From 9f4736fe7ca804aa79b5916221bb13dfc6221a0f Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Thu, 28 Jan 2016 20:13:39 -0800 Subject: block: revert runtime dax control of the raw block device Dynamically enabling DAX requires that the page cache first be flushed and invalidated. This must occur atomically with the change of DAX mode otherwise we confuse the fsync/msync tracking and violate data durability guarantees. Eliminate the possibilty of DAX-disabled to DAX-enabled transitions for now and revisit this for the next cycle. Cc: Jan Kara Cc: Jeff Moyer Cc: Christoph Hellwig Cc: Dave Chinner Cc: Matthew Wilcox Cc: Andrew Morton Cc: Ross Zwisler Signed-off-by: Dan Williams --- block/ioctl.c | 38 -------------------------------------- 1 file changed, 38 deletions(-) (limited to 'block') diff --git a/block/ioctl.c b/block/ioctl.c index 77f5d17779d6..d8996bbd7f12 100644 --- a/block/ioctl.c +++ b/block/ioctl.c @@ -434,42 +434,6 @@ bool blkdev_dax_capable(struct block_device *bdev) return true; } - -static int blkdev_daxset(struct block_device *bdev, unsigned long argp) -{ - unsigned long arg; - int rc = 0; - - if (!capable(CAP_SYS_ADMIN)) - return -EACCES; - - if (get_user(arg, (int __user *)(argp))) - return -EFAULT; - arg = !!arg; - if (arg == !!(bdev->bd_inode->i_flags & S_DAX)) - return 0; - - if (arg) - arg = S_DAX; - - if (arg && !blkdev_dax_capable(bdev)) - return -ENOTTY; - - inode_lock(bdev->bd_inode); - if (bdev->bd_map_count == 0) - inode_set_flags(bdev->bd_inode, arg, S_DAX); - else - rc = -EBUSY; - inode_unlock(bdev->bd_inode); - return rc; -} -#else -static int blkdev_daxset(struct block_device *bdev, int arg) -{ - if (arg) - return -ENOTTY; - return 0; -} #endif static int blkdev_flushbuf(struct block_device *bdev, fmode_t mode, @@ -634,8 +598,6 @@ int blkdev_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, case BLKTRACESETUP: case BLKTRACETEARDOWN: return blk_trace_ioctl(bdev, cmd, argp); - case BLKDAXSET: - return blkdev_daxset(bdev, arg); case BLKDAXGET: return put_int(arg, !!(bdev->bd_inode->i_flags & S_DAX)); break; -- cgit v1.2.3 From d1a5f2b4d8a125943dcb6b032fc7eaefc2c78296 Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Thu, 28 Jan 2016 20:25:31 -0800 Subject: block: use DAX for partition table reads Avoid populating pagecache when the block device is in DAX mode. Otherwise these page cache entries collide with the fsync/msync implementation and break data durability guarantees. Cc: Jan Kara Cc: Jeff Moyer Cc: Christoph Hellwig Cc: Dave Chinner Cc: Andrew Morton Reported-by: Ross Zwisler Tested-by: Ross Zwisler Reviewed-by: Matthew Wilcox Signed-off-by: Dan Williams --- block/partition-generic.c | 18 +++++++++++++++--- fs/dax.c | 20 ++++++++++++++++++++ include/linux/dax.h | 11 +++++++++++ 3 files changed, 46 insertions(+), 3 deletions(-) (limited to 'block') diff --git a/block/partition-generic.c b/block/partition-generic.c index 746935a5973c..fefd01b496a0 100644 --- a/block/partition-generic.c +++ b/block/partition-generic.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include "partitions/check.h" @@ -550,13 +551,24 @@ int invalidate_partitions(struct gendisk *disk, struct block_device *bdev) return 0; } -unsigned char *read_dev_sector(struct block_device *bdev, sector_t n, Sector *p) +static struct page *read_pagecache_sector(struct block_device *bdev, sector_t n) { struct address_space *mapping = bdev->bd_inode->i_mapping; + + return read_mapping_page(mapping, (pgoff_t)(n >> (PAGE_CACHE_SHIFT-9)), + NULL); +} + +unsigned char *read_dev_sector(struct block_device *bdev, sector_t n, Sector *p) +{ struct page *page; - page = read_mapping_page(mapping, (pgoff_t)(n >> (PAGE_CACHE_SHIFT-9)), - NULL); + /* don't populate page cache for dax capable devices */ + if (IS_DAX(bdev->bd_inode)) + page = read_dax_sector(bdev, n); + else + page = read_pagecache_sector(bdev, n); + if (!IS_ERR(page)) { if (PageError(page)) goto fail; diff --git a/fs/dax.c b/fs/dax.c index 4fd6b0c5c6b5..e0e9358baf35 100644 --- a/fs/dax.c +++ b/fs/dax.c @@ -58,6 +58,26 @@ static void dax_unmap_atomic(struct block_device *bdev, blk_queue_exit(bdev->bd_queue); } +struct page *read_dax_sector(struct block_device *bdev, sector_t n) +{ + struct page *page = alloc_pages(GFP_KERNEL, 0); + struct blk_dax_ctl dax = { + .size = PAGE_SIZE, + .sector = n & ~((((int) PAGE_SIZE) / 512) - 1), + }; + long rc; + + if (!page) + return ERR_PTR(-ENOMEM); + + rc = dax_map_atomic(bdev, &dax); + if (rc < 0) + return ERR_PTR(rc); + memcpy_from_pmem(page_address(page), dax.addr, PAGE_SIZE); + dax_unmap_atomic(bdev, &dax); + return page; +} + /* * dax_clear_blocks() is called from within transaction context from XFS, * and hence this means the stack from this point must follow GFP_NOFS diff --git a/include/linux/dax.h b/include/linux/dax.h index 8204c3dc3800..818e45078929 100644 --- a/include/linux/dax.h +++ b/include/linux/dax.h @@ -14,6 +14,17 @@ int dax_fault(struct vm_area_struct *, struct vm_fault *, get_block_t, dax_iodone_t); int __dax_fault(struct vm_area_struct *, struct vm_fault *, get_block_t, dax_iodone_t); + +#ifdef CONFIG_FS_DAX +struct page *read_dax_sector(struct block_device *bdev, sector_t n); +#else +static inline struct page *read_dax_sector(struct block_device *bdev, + sector_t n) +{ + return ERR_PTR(-ENXIO); +} +#endif + #ifdef CONFIG_TRANSPARENT_HUGEPAGE int dax_pmd_fault(struct vm_area_struct *, unsigned long addr, pmd_t *, unsigned int flags, get_block_t, dax_iodone_t); -- cgit v1.2.3