#include "os.h"
#include "mem.h"
#include "mem_kern.h"
+#include "cow.h"
+
+enum ubd_req { UBD_READ, UBD_WRITE, UBD_MMAP };
+
+struct io_thread_req {
+ enum ubd_req op;
+ int fds[2];
+ unsigned long offsets[2];
+ unsigned long long offset;
+ unsigned long length;
+ char *buffer;
+ int sectorsize;
+ unsigned long sector_mask;
+ unsigned long long cow_offset;
+ unsigned long bitmap_words[2];
+ int map_fd;
+ unsigned long long map_offset;
+ int error;
+};
+
+extern int open_ubd_file(char *file, struct openflags *openflags,
+ char **backing_file_out, int *bitmap_offset_out,
+ unsigned long *bitmap_len_out, int *data_offset_out,
+ int *create_cow_out);
+extern int create_cow_file(char *cow_file, char *backing_file,
+ struct openflags flags, int sectorsize,
+ int alignment, int *bitmap_offset_out,
+ unsigned long *bitmap_len_out,
+ int *data_offset_out);
+extern int read_cow_bitmap(int fd, void *buf, int offset, int len);
+extern void do_io(struct io_thread_req *req);
+
+static inline int ubd_test_bit(__u64 bit, unsigned char *data)
+{
+ __u64 n;
+ int bits, off;
+
+ bits = sizeof(data[0]) * 8;
+ n = bit / bits;
+ off = bit % bits;
+ return((data[n] & (1 << off)) != 0);
+}
+
+static inline void ubd_set_bit(__u64 bit, unsigned char *data)
+{
+ __u64 n;
+ int bits, off;
+
+ bits = sizeof(data[0]) * 8;
+ n = bit / bits;
+ off = bit % bits;
+ data[n] |= (1 << off);
+}
+/*End stuff from ubd_user.h*/
-static spinlock_t ubd_io_lock = SPIN_LOCK_UNLOCKED;
-static spinlock_t ubd_lock = SPIN_LOCK_UNLOCKED;
+#define DRIVER_NAME "uml-blkdev"
+
+static DEFINE_SPINLOCK(ubd_io_lock);
+static DEFINE_SPINLOCK(ubd_lock);
static void (*do_ubd)(void);
struct openflags openflags;
int no_cow;
struct cow cow;
+ struct platform_device pdev;
int map_writes;
int map_reads;
static void make_proc_ide(void)
{
- proc_ide_root = proc_mkdir("ide", 0);
+ proc_ide_root = proc_mkdir("ide", NULL);
proc_ide = proc_mkdir("ide0", proc_ide_root);
}
struct ubd *dev;
struct openflags flags = global_openflags;
char *backing_file;
- int n, err;
+ int n, err, i;
if(index_out) *index_out = -1;
n = *str;
}
if(!strcmp(str, "sync")){
- global_openflags.s = 1;
+ global_openflags = of_sync(global_openflags);
return(0);
}
major = simple_strtoul(str, &end, 0);
dev = &ubd_dev[n];
if(dev->file != NULL){
printk(KERN_ERR "ubd_setup : device already configured\n");
- goto out2;
+ goto out;
}
- if(index_out) *index_out = n;
-
- if (*str == 'r'){
- flags.w = 0;
- str++;
- }
- if (*str == 's'){
- flags.s = 1;
- str++;
- }
- if (*str == 'd'){
- dev->no_cow = 1;
+ if (index_out)
+ *index_out = n;
+
+ for (i = 0; i < 4; i++) {
+ switch (*str) {
+ case 'r':
+ flags.w = 0;
+ break;
+ case 's':
+ flags.s = 1;
+ break;
+ case 'd':
+ dev->no_cow = 1;
+ break;
+ case '=':
+ str++;
+ goto break_loop;
+ default:
+ printk(KERN_ERR "ubd_setup : Expected '=' or flag letter (r,s or d)\n");
+ goto out;
+ }
str++;
}
- if(*str++ != '='){
+ if (*str == '=')
+ printk(KERN_ERR "ubd_setup : Too many flags specified\n");
+ else
printk(KERN_ERR "ubd_setup : Expected '='\n");
- goto out2;
- }
+ goto out;
+break_loop:
err = 0;
backing_file = strchr(str, ',');
+
+ if (!backing_file) {
+ backing_file = strchr(str, ':');
+ }
+
if(backing_file){
if(dev->no_cow)
printk(KERN_ERR "Can't specify both 'd' and a "
dev->file = str;
dev->cow.file = backing_file;
dev->boot_openflags = flags;
- out2:
+out:
spin_unlock(&ubd_lock);
return(err);
}
__setup("ubd", ubd_setup);
__uml_help(ubd_setup,
-"ubd<n>=<filename>\n"
+"ubd<n><flags>=<filename>[(:|,)<filename2>]\n"
" This is used to associate a device with a file in the underlying\n"
-" filesystem. Usually, there is a filesystem in the file, but \n"
+" filesystem. When specifying two filenames, the first one is the\n"
+" COW name and the second is the backing file name. As separator you can\n"
+" use either a ':' or a ',': the first one allows writing things like;\n"
+" ubd0=~/Uml/root_cow:~/Uml/root_backing_file\n"
+" while with a ',' the shell would not expand the 2nd '~'.\n"
+" When using only one filename, UML will detect whether to thread it like\n"
+" a COW file or a backing file. To override this detection, add the 'd'\n"
+" flag:\n"
+" ubd0d=BackingFile\n"
+" Usually, there is a filesystem in the file, but \n"
" that's not required. Swap devices containing swap files can be\n"
" specified like this. Also, a file which doesn't contain a\n"
" filesystem can have its contents read in the virtual \n"
-" machine by running dd on the device. n must be in the range\n"
+" machine by running 'dd' on the device. <n> must be in the range\n"
" 0 to 7. Appending an 'r' to the number will cause that device\n"
" to be mounted read-only. For example ubd1r=./ext_fs. Appending\n"
-" an 's' (has to be _after_ 'r', if there is one) will cause data\n"
-" to be written to disk on the host immediately.\n\n"
+" an 's' will cause data to be written to disk on the host immediately.\n\n"
);
static int fakehd_set = 0;
do_ubd = NULL;
intr_count++;
- n = read_ubd_fs(thread_fd, &req, sizeof(req));
+ n = os_read_file(thread_fd, &req, sizeof(req));
if(n != sizeof(req)){
printk(KERN_ERR "Pid %d - spurious interrupt in ubd_handler, "
"err = %d\n", os_getpid(), -n);
sprintf(disk->devfs_name, "ubd_fake/disc%d", unit);
}
+ /* sysfs register (not for ide fake devices) */
+ if (major == MAJOR_NR) {
+ ubd_dev[unit].pdev.id = unit;
+ ubd_dev[unit].pdev.name = DRIVER_NAME;
+ platform_device_register(&ubd_dev[unit].pdev);
+ disk->driverfs_dev = &ubd_dev[unit].pdev.dev;
+ }
+
disk->private_data = &ubd_dev[unit];
disk->queue = ubd_queue;
add_disk(disk);
return 0;
}
+#define ROUND_BLOCK(n) ((n + ((1 << 9) - 1)) & (-1 << 9))
+
static int ubd_add(int n)
{
struct ubd *dev = &ubd_dev[n];
if(err < 0)
return(err);
+ dev->size = ROUND_BLOCK(dev->size);
+
err = ubd_new_disk(MAJOR_NR, dev->size, n, &ubd_gendisk[n]);
if(err)
return(err);
fake_gendisk[n] = NULL;
}
+ platform_device_unregister(&dev->pdev);
*dev = ((struct ubd) DEFAULT_UBD);
err = 0;
out:
__initcall(ubd_mc_init);
+static struct device_driver ubd_driver = {
+ .name = DRIVER_NAME,
+ .bus = &platform_bus_type,
+};
+
int ubd_init(void)
{
int i;
if (register_blkdev(fake_major, "ubd"))
return -1;
}
+ driver_register(&ubd_driver);
for (i = 0; i < MAX_DEV; i++)
ubd_add(i);
return 0;
err = prepare_request(req, &io_req);
if(!err){
do_ubd = ubd_handler;
- n = write_ubd_fs(thread_fd, (char *) &io_req,
+ n = os_write_file(thread_fd, (char *) &io_req,
sizeof(io_req));
if(n != sizeof(io_req))
printk("write to io thread failed, "
static int ubd_ioctl(struct inode * inode, struct file * file,
unsigned int cmd, unsigned long arg)
{
- struct hd_geometry *loc = (struct hd_geometry *) arg;
+ struct hd_geometry __user *loc = (struct hd_geometry __user *) arg;
struct ubd *dev = inode->i_bdev->bd_disk->private_data;
struct hd_driveid ubd_id = {
.cyls = 0,
case HDIO_GET_IDENTITY:
ubd_id.cyls = dev->size / (128 * 32 * 512);
- if(copy_to_user((char *) arg, (char *) &ubd_id,
+ if(copy_to_user((char __user *) arg, (char *) &ubd_id,
sizeof(ubd_id)))
return(-EFAULT);
return(0);
case CDROMVOLREAD:
- if(copy_from_user(&volume, (char *) arg, sizeof(volume)))
+ if(copy_from_user(&volume, (char __user *) arg, sizeof(volume)))
return(-EFAULT);
volume.channel0 = 255;
volume.channel1 = 255;
volume.channel2 = 255;
volume.channel3 = 255;
- if(copy_to_user((char *) arg, &volume, sizeof(volume)))
+ if(copy_to_user((char __user *) arg, &volume, sizeof(volume)))
return(-EFAULT);
return(0);
}
__initcall(ubd_remapper_setup);
+static int same_backing_files(char *from_cmdline, char *from_cow, char *cow)
+{
+ struct uml_stat buf1, buf2;
+ int err;
+
+ if(from_cmdline == NULL) return(1);
+ if(!strcmp(from_cmdline, from_cow)) return(1);
+
+ err = os_stat_file(from_cmdline, &buf1);
+ if(err < 0){
+ printk("Couldn't stat '%s', err = %d\n", from_cmdline, -err);
+ return(1);
+ }
+ err = os_stat_file(from_cow, &buf2);
+ if(err < 0){
+ printk("Couldn't stat '%s', err = %d\n", from_cow, -err);
+ return(1);
+ }
+ if((buf1.ust_dev == buf2.ust_dev) && (buf1.ust_ino == buf2.ust_ino))
+ return(1);
+
+ printk("Backing file mismatch - \"%s\" requested,\n"
+ "\"%s\" specified in COW header of \"%s\"\n",
+ from_cmdline, from_cow, cow);
+ return(0);
+}
+
+static int backing_file_mismatch(char *file, __u64 size, time_t mtime)
+{
+ unsigned long modtime;
+ long long actual;
+ int err;
+
+ err = os_file_modtime(file, &modtime);
+ if(err < 0){
+ printk("Failed to get modification time of backing file "
+ "\"%s\", err = %d\n", file, -err);
+ return(err);
+ }
+
+ err = os_file_size(file, &actual);
+ if(err < 0){
+ printk("Failed to get size of backing file \"%s\", "
+ "err = %d\n", file, -err);
+ return(err);
+ }
+
+ if(actual != size){
+ /*__u64 can be a long on AMD64 and with %lu GCC complains; so
+ * the typecast.*/
+ printk("Size mismatch (%llu vs %llu) of COW header vs backing "
+ "file\n", (unsigned long long) size, actual);
+ return(-EINVAL);
+ }
+ if(modtime != mtime){
+ printk("mtime mismatch (%ld vs %ld) of COW header vs backing "
+ "file\n", mtime, modtime);
+ return(-EINVAL);
+ }
+ return(0);
+}
+
+int read_cow_bitmap(int fd, void *buf, int offset, int len)
+{
+ int err;
+
+ err = os_seek_file(fd, offset);
+ if(err < 0)
+ return(err);
+
+ err = os_read_file(fd, buf, len);
+ if(err < 0)
+ return(err);
+
+ return(0);
+}
+
+int open_ubd_file(char *file, struct openflags *openflags,
+ char **backing_file_out, int *bitmap_offset_out,
+ unsigned long *bitmap_len_out, int *data_offset_out,
+ int *create_cow_out)
+{
+ time_t mtime;
+ unsigned long long size;
+ __u32 version, align;
+ char *backing_file;
+ int fd, err, sectorsize, same, mode = 0644;
+
+ fd = os_open_file(file, *openflags, mode);
+ if(fd < 0){
+ if((fd == -ENOENT) && (create_cow_out != NULL))
+ *create_cow_out = 1;
+ if(!openflags->w ||
+ ((fd != -EROFS) && (fd != -EACCES))) return(fd);
+ openflags->w = 0;
+ fd = os_open_file(file, *openflags, mode);
+ if(fd < 0)
+ return(fd);
+ }
+
+ err = os_lock_file(fd, openflags->w);
+ if(err < 0){
+ printk("Failed to lock '%s', err = %d\n", file, -err);
+ goto out_close;
+ }
+
+ if(backing_file_out == NULL) return(fd);
+
+ err = read_cow_header(file_reader, &fd, &version, &backing_file, &mtime,
+ &size, §orsize, &align, bitmap_offset_out);
+ if(err && (*backing_file_out != NULL)){
+ printk("Failed to read COW header from COW file \"%s\", "
+ "errno = %d\n", file, -err);
+ goto out_close;
+ }
+ if(err) return(fd);
+
+ if(backing_file_out == NULL) return(fd);
+
+ same = same_backing_files(*backing_file_out, backing_file, file);
+
+ if(!same && !backing_file_mismatch(*backing_file_out, size, mtime)){
+ printk("Switching backing file to '%s'\n", *backing_file_out);
+ err = write_cow_header(file, fd, *backing_file_out,
+ sectorsize, align, &size);
+ if(err){
+ printk("Switch failed, errno = %d\n", -err);
+ return(err);
+ }
+ }
+ else {
+ *backing_file_out = backing_file;
+ err = backing_file_mismatch(*backing_file_out, size, mtime);
+ if(err) goto out_close;
+ }
+
+ cow_sizes(version, size, sectorsize, align, *bitmap_offset_out,
+ bitmap_len_out, data_offset_out);
+
+ return(fd);
+ out_close:
+ os_close_file(fd);
+ return(err);
+}
+
+int create_cow_file(char *cow_file, char *backing_file, struct openflags flags,
+ int sectorsize, int alignment, int *bitmap_offset_out,
+ unsigned long *bitmap_len_out, int *data_offset_out)
+{
+ int err, fd;
+
+ flags.c = 1;
+ fd = open_ubd_file(cow_file, &flags, NULL, NULL, NULL, NULL, NULL);
+ if(fd < 0){
+ err = fd;
+ printk("Open of COW file '%s' failed, errno = %d\n", cow_file,
+ -err);
+ goto out;
+ }
+
+ err = init_cow_file(fd, cow_file, backing_file, sectorsize, alignment,
+ bitmap_offset_out, bitmap_len_out,
+ data_offset_out);
+ if(!err)
+ return(fd);
+ os_close_file(fd);
+ out:
+ return(err);
+}
+
+static int update_bitmap(struct io_thread_req *req)
+{
+ int n;
+
+ if(req->cow_offset == -1)
+ return(0);
+
+ n = os_seek_file(req->fds[1], req->cow_offset);
+ if(n < 0){
+ printk("do_io - bitmap lseek failed : err = %d\n", -n);
+ return(1);
+ }
+
+ n = os_write_file(req->fds[1], &req->bitmap_words,
+ sizeof(req->bitmap_words));
+ if(n != sizeof(req->bitmap_words)){
+ printk("do_io - bitmap update failed, err = %d fd = %d\n", -n,
+ req->fds[1]);
+ return(1);
+ }
+
+ return(0);
+}
+
+void do_io(struct io_thread_req *req)
+{
+ char *buf;
+ unsigned long len;
+ int n, nsectors, start, end, bit;
+ int err;
+ __u64 off;
+
+ if(req->op == UBD_MMAP){
+ /* Touch the page to force the host to do any necessary IO to
+ * get it into memory
+ */
+ n = *((volatile int *) req->buffer);
+ req->error = update_bitmap(req);
+ return;
+ }
+
+ nsectors = req->length / req->sectorsize;
+ start = 0;
+ do {
+ bit = ubd_test_bit(start, (unsigned char *) &req->sector_mask);
+ end = start;
+ while((end < nsectors) &&
+ (ubd_test_bit(end, (unsigned char *)
+ &req->sector_mask) == bit))
+ end++;
+
+ off = req->offset + req->offsets[bit] +
+ start * req->sectorsize;
+ len = (end - start) * req->sectorsize;
+ buf = &req->buffer[start * req->sectorsize];
+
+ err = os_seek_file(req->fds[bit], off);
+ if(err < 0){
+ printk("do_io - lseek failed : err = %d\n", -err);
+ req->error = 1;
+ return;
+ }
+ if(req->op == UBD_READ){
+ n = 0;
+ do {
+ buf = &buf[n];
+ len -= n;
+ n = os_read_file(req->fds[bit], buf, len);
+ if (n < 0) {
+ printk("do_io - read failed, err = %d "
+ "fd = %d\n", -n, req->fds[bit]);
+ req->error = 1;
+ return;
+ }
+ } while((n < len) && (n != 0));
+ if (n < len) memset(&buf[n], 0, len - n);
+ }
+ else {
+ n = os_write_file(req->fds[bit], buf, len);
+ if(n != len){
+ printk("do_io - write failed err = %d "
+ "fd = %d\n", -n, req->fds[bit]);
+ req->error = 1;
+ return;
+ }
+ }
+
+ start = end;
+ } while(start < nsectors);
+
+ req->error = update_bitmap(req);
+}
+
+/* Changed in start_io_thread, which is serialized by being called only
+ * from ubd_init, which is an initcall.
+ */
+int kernel_fd = -1;
+
+/* Only changed by the io thread */
+int io_count = 0;
+
+int io_thread(void *arg)
+{
+ struct io_thread_req req;
+ int n;
+
+ ignore_sigwinch_sig();
+ while(1){
+ n = os_read_file(kernel_fd, &req, sizeof(req));
+ if(n != sizeof(req)){
+ if(n < 0)
+ printk("io_thread - read failed, fd = %d, "
+ "err = %d\n", kernel_fd, -n);
+ else {
+ printk("io_thread - short read, fd = %d, "
+ "length = %d\n", kernel_fd, n);
+ }
+ continue;
+ }
+ io_count++;
+ do_io(&req);
+ n = os_write_file(kernel_fd, &req, sizeof(req));
+ if(n != sizeof(req))
+ printk("io_thread - write failed, fd = %d, err = %d\n",
+ kernel_fd, -n);
+ }
+}
+
/*
* Overrides for Emacs so that we follow Linus's tabbing style.
* Emacs will notice this stuff at the end of the file and automatically