All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH 1/2] btrfs-progs: Return more meaningful value for btrfs_read_dev_super
@ 2016-09-20  2:19 Qu Wenruo
  2016-09-20  2:19 ` [PATCH 2/2] btrfs-progs: super-recover: Reuse btrfs_read_dev_super function Qu Wenruo
  2016-09-23 13:22 ` [PATCH 1/2] btrfs-progs: Return more meaningful value for btrfs_read_dev_super David Sterba
  0 siblings, 2 replies; 5+ messages in thread
From: Qu Wenruo @ 2016-09-20  2:19 UTC (permalink / raw)
  To: linux-btrfs; +Cc: hawken

btrfs_read_dev_super() only returns 0 or -1, which doesn't really help,
caller won't know if it's caused by bad superblock or superblock out of
range.

Change it to return -ENOENT for superblock which is out of range, and
return -EIO for bad superblocks.

Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
---
 disk-io.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/disk-io.c b/disk-io.c
index 854c285..26852ee 100644
--- a/disk-io.c
+++ b/disk-io.c
@@ -1590,14 +1590,15 @@ int btrfs_read_dev_super(int fd, struct btrfs_super_block *sb, u64 sb_bytenr,
 
 	if (sb_bytenr != BTRFS_SUPER_INFO_OFFSET) {
 		ret = pread64(fd, buf, BTRFS_SUPER_INFO_SIZE, sb_bytenr);
+		/* Special return value for out of dev/file range */
 		if (ret < BTRFS_SUPER_INFO_SIZE)
-			return -1;
+			return -ENOENT;
 
 		if (btrfs_super_bytenr(buf) != sb_bytenr)
-			return -1;
+			return -EIO;
 
 		if (check_super(buf, sbflags))
-			return -1;
+			return -EIO;
 		memcpy(sb, buf, BTRFS_SUPER_INFO_SIZE);
 		return 0;
 	}
@@ -1641,7 +1642,7 @@ int btrfs_read_dev_super(int fd, struct btrfs_super_block *sb, u64 sb_bytenr,
 		}
 	}
 
-	return transid > 0 ? 0 : -1;
+	return transid > 0 ? 0 : -EIO;
 }
 
 static int write_dev_supers(struct btrfs_root *root,
-- 
2.10.0




^ permalink raw reply related	[flat|nested] 5+ messages in thread

* [PATCH 2/2] btrfs-progs: super-recover: Reuse btrfs_read_dev_super function
  2016-09-20  2:19 [PATCH 1/2] btrfs-progs: Return more meaningful value for btrfs_read_dev_super Qu Wenruo
@ 2016-09-20  2:19 ` Qu Wenruo
  2016-09-23 13:26   ` David Sterba
  2016-09-23 13:22 ` [PATCH 1/2] btrfs-progs: Return more meaningful value for btrfs_read_dev_super David Sterba
  1 sibling, 1 reply; 5+ messages in thread
From: Qu Wenruo @ 2016-09-20  2:19 UTC (permalink / raw)
  To: linux-btrfs; +Cc: hawken

We have enhanced super block validation check in disk_io.c, while
super-recover doesn't use that function and use a custom one.

That is duplicated, so reuse btrfs_read_dev_super to handle it, which
not only enhanced super check, but also reduce the code length.

Reported-by: hawken <hawken@thehawken.org>
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
---
 super-recover.c | 39 +++++++++++----------------------------
 1 file changed, 11 insertions(+), 28 deletions(-)

diff --git a/super-recover.c b/super-recover.c
index ecd1215..5298d46 100644
--- a/super-recover.c
+++ b/super-recover.c
@@ -88,24 +88,6 @@ void free_recover_superblock(struct btrfs_recover_superblock *recover)
 	}
 }
 
-static int check_super(u64 bytenr, struct btrfs_super_block *sb)
-{
-	int csum_size = btrfs_super_csum_size(sb);
-	u8 result[csum_size];
-	u32 crc = ~(u32)0;
-
-	if (btrfs_super_bytenr(sb) != bytenr)
-		return 0;
-	if (sb->magic != cpu_to_le64(BTRFS_MAGIC))
-		return 0;
-
-	crc = btrfs_csum_data(NULL, (char *)sb + BTRFS_CSUM_SIZE,
-			crc, BTRFS_SUPER_INFO_SIZE - BTRFS_CSUM_SIZE);
-	btrfs_csum_final(crc, result);
-
-	return !memcmp(sb, &result, csum_size);
-}
-
 static int add_superblock_record(struct btrfs_super_block *sb, char *fname,
 			u64 bytenr, struct list_head *head)
 {
@@ -133,24 +115,20 @@ read_dev_supers(char *filename, struct btrfs_recover_superblock *recover)
 	int i, ret, fd;
 	u8 buf[BTRFS_SUPER_INFO_SIZE];
 	u64 max_gen, bytenr;
+	struct btrfs_super_block *sb = (struct btrfs_super_block *)buf;
+
 	/* just ignore errno that were set in btrfs_scan_fs_devices() */
 	errno = 0;
 
-	struct btrfs_super_block *sb = (struct btrfs_super_block *)buf;
-
 	fd = open(filename, O_RDONLY);
 	if (fd < 0)
 		return -errno;
 
 	for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) {
 		bytenr = btrfs_sb_offset(i);
-		ret = pread64(fd, buf, sizeof(buf), bytenr);
-		if (ret < sizeof(buf)) {
-			ret = -errno;
-			goto out;
-		}
-		ret = check_super(bytenr, sb);
-		if (ret) {
+
+		ret = btrfs_read_dev_super(fd, sb, bytenr, SBREAD_DEFAULT);
+		if (!ret) {
 			ret = add_superblock_record(sb, filename, bytenr,
 							&recover->good_supers);
 			if (ret)
@@ -158,13 +136,18 @@ read_dev_supers(char *filename, struct btrfs_recover_superblock *recover)
 			max_gen = btrfs_super_generation(sb);
 			if (max_gen > recover->max_generation)
 				recover->max_generation = max_gen;
-		} else {
+		} else if (ret == -EIO){
+			/*
+			 * Skip superblock which doesn't exist, only adds
+			 * really corrupted superblock
+			 */
 			ret = add_superblock_record(sb, filename, bytenr,
 						&recover->bad_supers);
 			if (ret)
 				goto out;
 		}
 	}
+	ret = 0;
 out:
 	close(fd);
 	return ret;
-- 
2.10.0




^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Re: [PATCH 1/2] btrfs-progs: Return more meaningful value for btrfs_read_dev_super
  2016-09-20  2:19 [PATCH 1/2] btrfs-progs: Return more meaningful value for btrfs_read_dev_super Qu Wenruo
  2016-09-20  2:19 ` [PATCH 2/2] btrfs-progs: super-recover: Reuse btrfs_read_dev_super function Qu Wenruo
@ 2016-09-23 13:22 ` David Sterba
  1 sibling, 0 replies; 5+ messages in thread
From: David Sterba @ 2016-09-23 13:22 UTC (permalink / raw)
  To: Qu Wenruo; +Cc: linux-btrfs, hawken

On Tue, Sep 20, 2016 at 10:19:01AM +0800, Qu Wenruo wrote:
> btrfs_read_dev_super() only returns 0 or -1, which doesn't really help,
> caller won't know if it's caused by bad superblock or superblock out of
> range.
> 
> Change it to return -ENOENT for superblock which is out of range, and
> return -EIO for bad superblocks.
> 
> Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
> ---
>  disk-io.c | 9 +++++----
>  1 file changed, 5 insertions(+), 4 deletions(-)
> 
> diff --git a/disk-io.c b/disk-io.c
> index 854c285..26852ee 100644
> --- a/disk-io.c
> +++ b/disk-io.c
> @@ -1590,14 +1590,15 @@ int btrfs_read_dev_super(int fd, struct btrfs_super_block *sb, u64 sb_bytenr,
>  
>  	if (sb_bytenr != BTRFS_SUPER_INFO_OFFSET) {
>  		ret = pread64(fd, buf, BTRFS_SUPER_INFO_SIZE, sb_bytenr);

pread64 sets errno, should be returned

> +		/* Special return value for out of dev/file range */
>  		if (ret < BTRFS_SUPER_INFO_SIZE)
> -			return -1;
> +			return -ENOENT;
>  
>  		if (btrfs_super_bytenr(buf) != sb_bytenr)
> -			return -1;
> +			return -EIO;
>  
>  		if (check_super(buf, sbflags))
> -			return -1;
> +			return -EIO;

and we should return what check_super returns, though it's -EIO in most
cases.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH 2/2] btrfs-progs: super-recover: Reuse btrfs_read_dev_super function
  2016-09-20  2:19 ` [PATCH 2/2] btrfs-progs: super-recover: Reuse btrfs_read_dev_super function Qu Wenruo
@ 2016-09-23 13:26   ` David Sterba
  2016-09-26  3:05     ` Qu Wenruo
  0 siblings, 1 reply; 5+ messages in thread
From: David Sterba @ 2016-09-23 13:26 UTC (permalink / raw)
  To: Qu Wenruo; +Cc: linux-btrfs, hawken

On Tue, Sep 20, 2016 at 10:19:02AM +0800, Qu Wenruo wrote:
> We have enhanced super block validation check in disk_io.c, while
> super-recover doesn't use that function and use a custom one.
> 
> That is duplicated, so reuse btrfs_read_dev_super to handle it, which
> not only enhanced super check, but also reduce the code length.

Yeah, the existing check_super does way more checks compared to this
one. Applied, thanks.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH 2/2] btrfs-progs: super-recover: Reuse btrfs_read_dev_super function
  2016-09-23 13:26   ` David Sterba
@ 2016-09-26  3:05     ` Qu Wenruo
  0 siblings, 0 replies; 5+ messages in thread
From: Qu Wenruo @ 2016-09-26  3:05 UTC (permalink / raw)
  To: dsterba, linux-btrfs, hawken



At 09/23/2016 09:26 PM, David Sterba wrote:
> On Tue, Sep 20, 2016 at 10:19:02AM +0800, Qu Wenruo wrote:
>> We have enhanced super block validation check in disk_io.c, while
>> super-recover doesn't use that function and use a custom one.
>>
>> That is duplicated, so reuse btrfs_read_dev_super to handle it, which
>> not only enhanced super check, but also reduce the code length.
>
> Yeah, the existing check_super does way more checks compared to this
> one. Applied, thanks.
>
>
Sorry David, this patch can't be applied without previous patch.

As this patch relies on previous patch to get meaningful return value, 
to check if it's a real error or that superblock just doesn't exist.

I'll update previous patch soon.

Thanks,
Qu



^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2016-09-26  3:06 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-09-20  2:19 [PATCH 1/2] btrfs-progs: Return more meaningful value for btrfs_read_dev_super Qu Wenruo
2016-09-20  2:19 ` [PATCH 2/2] btrfs-progs: super-recover: Reuse btrfs_read_dev_super function Qu Wenruo
2016-09-23 13:26   ` David Sterba
2016-09-26  3:05     ` Qu Wenruo
2016-09-23 13:22 ` [PATCH 1/2] btrfs-progs: Return more meaningful value for btrfs_read_dev_super David Sterba

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.