2005-05-18 04:27:41

by Jeff Dike

[permalink] [raw]
Subject: [PATCH 7/9] UML - Remove ubd-mmap support

Finally rip out the ubd-mmap code, which turned out to be broken by design.

Signed-off-by: Jeff Dike <[email protected]>

Index: linux-2.6.11/arch/um/drivers/ubd_kern.c
===================================================================
--- linux-2.6.11.orig/arch/um/drivers/ubd_kern.c 2005-05-17 22:19:48.000000000 -0400
+++ linux-2.6.11/arch/um/drivers/ubd_kern.c 2005-05-17 22:21:58.000000000 -0400
@@ -55,7 +55,7 @@
#include "mem_kern.h"
#include "cow.h"

-enum ubd_req { UBD_READ, UBD_WRITE, UBD_MMAP };
+enum ubd_req { UBD_READ, UBD_WRITE };

struct io_thread_req {
enum ubd_req op;
@@ -68,8 +68,6 @@ struct io_thread_req {
unsigned long sector_mask;
unsigned long long cow_offset;
unsigned long bitmap_words[2];
- int map_fd;
- unsigned long long map_offset;
int error;
};

@@ -122,10 +120,6 @@ static int ubd_ioctl(struct inode * inod

#define MAX_DEV (8)

-/* Changed in early boot */
-static int ubd_do_mmap = 0;
-#define UBD_MMAP_BLOCK_SIZE PAGE_SIZE
-
static struct block_device_operations ubd_blops = {
.owner = THIS_MODULE,
.open = ubd_open,
@@ -175,12 +169,6 @@ struct ubd {
int no_cow;
struct cow cow;
struct platform_device pdev;
-
- int map_writes;
- int map_reads;
- int nomap_writes;
- int nomap_reads;
- int write_maps;
};

#define DEFAULT_COW { \
@@ -200,11 +188,6 @@ struct ubd {
.openflags = OPEN_FLAGS, \
.no_cow = 0, \
.cow = DEFAULT_COW, \
- .map_writes = 0, \
- .map_reads = 0, \
- .nomap_writes = 0, \
- .nomap_reads = 0, \
- .write_maps = 0, \
}

struct ubd ubd_dev[MAX_DEV] = { [ 0 ... MAX_DEV - 1 ] = DEFAULT_UBD };
@@ -314,13 +297,6 @@ static int ubd_setup_common(char *str, i
int major;

str++;
- if(!strcmp(str, "mmap")){
- CHOOSE_MODE(printk("mmap not supported by the ubd "
- "driver in tt mode\n"),
- ubd_do_mmap = 1);
- return(0);
- }
-
if(!strcmp(str, "sync")){
global_openflags = of_sync(global_openflags);
return(0);
@@ -524,7 +500,7 @@ static void ubd_handler(void)
{
struct io_thread_req req;
struct request *rq = elv_next_request(ubd_queue);
- int n, err;
+ int n;

do_ubd = NULL;
intr_count++;
@@ -538,19 +514,6 @@ static void ubd_handler(void)
return;
}

- if((req.op != UBD_MMAP) &&
- ((req.offset != ((__u64) (rq->sector)) << 9) ||
- (req.length != (rq->current_nr_sectors) << 9)))
- panic("I/O op mismatch");
-
- if(req.map_fd != -1){
- err = physmem_subst_mapping(req.buffer, req.map_fd,
- req.map_offset, 1);
- if(err)
- printk("ubd_handler - physmem_subst_mapping failed, "
- "err = %d\n", -err);
- }
-
ubd_finish(rq, req.error);
reactivate_fd(thread_fd, UBD_IRQ);
do_ubd_request(ubd_queue);
@@ -583,14 +546,10 @@ static int ubd_file_size(struct ubd *dev

static void ubd_close(struct ubd *dev)
{
- if(ubd_do_mmap)
- physmem_forget_descriptor(dev->fd);
os_close_file(dev->fd);
if(dev->cow.file == NULL)
return;

- if(ubd_do_mmap)
- physmem_forget_descriptor(dev->cow.fd);
os_close_file(dev->cow.fd);
vfree(dev->cow.bitmap);
dev->cow.bitmap = NULL;
@@ -1010,94 +969,13 @@ static void cowify_req(struct io_thread_
req->bitmap_words, bitmap_len);
}

-static int mmap_fd(struct request *req, struct ubd *dev, __u64 offset)
-{
- __u64 sector;
- unsigned char *bitmap;
- int bit, i;
-
- /* mmap must have been requested on the command line */
- if(!ubd_do_mmap)
- return(-1);
-
- /* The buffer must be page aligned */
- if(((unsigned long) req->buffer % UBD_MMAP_BLOCK_SIZE) != 0)
- return(-1);
-
- /* The request must be a page long */
- if((req->current_nr_sectors << 9) != PAGE_SIZE)
- return(-1);
-
- if(dev->cow.file == NULL)
- return(dev->fd);
-
- sector = offset >> 9;
- bitmap = (unsigned char *) dev->cow.bitmap;
- bit = ubd_test_bit(sector, bitmap);
-
- for(i = 1; i < req->current_nr_sectors; i++){
- if(ubd_test_bit(sector + i, bitmap) != bit)
- return(-1);
- }
-
- if(bit || (rq_data_dir(req) == WRITE))
- offset += dev->cow.data_offset;
-
- /* The data on disk must be page aligned */
- if((offset % UBD_MMAP_BLOCK_SIZE) != 0)
- return(-1);
-
- return(bit ? dev->fd : dev->cow.fd);
-}
-
-static int prepare_mmap_request(struct ubd *dev, int fd, __u64 offset,
- struct request *req,
- struct io_thread_req *io_req)
-{
- int err;
-
- if(rq_data_dir(req) == WRITE){
- /* Writes are almost no-ops since the new data is already in the
- * host page cache
- */
- dev->map_writes++;
- if(dev->cow.file != NULL)
- cowify_bitmap(io_req->offset, io_req->length,
- &io_req->sector_mask, &io_req->cow_offset,
- dev->cow.bitmap, dev->cow.bitmap_offset,
- io_req->bitmap_words,
- dev->cow.bitmap_len);
- }
- else {
- int w;
-
- if((dev->cow.file != NULL) && (fd == dev->cow.fd))
- w = 0;
- else w = dev->openflags.w;
-
- if((dev->cow.file != NULL) && (fd == dev->fd))
- offset += dev->cow.data_offset;
-
- err = physmem_subst_mapping(req->buffer, fd, offset, w);
- if(err){
- printk("physmem_subst_mapping failed, err = %d\n",
- -err);
- return(1);
- }
- dev->map_reads++;
- }
- io_req->op = UBD_MMAP;
- io_req->buffer = req->buffer;
- return(0);
-}
-
/* Called with ubd_io_lock held */
static int prepare_request(struct request *req, struct io_thread_req *io_req)
{
struct gendisk *disk = req->rq_disk;
struct ubd *dev = disk->private_data;
__u64 offset;
- int len, fd;
+ int len;

if(req->rq_status == RQ_INACTIVE) return(1);

@@ -1114,34 +992,12 @@ static int prepare_request(struct reques

io_req->fds[0] = (dev->cow.file != NULL) ? dev->cow.fd : dev->fd;
io_req->fds[1] = dev->fd;
- io_req->map_fd = -1;
io_req->cow_offset = -1;
io_req->offset = offset;
io_req->length = len;
io_req->error = 0;
io_req->sector_mask = 0;

- fd = mmap_fd(req, dev, io_req->offset);
- if(fd > 0){
- /* If mmapping is otherwise OK, but the first access to the
- * page is a write, then it's not mapped in yet. So we have
- * to write the data to disk first, then we can map the disk
- * page in and continue normally from there.
- */
- if((rq_data_dir(req) == WRITE) && !is_remapped(req->buffer)){
- io_req->map_fd = dev->fd;
- io_req->map_offset = io_req->offset +
- dev->cow.data_offset;
- dev->write_maps++;
- }
- else return(prepare_mmap_request(dev, fd, io_req->offset, req,
- io_req));
- }
-
- if(rq_data_dir(req) == READ)
- dev->nomap_reads++;
- else dev->nomap_writes++;
-
io_req->op = (rq_data_dir(req) == READ) ? UBD_READ : UBD_WRITE;
io_req->offsets[0] = 0;
io_req->offsets[1] = dev->cow.data_offset;
@@ -1229,143 +1085,6 @@ static int ubd_ioctl(struct inode * inod
return(-EINVAL);
}

-static int ubd_check_remapped(int fd, unsigned long address, int is_write,
- __u64 offset)
-{
- __u64 bitmap_offset;
- unsigned long new_bitmap[2];
- int i, err, n;
-
- /* If it's not a write access, we can't do anything about it */
- if(!is_write)
- return(0);
-
- /* We have a write */
- for(i = 0; i < sizeof(ubd_dev) / sizeof(ubd_dev[0]); i++){
- struct ubd *dev = &ubd_dev[i];
-
- if((dev->fd != fd) && (dev->cow.fd != fd))
- continue;
-
- /* It's a write to a ubd device */
-
- /* This should be impossible now */
- if(!dev->openflags.w){
- /* It's a write access on a read-only device - probably
- * shouldn't happen. If the kernel is trying to change
- * something with no intention of writing it back out,
- * then this message will clue us in that this needs
- * fixing
- */
- printk("Write access to mapped page from readonly ubd "
- "device %d\n", i);
- return(0);
- }
-
- /* It's a write to a writeable ubd device - it must be COWed
- * because, otherwise, the page would have been mapped in
- * writeable
- */
-
- if(!dev->cow.file)
- panic("Write fault on writeable non-COW ubd device %d",
- i);
-
- /* It should also be an access to the backing file since the
- * COW pages should be mapped in read-write
- */
-
- if(fd == dev->fd)
- panic("Write fault on a backing page of ubd "
- "device %d\n", i);
-
- /* So, we do the write, copying the backing data to the COW
- * file...
- */
-
- err = os_seek_file(dev->fd, offset + dev->cow.data_offset);
- if(err < 0)
- panic("Couldn't seek to %lld in COW file of ubd "
- "device %d, err = %d",
- offset + dev->cow.data_offset, i, -err);
-
- n = os_write_file(dev->fd, (void *) address, PAGE_SIZE);
- if(n != PAGE_SIZE)
- panic("Couldn't copy data to COW file of ubd "
- "device %d, err = %d", i, -n);
-
- /* ... updating the COW bitmap... */
-
- cowify_bitmap(offset, PAGE_SIZE, NULL, &bitmap_offset,
- dev->cow.bitmap, dev->cow.bitmap_offset,
- new_bitmap, dev->cow.bitmap_len);
-
- err = os_seek_file(dev->fd, bitmap_offset);
- if(err < 0)
- panic("Couldn't seek to %lld in COW file of ubd "
- "device %d, err = %d", bitmap_offset, i, -err);
-
- n = os_write_file(dev->fd, new_bitmap, sizeof(new_bitmap));
- if(n != sizeof(new_bitmap))
- panic("Couldn't update bitmap of ubd device %d, "
- "err = %d", i, -n);
-
- /* Maybe we can map the COW page in, and maybe we can't. If
- * it is a pre-V3 COW file, we can't, since the alignment will
- * be wrong. If it is a V3 or later COW file which has been
- * moved to a system with a larger page size, then maybe we
- * can't, depending on the exact location of the page.
- */
-
- offset += dev->cow.data_offset;
-
- /* Remove the remapping, putting the original anonymous page
- * back. If the COW file can be mapped in, that is done.
- * Otherwise, the COW page is read in.
- */
-
- if(!physmem_remove_mapping((void *) address))
- panic("Address 0x%lx not remapped by ubd device %d",
- address, i);
- if((offset % UBD_MMAP_BLOCK_SIZE) == 0)
- physmem_subst_mapping((void *) address, dev->fd,
- offset, 1);
- else {
- err = os_seek_file(dev->fd, offset);
- if(err < 0)
- panic("Couldn't seek to %lld in COW file of "
- "ubd device %d, err = %d", offset, i,
- -err);
-
- n = os_read_file(dev->fd, (void *) address, PAGE_SIZE);
- if(n != PAGE_SIZE)
- panic("Failed to read page from offset %llx of "
- "COW file of ubd device %d, err = %d",
- offset, i, -n);
- }
-
- return(1);
- }
-
- /* It's not a write on a ubd device */
- return(0);
-}
-
-static struct remapper ubd_remapper = {
- .list = LIST_HEAD_INIT(ubd_remapper.list),
- .proc = ubd_check_remapped,
-};
-
-static int ubd_remapper_setup(void)
-{
- if(ubd_do_mmap)
- register_remapper(&ubd_remapper);
-
- return(0);
-}
-
-__initcall(ubd_remapper_setup);
-
static int same_backing_files(char *from_cmdline, char *from_cow, char *cow)
{
struct uml_stat buf1, buf2;
@@ -1568,15 +1287,6 @@ void do_io(struct io_thread_req *req)
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 {
Index: linux-2.6.11/arch/um/kernel/trap_kern.c
===================================================================
--- linux-2.6.11.orig/arch/um/kernel/trap_kern.c 2005-05-17 22:21:51.000000000 -0400
+++ linux-2.6.11/arch/um/kernel/trap_kern.c 2005-05-17 22:21:58.000000000 -0400
@@ -107,33 +107,6 @@ out_of_memory:
goto out;
}

-LIST_HEAD(physmem_remappers);
-
-void register_remapper(struct remapper *info)
-{
- list_add(&info->list, &physmem_remappers);
-}
-
-static int check_remapped_addr(unsigned long address, int is_write)
-{
- struct remapper *remapper;
- struct list_head *ele;
- __u64 offset;
- int fd;
-
- fd = phys_mapping(__pa(address), &offset);
- if(fd == -1)
- return(0);
-
- list_for_each(ele, &physmem_remappers){
- remapper = list_entry(ele, struct remapper, list);
- if((*remapper->proc)(fd, address, is_write, offset))
- return(1);
- }
-
- return(0);
-}
-
/*
* We give a *copy* of the faultinfo in the regs to segv.
* This must be done, since nesting SEGVs could overwrite
@@ -152,8 +125,6 @@ unsigned long segv(struct faultinfo fi,
flush_tlb_kernel_vm();
return(0);
}
- else if(check_remapped_addr(address & PAGE_MASK, is_write))
- return(0);
else if(current->mm == NULL)
panic("Segfault with no mm");
err = handle_page_fault(address, ip, is_write, is_user, &si.si_code);