite
+ * READ: 0 0
+ * WRITE: 1 1
+ * MAPREAD: 2 2
+ * MAPWRITE: 3 3
+ * TRUNCATE: - 4
+ * FALLOCATE: - 5
+ * PUNCH HOLE: - 6
+ * ZERO RANGE: - 7
+ *
+ * When mapped read/writes are disabled, they are simply converted to normal
+ * reads and writes. When fallocate/fpunch calls are disabled, they are
+ * converted to OP_SKIPPED. Hence OP_SKIPPED needs to have a number higher than
+ * the operation selction matrix, as does the OP_CLOSEOPEN which is an
+ * operation modifier rather than an operation in itself.
+ *
+ * Because of the "lite" version, we also need to have different "maximum
+ * operation" defines to allow the ops to be selected correctly based on the
+ * mode being run.
*/
-#define OP_READ 1
-#define OP_WRITE 2
-#define OP_TRUNCATE 3
-#define OP_CLOSEOPEN 4
-#define OP_MAPREAD 5
-#define OP_MAPWRITE 6
-#define OP_SKIPPED 7
-
-#ifndef PAGE_SIZE
+/* common operations */
+#define OP_READ 0
+#define OP_WRITE 1
+#define OP_MAPREAD 2
+#define OP_MAPWRITE 3
+#define OP_MAX_LITE 4
+
+/* !lite operations */
+#define OP_TRUNCATE 4
+#define OP_FALLOCATE 5
+#define OP_PUNCH_HOLE 6
+#define OP_ZERO_RANGE 7
+#define OP_COLLAPSE_RANGE 8
+#define OP_MAX_FULL 9
+
+/* operation modifiers */
+#define OP_CLOSEOPEN 100
+#define OP_SKIPPED 101
+
+#undef PAGE_SIZE
#define PAGE_SIZE getpagesize()
-#endif
+#undef PAGE_MASK
#define PAGE_MASK (PAGE_SIZE - 1)
char *original_buf; /* a pointer to the original data */
char *fname; /* name of our test file */
int fd; /* fd for our test file */
+blksize_t block_size = 0;
off_t file_size = 0;
off_t biggest = 0;
char state[256];
int closeprob = 0; /* -c flag */
int debug = 0; /* -d flag */
unsigned long debugstart = 0; /* -D flag */
+int flush = 0; /* -f flag */
+int do_fsync = 0; /* -y flag */
unsigned long maxfilelen = 256 * 1024; /* -l flag */
int sizechecks = 1; /* -n flag disables them */
int maxoplen = 64 * 1024; /* -o flag */
unsigned long progressinterval = 0; /* -p flag */
int readbdy = 1; /* -r flag */
int style = 0; /* -s flag */
+int prealloc = 0; /* -x flag */
int truncbdy = 1; /* -t flag */
int writebdy = 1; /* -w flag */
long monitorstart = -1; /* -m flag */
int randomoplen = 1; /* -O flag disables it */
int seed = 1; /* -S flag */
int mapped_writes = 1; /* -W flag disables */
+int fallocate_calls = 1; /* -F flag disables */
+int punch_hole_calls = 1; /* -H flag disables */
+int zero_range_calls = 1; /* -z flag disables */
+int collapse_range_calls = 1; /* -C flag disables */
int mapped_reads = 1; /* -R flag disables it */
int fsxgoodfd = 0;
+int o_direct; /* -Z */
+int aio = 0;
+
+int page_size;
+int page_mask;
+int mmap_mask;
+#ifdef AIO
+int aio_rw(int rw, int fd, char *buf, unsigned len, unsigned offset);
+#define READ 0
+#define WRITE 1
+#define fsxread(a,b,c,d) aio_rw(READ, a,b,c,d)
+#define fsxwrite(a,b,c,d) aio_rw(WRITE, a,b,c,d)
+#else
+#define fsxread(a,b,c,d) read(a,b,c)
+#define fsxwrite(a,b,c,d) write(a,b,c)
+#endif
+
FILE * fsxlogf = NULL;
int badoff = -1;
int closeopen = 0;
+static void *round_ptr_up(void *ptr, unsigned long align, unsigned long offset)
+{
+ unsigned long ret = (unsigned long)ptr;
+
+ ret = ((ret + align - 1) & ~(align - 1));
+ ret += offset;
+ return (void *)ret;
+}
+
+void
+vwarnc(int code, const char *fmt, va_list ap) {
+ fprintf(stderr, "fsx: ");
+ if (fmt != NULL) {
+ vfprintf(stderr, fmt, ap);
+ fprintf(stderr, ": ");
+ }
+ fprintf(stderr, "%s\n", strerror(code));
+}
+
+void
+warn(const char * fmt, ...) {
+ va_list ap;
+ va_start(ap, fmt);
+ vwarnc(errno, fmt, ap);
+ va_end(ap);
+}
+
+#define BUF_SIZE 1024
void
prt(char *fmt, ...)
{
va_list args;
+ char buffer[BUF_SIZE];
va_start(args, fmt);
- vfprintf(stdout, fmt, args);
- if (fsxlogf)
- vfprintf(fsxlogf, fmt, args);
+ vsnprintf(buffer, BUF_SIZE, fmt, args);
va_end(args);
+ fprintf(stdout, buffer);
+ if (fsxlogf)
+ fprintf(fsxlogf, buffer);
}
void
{
int i, count, down;
struct log_entry *lp;
+ char *falloc_type[3] = {"PAST_EOF", "EXTENDING", "INTERIOR"};
prt("LOG DUMP (%d total operations):\n", logcount);
if (logcount < LOGSIZE) {
int opnum;
opnum = i+1 + (logcount/LOGSIZE)*LOGSIZE;
- prt("%d(%d mod 256): ", opnum, opnum%256);
+ prt("%d(%3d mod 256): ", opnum, opnum%256);
lp = &oplog[i];
if ((closeopen = lp->operation < 0))
lp->operation = ~ lp->operation;
switch (lp->operation) {
case OP_MAPREAD:
- prt("MAPREAD\t0x%x thru 0x%x\t(0x%x bytes)",
+ prt("MAPREAD 0x%x thru 0x%x\t(0x%x bytes)",
lp->args[0], lp->args[0] + lp->args[1] - 1,
lp->args[1]);
if (badoff >= lp->args[0] && badoff <
prt("\t******WWWW");
break;
case OP_READ:
- prt("READ\t0x%x thru 0x%x\t(0x%x bytes)",
+ prt("READ 0x%x thru 0x%x\t(0x%x bytes)",
lp->args[0], lp->args[0] + lp->args[1] - 1,
lp->args[1]);
if (badoff >= lp->args[0] &&
prt("\t***RRRR***");
break;
case OP_WRITE:
- prt("WRITE\t0x%x thru 0x%x\t(0x%x bytes)",
+ prt("WRITE 0x%x thru 0x%x\t(0x%x bytes)",
lp->args[0], lp->args[0] + lp->args[1] - 1,
lp->args[1]);
if (lp->args[0] > lp->args[2])
badoff < lp->args[!!down])
prt("\t******WWWW");
break;
+ case OP_FALLOCATE:
+ /* 0: offset 1: length 2: where alloced */
+ prt("FALLOC 0x%x thru 0x%x\t(0x%x bytes) %s",
+ lp->args[0], lp->args[0] + lp->args[1],
+ lp->args[1], falloc_type[lp->args[2]]);
+ if (badoff >= lp->args[0] &&
+ badoff < lp->args[0] + lp->args[1])
+ prt("\t******FFFF");
+ break;
+ case OP_PUNCH_HOLE:
+ prt("PUNCH 0x%x thru 0x%x\t(0x%x bytes)",
+ lp->args[0], lp->args[0] + lp->args[1] - 1,
+ lp->args[1]);
+ if (badoff >= lp->args[0] && badoff <
+ lp->args[0] + lp->args[1])
+ prt("\t******PPPP");
+ break;
+ case OP_ZERO_RANGE:
+ prt("ZERO 0x%x thru 0x%x\t(0x%x bytes)",
+ lp->args[0], lp->args[0] + lp->args[1] - 1,
+ lp->args[1]);
+ if (badoff >= lp->args[0] && badoff <
+ lp->args[0] + lp->args[1])
+ prt("\t******ZZZZ");
+ break;
+ case OP_COLLAPSE_RANGE:
+ prt("COLLAPSE 0x%x thru 0x%x\t(0x%x bytes)",
+ lp->args[0], lp->args[0] + lp->args[1] - 1,
+ lp->args[1]);
+ if (badoff >= lp->args[0] && badoff <
+ lp->args[0] + lp->args[1])
+ prt("\t******CCCC");
+ break;
case OP_SKIPPED:
prt("SKIPPED (no operation)");
break;
exit(67);
}
if (lite) {
- off_t size_by_seek = lseek(fd, (off_t)0, L_XTND);
+ off_t size_by_seek = lseek(fd, (off_t)0, SEEK_END);
if (size_by_seek == (off_t)-1)
prterr("save_buffer: lseek eof");
else if (bufferlength > size_by_seek) {
- fprintf(stderr, "save_buffer: .fsxgood file too short... will save 0x%qx bytes instead of 0x%qx\n", (unsigned long long)size_by_seek,
+ warn("save_buffer: .fsxgood file too short... will save 0x%llx bytes instead of 0x%llx\n", (unsigned long long)size_by_seek,
(unsigned long long)bufferlength);
bufferlength = size_by_seek;
}
if (byteswritten == -1)
prterr("save_buffer write");
else
- fprintf(stderr, "save_buffer: short write, 0x%x bytes instead of 0x%qx\n",
+ warn("save_buffer: short write, 0x%x bytes instead of 0x%llx\n",
(unsigned)byteswritten,
(unsigned long long)bufferlength);
}
unsigned op = 0;
unsigned bad = 0;
- if (bcmp(good_buf + offset, temp_buf, size) != 0) {
- prt("READ BAD DATA: offset = 0x%x, size = 0x%x\n",
- offset, size);
+ if (memcmp(good_buf + offset, temp_buf, size) != 0) {
+ prt("READ BAD DATA: offset = 0x%x, size = 0x%x, fname = %s\n",
+ offset, size, fname);
prt("OFFSET\tGOOD\tBAD\tRANGE\n");
while (size > 0) {
c = good_buf[offset];
t = temp_buf[i];
if (c != t) {
- if (n == 0) {
+ if (n < 16) {
bad = short_at(&temp_buf[i]);
prt("0x%5x\t0x%04x\t0x%04x", offset,
short_at(&good_buf[offset]), bad);
op = temp_buf[offset & 1 ? i+1 : i];
+ prt("\t0x%5x\n", n);
+ if (op)
+ prt("operation# (mod 256) for "
+ "the bad data may be %u\n",
+ ((unsigned)op & 0xff));
+ else
+ prt("operation# (mod 256) for "
+ "the bad data unknown, check"
+ " HOLE and EXTEND ops\n");
}
n++;
badoff = offset;
i++;
size--;
}
- if (n) {
- prt("\t0x%5x\n", n);
- if (bad)
- prt("operation# (mod 256) for the bad data may be %u\n", ((unsigned)op & 0xff));
- else
- prt("operation# (mod 256) for the bad data unknown, check HOLE and EXTEND ops\n");
- } else
- prt("????????????????\n");
report_failure(110);
}
}
prterr("check_size: fstat");
statbuf.st_size = -1;
}
- size_by_seek = lseek(fd, (off_t)0, L_XTND);
+ size_by_seek = lseek(fd, (off_t)0, SEEK_END);
if (file_size != statbuf.st_size || file_size != size_by_seek) {
- prt("Size error: expected 0x%qx stat 0x%qx seek 0x%qx\n",
+ prt("Size error: expected 0x%llx stat 0x%llx seek 0x%llx\n",
(unsigned long long)file_size,
(unsigned long long)statbuf.st_size,
(unsigned long long)size_by_seek);
ftruncate(fd, 0);
}
+void
+doflush(unsigned offset, unsigned size)
+{
+ unsigned pg_offset;
+ unsigned map_size;
+ char *p;
+
+ if (o_direct == O_DIRECT)
+ return;
+
+ pg_offset = offset & mmap_mask;
+ map_size = pg_offset + size;
+
+ if ((p = (char *)mmap(0, map_size, PROT_READ | PROT_WRITE,
+ MAP_FILE | MAP_SHARED, fd,
+ (off_t)(offset - pg_offset))) == (char *)-1) {
+ prterr("doflush: mmap");
+ report_failure(202);
+ }
+ if (msync(p, map_size, MS_INVALIDATE) != 0) {
+ prterr("doflush: msync");
+ report_failure(203);
+ }
+ if (munmap(p, map_size) != 0) {
+ prterr("doflush: munmap");
+ report_failure(204);
+ }
+}
void
doread(unsigned offset, unsigned size)
unsigned iret;
offset -= offset % readbdy;
+ if (o_direct)
+ size -= size % readbdy;
if (size == 0) {
- if (!quiet && testcalls > simulatedopcount)
+ if (!quiet && testcalls > simulatedopcount && !o_direct)
prt("skipping zero size read\n");
log4(OP_SKIPPED, OP_READ, offset, size);
return;
if (testcalls <= simulatedopcount)
return;
- if (!quiet && ((progressinterval && !(testcalls % progressinterval)) ||
- (debug &&
- (monitorstart == -1 ||
- (offset + size > monitorstart &&
- (monitorend == -1 || offset <= monitorend))))))
+ if (!quiet &&
+ ((progressinterval && testcalls % progressinterval == 0) ||
+ (debug &&
+ (monitorstart == -1 ||
+ (offset + size > monitorstart &&
+ (monitorend == -1 || offset <= monitorend))))))
prt("%lu read\t0x%x thru\t0x%x\t(0x%x bytes)\n", testcalls,
offset, offset + size - 1, size);
ret = lseek(fd, (off_t)offset, SEEK_SET);
prterr("doread: lseek");
report_failure(140);
}
- iret = read(fd, temp_buf, size);
+ iret = fsxread(fd, temp_buf, size, offset);
if (iret != size) {
if (iret == -1)
prterr("doread: read");
}
+void
+check_eofpage(char *s, unsigned offset, char *p, int size)
+{
+ unsigned long last_page, should_be_zero;
+
+ if (offset + size <= (file_size & ~page_mask))
+ return;
+ /*
+ * we landed in the last page of the file
+ * test to make sure the VM system provided 0's
+ * beyond the true end of the file mapping
+ * (as required by mmap def in 1996 posix 1003.1)
+ */
+ last_page = ((unsigned long)p + (offset & page_mask) + size) & ~page_mask;
+
+ for (should_be_zero = last_page + (file_size & page_mask);
+ should_be_zero < last_page + page_size;
+ should_be_zero++)
+ if (*(char *)should_be_zero) {
+ prt("Mapped %s: non-zero data past EOF (0x%llx) page offset 0x%x is 0x%04x\n",
+ s, file_size - 1, should_be_zero & page_mask,
+ short_at(should_be_zero));
+ report_failure(205);
+ }
+}
+
+
void
domapread(unsigned offset, unsigned size)
{
if (testcalls <= simulatedopcount)
return;
- if (!quiet && ((progressinterval && !(testcalls % progressinterval)) ||
+ if (!quiet &&
+ ((progressinterval && testcalls % progressinterval == 0) ||
(debug &&
- (monitorstart == -1 ||
- (offset + size > monitorstart &&
- ((monitorend == -1 || offset <= monitorend)))))))
+ (monitorstart == -1 ||
+ (offset + size > monitorstart &&
+ (monitorend == -1 || offset <= monitorend))))))
prt("%lu mapread\t0x%x thru\t0x%x\t(0x%x bytes)\n", testcalls,
offset, offset + size - 1, size);
report_failure(190);
}
memcpy(temp_buf, p + pg_offset, size);
+
+ check_eofpage("Read", offset, p, size);
+
if (munmap(p, map_size) != 0) {
prterr("domapread: munmap");
report_failure(191);
unsigned iret;
offset -= offset % writebdy;
+ if (o_direct)
+ size -= size % writebdy;
if (size == 0) {
- if (!quiet && testcalls > simulatedopcount)
+ if (!quiet && testcalls > simulatedopcount && !o_direct)
prt("skipping zero size write\n");
log4(OP_SKIPPED, OP_WRITE, offset, size);
return;
gendata(original_buf, good_buf, offset, size);
if (file_size < offset + size) {
if (file_size < offset)
- bzero(good_buf + file_size, offset - file_size);
+ memset(good_buf + file_size, '\0', offset - file_size);
file_size = offset + size;
if (lite) {
- fprintf(stderr, "Lite file size bug in fsx!");
+ warn("Lite file size bug in fsx!");
report_failure(149);
}
}
if (testcalls <= simulatedopcount)
return;
- if (!quiet && ((progressinterval && !(testcalls % progressinterval)) ||
+
+ if (!quiet &&
+ ((progressinterval && testcalls % progressinterval == 0) ||
(debug &&
- (monitorstart == -1 ||
- (offset + size > monitorstart &&
- ((monitorend == -1 || offset <= monitorend)))))))
+ (monitorstart == -1 ||
+ (offset + size > monitorstart &&
+ (monitorend == -1 || offset <= monitorend))))))
prt("%lu write\t0x%x thru\t0x%x\t(0x%x bytes)\n", testcalls,
offset, offset + size - 1, size);
ret = lseek(fd, (off_t)offset, SEEK_SET);
prterr("dowrite: lseek");
report_failure(150);
}
- iret = write(fd, good_buf + offset, size);
+ iret = fsxwrite(fd, good_buf + offset, size, offset);
if (iret != size) {
if (iret == -1)
prterr("dowrite: write");
iret, size);
report_failure(151);
}
+ if (do_fsync) {
+ if (fsync(fd)) {
+ prt("fsync() failed: %s\n", strerror(errno));
+ report_failure(152);
+ }
+ }
+ if (flush) {
+ doflush(offset, size);
+ }
}
gendata(original_buf, good_buf, offset, size);
if (file_size < offset + size) {
if (file_size < offset)
- bzero(good_buf + file_size, offset - file_size);
+ memset(good_buf + file_size, '\0', offset - file_size);
file_size = offset + size;
if (lite) {
- fprintf(stderr, "Lite file size bug in fsx!");
+ warn("Lite file size bug in fsx!");
report_failure(200);
}
}
if (testcalls <= simulatedopcount)
return;
- if (!quiet && ((progressinterval && !(testcalls % progressinterval)) ||
+ if (!quiet &&
+ ((progressinterval && testcalls % progressinterval == 0) ||
(debug &&
- (monitorstart == -1 ||
- (offset + size > monitorstart &&
- (monitorend == -1 || offset <= monitorend))))))
+ (monitorstart == -1 ||
+ (offset + size > monitorstart &&
+ (monitorend == -1 || offset <= monitorend))))))
prt("%lu mapwrite\t0x%x thru\t0x%x\t(0x%x bytes)\n", testcalls,
offset, offset + size - 1, size);
report_failure(202);
}
memcpy(p + pg_offset, good_buf + offset, size);
- if (msync(p, map_size, 0) != 0) {
+ if (msync(p, map_size, MS_SYNC) != 0) {
prterr("domapwrite: msync");
report_failure(203);
}
+
+ check_eofpage("Write", offset, p, size);
+
if (munmap(p, map_size) != 0) {
prterr("domapwrite: munmap");
report_failure(204);
log4(OP_TRUNCATE, size, (unsigned)file_size, 0);
if (size > file_size)
- bzero(good_buf + file_size, size - file_size);
+ memset(good_buf + file_size, '\0', size - file_size);
file_size = size;
if (testcalls <= simulatedopcount)
return;
- if ((progressinterval && !(testcalls % progressinterval)) ||
+ if ((progressinterval && testcalls % progressinterval == 0) ||
(debug && (monitorstart == -1 || monitorend == -1 ||
size <= monitorend)))
prt("%lu trunc\tfrom 0x%x to 0x%x\n", testcalls, oldsize, size);
}
}
+#ifdef FALLOC_FL_PUNCH_HOLE
+void
+do_punch_hole(unsigned offset, unsigned length)
+{
+ unsigned end_offset;
+ int max_offset = 0;
+ int max_len = 0;
+ int mode = FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE;
+
+ if (length == 0) {
+ if (!quiet && testcalls > simulatedopcount)
+ prt("skipping zero length punch hole\n");
+ log4(OP_SKIPPED, OP_PUNCH_HOLE, offset, length);
+ return;
+ }
+
+ if (file_size <= (loff_t)offset) {
+ if (!quiet && testcalls > simulatedopcount)
+ prt("skipping hole punch off the end of the file\n");
+ log4(OP_SKIPPED, OP_PUNCH_HOLE, offset, length);
+ return;
+ }
+
+ end_offset = offset + length;
+
+ log4(OP_PUNCH_HOLE, offset, length, 0);
+
+ if (testcalls <= simulatedopcount)
+ return;
+
+ if ((progressinterval && testcalls % progressinterval == 0) ||
+ (debug && (monitorstart == -1 || monitorend == -1 ||
+ end_offset <= monitorend))) {
+ prt("%lu punch\tfrom 0x%x to 0x%x, (0x%x bytes)\n", testcalls,
+ offset, offset+length, length);
+ }
+ if (fallocate(fd, mode, (loff_t)offset, (loff_t)length) == -1) {
+ prt("%punch hole: %x to %x\n", offset, length);
+ prterr("do_punch_hole: fallocate");
+ report_failure(161);
+ }
+
+
+ max_offset = offset < file_size ? offset : file_size;
+ max_len = max_offset + length <= file_size ? length :
+ file_size - max_offset;
+ memset(good_buf + max_offset, '\0', max_len);
+}
+
+#else
+void
+do_punch_hole(unsigned offset, unsigned length)
+{
+ return;
+}
+#endif
+
+#ifdef FALLOC_FL_ZERO_RANGE
+void
+do_zero_range(unsigned offset, unsigned length)
+{
+ unsigned end_offset;
+ int mode = FALLOC_FL_ZERO_RANGE;
+ int keep_size;
+
+ if (length == 0) {
+ if (!quiet && testcalls > simulatedopcount)
+ prt("skipping zero length zero range\n");
+ log4(OP_SKIPPED, OP_ZERO_RANGE, offset, length);
+ return;
+ }
+
+ keep_size = random() % 2;
+
+ end_offset = keep_size ? 0 : offset + length;
+
+ if (end_offset > biggest) {
+ biggest = end_offset;
+ if (!quiet && testcalls > simulatedopcount)
+ prt("zero_range to largest ever: 0x%x\n", end_offset);
+ }
+
+ /*
+ * last arg matches fallocate string array index in logdump:
+ * 0: allocate past EOF
+ * 1: extending prealloc
+ * 2: interior prealloc
+ */
+ log4(OP_ZERO_RANGE, offset, length, (end_offset > file_size) ? (keep_size ? 0 : 1) : 2);
+
+ if (testcalls <= simulatedopcount)
+ return;
+
+ if ((progressinterval && testcalls % progressinterval == 0) ||
+ (debug && (monitorstart == -1 || monitorend == -1 ||
+ end_offset <= monitorend))) {
+ prt("%lu zero\tfrom 0x%x to 0x%x, (0x%x bytes)\n", testcalls,
+ offset, offset+length, length);
+ }
+ if (fallocate(fd, mode, (loff_t)offset, (loff_t)length) == -1) {
+ prt("%pzero range: %x to %x\n", offset, length);
+ prterr("do_zero_range: fallocate");
+ report_failure(161);
+ }
+
+ memset(good_buf + offset, '\0', length);
+}
+
+#else
+void
+do_zero_range(unsigned offset, unsigned length)
+{
+ return;
+}
+#endif
+
+#ifdef FALLOC_FL_COLLAPSE_RANGE
+void
+do_collapse_range(unsigned offset, unsigned length)
+{
+ unsigned end_offset;
+ int mode = FALLOC_FL_COLLAPSE_RANGE;
+
+ if (length == 0) {
+ if (!quiet && testcalls > simulatedopcount)
+ prt("skipping zero length collapse range\n");
+ log4(OP_SKIPPED, OP_COLLAPSE_RANGE, offset, length);
+ return;
+ }
+
+ end_offset = offset + length;
+ if ((loff_t)end_offset >= file_size) {
+ if (!quiet && testcalls > simulatedopcount)
+ prt("skipping collapse range behind EOF\n");
+ log4(OP_SKIPPED, OP_COLLAPSE_RANGE, offset, length);
+ return;
+ }
+
+ log4(OP_COLLAPSE_RANGE, offset, length, 0);
+
+ if (testcalls <= simulatedopcount)
+ return;
+
+ if ((progressinterval && testcalls % progressinterval == 0) ||
+ (debug && (monitorstart == -1 || monitorend == -1 ||
+ end_offset <= monitorend))) {
+ prt("%lu collapse\tfrom 0x%x to 0x%x, (0x%x bytes)\n", testcalls,
+ offset, offset+length, length);
+ }
+ if (fallocate(fd, mode, (loff_t)offset, (loff_t)length) == -1) {
+ prt("collapse range: %x to %x\n", offset, length);
+ prterr("do_collapse_range: fallocate");
+ report_failure(161);
+ }
+
+ memmove(good_buf + offset, good_buf + end_offset,
+ file_size - end_offset);
+ file_size -= length;
+}
+
+#else
+void
+do_collapse_range(unsigned offset, unsigned length)
+{
+ return;
+}
+#endif
+
+#ifdef HAVE_LINUX_FALLOC_H
+/* fallocate is basically a no-op unless extending, then a lot like a truncate */
+void
+do_preallocate(unsigned offset, unsigned length)
+{
+ unsigned end_offset;
+ int keep_size;
+
+ if (length == 0) {
+ if (!quiet && testcalls > simulatedopcount)
+ prt("skipping zero length fallocate\n");
+ log4(OP_SKIPPED, OP_FALLOCATE, offset, length);
+ return;
+ }
+
+ keep_size = random() % 2;
+
+ end_offset = keep_size ? 0 : offset + length;
+
+ if (end_offset > biggest) {
+ biggest = end_offset;
+ if (!quiet && testcalls > simulatedopcount)
+ prt("fallocating to largest ever: 0x%x\n", end_offset);
+ }
+
+ /*
+ * last arg matches fallocate string array index in logdump:
+ * 0: allocate past EOF
+ * 1: extending prealloc
+ * 2: interior prealloc
+ */
+ log4(OP_FALLOCATE, offset, length, (end_offset > file_size) ? (keep_size ? 0 : 1) : 2);
+
+ if (end_offset > file_size) {
+ memset(good_buf + file_size, '\0', end_offset - file_size);
+ file_size = end_offset;
+ }
+
+ if (testcalls <= simulatedopcount)
+ return;
+
+ if ((progressinterval && testcalls % progressinterval == 0) ||
+ (debug && (monitorstart == -1 || monitorend == -1 ||
+ end_offset <= monitorend)))
+ prt("%lu falloc\tfrom 0x%x to 0x%x (0x%x bytes)\n", testcalls,
+ offset, offset + length, length);
+ if (fallocate(fd, keep_size ? FALLOC_FL_KEEP_SIZE : 0, (loff_t)offset, (loff_t)length) == -1) {
+ prt("fallocate: %x to %x\n", offset, length);
+ prterr("do_preallocate: fallocate");
+ report_failure(161);
+ }
+}
+#else
+void
+do_preallocate(unsigned offset, unsigned length)
+{
+ return;
+}
+#endif
void
writefileimage()
if (iret == -1)
prterr("writefileimage: write");
else
- prt("short write: 0x%x bytes instead of 0x%qx\n",
+ prt("short write: 0x%x bytes instead of 0x%llx\n",
iret, (unsigned long long)file_size);
report_failure(172);
}
if (lite ? 0 : ftruncate(fd, file_size) == -1) {
- prt("ftruncate2: %qx\n", (unsigned long long)file_size);
+ prt("ftruncate2: %llx\n", (unsigned long long)file_size);
prterr("writefileimage: ftruncate");
report_failure(173);
}
prterr("docloseopen: close");
report_failure(180);
}
- fd = open(fname, O_RDWR, 0);
+ fd = open(fname, O_RDWR|o_direct, 0);
if (fd < 0) {
prterr("docloseopen: open");
report_failure(181);
}
}
+#define TRIM_OFF_LEN(off, len, size) \
+do { \
+ if (size) \
+ (off) %= (size); \
+ else \
+ (off) = 0; \
+ if ((off) + (len) > (size)) \
+ (len) = (size) - (off); \
+} while (0)
void
test(void)
unsigned long offset;
unsigned long size = maxoplen;
unsigned long rv = random();
- unsigned long op = rv % (3 + !lite + mapped_writes);
-
- /* turn off the map read if necessary */
-
- if (op == 2 && !mapped_reads)
- op = 0;
+ unsigned long op;
if (simulatedopcount > 0 && testcalls == simulatedopcount)
writefileimage();
if (!quiet && testcalls < simulatedopcount && testcalls % 100000 == 0)
prt("%lu...\n", testcalls);
- /*
- * READ: op = 0
- * WRITE: op = 1
- * MAPREAD: op = 2
- * TRUNCATE: op = 3
- * MAPWRITE: op = 3 or 4
- */
- if (lite ? 0 : op == 3 && (style & 1) == 0) /* vanilla truncate? */
- dotruncate(random() % maxfilelen);
- else {
- if (randomoplen)
- size = random() % (maxoplen+1);
- if (lite ? 0 : op == 3)
- dotruncate(size);
- else {
- offset = random();
- if (op == 1 || op == (lite ? 3 : 4)) {
- offset %= maxfilelen;
- if (offset + size > maxfilelen)
- size = maxfilelen - offset;
- if (op != 1)
- domapwrite(offset, size);
- else
- dowrite(offset, size);
- } else {
- if (file_size)
- offset %= file_size;
- else
- offset = 0;
- if (offset + size > file_size)
- size = file_size - offset;
- if (op != 0)
- domapread(offset, size);
- else
- doread(offset, size);
- }
+ offset = random();
+ if (randomoplen)
+ size = random() % (maxoplen + 1);
+
+ /* calculate appropriate op to run */
+ if (lite)
+ op = rv % OP_MAX_LITE;
+ else
+ op = rv % OP_MAX_FULL;
+
+ switch (op) {
+ case OP_MAPREAD:
+ if (!mapped_reads)
+ op = OP_READ;
+ break;
+ case OP_MAPWRITE:
+ if (!mapped_writes)
+ op = OP_WRITE;
+ break;
+ case OP_FALLOCATE:
+ if (!fallocate_calls) {
+ log4(OP_SKIPPED, OP_FALLOCATE, offset, size);
+ goto out;
+ }
+ break;
+ case OP_PUNCH_HOLE:
+ if (!punch_hole_calls) {
+ log4(OP_SKIPPED, OP_PUNCH_HOLE, offset, size);
+ goto out;
+ }
+ break;
+ case OP_ZERO_RANGE:
+ if (!zero_range_calls) {
+ log4(OP_SKIPPED, OP_ZERO_RANGE, offset, size);
+ goto out;
+ }
+ break;
+ case OP_COLLAPSE_RANGE:
+ if (!collapse_range_calls) {
+ log4(OP_SKIPPED, OP_COLLAPSE_RANGE, offset, size);
+ goto out;
}
+ break;
}
+
+ switch (op) {
+ case OP_READ:
+ TRIM_OFF_LEN(offset, size, file_size);
+ doread(offset, size);
+ break;
+
+ case OP_WRITE:
+ TRIM_OFF_LEN(offset, size, maxfilelen);
+ dowrite(offset, size);
+ break;
+
+ case OP_MAPREAD:
+ TRIM_OFF_LEN(offset, size, file_size);
+ domapread(offset, size);
+ break;
+
+ case OP_MAPWRITE:
+ TRIM_OFF_LEN(offset, size, maxfilelen);
+ domapwrite(offset, size);
+ break;
+
+ case OP_TRUNCATE:
+ if (!style)
+ size = random() % maxfilelen;
+ dotruncate(size);
+ break;
+
+ case OP_FALLOCATE:
+ TRIM_OFF_LEN(offset, size, maxfilelen);
+ do_preallocate(offset, size);
+ break;
+
+ case OP_PUNCH_HOLE:
+ TRIM_OFF_LEN(offset, size, file_size);
+ do_punch_hole(offset, size);
+ break;
+ case OP_ZERO_RANGE:
+ TRIM_OFF_LEN(offset, size, file_size);
+ do_zero_range(offset, size);
+ break;
+ case OP_COLLAPSE_RANGE:
+ TRIM_OFF_LEN(offset, size, file_size - 1);
+ offset = offset & ~(block_size - 1);
+ size = size & ~(block_size - 1);
+ if (size == 0) {
+ log4(OP_SKIPPED, OP_COLLAPSE_RANGE, offset, size);
+ goto out;
+ }
+ do_collapse_range(offset, size);
+ break;
+ default:
+ prterr("test: unknown operation");
+ report_failure(42);
+ break;
+ }
+
+out:
if (sizechecks && testcalls > simulatedopcount)
check_size();
if (closeopen)
usage(void)
{
fprintf(stdout, "usage: %s",
- "fsx [-dnqLOW] [-b opnum] [-c Prob] [-l flen] [-m start:end] [-o oplen] [-p progressinterval] [-r readbdy] [-s style] [-t truncbdy] [-w writebdy] [-D startingop] [-N numops] [-P dirpath] [-S seed] fname\n\
+ "fsx [-dnqxAFLOWZ] [-b opnum] [-c Prob] [-l flen] [-m start:end] [-o oplen] [-p progressinterval] [-r readbdy] [-s style] [-t truncbdy] [-w writebdy] [-D startingop] [-N numops] [-P dirpath] [-S seed] fname\n\
-b opnum: beginning operation number (default 1)\n\
-c P: 1 in P chance of file close+open at each op (default infinity)\n\
-d: debug output for all operations\n\
+ -f flush and invalidate cache after I/O\n\
-l flen: the upper bound on file size (default 262144)\n\
-m startop:endop: monitor (print debug output) specified byte range (default 0:infinity)\n\
-n: no verifications of file size\n\
-s style: 1 gives smaller truncates (default 0)\n\
-t truncbdy: 4096 would make truncates page aligned (default 1)\n\
-w writebdy: 4096 would make writes page aligned (default 1)\n\
- -D startingop: debug output starting at specified operation\n\
- -L: fsxLite - no file creations & no file size changes\n\
+ -x: preallocate file space before starting, XFS only (default 0)\n\
+ -y synchronize changes to a file\n"
+
+#ifdef AIO
+" -A: Use the AIO system calls\n"
+#endif
+" -D startingop: debug output starting at specified operation\n"
+#ifdef HAVE_LINUX_FALLOC_H
+" -F: Do not use fallocate (preallocation) calls\n"
+#endif
+#ifdef FALLOC_FL_PUNCH_HOLE
+" -H: Do not use punch hole calls\n"
+#endif
+#ifdef FALLOC_FL_ZERO_RANGE
+" -z: Do not use zero range calls\n"
+#endif
+#ifdef FALLOC_FL_COLLAPSE_RANGE
+" -C: Do not use collapse range calls\n"
+#endif
+" -L: fsxLite - no file creations & no file size changes\n\
-N numops: total # operations to do (default infinity)\n\
-O: use oplen (see -o flag) for every op (default random)\n\
-P: save .fsxlog and .fsxgood files in dirpath (default ./)\n\
-S seed: for random # generator (default 1) 0 gets timestamp\n\
-W: mapped write operations DISabled\n\
-R: read() system calls only (mapped reads disabled)\n\
+ -Z: O_DIRECT (use -R, -W, -r and -w too)\n\
fname: this filename is REQUIRED (no default)\n");
exit(90);
}
int
getnum(char *s, char **e)
{
- int ret = -1;
+ int ret;
*e = (char *) 0;
ret = strtol(s, e, 0);
return (ret);
}
+#ifdef AIO
+
+#define QSZ 1024
+io_context_t io_ctx;
+struct iocb iocb;
+
+int aio_setup()
+{
+ int ret;
+ ret = io_queue_init(QSZ, &io_ctx);
+ if (ret != 0) {
+ fprintf(stderr, "aio_setup: io_queue_init failed: %s\n",
+ strerror(ret));
+ return(-1);
+ }
+ return(0);
+}
+
+int
+__aio_rw(int rw, int fd, char *buf, unsigned len, unsigned offset)
+{
+ struct io_event event;
+ static struct timespec ts;
+ struct iocb *iocbs[] = { &iocb };
+ int ret;
+ long res;
+
+ if (rw == READ) {
+ io_prep_pread(&iocb, fd, buf, len, offset);
+ } else {
+ io_prep_pwrite(&iocb, fd, buf, len, offset);
+ }
+
+ ts.tv_sec = 30;
+ ts.tv_nsec = 0;
+ ret = io_submit(io_ctx, 1, iocbs);
+ if (ret != 1) {
+ fprintf(stderr, "errcode=%d\n", ret);
+ fprintf(stderr, "aio_rw: io_submit failed: %s\n",
+ strerror(ret));
+ goto out_error;
+ }
+
+ ret = io_getevents(io_ctx, 1, 1, &event, &ts);
+ if (ret != 1) {
+ if (ret == 0)
+ fprintf(stderr, "aio_rw: no events available\n");
+ else {
+ fprintf(stderr, "errcode=%d\n", -ret);
+ fprintf(stderr, "aio_rw: io_getevents failed: %s\n",
+ strerror(-ret));
+ }
+ goto out_error;
+ }
+ if (len != event.res) {
+ /*
+ * The b0rked libaio defines event.res as unsigned.
+ * However the kernel strucuture has it signed,
+ * and it's used to pass negated error value.
+ * Till the library is fixed use the temp var.
+ */
+ res = (long)event.res;
+ if (res >= 0)
+ fprintf(stderr, "bad io length: %lu instead of %u\n",
+ res, len);
+ else {
+ fprintf(stderr, "errcode=%ld\n", -res);
+ fprintf(stderr, "aio_rw: async io failed: %s\n",
+ strerror(-res));
+ ret = res;
+ goto out_error;
+ }
+
+ }
+ return event.res;
+
+out_error:
+ /*
+ * The caller expects error return in traditional libc
+ * convention, i.e. -1 and the errno set to error.
+ */
+ errno = -ret;
+ return -1;
+}
+
+int aio_rw(int rw, int fd, char *buf, unsigned len, unsigned offset)
+{
+ int ret;
+
+ if (aio) {
+ ret = __aio_rw(rw, fd, buf, len, offset);
+ } else {
+ if (rw == READ)
+ ret = read(fd, buf, len);
+ else
+ ret = write(fd, buf, len);
+ }
+ return ret;
+}
+
+#endif
+
+int
+test_fallocate(int mode)
+{
+#ifdef HAVE_LINUX_FALLOC_H
+ int ret = 0;
+ if (!lite) {
+ if (fallocate(fd, mode, 0, 1) && errno == EOPNOTSUPP) {
+ if(!quiet)
+ warn("main: filesystem does not support "
+ "fallocate mode 0x%x, disabling!\n", mode);
+ } else {
+ ret = 1;
+ ftruncate(fd, 0);
+ }
+ }
+ return ret;
+#endif
+}
int
main(int argc, char **argv)
char *endp;
char goodfile[1024];
char logfile[1024];
+ struct stat statbuf;
goodfile[0] = 0;
logfile[0] = 0;
+ page_size = getpagesize();
+ page_mask = page_size - 1;
+ mmap_mask = page_mask;
+
+
setvbuf(stdout, (char *)0, _IOLBF, 0); /* line buffered stdout */
- while ((ch = getopt(argc, argv, "b:c:dl:m:no:p:qr:s:t:w:D:LN:OP:RS:W"))
+ while ((ch = getopt(argc, argv, "b:c:dfl:m:no:p:qr:s:t:w:xyAD:FHzCLN:OP:RS:WZ"))
!= EOF)
switch (ch) {
case 'b':
case 'd':
debug = 1;
break;
+ case 'f':
+ flush = 1;
+ break;
case 'l':
maxfilelen = getnum(optarg, &endp);
if (maxfilelen <= 0)
break;
case 'p':
progressinterval = getnum(optarg, &endp);
- if (progressinterval < 0)
+ if (progressinterval == 0)
usage();
break;
case 'q':
if (writebdy <= 0)
usage();
break;
+ case 'x':
+ prealloc = 1;
+ break;
+ case 'y':
+ do_fsync = 1;
+ break;
+ case 'A':
+ aio = 1;
+ break;
case 'D':
debugstart = getnum(optarg, &endp);
if (debugstart < 1)
usage();
break;
+ case 'F':
+ fallocate_calls = 0;
+ break;
+ case 'H':
+ punch_hole_calls = 0;
+ break;
+ case 'z':
+ zero_range_calls = 0;
+ break;
+ case 'C':
+ collapse_range_calls = 0;
+ break;
case 'L':
lite = 1;
break;
if (!quiet)
fprintf(stdout, "mapped writes DISABLED\n");
break;
-
+ case 'Z':
+ o_direct = O_DIRECT;
+ break;
default:
usage();
/* NOTREACHED */
usage();
fname = argv[0];
-
signal(SIGHUP, cleanup);
signal(SIGINT, cleanup);
signal(SIGPIPE, cleanup);
initstate(seed, state, 256);
setstate(state);
- fd = open(fname, O_RDWR|(lite ? 0 : O_CREAT|O_TRUNC), 0666);
+ fd = open(fname,
+ O_RDWR|(lite ? 0 : O_CREAT|O_TRUNC)|o_direct, 0666);
if (fd < 0) {
prterr(fname);
exit(91);
}
+ if (fstat(fd, &statbuf)) {
+ prterr("check_size: fstat");
+ exit(91);
+ }
+ block_size = statbuf.st_blksize;
+#ifdef XFS
+ if (prealloc) {
+ xfs_flock64_t resv = { 0 };
+#ifdef HAVE_XFS_PLATFORM_DEFS_H
+ if (!platform_test_xfs_fd(fd)) {
+ prterr(fname);
+ fprintf(stderr, "main: cannot prealloc, non XFS\n");
+ exit(96);
+ }
+#endif
+ resv.l_len = maxfilelen;
+ if ((xfsctl(fname, fd, XFS_IOC_RESVSP, &resv)) < 0) {
+ prterr(fname);
+ exit(97);
+ }
+ }
+#endif
strncat(goodfile, fname, 256);
strcat (goodfile, ".fsxgood");
fsxgoodfd = open(goodfile, O_RDWR|O_CREAT|O_TRUNC, 0666);
prterr(logfile);
exit(93);
}
+
+#ifdef AIO
+ if (aio)
+ aio_setup();
+#endif
+
if (lite) {
off_t ret;
- file_size = maxfilelen = lseek(fd, (off_t)0, L_XTND);
+ file_size = maxfilelen = lseek(fd, (off_t)0, SEEK_END);
if (file_size == (off_t)-1) {
prterr(fname);
- fprintf(stderr, "main: lseek eof");
+ warn("main: lseek eof");
exit(94);
}
ret = lseek(fd, (off_t)0, SEEK_SET);
if (ret == (off_t)-1) {
prterr(fname);
- fprintf(stderr, "main: lseek 0");
+ warn("main: lseek 0");
exit(95);
}
}
original_buf = (char *) malloc(maxfilelen);
for (i = 0; i < maxfilelen; i++)
original_buf[i] = random() % 256;
- good_buf = (char *) malloc(maxfilelen);
- bzero(good_buf, maxfilelen);
- temp_buf = (char *) malloc(maxoplen);
- bzero(temp_buf, maxoplen);
+ good_buf = (char *) malloc(maxfilelen + writebdy);
+ good_buf = round_ptr_up(good_buf, writebdy, 0);
+ memset(good_buf, '\0', maxfilelen);
+ temp_buf = (char *) malloc(maxoplen + readbdy);
+ temp_buf = round_ptr_up(temp_buf, readbdy, 0);
+ memset(temp_buf, '\0', maxoplen);
if (lite) { /* zero entire existing file */
ssize_t written;
if (written != maxfilelen) {
if (written == -1) {
prterr(fname);
- fprintf(stderr, "main: error on write");
+ warn("main: error on write");
} else
- fprintf(stderr, "main: short write, 0x%x bytes instead of 0x%lx\n",
- (unsigned)written, maxfilelen);
+ warn("main: short write, 0x%x bytes instead "
+ "of 0x%lx\n",
+ (unsigned)written,
+ maxfilelen);
exit(98);
}
} else
check_trunc_hack();
+ if (fallocate_calls)
+ fallocate_calls = test_fallocate(0);
+ if (punch_hole_calls)
+ punch_hole_calls = test_fallocate(FALLOC_FL_PUNCH_HOLE |
+ FALLOC_FL_KEEP_SIZE);
+ if (zero_range_calls)
+ zero_range_calls = test_fallocate(FALLOC_FL_ZERO_RANGE);
+ if (collapse_range_calls)
+ collapse_range_calls = test_fallocate(FALLOC_FL_COLLAPSE_RANGE);
+
while (numops == -1 || numops--)
test();