| Message ID | 20241113094727.1497722-7-mcgrof@kernel.org (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | enable bs > ps for block devices | expand |
On 11/13/24 10:47, Luis Chamberlain wrote: > We now can support blocksizes larger than PAGE_SIZE, so lift > the restriction up to the max supported page cache order and > just bake this into a common helper used by the block layer. > > We bound ourselves to 64k, because beyond that we need more testing. > > Signed-off-by: Luis Chamberlain <mcgrof@kernel.org> > --- > block/bdev.c | 5 ++--- > include/linux/blkdev.h | 6 +++++- > 2 files changed, 7 insertions(+), 4 deletions(-) > > diff --git a/block/bdev.c b/block/bdev.c > index 167d82b46781..3a5fd65f6c8e 100644 > --- a/block/bdev.c > +++ b/block/bdev.c > @@ -157,8 +157,7 @@ int set_blocksize(struct file *file, int size) > struct inode *inode = file->f_mapping->host; > struct block_device *bdev = I_BDEV(inode); > > - /* Size must be a power of two, and between 512 and PAGE_SIZE */ > - if (size > PAGE_SIZE || size < 512 || !is_power_of_2(size)) > + if (blk_validate_block_size(size)) > return -EINVAL; > > /* Size cannot be smaller than the size supported by the device */ > @@ -185,7 +184,7 @@ int sb_set_blocksize(struct super_block *sb, int size) > if (set_blocksize(sb->s_bdev_file, size)) > return 0; > /* If we get here, we know size is power of two > - * and it's value is between 512 and PAGE_SIZE */ > + * and it's value is larger than 512 */ > sb->s_blocksize = size; > sb->s_blocksize_bits = blksize_bits(size); > return sb->s_blocksize; > diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h > index 50c3b959da28..cc9fca1fceaa 100644 > --- a/include/linux/blkdev.h > +++ b/include/linux/blkdev.h > @@ -25,6 +25,7 @@ > #include <linux/uuid.h> > #include <linux/xarray.h> > #include <linux/file.h> > +#include <linux/pagemap.h> > > struct module; > struct request_queue; > @@ -268,10 +269,13 @@ static inline dev_t disk_devt(struct gendisk *disk) > return MKDEV(disk->major, disk->first_minor); > } > > +/* We should strive for 1 << (PAGE_SHIFT + MAX_PAGECACHE_ORDER) */ > +#define BLK_MAX_BLOCK_SIZE (SZ_64K) > + Please make the comment a bit more descriptive, indicating that beyond 64k more testing is required, hence it's not enabled for now. We _could_ add a config option to make this conditional... > /* blk_validate_limits() validates bsize, so drivers don't usually need to */ > static inline int blk_validate_block_size(unsigned long bsize) > { > - if (bsize < 512 || bsize > PAGE_SIZE || !is_power_of_2(bsize)) > + if (bsize < 512 || bsize > BLK_MAX_BLOCK_SIZE || !is_power_of_2(bsize)) > return -EINVAL; > > return 0; Cheers, Hannes
On Wed, Nov 13, 2024 at 01:47:25AM -0800, Luis Chamberlain wrote: > @@ -185,7 +184,7 @@ int sb_set_blocksize(struct super_block *sb, int size) > if (set_blocksize(sb->s_bdev_file, size)) > return 0; > /* If we get here, we know size is power of two > - * and it's value is between 512 and PAGE_SIZE */ > + * and it's value is larger than 512 */ If you're changing this line, please delete the incorrect apostrophe. > sb->s_blocksize = size; > sb->s_blocksize_bits = blksize_bits(size); > return sb->s_blocksize; > diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h > index 50c3b959da28..cc9fca1fceaa 100644 > --- a/include/linux/blkdev.h > +++ b/include/linux/blkdev.h > @@ -25,6 +25,7 @@ > #include <linux/uuid.h> > #include <linux/xarray.h> > #include <linux/file.h> > +#include <linux/pagemap.h> Why do we need to add this include? > @@ -268,10 +269,13 @@ static inline dev_t disk_devt(struct gendisk *disk) > return MKDEV(disk->major, disk->first_minor); > } > > +/* We should strive for 1 << (PAGE_SHIFT + MAX_PAGECACHE_ORDER) */ > +#define BLK_MAX_BLOCK_SIZE (SZ_64K) I think we need CONFIG_TRANSPARENT_HUGEPAGE to go over PAGE_SIZE.
On 13/11/2024 09:47, Luis Chamberlain wrote: > #include <linux/uuid.h> > #include <linux/xarray.h> > #include <linux/file.h> > +#include <linux/pagemap.h> > > struct module; > struct request_queue; > @@ -268,10 +269,13 @@ static inline dev_t disk_devt(struct gendisk *disk) > return MKDEV(disk->major, disk->first_minor); > } > > +/* We should strive for 1 << (PAGE_SHIFT + MAX_PAGECACHE_ORDER) */ I fell that this comment can be reworked. I think that what we want to say is that hard limit is 1 << (PAGE_SHIFT + MAX_PAGECACHE_ORDER), but we set at sensible size of 64K. > +#define BLK_MAX_BLOCK_SIZE (SZ_64K) > +
diff --git a/block/bdev.c b/block/bdev.c index 167d82b46781..3a5fd65f6c8e 100644 --- a/block/bdev.c +++ b/block/bdev.c @@ -157,8 +157,7 @@ int set_blocksize(struct file *file, int size) struct inode *inode = file->f_mapping->host; struct block_device *bdev = I_BDEV(inode); - /* Size must be a power of two, and between 512 and PAGE_SIZE */ - if (size > PAGE_SIZE || size < 512 || !is_power_of_2(size)) + if (blk_validate_block_size(size)) return -EINVAL; /* Size cannot be smaller than the size supported by the device */ @@ -185,7 +184,7 @@ int sb_set_blocksize(struct super_block *sb, int size) if (set_blocksize(sb->s_bdev_file, size)) return 0; /* If we get here, we know size is power of two - * and it's value is between 512 and PAGE_SIZE */ + * and it's value is larger than 512 */ sb->s_blocksize = size; sb->s_blocksize_bits = blksize_bits(size); return sb->s_blocksize; diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 50c3b959da28..cc9fca1fceaa 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -25,6 +25,7 @@ #include <linux/uuid.h> #include <linux/xarray.h> #include <linux/file.h> +#include <linux/pagemap.h> struct module; struct request_queue; @@ -268,10 +269,13 @@ static inline dev_t disk_devt(struct gendisk *disk) return MKDEV(disk->major, disk->first_minor); } +/* We should strive for 1 << (PAGE_SHIFT + MAX_PAGECACHE_ORDER) */ +#define BLK_MAX_BLOCK_SIZE (SZ_64K) + /* blk_validate_limits() validates bsize, so drivers don't usually need to */ static inline int blk_validate_block_size(unsigned long bsize) { - if (bsize < 512 || bsize > PAGE_SIZE || !is_power_of_2(bsize)) + if (bsize < 512 || bsize > BLK_MAX_BLOCK_SIZE || !is_power_of_2(bsize)) return -EINVAL; return 0;
We now can support blocksizes larger than PAGE_SIZE, so lift the restriction up to the max supported page cache order and just bake this into a common helper used by the block layer. We bound ourselves to 64k, because beyond that we need more testing. Signed-off-by: Luis Chamberlain <mcgrof@kernel.org> --- block/bdev.c | 5 ++--- include/linux/blkdev.h | 6 +++++- 2 files changed, 7 insertions(+), 4 deletions(-)