All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] erofs-utils: erofs debug utility.
@ 2019-08-21 16:38 Pratik Shinde
  2019-08-21 16:56 ` Gao Xiang via Linux-erofs
  0 siblings, 1 reply; 5+ messages in thread
From: Pratik Shinde @ 2019-08-21 16:38 UTC (permalink / raw)
  To: linux-erofs, bluce.liguifu, miaoxie, fangwei1

Hello Maintainers,

After going through the recent mail thread between linux's filesystem folks
on erofs channel, I felt erofs needs an interactive debug utility (like xfs_db)
which can be used to examine erofs images & can also be used to inject errors OR
fuzzing for testing purpose & dumping different erofs meta data structures
for debugging.
In order to demonstrate above I wrote an experimental patch that simply dumps
the superblock of an image after mkfs completes.the full fletch utility will run
independently and be able to seek / print / modify any byte of an erofs image,
dump structures/lists/directory content of an image.

NOTE:This is an experimental patch just to demonstrate the purpose. The patch
lacks a lot of things like coding standard, and new code runs in the context
of mkfs itself.kindly ignore it.

kindly provide your feedback on this.

Signed-off-by: Pratik Shinde <pratikshinde320@gmail.com>
---
 include/erofs/io.h |  8 ++++++++
 lib/io.c           | 27 +++++++++++++++++++++++++++
 mkfs/main.c        | 36 ++++++++++++++++++++++++++++++++++++
 3 files changed, 71 insertions(+)

diff --git a/include/erofs/io.h b/include/erofs/io.h
index 4b574bd..e91d6ee 100644
--- a/include/erofs/io.h
+++ b/include/erofs/io.h
@@ -18,6 +18,7 @@
 
 int dev_open(const char *devname);
 void dev_close(void);
+int dev_read(void *buf, u64 offset, size_t len);
 int dev_write(const void *buf, u64 offset, size_t len);
 int dev_fillzero(u64 offset, size_t len, bool padding);
 int dev_fsync(void);
@@ -30,5 +31,12 @@ static inline int blk_write(const void *buf, erofs_blk_t blkaddr,
 			 blknr_to_addr(nblocks));
 }
 
+static inline int blk_read(void *buf, erofs_blk_t blkaddr,
+			   u32 nblocks)
+{
+	return dev_read(buf, blknr_to_addr(blkaddr),
+			blknr_to_addr(nblocks));
+}
+
 #endif
 
diff --git a/lib/io.c b/lib/io.c
index 15c5a35..87d7d6c 100644
--- a/lib/io.c
+++ b/lib/io.c
@@ -109,6 +109,33 @@ u64 dev_length(void)
 	return erofs_devsz;
 }
 
+int dev_read(void *buf, u64 offset, size_t len)
+{
+	int ret;
+
+	if (cfg.c_dry_run)
+		return 0;
+
+	if (!buf) {
+		erofs_err("buf is NULL");
+		return -EINVAL;
+	}
+	if (offset >= erofs_devsz || len > erofs_devsz ||
+	    offset > erofs_devsz - len) {
+		erofs_err("read posion[%" PRIu64 ", %zd] is too large beyond the end of device(%" PRIu64 ").",
+			  offset, len, erofs_devsz);
+		return -EINVAL;
+	}
+
+	ret = pread64(erofs_devfd, buf, len, (off64_t)offset);
+	if (ret != (int)len) {
+		erofs_err("Failed to read data from device - %s:[%" PRIu64 ", %zd].",
+			  erofs_devname, offset, len);
+		return -errno;
+	}
+	return 0;
+}
+
 int dev_write(const void *buf, u64 offset, size_t len)
 {
 	int ret;
diff --git a/mkfs/main.c b/mkfs/main.c
index f127fe1..109486e 100644
--- a/mkfs/main.c
+++ b/mkfs/main.c
@@ -182,6 +182,41 @@ int erofs_mkfs_update_super_block(struct erofs_buffer_head *bh,
 	return 0;
 }
 
+void erofs_dump_super(char *img_path)
+{
+	struct erofs_super_block *sb;
+	char buf[EROFS_BLKSIZ];
+	unsigned int blksz;
+	int ret = 0;
+
+	if (img_path == NULL) {
+		erofs_err("image path cannot be null");
+		return;
+	}
+	ret = blk_read(buf, 0, 1);
+	if (ret) {
+		erofs_err("error reading super-block structure");
+		return;
+	}
+
+	sb = (struct erofs_super_block *)((u8 *)buf + EROFS_SUPER_OFFSET);
+	if (le32_to_cpu(sb->magic) != EROFS_SUPER_MAGIC_V1) {
+		erofs_err("not a erofs image");
+		return;
+	}
+
+	erofs_dump("magic: 0x%x\n", le32_to_cpu(sb->magic));
+	blksz = 1 << sb->blkszbits;
+	erofs_dump("block size: %d\n", blksz);
+	erofs_dump("root inode: %d\n", le32_to_cpu(sb->root_nid));
+	erofs_dump("inodes: %llu\n", le64_to_cpu(sb->inos));
+	erofs_dump("build time: %u\n", le32_to_cpu(sb->build_time));
+	erofs_dump("blocks: %u\n", le32_to_cpu(sb->blocks));
+	erofs_dump("meta block: %u\n", le32_to_cpu(sb->meta_blkaddr));
+	erofs_dump("xattr block: %u\n", le32_to_cpu(sb->xattr_blkaddr));
+	erofs_dump("requirements: 0x%x\n", le32_to_cpu(sb->requirements));
+}
+
 int main(int argc, char **argv)
 {
 	int err = 0;
@@ -268,6 +303,7 @@ int main(int argc, char **argv)
 		err = -EIO;
 exit:
 	z_erofs_compress_exit();
+	erofs_dump_super("dummy");
 	dev_close();
 	erofs_exit_configure();
 
-- 
2.9.3


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

* Re: [PATCH] erofs-utils: erofs debug utility.
  2019-08-21 16:38 [PATCH] erofs-utils: erofs debug utility Pratik Shinde
@ 2019-08-21 16:56 ` Gao Xiang via Linux-erofs
  2019-08-21 18:15   ` Pratik Shinde
  0 siblings, 1 reply; 5+ messages in thread
From: Gao Xiang via Linux-erofs @ 2019-08-21 16:56 UTC (permalink / raw)
  To: Pratik Shinde; +Cc: miaoxie, linux-erofs

Hi Pratik,

On Wed, Aug 21, 2019 at 10:08:08PM +0530, Pratik Shinde wrote:
> Hello Maintainers,
> 
> After going through the recent mail thread between linux's filesystem folks
> on erofs channel, I felt erofs needs an interactive debug utility (like xfs_db)
> which can be used to examine erofs images & can also be used to inject errors OR
> fuzzing for testing purpose & dumping different erofs meta data structures
> for debugging.
> In order to demonstrate above I wrote an experimental patch that simply dumps
> the superblock of an image after mkfs completes.the full fletch utility will run
> independently and be able to seek / print / modify any byte of an erofs image,
> dump structures/lists/directory content of an image.

Yes, I think we really need that interactive tools, actually I'm stuggling in
modifing Guifu's erofs-fuse now, we need to add the parsing ability to "lib/"
first.

I mean, first, I will add a "fuse" field to "cfg". If it is false, it will
generate a image, or it will parse a image...
And then we need to add parsing logic into "lib/" as well, and use 
"if (cfg.fuse)" to differnate whether it should read or write data.

That is my prelimitary thought.. I will work on this framework in this weekend.
and then we can work together on it. :)

p.s. Pratik, if you have some time, could you take some extra time adding the
super checksum calulation to EROFS? I mean we can add EROFS_FEATURE_SB_CHKSUM
to the compat superblock field ("features"), and do crc32_le on kernel and mkfs...
If you dont have time, I will do it later instead... (since we are using EROFS
on the top of dm-verity, but completing the superblock chksum is also a good idea.)

And then we can add block-based verification layer to EROFS, it can be seen
as a hash tree like dm-verity or just simply CRC32 arrays for user to choise.

Thanks,
Gao Xiang

> 
> NOTE:This is an experimental patch just to demonstrate the purpose. The patch
> lacks a lot of things like coding standard, and new code runs in the context
> of mkfs itself.kindly ignore it.
> 
> kindly provide your feedback on this.
> 
> Signed-off-by: Pratik Shinde <pratikshinde320@gmail.com>
> ---
>  include/erofs/io.h |  8 ++++++++
>  lib/io.c           | 27 +++++++++++++++++++++++++++
>  mkfs/main.c        | 36 ++++++++++++++++++++++++++++++++++++
>  3 files changed, 71 insertions(+)
> 
> diff --git a/include/erofs/io.h b/include/erofs/io.h
> index 4b574bd..e91d6ee 100644
> --- a/include/erofs/io.h
> +++ b/include/erofs/io.h
> @@ -18,6 +18,7 @@
>  
>  int dev_open(const char *devname);
>  void dev_close(void);
> +int dev_read(void *buf, u64 offset, size_t len);
>  int dev_write(const void *buf, u64 offset, size_t len);
>  int dev_fillzero(u64 offset, size_t len, bool padding);
>  int dev_fsync(void);
> @@ -30,5 +31,12 @@ static inline int blk_write(const void *buf, erofs_blk_t blkaddr,
>  			 blknr_to_addr(nblocks));
>  }
>  
> +static inline int blk_read(void *buf, erofs_blk_t blkaddr,
> +			   u32 nblocks)
> +{
> +	return dev_read(buf, blknr_to_addr(blkaddr),
> +			blknr_to_addr(nblocks));
> +}
> +
>  #endif
>  
> diff --git a/lib/io.c b/lib/io.c
> index 15c5a35..87d7d6c 100644
> --- a/lib/io.c
> +++ b/lib/io.c
> @@ -109,6 +109,33 @@ u64 dev_length(void)
>  	return erofs_devsz;
>  }
>  
> +int dev_read(void *buf, u64 offset, size_t len)
> +{
> +	int ret;
> +
> +	if (cfg.c_dry_run)
> +		return 0;
> +
> +	if (!buf) {
> +		erofs_err("buf is NULL");
> +		return -EINVAL;
> +	}
> +	if (offset >= erofs_devsz || len > erofs_devsz ||
> +	    offset > erofs_devsz - len) {
> +		erofs_err("read posion[%" PRIu64 ", %zd] is too large beyond the end of device(%" PRIu64 ").",
> +			  offset, len, erofs_devsz);
> +		return -EINVAL;
> +	}
> +
> +	ret = pread64(erofs_devfd, buf, len, (off64_t)offset);
> +	if (ret != (int)len) {
> +		erofs_err("Failed to read data from device - %s:[%" PRIu64 ", %zd].",
> +			  erofs_devname, offset, len);
> +		return -errno;
> +	}
> +	return 0;
> +}
> +
>  int dev_write(const void *buf, u64 offset, size_t len)
>  {
>  	int ret;
> diff --git a/mkfs/main.c b/mkfs/main.c
> index f127fe1..109486e 100644
> --- a/mkfs/main.c
> +++ b/mkfs/main.c
> @@ -182,6 +182,41 @@ int erofs_mkfs_update_super_block(struct erofs_buffer_head *bh,
>  	return 0;
>  }
>  
> +void erofs_dump_super(char *img_path)
> +{
> +	struct erofs_super_block *sb;
> +	char buf[EROFS_BLKSIZ];
> +	unsigned int blksz;
> +	int ret = 0;
> +
> +	if (img_path == NULL) {
> +		erofs_err("image path cannot be null");
> +		return;
> +	}
> +	ret = blk_read(buf, 0, 1);
> +	if (ret) {
> +		erofs_err("error reading super-block structure");
> +		return;
> +	}
> +
> +	sb = (struct erofs_super_block *)((u8 *)buf + EROFS_SUPER_OFFSET);
> +	if (le32_to_cpu(sb->magic) != EROFS_SUPER_MAGIC_V1) {
> +		erofs_err("not a erofs image");
> +		return;
> +	}
> +
> +	erofs_dump("magic: 0x%x\n", le32_to_cpu(sb->magic));
> +	blksz = 1 << sb->blkszbits;
> +	erofs_dump("block size: %d\n", blksz);
> +	erofs_dump("root inode: %d\n", le32_to_cpu(sb->root_nid));
> +	erofs_dump("inodes: %llu\n", le64_to_cpu(sb->inos));
> +	erofs_dump("build time: %u\n", le32_to_cpu(sb->build_time));
> +	erofs_dump("blocks: %u\n", le32_to_cpu(sb->blocks));
> +	erofs_dump("meta block: %u\n", le32_to_cpu(sb->meta_blkaddr));
> +	erofs_dump("xattr block: %u\n", le32_to_cpu(sb->xattr_blkaddr));
> +	erofs_dump("requirements: 0x%x\n", le32_to_cpu(sb->requirements));
> +}
> +
>  int main(int argc, char **argv)
>  {
>  	int err = 0;
> @@ -268,6 +303,7 @@ int main(int argc, char **argv)
>  		err = -EIO;
>  exit:
>  	z_erofs_compress_exit();
> +	erofs_dump_super("dummy");
>  	dev_close();
>  	erofs_exit_configure();
>  
> -- 
> 2.9.3
> 

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

* Re: [PATCH] erofs-utils: erofs debug utility.
  2019-08-21 16:56 ` Gao Xiang via Linux-erofs
@ 2019-08-21 18:15   ` Pratik Shinde
  2019-08-21 20:08     ` Gao Xiang via Linux-erofs
  0 siblings, 1 reply; 5+ messages in thread
From: Pratik Shinde @ 2019-08-21 18:15 UTC (permalink / raw)
  To: Gao Xiang; +Cc: miaoxie, linux-erofs

[-- Attachment #1: Type: text/plain, Size: 6788 bytes --]

Thanks Gao,

Let me know when we can work on debug utility OR atleast list out things we
want to achieve through the utility.
Regarding the superblock checksum calculations, Yes I will dedicate
sometime this week for it. will start exploring it.

--Pratik.

On Wed, Aug 21, 2019 at 10:26 PM Gao Xiang <hsiangkao@aol.com> wrote:

> Hi Pratik,
>
> On Wed, Aug 21, 2019 at 10:08:08PM +0530, Pratik Shinde wrote:
> > Hello Maintainers,
> >
> > After going through the recent mail thread between linux's filesystem
> folks
> > on erofs channel, I felt erofs needs an interactive debug utility (like
> xfs_db)
> > which can be used to examine erofs images & can also be used to inject
> errors OR
> > fuzzing for testing purpose & dumping different erofs meta data
> structures
> > for debugging.
> > In order to demonstrate above I wrote an experimental patch that simply
> dumps
> > the superblock of an image after mkfs completes.the full fletch utility
> will run
> > independently and be able to seek / print / modify any byte of an erofs
> image,
> > dump structures/lists/directory content of an image.
>
> Yes, I think we really need that interactive tools, actually I'm stuggling
> in
> modifing Guifu's erofs-fuse now, we need to add the parsing ability to
> "lib/"
> first.
>
> I mean, first, I will add a "fuse" field to "cfg". If it is false, it will
> generate a image, or it will parse a image...
> And then we need to add parsing logic into "lib/" as well, and use
> "if (cfg.fuse)" to differnate whether it should read or write data.
>
> That is my prelimitary thought.. I will work on this framework in this
> weekend.
> and then we can work together on it. :)
>
> p.s. Pratik, if you have some time, could you take some extra time adding
> the
> super checksum calulation to EROFS? I mean we can add
> EROFS_FEATURE_SB_CHKSUM
> to the compat superblock field ("features"), and do crc32_le on kernel and
> mkfs...
> If you dont have time, I will do it later instead... (since we are using
> EROFS
> on the top of dm-verity, but completing the superblock chksum is also a
> good idea.)
>
> And then we can add block-based verification layer to EROFS, it can be seen
> as a hash tree like dm-verity or just simply CRC32 arrays for user to
> choise.
>
> Thanks,
> Gao Xiang
>
> >
> > NOTE:This is an experimental patch just to demonstrate the purpose. The
> patch
> > lacks a lot of things like coding standard, and new code runs in the
> context
> > of mkfs itself.kindly ignore it.
> >
> > kindly provide your feedback on this.
> >
> > Signed-off-by: Pratik Shinde <pratikshinde320@gmail.com>
> > ---
> >  include/erofs/io.h |  8 ++++++++
> >  lib/io.c           | 27 +++++++++++++++++++++++++++
> >  mkfs/main.c        | 36 ++++++++++++++++++++++++++++++++++++
> >  3 files changed, 71 insertions(+)
> >
> > diff --git a/include/erofs/io.h b/include/erofs/io.h
> > index 4b574bd..e91d6ee 100644
> > --- a/include/erofs/io.h
> > +++ b/include/erofs/io.h
> > @@ -18,6 +18,7 @@
> >
> >  int dev_open(const char *devname);
> >  void dev_close(void);
> > +int dev_read(void *buf, u64 offset, size_t len);
> >  int dev_write(const void *buf, u64 offset, size_t len);
> >  int dev_fillzero(u64 offset, size_t len, bool padding);
> >  int dev_fsync(void);
> > @@ -30,5 +31,12 @@ static inline int blk_write(const void *buf,
> erofs_blk_t blkaddr,
> >                        blknr_to_addr(nblocks));
> >  }
> >
> > +static inline int blk_read(void *buf, erofs_blk_t blkaddr,
> > +                        u32 nblocks)
> > +{
> > +     return dev_read(buf, blknr_to_addr(blkaddr),
> > +                     blknr_to_addr(nblocks));
> > +}
> > +
> >  #endif
> >
> > diff --git a/lib/io.c b/lib/io.c
> > index 15c5a35..87d7d6c 100644
> > --- a/lib/io.c
> > +++ b/lib/io.c
> > @@ -109,6 +109,33 @@ u64 dev_length(void)
> >       return erofs_devsz;
> >  }
> >
> > +int dev_read(void *buf, u64 offset, size_t len)
> > +{
> > +     int ret;
> > +
> > +     if (cfg.c_dry_run)
> > +             return 0;
> > +
> > +     if (!buf) {
> > +             erofs_err("buf is NULL");
> > +             return -EINVAL;
> > +     }
> > +     if (offset >= erofs_devsz || len > erofs_devsz ||
> > +         offset > erofs_devsz - len) {
> > +             erofs_err("read posion[%" PRIu64 ", %zd] is too large
> beyond the end of device(%" PRIu64 ").",
> > +                       offset, len, erofs_devsz);
> > +             return -EINVAL;
> > +     }
> > +
> > +     ret = pread64(erofs_devfd, buf, len, (off64_t)offset);
> > +     if (ret != (int)len) {
> > +             erofs_err("Failed to read data from device - %s:[%" PRIu64
> ", %zd].",
> > +                       erofs_devname, offset, len);
> > +             return -errno;
> > +     }
> > +     return 0;
> > +}
> > +
> >  int dev_write(const void *buf, u64 offset, size_t len)
> >  {
> >       int ret;
> > diff --git a/mkfs/main.c b/mkfs/main.c
> > index f127fe1..109486e 100644
> > --- a/mkfs/main.c
> > +++ b/mkfs/main.c
> > @@ -182,6 +182,41 @@ int erofs_mkfs_update_super_block(struct
> erofs_buffer_head *bh,
> >       return 0;
> >  }
> >
> > +void erofs_dump_super(char *img_path)
> > +{
> > +     struct erofs_super_block *sb;
> > +     char buf[EROFS_BLKSIZ];
> > +     unsigned int blksz;
> > +     int ret = 0;
> > +
> > +     if (img_path == NULL) {
> > +             erofs_err("image path cannot be null");
> > +             return;
> > +     }
> > +     ret = blk_read(buf, 0, 1);
> > +     if (ret) {
> > +             erofs_err("error reading super-block structure");
> > +             return;
> > +     }
> > +
> > +     sb = (struct erofs_super_block *)((u8 *)buf + EROFS_SUPER_OFFSET);
> > +     if (le32_to_cpu(sb->magic) != EROFS_SUPER_MAGIC_V1) {
> > +             erofs_err("not a erofs image");
> > +             return;
> > +     }
> > +
> > +     erofs_dump("magic: 0x%x\n", le32_to_cpu(sb->magic));
> > +     blksz = 1 << sb->blkszbits;
> > +     erofs_dump("block size: %d\n", blksz);
> > +     erofs_dump("root inode: %d\n", le32_to_cpu(sb->root_nid));
> > +     erofs_dump("inodes: %llu\n", le64_to_cpu(sb->inos));
> > +     erofs_dump("build time: %u\n", le32_to_cpu(sb->build_time));
> > +     erofs_dump("blocks: %u\n", le32_to_cpu(sb->blocks));
> > +     erofs_dump("meta block: %u\n", le32_to_cpu(sb->meta_blkaddr));
> > +     erofs_dump("xattr block: %u\n", le32_to_cpu(sb->xattr_blkaddr));
> > +     erofs_dump("requirements: 0x%x\n", le32_to_cpu(sb->requirements));
> > +}
> > +
> >  int main(int argc, char **argv)
> >  {
> >       int err = 0;
> > @@ -268,6 +303,7 @@ int main(int argc, char **argv)
> >               err = -EIO;
> >  exit:
> >       z_erofs_compress_exit();
> > +     erofs_dump_super("dummy");
> >       dev_close();
> >       erofs_exit_configure();
> >
> > --
> > 2.9.3
> >
>

[-- Attachment #2: Type: text/html, Size: 8786 bytes --]

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

* Re: [PATCH] erofs-utils: erofs debug utility.
  2019-08-21 18:15   ` Pratik Shinde
@ 2019-08-21 20:08     ` Gao Xiang via Linux-erofs
  2019-08-22  0:09       ` Gao Xiang
  0 siblings, 1 reply; 5+ messages in thread
From: Gao Xiang via Linux-erofs @ 2019-08-21 20:08 UTC (permalink / raw)
  To: Pratik Shinde; +Cc: miaoxie, linux-erofs

Hi Pratik,

On Wed, Aug 21, 2019 at 11:45:48PM +0530, Pratik Shinde wrote:
> Thanks Gao,
> 
> Let me know when we can work on debug utility OR atleast list out things we
> want to achieve through the utility.

I think we need to implement
 - dump super block
 - dump a specfic file
 - dump a physical block or a physical compressed cluster from nid and logical offset
 - dump a specfic metadata (I mean dump the inode meta of a file)
 - ...

Thanks,
Gao Xiang

> Regarding the superblock checksum calculations, Yes I will dedicate
> sometime this week for it. will start exploring it.
> 
> --Pratik.
> 
> On Wed, Aug 21, 2019 at 10:26 PM Gao Xiang <hsiangkao@aol.com> wrote:
> 
> > Hi Pratik,
> >
> > On Wed, Aug 21, 2019 at 10:08:08PM +0530, Pratik Shinde wrote:
> > > Hello Maintainers,
> > >
> > > After going through the recent mail thread between linux's filesystem
> > folks
> > > on erofs channel, I felt erofs needs an interactive debug utility (like
> > xfs_db)
> > > which can be used to examine erofs images & can also be used to inject
> > errors OR
> > > fuzzing for testing purpose & dumping different erofs meta data
> > structures
> > > for debugging.
> > > In order to demonstrate above I wrote an experimental patch that simply
> > dumps
> > > the superblock of an image after mkfs completes.the full fletch utility
> > will run
> > > independently and be able to seek / print / modify any byte of an erofs
> > image,
> > > dump structures/lists/directory content of an image.
> >
> > Yes, I think we really need that interactive tools, actually I'm stuggling
> > in
> > modifing Guifu's erofs-fuse now, we need to add the parsing ability to
> > "lib/"
> > first.
> >
> > I mean, first, I will add a "fuse" field to "cfg". If it is false, it will
> > generate a image, or it will parse a image...
> > And then we need to add parsing logic into "lib/" as well, and use
> > "if (cfg.fuse)" to differnate whether it should read or write data.
> >
> > That is my prelimitary thought.. I will work on this framework in this
> > weekend.
> > and then we can work together on it. :)
> >
> > p.s. Pratik, if you have some time, could you take some extra time adding
> > the
> > super checksum calulation to EROFS? I mean we can add
> > EROFS_FEATURE_SB_CHKSUM
> > to the compat superblock field ("features"), and do crc32_le on kernel and
> > mkfs...
> > If you dont have time, I will do it later instead... (since we are using
> > EROFS
> > on the top of dm-verity, but completing the superblock chksum is also a
> > good idea.)
> >
> > And then we can add block-based verification layer to EROFS, it can be seen
> > as a hash tree like dm-verity or just simply CRC32 arrays for user to
> > choise.
> >
> > Thanks,
> > Gao Xiang
> >
> > >
> > > NOTE:This is an experimental patch just to demonstrate the purpose. The
> > patch
> > > lacks a lot of things like coding standard, and new code runs in the
> > context
> > > of mkfs itself.kindly ignore it.
> > >
> > > kindly provide your feedback on this.
> > >
> > > Signed-off-by: Pratik Shinde <pratikshinde320@gmail.com>
> > > ---
> > >  include/erofs/io.h |  8 ++++++++
> > >  lib/io.c           | 27 +++++++++++++++++++++++++++
> > >  mkfs/main.c        | 36 ++++++++++++++++++++++++++++++++++++
> > >  3 files changed, 71 insertions(+)
> > >
> > > diff --git a/include/erofs/io.h b/include/erofs/io.h
> > > index 4b574bd..e91d6ee 100644
> > > --- a/include/erofs/io.h
> > > +++ b/include/erofs/io.h
> > > @@ -18,6 +18,7 @@
> > >
> > >  int dev_open(const char *devname);
> > >  void dev_close(void);
> > > +int dev_read(void *buf, u64 offset, size_t len);
> > >  int dev_write(const void *buf, u64 offset, size_t len);
> > >  int dev_fillzero(u64 offset, size_t len, bool padding);
> > >  int dev_fsync(void);
> > > @@ -30,5 +31,12 @@ static inline int blk_write(const void *buf,
> > erofs_blk_t blkaddr,
> > >                        blknr_to_addr(nblocks));
> > >  }
> > >
> > > +static inline int blk_read(void *buf, erofs_blk_t blkaddr,
> > > +                        u32 nblocks)
> > > +{
> > > +     return dev_read(buf, blknr_to_addr(blkaddr),
> > > +                     blknr_to_addr(nblocks));
> > > +}
> > > +
> > >  #endif
> > >
> > > diff --git a/lib/io.c b/lib/io.c
> > > index 15c5a35..87d7d6c 100644
> > > --- a/lib/io.c
> > > +++ b/lib/io.c
> > > @@ -109,6 +109,33 @@ u64 dev_length(void)
> > >       return erofs_devsz;
> > >  }
> > >
> > > +int dev_read(void *buf, u64 offset, size_t len)
> > > +{
> > > +     int ret;
> > > +
> > > +     if (cfg.c_dry_run)
> > > +             return 0;
> > > +
> > > +     if (!buf) {
> > > +             erofs_err("buf is NULL");
> > > +             return -EINVAL;
> > > +     }
> > > +     if (offset >= erofs_devsz || len > erofs_devsz ||
> > > +         offset > erofs_devsz - len) {
> > > +             erofs_err("read posion[%" PRIu64 ", %zd] is too large
> > beyond the end of device(%" PRIu64 ").",
> > > +                       offset, len, erofs_devsz);
> > > +             return -EINVAL;
> > > +     }
> > > +
> > > +     ret = pread64(erofs_devfd, buf, len, (off64_t)offset);
> > > +     if (ret != (int)len) {
> > > +             erofs_err("Failed to read data from device - %s:[%" PRIu64
> > ", %zd].",
> > > +                       erofs_devname, offset, len);
> > > +             return -errno;
> > > +     }
> > > +     return 0;
> > > +}
> > > +
> > >  int dev_write(const void *buf, u64 offset, size_t len)
> > >  {
> > >       int ret;
> > > diff --git a/mkfs/main.c b/mkfs/main.c
> > > index f127fe1..109486e 100644
> > > --- a/mkfs/main.c
> > > +++ b/mkfs/main.c
> > > @@ -182,6 +182,41 @@ int erofs_mkfs_update_super_block(struct
> > erofs_buffer_head *bh,
> > >       return 0;
> > >  }
> > >
> > > +void erofs_dump_super(char *img_path)
> > > +{
> > > +     struct erofs_super_block *sb;
> > > +     char buf[EROFS_BLKSIZ];
> > > +     unsigned int blksz;
> > > +     int ret = 0;
> > > +
> > > +     if (img_path == NULL) {
> > > +             erofs_err("image path cannot be null");
> > > +             return;
> > > +     }
> > > +     ret = blk_read(buf, 0, 1);
> > > +     if (ret) {
> > > +             erofs_err("error reading super-block structure");
> > > +             return;
> > > +     }
> > > +
> > > +     sb = (struct erofs_super_block *)((u8 *)buf + EROFS_SUPER_OFFSET);
> > > +     if (le32_to_cpu(sb->magic) != EROFS_SUPER_MAGIC_V1) {
> > > +             erofs_err("not a erofs image");
> > > +             return;
> > > +     }
> > > +
> > > +     erofs_dump("magic: 0x%x\n", le32_to_cpu(sb->magic));
> > > +     blksz = 1 << sb->blkszbits;
> > > +     erofs_dump("block size: %d\n", blksz);
> > > +     erofs_dump("root inode: %d\n", le32_to_cpu(sb->root_nid));
> > > +     erofs_dump("inodes: %llu\n", le64_to_cpu(sb->inos));
> > > +     erofs_dump("build time: %u\n", le32_to_cpu(sb->build_time));
> > > +     erofs_dump("blocks: %u\n", le32_to_cpu(sb->blocks));
> > > +     erofs_dump("meta block: %u\n", le32_to_cpu(sb->meta_blkaddr));
> > > +     erofs_dump("xattr block: %u\n", le32_to_cpu(sb->xattr_blkaddr));
> > > +     erofs_dump("requirements: 0x%x\n", le32_to_cpu(sb->requirements));
> > > +}
> > > +
> > >  int main(int argc, char **argv)
> > >  {
> > >       int err = 0;
> > > @@ -268,6 +303,7 @@ int main(int argc, char **argv)
> > >               err = -EIO;
> > >  exit:
> > >       z_erofs_compress_exit();
> > > +     erofs_dump_super("dummy");
> > >       dev_close();
> > >       erofs_exit_configure();
> > >
> > > --
> > > 2.9.3
> > >
> >

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

* Re: [PATCH] erofs-utils: erofs debug utility.
  2019-08-21 20:08     ` Gao Xiang via Linux-erofs
@ 2019-08-22  0:09       ` Gao Xiang
  0 siblings, 0 replies; 5+ messages in thread
From: Gao Xiang @ 2019-08-22  0:09 UTC (permalink / raw)
  To: Pratik Shinde; +Cc: linux-erofs, miaoxie

On Thu, Aug 22, 2019 at 04:08:24AM +0800, Gao Xiang via Linux-erofs wrote:
> Hi Pratik,
> 
> On Wed, Aug 21, 2019 at 11:45:48PM +0530, Pratik Shinde wrote:
> > Thanks Gao,
> > 
> > Let me know when we can work on debug utility OR atleast list out things we
> > want to achieve through the utility.

Sorry, it seems I misunderstand that, I think I will do a prelimitary parsing
version this weekend, but you can develop it now :) and then we can merge
our code later, that's also fine of course.

Thanks,
Gao Xiang

> 
> I think we need to implement
>  - dump super block
>  - dump a specfic file
>  - dump a physical block or a physical compressed cluster from nid and logical offset
>  - dump a specfic metadata (I mean dump the inode meta of a file)
>  - ...
> 
> Thanks,
> Gao Xiang
> 
> > Regarding the superblock checksum calculations, Yes I will dedicate
> > sometime this week for it. will start exploring it.
> > 
> > --Pratik.
> > 
> > On Wed, Aug 21, 2019 at 10:26 PM Gao Xiang <hsiangkao@aol.com> wrote:
> > 
> > > Hi Pratik,
> > >
> > > On Wed, Aug 21, 2019 at 10:08:08PM +0530, Pratik Shinde wrote:
> > > > Hello Maintainers,
> > > >
> > > > After going through the recent mail thread between linux's filesystem
> > > folks
> > > > on erofs channel, I felt erofs needs an interactive debug utility (like
> > > xfs_db)
> > > > which can be used to examine erofs images & can also be used to inject
> > > errors OR
> > > > fuzzing for testing purpose & dumping different erofs meta data
> > > structures
> > > > for debugging.
> > > > In order to demonstrate above I wrote an experimental patch that simply
> > > dumps
> > > > the superblock of an image after mkfs completes.the full fletch utility
> > > will run
> > > > independently and be able to seek / print / modify any byte of an erofs
> > > image,
> > > > dump structures/lists/directory content of an image.
> > >
> > > Yes, I think we really need that interactive tools, actually I'm stuggling
> > > in
> > > modifing Guifu's erofs-fuse now, we need to add the parsing ability to
> > > "lib/"
> > > first.
> > >
> > > I mean, first, I will add a "fuse" field to "cfg". If it is false, it will
> > > generate a image, or it will parse a image...
> > > And then we need to add parsing logic into "lib/" as well, and use
> > > "if (cfg.fuse)" to differnate whether it should read or write data.
> > >
> > > That is my prelimitary thought.. I will work on this framework in this
> > > weekend.
> > > and then we can work together on it. :)
> > >
> > > p.s. Pratik, if you have some time, could you take some extra time adding
> > > the
> > > super checksum calulation to EROFS? I mean we can add
> > > EROFS_FEATURE_SB_CHKSUM
> > > to the compat superblock field ("features"), and do crc32_le on kernel and
> > > mkfs...
> > > If you dont have time, I will do it later instead... (since we are using
> > > EROFS
> > > on the top of dm-verity, but completing the superblock chksum is also a
> > > good idea.)
> > >
> > > And then we can add block-based verification layer to EROFS, it can be seen
> > > as a hash tree like dm-verity or just simply CRC32 arrays for user to
> > > choise.
> > >
> > > Thanks,
> > > Gao Xiang
> > >
> > > >
> > > > NOTE:This is an experimental patch just to demonstrate the purpose. The
> > > patch
> > > > lacks a lot of things like coding standard, and new code runs in the
> > > context
> > > > of mkfs itself.kindly ignore it.
> > > >
> > > > kindly provide your feedback on this.
> > > >
> > > > Signed-off-by: Pratik Shinde <pratikshinde320@gmail.com>
> > > > ---
> > > >  include/erofs/io.h |  8 ++++++++
> > > >  lib/io.c           | 27 +++++++++++++++++++++++++++
> > > >  mkfs/main.c        | 36 ++++++++++++++++++++++++++++++++++++
> > > >  3 files changed, 71 insertions(+)
> > > >
> > > > diff --git a/include/erofs/io.h b/include/erofs/io.h
> > > > index 4b574bd..e91d6ee 100644
> > > > --- a/include/erofs/io.h
> > > > +++ b/include/erofs/io.h
> > > > @@ -18,6 +18,7 @@
> > > >
> > > >  int dev_open(const char *devname);
> > > >  void dev_close(void);
> > > > +int dev_read(void *buf, u64 offset, size_t len);
> > > >  int dev_write(const void *buf, u64 offset, size_t len);
> > > >  int dev_fillzero(u64 offset, size_t len, bool padding);
> > > >  int dev_fsync(void);
> > > > @@ -30,5 +31,12 @@ static inline int blk_write(const void *buf,
> > > erofs_blk_t blkaddr,
> > > >                        blknr_to_addr(nblocks));
> > > >  }
> > > >
> > > > +static inline int blk_read(void *buf, erofs_blk_t blkaddr,
> > > > +                        u32 nblocks)
> > > > +{
> > > > +     return dev_read(buf, blknr_to_addr(blkaddr),
> > > > +                     blknr_to_addr(nblocks));
> > > > +}
> > > > +
> > > >  #endif
> > > >
> > > > diff --git a/lib/io.c b/lib/io.c
> > > > index 15c5a35..87d7d6c 100644
> > > > --- a/lib/io.c
> > > > +++ b/lib/io.c
> > > > @@ -109,6 +109,33 @@ u64 dev_length(void)
> > > >       return erofs_devsz;
> > > >  }
> > > >
> > > > +int dev_read(void *buf, u64 offset, size_t len)
> > > > +{
> > > > +     int ret;
> > > > +
> > > > +     if (cfg.c_dry_run)
> > > > +             return 0;
> > > > +
> > > > +     if (!buf) {
> > > > +             erofs_err("buf is NULL");
> > > > +             return -EINVAL;
> > > > +     }
> > > > +     if (offset >= erofs_devsz || len > erofs_devsz ||
> > > > +         offset > erofs_devsz - len) {
> > > > +             erofs_err("read posion[%" PRIu64 ", %zd] is too large
> > > beyond the end of device(%" PRIu64 ").",
> > > > +                       offset, len, erofs_devsz);
> > > > +             return -EINVAL;
> > > > +     }
> > > > +
> > > > +     ret = pread64(erofs_devfd, buf, len, (off64_t)offset);
> > > > +     if (ret != (int)len) {
> > > > +             erofs_err("Failed to read data from device - %s:[%" PRIu64
> > > ", %zd].",
> > > > +                       erofs_devname, offset, len);
> > > > +             return -errno;
> > > > +     }
> > > > +     return 0;
> > > > +}
> > > > +
> > > >  int dev_write(const void *buf, u64 offset, size_t len)
> > > >  {
> > > >       int ret;
> > > > diff --git a/mkfs/main.c b/mkfs/main.c
> > > > index f127fe1..109486e 100644
> > > > --- a/mkfs/main.c
> > > > +++ b/mkfs/main.c
> > > > @@ -182,6 +182,41 @@ int erofs_mkfs_update_super_block(struct
> > > erofs_buffer_head *bh,
> > > >       return 0;
> > > >  }
> > > >
> > > > +void erofs_dump_super(char *img_path)
> > > > +{
> > > > +     struct erofs_super_block *sb;
> > > > +     char buf[EROFS_BLKSIZ];
> > > > +     unsigned int blksz;
> > > > +     int ret = 0;
> > > > +
> > > > +     if (img_path == NULL) {
> > > > +             erofs_err("image path cannot be null");
> > > > +             return;
> > > > +     }
> > > > +     ret = blk_read(buf, 0, 1);
> > > > +     if (ret) {
> > > > +             erofs_err("error reading super-block structure");
> > > > +             return;
> > > > +     }
> > > > +
> > > > +     sb = (struct erofs_super_block *)((u8 *)buf + EROFS_SUPER_OFFSET);
> > > > +     if (le32_to_cpu(sb->magic) != EROFS_SUPER_MAGIC_V1) {
> > > > +             erofs_err("not a erofs image");
> > > > +             return;
> > > > +     }
> > > > +
> > > > +     erofs_dump("magic: 0x%x\n", le32_to_cpu(sb->magic));
> > > > +     blksz = 1 << sb->blkszbits;
> > > > +     erofs_dump("block size: %d\n", blksz);
> > > > +     erofs_dump("root inode: %d\n", le32_to_cpu(sb->root_nid));
> > > > +     erofs_dump("inodes: %llu\n", le64_to_cpu(sb->inos));
> > > > +     erofs_dump("build time: %u\n", le32_to_cpu(sb->build_time));
> > > > +     erofs_dump("blocks: %u\n", le32_to_cpu(sb->blocks));
> > > > +     erofs_dump("meta block: %u\n", le32_to_cpu(sb->meta_blkaddr));
> > > > +     erofs_dump("xattr block: %u\n", le32_to_cpu(sb->xattr_blkaddr));
> > > > +     erofs_dump("requirements: 0x%x\n", le32_to_cpu(sb->requirements));
> > > > +}
> > > > +
> > > >  int main(int argc, char **argv)
> > > >  {
> > > >       int err = 0;
> > > > @@ -268,6 +303,7 @@ int main(int argc, char **argv)
> > > >               err = -EIO;
> > > >  exit:
> > > >       z_erofs_compress_exit();
> > > > +     erofs_dump_super("dummy");
> > > >       dev_close();
> > > >       erofs_exit_configure();
> > > >
> > > > --
> > > > 2.9.3
> > > >
> > >

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

end of thread, other threads:[~2019-08-22  0:10 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-08-21 16:38 [PATCH] erofs-utils: erofs debug utility Pratik Shinde
2019-08-21 16:56 ` Gao Xiang via Linux-erofs
2019-08-21 18:15   ` Pratik Shinde
2019-08-21 20:08     ` Gao Xiang via Linux-erofs
2019-08-22  0:09       ` Gao Xiang

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.