md/bitmap: Optimal last page size

PATCH Superseded Jonathan Derrick 2 participants 6 messages Feb 23, 2023 – Feb 24, 2023 MD RAID

lore

Share & tools

lore

Jonathan Derrick

author

2023-02-23 19:53 UTC #0 lore raw

v4 cover md/bitmap: Optimal last page size

From: Jon Derrick <[email protected]>

Currently the last bitmap page write will size itself down to the logical block size. This could cause less performance for devices which have atomic write units larger than the block size, such as many NVMe devices with 4kB write units and 512B block sizes. There is usually a large amount of space after the bitmap and using the optimal I/O size could favor speed over size.

This was tested on an Intel/Solidigm P5520 drive with lba format 512B, optimal I/O size of 4kB, resulting in a > 10x IOPS increase.

See patch 3 log for results.

v3->v4: Fixed reviewers concerns
v2->v3: Prep patch added and types fixed
Added helpers for optimal I/O sizes

Jon Derrick (3):
md: Move sb writer loop to its own function
md: Fix types in sb writer
md: Use optimal I/O size for last bitmap page

 drivers/md/md-bitmap.c | 142 ++++++++++++++++++++++++-----------------
 1 file changed, 82 insertions(+), 60 deletions(-)

Signature · 1 line

2.27.0

Jonathan Derrick

author

2023-02-23 19:53 UTC #1 lore raw

PATCH v4 1/3 md: Move sb writer loop to its own function

From: Jon Derrick <[email protected]>

Preparatory patch for optimal I/O size calculation. Move the sb writer loop routine into its own function for clarity.

Reviewed-by Christoph Hellwig <[email protected]>

Signed-off-by Jon Derrick <[email protected]>

---
 drivers/md/md-bitmap.c | 124 +++++++++++++++++++++--------------------
 1 file changed, 64 insertions(+), 60 deletions(-)

drivers/md/md-bitmap.c +64 ~~-60~~

index e7cc6ba1b657..2d161d24c51c 100644
@@ -209,76 +209,80 @@ static struct md_rdev *next_active_rdev(struct md_rdev *rdev, struct mddev *mdde
 	return NULL;
 }
 
-static int write_sb_page(struct bitmap *bitmap, struct page *page, int wait)
+static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap,
+			   struct page *page)
 {
-	struct md_rdev *rdev;
 	struct block_device *bdev;
 	struct mddev *mddev = bitmap->mddev;
 	struct bitmap_storage *store = &bitmap->storage;
+	loff_t offset = mddev->bitmap_info.offset;
+	int size = PAGE_SIZE;
+
+	bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev;
+	if (page->index == store->file_pages - 1) {
+		int last_page_size = store->bytes & (PAGE_SIZE - 1);
+
+		if (last_page_size == 0)
+			last_page_size = PAGE_SIZE;
+		size = roundup(last_page_size,
+			       bdev_logical_block_size(bdev));
+	}
+
+	/* Just make sure we aren't corrupting data or metadata */
+	if (mddev->external) {
+		/* Bitmap could be anywhere. */
+		if (rdev->sb_start + offset
+		    + (page->index * (PAGE_SIZE / SECTOR_SIZE))
+		    > rdev->data_offset &&
+		    rdev->sb_start + offset
+		    < (rdev->data_offset + mddev->dev_sectors
+		     + (PAGE_SIZE / SECTOR_SIZE)))
+			return -EINVAL;
+	} else if (offset < 0) {
+		/* DATA  BITMAP METADATA  */
+		if (offset
+		    + (long)(page->index * (PAGE_SIZE / SECTOR_SIZE))
+		    + size / SECTOR_SIZE > 0)
+			/* bitmap runs in to metadata */
+			return -EINVAL;
+
+		if (rdev->data_offset + mddev->dev_sectors
+		    > rdev->sb_start + offset)
+			/* data runs in to bitmap */
+			return -EINVAL;
+	} else if (rdev->sb_start < rdev->data_offset) {
+		/* METADATA BITMAP DATA */
+		if (rdev->sb_start + offset
+		    + page->index * (PAGE_SIZE / SECTOR_SIZE)
+		    + size / SECTOR_SIZE > rdev->data_offset)
+			/* bitmap runs in to data */
+			return -EINVAL;
+	} else {
+		/* DATA METADATA BITMAP - no problems */
+	}
 
-restart:
-	rdev = NULL;
-	while ((rdev = next_active_rdev(rdev, mddev)) != NULL) {
-		int size = PAGE_SIZE;
-		loff_t offset = mddev->bitmap_info.offset;
+	md_super_write(mddev, rdev,
+		       rdev->sb_start + offset
+		       + page->index * (PAGE_SIZE / SECTOR_SIZE),
+		       size, page);
+	return 0;
+}
 
-		bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev;
+static int write_sb_page(struct bitmap *bitmap, struct page *page, int wait)
+{
+	struct md_rdev *rdev;
+	struct mddev *mddev = bitmap->mddev;
+	int ret;
 
-		if (page->index == store->file_pages-1) {
-			int last_page_size = store->bytes & (PAGE_SIZE-1);
-			if (last_page_size == 0)
-				last_page_size = PAGE_SIZE;
-			size = roundup(last_page_size,
-				       bdev_logical_block_size(bdev));
-		}
-		/* Just make sure we aren't corrupting data or
-		 * metadata
-		 */
-		if (mddev->external) {
-			/* Bitmap could be anywhere. */
-			if (rdev->sb_start + offset + (page->index
-						       * (PAGE_SIZE/512))
-			    > rdev->data_offset
-			    &&
-			    rdev->sb_start + offset
-			    < (rdev->data_offset + mddev->dev_sectors
-			     + (PAGE_SIZE/512)))
-				goto bad_alignment;
-		} else if (offset < 0) {
-			/* DATA  BITMAP METADATA  */
-			if (offset
-			    + (long)(page->index * (PAGE_SIZE/512))
-			    + size/512 > 0)
-				/* bitmap runs in to metadata */
-				goto bad_alignment;
-			if (rdev->data_offset + mddev->dev_sectors
-			    > rdev->sb_start + offset)
-				/* data runs in to bitmap */
-				goto bad_alignment;
-		} else if (rdev->sb_start < rdev->data_offset) {
-			/* METADATA BITMAP DATA */
-			if (rdev->sb_start
-			    + offset
-			    + page->index*(PAGE_SIZE/512) + size/512
-			    > rdev->data_offset)
-				/* bitmap runs in to data */
-				goto bad_alignment;
-		} else {
-			/* DATA METADATA BITMAP - no problems */
+	do {
+		while ((rdev = next_active_rdev(rdev, mddev)) != NULL) {
+			ret = __write_sb_page(rdev, bitmap, page);
+			if (ret)
+				return ret;
 		}
-		md_super_write(mddev, rdev,
-			       rdev->sb_start + offset
-			       + page->index * (PAGE_SIZE/512),
-			       size,
-			       page);
-	}
+	} while (wait && md_super_wait(mddev) < 0);
 
-	if (wait && md_super_wait(mddev) < 0)
-		goto restart;
 	return 0;
-
- bad_alignment:
-	return -EINVAL;
 }
 
 static void md_bitmap_file_kick(struct bitmap *bitmap);

Signature · 1 line

2.27.0

kernel test robot

build bot

2023-02-24 07:23 UTC #2 lore raw

Re: [PATCH v4 1/3] md: Move sb writer loop to its own function

Thank you for the patch! Perhaps something to improve:

Hi Jonathan,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on song-md/md-next]
[also build test WARNING on linus/master v6.2 next-20230224] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Jonathan-Derrick/md-Move-sb-writer-loop-to-its-own-function/20230224-035451
base:   git://git.kernel.org/pub/scm/linux/kernel/git/song/md.git md-next
patch link:    https://lore.kernel.org/r/20230223195225.534-2-jonathan.derrick%40linux.dev
patch subject: [PATCH v4 1/3] md: Move sb writer loop to its own function
config: s390-randconfig-r036-20230223 (https://download.01.org/0day-ci/archive/20230224/[email protected]/config)
compiler: clang version 17.0.0 (https://github.com/llvm/llvm-project db89896bbbd2251fff457699635acbbedeead27f)

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# install s390 cross compiling tool for clang build
# apt-get install binutils-s390x-linux-gnu
# https://github.com/intel-lab-lkp/linux/commit/93818e631c76938f9a3ad87d1baf681645d02ba7
git remote add linux-review https://github.com/intel-lab-lkp/linux
git fetch --no-tags linux-review Jonathan-Derrick/md-Move-sb-writer-loop-to-its-own-function/20230224-035451 git checkout 93818e631c76938f9a3ad87d1baf681645d02ba7
# save the config file
mkdir build_dir && cp config build_dir/.config
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir ARCH=s390 olddefconfig COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir ARCH=s390 SHELL=/bin/bash

If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <[email protected]>
| Link: https://lore.kernel.org/oe-kbuild-all/[email protected]/

All warnings (new ones prefixed by >>):

In file included from drivers/md/md-bitmap.c:31:
In file included from include/trace/events/block.h:8:
In file included from include/linux/blktrace_api.h:5:
In file included from include/linux/blk-mq.h:8:
In file included from include/linux/scatterlist.h:9:
In file included from arch/s390/include/asm/io.h:75:
include/asm-generic/io.h:547:31: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]

           val = __raw_readb(PCI_IOBASE + addr);
                             ~~~~~~~~~~ ^
   include/asm-generic/io.h:560:61: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           val = __le16_to_cpu((__le16 __force)__raw_readw(PCI_IOBASE + addr));
                                                           ~~~~~~~~~~ ^
   include/uapi/linux/byteorder/big_endian.h:37:59: note: expanded from macro '__le16_to_cpu'
   #define __le16_to_cpu(x) __swab16((__force __u16)(__le16)(x))
                                                             ^
   include/uapi/linux/swab.h:102:54: note: expanded from macro '__swab16'
   #define __swab16(x) (__u16)__builtin_bswap16((__u16)(x))
                                                        ^
   In file included from drivers/md/md-bitmap.c:31:
   In file included from include/trace/events/block.h:8:
   In file included from include/linux/blktrace_api.h:5:
   In file included from include/linux/blk-mq.h:8:
   In file included from include/linux/scatterlist.h:9:
   In file included from arch/s390/include/asm/io.h:75:
   include/asm-generic/io.h:573:61: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           val = __le32_to_cpu((__le32 __force)__raw_readl(PCI_IOBASE + addr));
                                                           ~~~~~~~~~~ ^
   include/uapi/linux/byteorder/big_endian.h:35:59: note: expanded from macro '__le32_to_cpu'
   #define __le32_to_cpu(x) __swab32((__force __u32)(__le32)(x))
                                                             ^
   include/uapi/linux/swab.h:115:54: note: expanded from macro '__swab32'
   #define __swab32(x) (__u32)__builtin_bswap32((__u32)(x))
                                                        ^
   In file included from drivers/md/md-bitmap.c:31:
   In file included from include/trace/events/block.h:8:
   In file included from include/linux/blktrace_api.h:5:
   In file included from include/linux/blk-mq.h:8:
   In file included from include/linux/scatterlist.h:9:
   In file included from arch/s390/include/asm/io.h:75:
   include/asm-generic/io.h:584:33: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           __raw_writeb(value, PCI_IOBASE + addr);
                               ~~~~~~~~~~ ^
   include/asm-generic/io.h:594:59: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           __raw_writew((u16 __force)cpu_to_le16(value), PCI_IOBASE + addr);
                                                         ~~~~~~~~~~ ^
   include/asm-generic/io.h:604:59: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           __raw_writel((u32 __force)cpu_to_le32(value), PCI_IOBASE + addr);
                                                         ~~~~~~~~~~ ^
   include/asm-generic/io.h:692:20: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           readsb(PCI_IOBASE + addr, buffer, count);
                  ~~~~~~~~~~ ^
   include/asm-generic/io.h:700:20: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           readsw(PCI_IOBASE + addr, buffer, count);
                  ~~~~~~~~~~ ^
   include/asm-generic/io.h:708:20: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           readsl(PCI_IOBASE + addr, buffer, count);
                  ~~~~~~~~~~ ^
   include/asm-generic/io.h:717:21: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           writesb(PCI_IOBASE + addr, buffer, count);
                   ~~~~~~~~~~ ^
   include/asm-generic/io.h:726:21: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           writesw(PCI_IOBASE + addr, buffer, count);
                   ~~~~~~~~~~ ^
   include/asm-generic/io.h:735:21: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           writesl(PCI_IOBASE + addr, buffer, count);
                   ~~~~~~~~~~ ^
>> drivers/md/md-bitmap.c:273:18: warning: variable 'rdev' is used uninitialized whenever function 'write_sb_page' is called [-Wsometimes-uninitialized]
           struct md_rdev *rdev;
           ~~~~~~~~~~~~~~~~^~~~
   drivers/md/md-bitmap.c:278:35: note: uninitialized use occurs here
                   while ((rdev = next_active_rdev(rdev, mddev)) != NULL) {
                                                   ^~~~
   drivers/md/md-bitmap.c:273:22: note: initialize the variable 'rdev' to silence this warning
           struct md_rdev *rdev;
                               ^
                                = NULL

13 warnings generated.

vim +273 drivers/md/md-bitmap.c

270
271 static int write_sb_page(struct bitmap *bitmap, struct page *page, int wait)
272 {

 > 273		struct md_rdev *rdev;

274 struct mddev *mddev = bitmap->mddev;
275 int ret;
276
277 do {
278 while ((rdev = next_active_rdev(rdev, mddev)) != NULL) {
279 ret = __write_sb_page(rdev, bitmap, page);
280 if (ret)
281 return ret;
282 }
283 } while (wait && md_super_wait(mddev) < 0);
284
285 return 0;
286 }
287

Signature · 2 lines

0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests

Jonathan Derrick

author

2023-02-24 18:27 UTC #3 lore raw

Re: [PATCH v4 1/3] md: Move sb writer loop to its own function

Quoted 102 lines · kernel test robot

Hi Jonathan,

Thank you for the patch! Perhaps something to improve:

url:    https://github.com/intel-lab-lkp/linux/commits/Jonathan-Derrick/md-Move-sb-writer-loop-to-its-own-function/20230224-035451
base:   git://git.kernel.org/pub/scm/linux/kernel/git/song/md.git md-next
patch link:    https://lore.kernel.org/r/20230223195225.534-2-jonathan.derrick%40linux.dev
patch subject: [PATCH v4 1/3] md: Move sb writer loop to its own function
config: s390-randconfig-r036-20230223 (https://download.01.org/0day-ci/archive/20230224/[email protected]/config)
compiler: clang version 17.0.0 (https://github.com/llvm/llvm-project db89896bbbd2251fff457699635acbbedeead27f)

If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <[email protected]>
| Link: https://lore.kernel.org/oe-kbuild-all/[email protected]/

All warnings (new ones prefixed by >>):

   In file included from drivers/md/md-bitmap.c:31:
   In file included from include/trace/events/block.h:8:
   In file included from include/linux/blktrace_api.h:5:
   In file included from include/linux/blk-mq.h:8:
   In file included from include/linux/scatterlist.h:9:
   In file included from arch/s390/include/asm/io.h:75:
   include/asm-generic/io.h:547:31: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           val = __raw_readb(PCI_IOBASE + addr);
                             ~~~~~~~~~~ ^
   include/asm-generic/io.h:560:61: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           val = __le16_to_cpu((__le16 __force)__raw_readw(PCI_IOBASE + addr));
                                                           ~~~~~~~~~~ ^
   include/uapi/linux/byteorder/big_endian.h:37:59: note: expanded from macro '__le16_to_cpu'
   #define __le16_to_cpu(x) __swab16((__force __u16)(__le16)(x))
                                                             ^
   include/uapi/linux/swab.h:102:54: note: expanded from macro '__swab16'
   #define __swab16(x) (__u16)__builtin_bswap16((__u16)(x))
                                                        ^
   In file included from drivers/md/md-bitmap.c:31:
   In file included from include/trace/events/block.h:8:
   In file included from include/linux/blktrace_api.h:5:
   In file included from include/linux/blk-mq.h:8:
   In file included from include/linux/scatterlist.h:9:
   In file included from arch/s390/include/asm/io.h:75:
   include/asm-generic/io.h:573:61: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           val = __le32_to_cpu((__le32 __force)__raw_readl(PCI_IOBASE + addr));
                                                           ~~~~~~~~~~ ^
   include/uapi/linux/byteorder/big_endian.h:35:59: note: expanded from macro '__le32_to_cpu'
   #define __le32_to_cpu(x) __swab32((__force __u32)(__le32)(x))
                                                             ^
   include/uapi/linux/swab.h:115:54: note: expanded from macro '__swab32'
   #define __swab32(x) (__u32)__builtin_bswap32((__u32)(x))
                                                        ^
   In file included from drivers/md/md-bitmap.c:31:
   In file included from include/trace/events/block.h:8:
   In file included from include/linux/blktrace_api.h:5:
   In file included from include/linux/blk-mq.h:8:
   In file included from include/linux/scatterlist.h:9:
   In file included from arch/s390/include/asm/io.h:75:
   include/asm-generic/io.h:584:33: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           __raw_writeb(value, PCI_IOBASE + addr);
                               ~~~~~~~~~~ ^
   include/asm-generic/io.h:594:59: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           __raw_writew((u16 __force)cpu_to_le16(value), PCI_IOBASE + addr);
                                                         ~~~~~~~~~~ ^
   include/asm-generic/io.h:604:59: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           __raw_writel((u32 __force)cpu_to_le32(value), PCI_IOBASE + addr);
                                                         ~~~~~~~~~~ ^
   include/asm-generic/io.h:692:20: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           readsb(PCI_IOBASE + addr, buffer, count);
                  ~~~~~~~~~~ ^
   include/asm-generic/io.h:700:20: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           readsw(PCI_IOBASE + addr, buffer, count);
                  ~~~~~~~~~~ ^
   include/asm-generic/io.h:708:20: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           readsl(PCI_IOBASE + addr, buffer, count);
                  ~~~~~~~~~~ ^
   include/asm-generic/io.h:717:21: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           writesb(PCI_IOBASE + addr, buffer, count);
                   ~~~~~~~~~~ ^
   include/asm-generic/io.h:726:21: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           writesw(PCI_IOBASE + addr, buffer, count);
                   ~~~~~~~~~~ ^
   include/asm-generic/io.h:735:21: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
           writesl(PCI_IOBASE + addr, buffer, count);
                   ~~~~~~~~~~ ^

Quoted 7 lines

Quoted 1 line
Quoted 1 line
drivers/md/md-bitmap.c:273:18: warning: variable 'rdev' is used uninitialized whenever function 'write_sb_page' is called [-Wsometimes-uninitialized]
           struct md_rdev *rdev;
           ~~~~~~~~~~~~~~~~^~~~
   drivers/md/md-bitmap.c:278:35: note: uninitialized use occurs here
                   while ((rdev = next_active_rdev(rdev, mddev)) != NULL) {
                                                   ^~~~
drivers/md/md-bitmap.c:273:22: note: initialize the variable 'rdev' to silence this warning

Oops, I missed the rdev = NULL;
V5 incoming

Quoted 27 lines

           struct md_rdev *rdev;
                               ^
                                = NULL

13 warnings generated.

vim +273 drivers/md/md-bitmap.c

   270	
   271	static int write_sb_page(struct bitmap *bitmap, struct page *page, int wait)
   272	{

Quoted 1 line

273 struct md_rdev *rdev;

   274		struct mddev *mddev = bitmap->mddev;
   275		int ret;
   276	
   277		do {
   278			while ((rdev = next_active_rdev(rdev, mddev)) != NULL) {
   279				ret = __write_sb_page(rdev, bitmap, page);
   280				if (ret)
   281					return ret;
   282			}
   283		} while (wait && md_super_wait(mddev) < 0);
   284	
   285		return 0;
   286	}
   287

Jonathan Derrick

author

2023-02-23 19:53 UTC #4 lore raw

PATCH v4 2/3 md: Fix types in sb writer

In reply to Jonathan Derrick

From: Jon Derrick <[email protected]>

Page->index is a pgoff_t and multiplying could cause overflows on a 32-bit architecture. In the sb writer, this is used to calculate and verify the sector being used, and is multiplied by a sector value. Using sector_t will cast it to a u64 type and is the more appropriate type for the unit. Additionally, the integer size unit is converted to a sector unit in later calculations, and is now corrected to be an unsigned type.

Finally, clean up the calculations using variable aliases to improve readabiliy.

Reviewed-by Christoph Hellwig <[email protected]>

Signed-off-by Jon Derrick <[email protected]>

---
 drivers/md/md-bitmap.c | 35 ++++++++++++++---------------------
 1 file changed, 14 insertions(+), 21 deletions(-)

drivers/md/md-bitmap.c +14 ~~-21~~

index 2d161d24c51c..20791d5e8903 100644
@@ -215,12 +215,13 @@ static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap,
 	struct block_device *bdev;
 	struct mddev *mddev = bitmap->mddev;
 	struct bitmap_storage *store = &bitmap->storage;
-	loff_t offset = mddev->bitmap_info.offset;
-	int size = PAGE_SIZE;
+	sector_t offset = mddev->bitmap_info.offset;
+	sector_t ps, sboff, doff;
+	unsigned int size = PAGE_SIZE;
 
 	bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev;
 	if (page->index == store->file_pages - 1) {
-		int last_page_size = store->bytes & (PAGE_SIZE - 1);
+		unsigned int last_page_size = store->bytes & (PAGE_SIZE - 1);
 
 		if (last_page_size == 0)
 			last_page_size = PAGE_SIZE;

@@ -228,43 +229,35 @@ static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap,
 			       bdev_logical_block_size(bdev));
 	}
 
+	ps = page->index * PAGE_SIZE / SECTOR_SIZE;
+	sboff = rdev->sb_start + offset;
+	doff = rdev->data_offset;
+
 	/* Just make sure we aren't corrupting data or metadata */
 	if (mddev->external) {
 		/* Bitmap could be anywhere. */
-		if (rdev->sb_start + offset
-		    + (page->index * (PAGE_SIZE / SECTOR_SIZE))
-		    > rdev->data_offset &&
-		    rdev->sb_start + offset
-		    < (rdev->data_offset + mddev->dev_sectors
-		     + (PAGE_SIZE / SECTOR_SIZE)))
+		if (sboff + ps > doff &&
+		    sboff < (doff + mddev->dev_sectors + PAGE_SIZE / SECTOR_SIZE))
 			return -EINVAL;
 	} else if (offset < 0) {
 		/* DATA  BITMAP METADATA  */
-		if (offset
-		    + (long)(page->index * (PAGE_SIZE / SECTOR_SIZE))
-		    + size / SECTOR_SIZE > 0)
+		if (offset + ps + size / SECTOR_SIZE > 0)
 			/* bitmap runs in to metadata */
 			return -EINVAL;
 
-		if (rdev->data_offset + mddev->dev_sectors
-		    > rdev->sb_start + offset)
+		if (doff + mddev->dev_sectors > sboff)
 			/* data runs in to bitmap */
 			return -EINVAL;
 	} else if (rdev->sb_start < rdev->data_offset) {
 		/* METADATA BITMAP DATA */
-		if (rdev->sb_start + offset
-		    + page->index * (PAGE_SIZE / SECTOR_SIZE)
-		    + size / SECTOR_SIZE > rdev->data_offset)
+		if (sboff + ps + size / SECTOR_SIZE > doff)
 			/* bitmap runs in to data */
 			return -EINVAL;
 	} else {
 		/* DATA METADATA BITMAP - no problems */
 	}
 
-	md_super_write(mddev, rdev,
-		       rdev->sb_start + offset
-		       + page->index * (PAGE_SIZE / SECTOR_SIZE),
-		       size, page);
+	md_super_write(mddev, rdev, sboff + ps, (int) size, page);
 	return 0;
 }

Signature · 1 line

2.27.0

Jonathan Derrick

author

2023-02-23 19:53 UTC #5 lore raw

PATCH v4 3/3 md: Use optimal I/O size for last bitmap page

In reply to Jonathan Derrick

From: Jon Derrick <[email protected]>

If the bitmap space has enough room, size the I/O for the last bitmap page write to the optimal I/O size for the storage device. The expanded write is checked that it won't overrun the data or metadata.

The drive this was tested against has higher latencies when there are sub-4k writes due to device-side read-mod-writes of its atomic 4k write unit. This change helps increase performance by sizing the last bitmap page I/O for the device's preferred write unit, if it is given.

Example Intel/Solidigm P5520
Raid10, Chunk-size 64M, bitmap-size 57228 bits

$ mdadm --create /dev/md0 --level=10 --raid-devices=4 /dev/nvme{0,1,2,3}n1
        --assume-clean --bitmap=internal --bitmap-chunk=64M
$ fio --name=test --direct=1 --filename=/dev/md0 --rw=randwrite --bs=4k --runtime=60

Without patch:
write: IOPS=1676, BW=6708KiB/s (6869kB/s)(393MiB/60001msec); 0 zone resets

With patch:
write: IOPS=15.7k, BW=61.4MiB/s (64.4MB/s)(3683MiB/60001msec); 0 zone resets

Biosnoop:
Without patch:

Time        Process        PID     Device      LBA        Size      Lat
1.410377    md0_raid10     6900    nvme0n1   W 16         4096      0.02
1.410387    md0_raid10     6900    nvme2n1   W 16         4096      0.02
1.410374    md0_raid10     6900    nvme3n1   W 16         4096      0.01
1.410381    md0_raid10     6900    nvme1n1   W 16         4096      0.02
1.410411    md0_raid10     6900    nvme1n1   W 115346512  4096      0.01
1.410418    md0_raid10     6900    nvme0n1   W 115346512  4096      0.02
1.410915    md0_raid10     6900    nvme2n1   W 24         3584      0.43 <--
1.410935    md0_raid10     6900    nvme3n1   W 24         3584      0.45 <--
1.411124    md0_raid10     6900    nvme1n1   W 24         3584      0.64 <--
1.411147    md0_raid10     6900    nvme0n1   W 24         3584      0.66 <--
1.411176    md0_raid10     6900    nvme3n1   W 2019022184 4096      0.01
1.411189    md0_raid10     6900    nvme2n1   W 2019022184 4096      0.02

With patch:

Time        Process        PID     Device      LBA        Size      Lat
5.747193    md0_raid10     727     nvme0n1   W 16         4096      0.01
5.747192    md0_raid10     727     nvme1n1   W 16         4096      0.02
5.747195    md0_raid10     727     nvme3n1   W 16         4096      0.01
5.747202    md0_raid10     727     nvme2n1   W 16         4096      0.02
5.747229    md0_raid10     727     nvme3n1   W 1196223704 4096      0.02
5.747224    md0_raid10     727     nvme0n1   W 1196223704 4096      0.01
5.747279    md0_raid10     727     nvme0n1   W 24         4096      0.01 <--
5.747279    md0_raid10     727     nvme1n1   W 24         4096      0.02 <--
5.747284    md0_raid10     727     nvme3n1   W 24         4096      0.02 <--
5.747291    md0_raid10     727     nvme2n1   W 24         4096      0.02 <--
5.747314    md0_raid10     727     nvme2n1   W 2234636712 4096      0.01
5.747317    md0_raid10     727     nvme1n1   W 2234636712 4096      0.02

Reviewed-by Christoph Hellwig <[email protected]>

Signed-off-by Jon Derrick <[email protected]>

---
 drivers/md/md-bitmap.c | 33 +++++++++++++++++++++++++++++----
 1 file changed, 29 insertions(+), 4 deletions(-)

drivers/md/md-bitmap.c +29 -4

index 20791d5e8903..06b2d7da57da 100644
@@ -209,6 +209,28 @@ static struct md_rdev *next_active_rdev(struct md_rdev *rdev, struct mddev *mdde
 	return NULL;
 }
 
+static unsigned int optimal_io_size(struct block_device *bdev,
+				    unsigned int last_page_size,
+				    unsigned int io_size)
+{
+	if (bdev_io_opt(bdev) > bdev_logical_block_size(bdev))
+		return roundup(last_page_size, bdev_io_opt(bdev));
+	return io_size;
+}
+
+static unsigned int bitmap_io_size(unsigned int io_size, unsigned int opt_size,
+				   sector_t start, sector_t boundary)
+{
+	if (io_size != opt_size &&
+	    start + opt_size / SECTOR_SIZE <= boundary)
+		return opt_size;
+	if (start + io_size / SECTOR_SIZE <= boundary)
+		return io_size;
+
+	/* Overflows boundary */
+	return 0;
+}
+
 static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap,
 			   struct page *page)
 {

@@ -218,6 +240,7 @@ static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap,
 	sector_t offset = mddev->bitmap_info.offset;
 	sector_t ps, sboff, doff;
 	unsigned int size = PAGE_SIZE;
+	unsigned int opt_size = PAGE_SIZE;
 
 	bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev;
 	if (page->index == store->file_pages - 1) {

@@ -225,8 +248,8 @@ static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap,
 
 		if (last_page_size == 0)
 			last_page_size = PAGE_SIZE;
-		size = roundup(last_page_size,
-			       bdev_logical_block_size(bdev));
+		size = roundup(last_page_size, bdev_logical_block_size(bdev));
+		opt_size = optimal_io_size(bdev, last_page_size, size);
 	}
 
 	ps = page->index * PAGE_SIZE / SECTOR_SIZE;

@@ -241,7 +264,8 @@ static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap,
 			return -EINVAL;
 	} else if (offset < 0) {
 		/* DATA  BITMAP METADATA  */
-		if (offset + ps + size / SECTOR_SIZE > 0)
+		size = bitmap_io_size(size, opt_size, offset + ps, 0);
+		if (size == 0)
 			/* bitmap runs in to metadata */
 			return -EINVAL;

@@ -250,7 +274,8 @@ static int __write_sb_page(struct md_rdev *rdev, struct bitmap *bitmap,
 			return -EINVAL;
 	} else if (rdev->sb_start < rdev->data_offset) {
 		/* METADATA BITMAP DATA */
-		if (sboff + ps + size / SECTOR_SIZE > doff)
+		size = bitmap_io_size(size, opt_size, sboff + ps, doff);
+		if (size == 0)
 			/* bitmap runs in to data */
 			return -EINVAL;
 	} else {

Signature · 1 line

2.27.0