[PATCH v2] mke2fs: warn about missing y2038 support when formatting fresh ext4 fs

[PATCH v2] mke2fs: warn about missing y2038 support when formatting fresh ext4 fs

[Darrick J. Wong]

From: Darrick J. Wong <djwong@kernel.org>

Filesystems with 128-byte inodes do not support timestamps beyond the
year 2038.  Since we're now less than 16.5 years away from that point,
it's time to start warning users about this lack of support when they
format an ext4 filesystem with small inodes.

First, change the mke2fs.conf file to specify 256-byte inodes even for
small filesystems, then add a warning to mke2fs itself if someone is
trying to make us format an ext4 filesystem with 128-byte inodes.

Note that we /don't/ warn about these things if the user has signalled
that they want an old format such as ext2, ext3, or hurd.  Everyone
should know by now that those are legacy.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
v2: fix the comments
---
 misc/mke2fs.c       |   39 +++++++++++++++++++++++++++++++++++++++
 misc/mke2fs.conf.in |    4 ++--
 2 files changed, 41 insertions(+), 2 deletions(-)

diff --git a/misc/mke2fs.c b/misc/mke2fs.c
index 92003e11..114a64f7 100644
--- a/misc/mke2fs.c
+++ b/misc/mke2fs.c
@@ -1537,6 +1537,35 @@ static int get_device_geometry(const char *file,
 }
 #endif
 
+/*
+ * Decide if the user is formatting with an old feature set (e.g. ext2, ext3).
+ *
+ * If there is no fs_types list, assume that the user's getting ext4 and return
+ * 0.  If we find 'ext4' anywhere in the fs_types list, take that as a sign
+ * that the user will get ext4 and return 0.  Any other case returns 1.
+ *
+ * Normally, 'ext4' will be the first item in fs_types, but the user can
+ * combine argv[0], -t, and -T options in such a way that fs_types will start
+ * with some other word and the 'ext4' will end up in a non-zero slot.  A
+ * simple way to do this is "mke2fs -T ext4 /dev/XXX".  The user is supposed to
+ * use -t for the fs type and not -T, but we've never enforced that.
+ */
+static inline int
+old_format_forced(char **fs_types)
+{
+	int found_ext4 = 0;
+	int i;
+
+	if (!fs_types)
+		return 0;
+
+	for (i = 0; fs_types[i]; i++)
+		if (!strcmp(fs_types[i], "ext4"))
+			found_ext4 = 1;
+
+	return !found_ext4;
+}
+
 static void PRS(int argc, char *argv[])
 {
 	int		b, c, flags;
@@ -2603,6 +2632,16 @@ static void PRS(int argc, char *argv[])
 		exit(1);
 	}
 
+	/*
+	 * If we're formatting with an ext4 feature set (and not an old ondisk
+	 * format), warn the user that filesystems with 128-byte inodes will
+	 * not work properly beyond 2038.
+	 */
+	if (!old_format_forced(fs_types) &&
+	    inode_size == EXT2_GOOD_OLD_INODE_SIZE)
+		printf(
+_("128-byte inodes cannot handle dates beyond 2038 and are deprecated\n"));
+
 	/* Make sure number of inodes specified will fit in 32 bits */
 	if (num_inodes == 0) {
 		unsigned long long n;
diff --git a/misc/mke2fs.conf.in b/misc/mke2fs.conf.in
index 01e35cf8..2fa1a824 100644
--- a/misc/mke2fs.conf.in
+++ b/misc/mke2fs.conf.in
@@ -16,12 +16,12 @@
 	}
 	small = {
 		blocksize = 1024
-		inode_size = 128
+		inode_size = 256
 		inode_ratio = 4096
 	}
 	floppy = {
 		blocksize = 1024
-		inode_size = 128
+		inode_size = 256
 		inode_ratio = 8192
 	}
 	big = {

Re: [PATCH v2] mke2fs: warn about missing y2038 support when formatting fresh ext4 fs

[Theodore Ts'o]

On Thu, Aug 12, 2021 at 04:22:22PM -0700, Darrick J. Wong wrote:
> Note that we /don't/ warn about these things if the user has signalled
> that they want an old format such as ext2, ext3, or hurd.  Everyone
> should know by now that those are legacy.

So I took a closer look, and it turns out that we changed the default
inode size for ext2 and ext3 file systems in 2008 (see commit
b1631cce648e ("Create new filesystems with 256-byte inodes by
default"), in e2fsprogs 1.40.4.  Even a positively antedeluvian distro
such as RHEL 7 uses e2fsprogs 1.42.9, and RHEL 6 EOL'ed November 30th,
2020.

There were only two cases where we created file systems with 128 byte
inodes --- "small" and "floppy" sized file systems, and for the GNU
Hurd, which only supports the original 128 byte inode.  What will GNU
Hurd do in 16.5 years?  ¯\_(ツ)_/¯

Given that, I think we can simplify the patch a little, and just use a
mke2fs.conf boolean to disable the warning message.

What do folks think?

						- Ted



commit f569fe86caddf964973ed35cdf36ed520ef23a0c
Author: Darrick J. Wong <djwong@kernel.org>
Date:   Thu Aug 12 16:22:22 2021 -0700

    mke2fs: warn about missing y2038 support when formatting fresh ext4 fs
    
    Filesystems with 128-byte inodes do not support timestamps beyond the
    year 2038.  Since we're now less than 16.5 years away from that point,
    it's time to start warning users about this lack of support when they
    format an ext4 filesystem with small inodes.
    
    (Note that even for ext2 and ext3, we changed the default for
    non-small file systems in 2008 in commit commit b1631cce648e ("Create
    new filesystems with 256-byte inodes by default").)
    
    So change the mke2fs.conf file to specify 256-byte inodes even for
    small filesystems, and then add a warning to mke2fs itself if someone
    is trying to make us format a file system with 128-byte inodes.  This
    can be suppressed by setting the boolean option warn_y2038_dates in
    the mke2fs.conf file to false, which we do in the case of GNU Hurd,
    since it only supports 128 byte inodes as of this writing.
    
    [ Patch reworked by tytso to only warn in the case of GNU Hurd, since
      the default for ext2/ext3 was changed for all but small file systems
      in 2008. ]
    
    Signed-off-by: Darrick J. Wong <djwong@kernel.org>
    Signed-off-by: Theodore Ts'o <tytso@mit.edu>

diff --git a/misc/mke2fs.c b/misc/mke2fs.c
index 92003e11..881ffd31 100644
--- a/misc/mke2fs.c
+++ b/misc/mke2fs.c
@@ -2603,6 +2603,17 @@ profile_error:
 		exit(1);
 	}
 
+	/*
+	 * Warn the user that filesystems with 128-byte inodes will
+	 * not work properly beyond 2038.  This can be suppressed via
+	 * a boolean in the mke2fs.conf file, and we will disable this
+	 * warning for ext2, ext3, and hurd file systems.
+	 */
+	if (inode_size == EXT2_GOOD_OLD_INODE_SIZE &&
+	    get_bool_from_profile(fs_types, "warn_y2038_dates", 1))
+		printf(
+_("128-byte inodes cannot handle dates beyond 2038 and are deprecated\n"));
+
 	/* Make sure number of inodes specified will fit in 32 bits */
 	if (num_inodes == 0) {
 		unsigned long long n;
diff --git a/misc/mke2fs.conf.5.in b/misc/mke2fs.conf.5.in
index 08bb9488..62d0fdb5 100644
--- a/misc/mke2fs.conf.5.in
+++ b/misc/mke2fs.conf.5.in
@@ -505,6 +505,13 @@ This relation specifies the base file name for the huge files.
 This relation specifies the (zero-padded) width of the field for the
 huge file number.
 .TP
+.I warn_y2038_dates
+This boolean relation specifies wheather mke2fs will issue a warning
+when creating a file system with 128 byte inodes (and so therefore will
+not support dates after January 19th, 2038.  The default value is true,
+except for file systems created for the GNU Hurd, which does not support
+inodes larger than 128 bytes.
+.TP
 .I zero_hugefiles
 This boolean relation specifies whether or not zero blocks will be
 written to the hugefiles while
diff --git a/misc/mke2fs.conf.in b/misc/mke2fs.conf.in
index 01e35cf8..05680992 100644
--- a/misc/mke2fs.conf.in
+++ b/misc/mke2fs.conf.in
@@ -12,16 +12,13 @@
 	}
 	ext4 = {
 		features = has_journal,extent,huge_file,flex_bg,metadata_csum,64bit,dir_nlink,extra_isize
-		inode_size = 256
 	}
 	small = {
 		blocksize = 1024
-		inode_size = 128
 		inode_ratio = 4096
 	}
 	floppy = {
 		blocksize = 1024
-		inode_size = 128
 		inode_ratio = 8192
 	}
 	big = {
@@ -44,4 +41,5 @@
 	hurd = {
 	     blocksize = 4096
 	     inode_size = 128
+	     warn_y2038_dates = 0
 	}

Re: [PATCH v2] mke2fs: warn about missing y2038 support when formatting fresh ext4 fs

[Darrick J. Wong]

On Fri, Aug 13, 2021 at 01:52:00PM -0400, Theodore Ts'o wrote:
> On Thu, Aug 12, 2021 at 04:22:22PM -0700, Darrick J. Wong wrote:
> > Note that we /don't/ warn about these things if the user has signalled
> > that they want an old format such as ext2, ext3, or hurd.  Everyone
> > should know by now that those are legacy.
> 
> So I took a closer look, and it turns out that we changed the default
> inode size for ext2 and ext3 file systems in 2008 (see commit
> b1631cce648e ("Create new filesystems with 256-byte inodes by
> default"), in e2fsprogs 1.40.4.  Even a positively antedeluvian distro
> such as RHEL 7 uses e2fsprogs 1.42.9, and RHEL 6 EOL'ed November 30th,
> 2020.
> 
> There were only two cases where we created file systems with 128 byte
> inodes --- "small" and "floppy" sized file systems, and for the GNU
> Hurd, which only supports the original 128 byte inode.  What will GNU
> Hurd do in 16.5 years?  ¯\_(ツ)_/¯

Perhaps in that time someone can donate a disused Opteron 140 system?
Assuming the motherboard capacitors haven't since lost their mojo.

> Given that, I think we can simplify the patch a little, and just use a
> mke2fs.conf boolean to disable the warning message.
> 
> What do folks think?
> 
> 						- Ted
> 
> 
> 
> commit f569fe86caddf964973ed35cdf36ed520ef23a0c
> Author: Darrick J. Wong <djwong@kernel.org>
> Date:   Thu Aug 12 16:22:22 2021 -0700
> 
>     mke2fs: warn about missing y2038 support when formatting fresh ext4 fs
>     
>     Filesystems with 128-byte inodes do not support timestamps beyond the
>     year 2038.  Since we're now less than 16.5 years away from that point,
>     it's time to start warning users about this lack of support when they
>     format an ext4 filesystem with small inodes.
>     
>     (Note that even for ext2 and ext3, we changed the default for
>     non-small file systems in 2008 in commit commit b1631cce648e ("Create
>     new filesystems with 256-byte inodes by default").)
>     
>     So change the mke2fs.conf file to specify 256-byte inodes even for
>     small filesystems, and then add a warning to mke2fs itself if someone
>     is trying to make us format a file system with 128-byte inodes.  This
>     can be suppressed by setting the boolean option warn_y2038_dates in
>     the mke2fs.conf file to false, which we do in the case of GNU Hurd,
>     since it only supports 128 byte inodes as of this writing.
>     
>     [ Patch reworked by tytso to only warn in the case of GNU Hurd, since
>       the default for ext2/ext3 was changed for all but small file systems
>       in 2008. ]

Even better!

>     
>     Signed-off-by: Darrick J. Wong <djwong@kernel.org>
>     Signed-off-by: Theodore Ts'o <tytso@mit.edu>
> 
> diff --git a/misc/mke2fs.c b/misc/mke2fs.c
> index 92003e11..881ffd31 100644
> --- a/misc/mke2fs.c
> +++ b/misc/mke2fs.c
> @@ -2603,6 +2603,17 @@ profile_error:
>  		exit(1);
>  	}
>  
> +	/*
> +	 * Warn the user that filesystems with 128-byte inodes will
> +	 * not work properly beyond 2038.  This can be suppressed via
> +	 * a boolean in the mke2fs.conf file, and we will disable this
> +	 * warning for ext2, ext3, and hurd file systems.

Um... the conffile changes only disable the warning for Hurd?

> +	 */
> +	if (inode_size == EXT2_GOOD_OLD_INODE_SIZE &&
> +	    get_bool_from_profile(fs_types, "warn_y2038_dates", 1))
> +		printf(
> +_("128-byte inodes cannot handle dates beyond 2038 and are deprecated\n"));
> +
>  	/* Make sure number of inodes specified will fit in 32 bits */
>  	if (num_inodes == 0) {
>  		unsigned long long n;
> diff --git a/misc/mke2fs.conf.5.in b/misc/mke2fs.conf.5.in
> index 08bb9488..62d0fdb5 100644
> --- a/misc/mke2fs.conf.5.in
> +++ b/misc/mke2fs.conf.5.in
> @@ -505,6 +505,13 @@ This relation specifies the base file name for the huge files.
>  This relation specifies the (zero-padded) width of the field for the
>  huge file number.
>  .TP
> +.I warn_y2038_dates
> +This boolean relation specifies wheather mke2fs will issue a warning
> +when creating a file system with 128 byte inodes (and so therefore will
> +not support dates after January 19th, 2038.  The default value is true,

Nit: need a closing parentheses after '2038' or no opening paren.

> +except for file systems created for the GNU Hurd, which does not support
> +inodes larger than 128 bytes.

I wonder if this statementt about Hurd this belongs in the conffile as a
comment in the hurd section?

--D

> +.TP
>  .I zero_hugefiles
>  This boolean relation specifies whether or not zero blocks will be
>  written to the hugefiles while
> diff --git a/misc/mke2fs.conf.in b/misc/mke2fs.conf.in
> index 01e35cf8..05680992 100644
> --- a/misc/mke2fs.conf.in
> +++ b/misc/mke2fs.conf.in
> @@ -12,16 +12,13 @@
>  	}
>  	ext4 = {
>  		features = has_journal,extent,huge_file,flex_bg,metadata_csum,64bit,dir_nlink,extra_isize
> -		inode_size = 256
>  	}
>  	small = {
>  		blocksize = 1024
> -		inode_size = 128
>  		inode_ratio = 4096
>  	}
>  	floppy = {
>  		blocksize = 1024
> -		inode_size = 128
>  		inode_ratio = 8192
>  	}
>  	big = {
> @@ -44,4 +41,5 @@
>  	hurd = {
>  	     blocksize = 4096
>  	     inode_size = 128
> +	     warn_y2038_dates = 0
>  	}

Re: [PATCH v2] mke2fs: warn about missing y2038 support when formatting fresh ext4 fs

[Theodore Ts'o]

On Fri, Aug 13, 2021 at 11:14:36AM -0700, Darrick J. Wong wrote:
> > There were only two cases where we created file systems with 128 byte
> > inodes --- "small" and "floppy" sized file systems, and for the GNU
> > Hurd, which only supports the original 128 byte inode.  What will GNU
> > Hurd do in 16.5 years?  ¯\_(ツ)_/¯
> 
> Perhaps in that time someone can donate a disused Opteron 140 system?
> Assuming the motherboard capacitors haven't since lost their mojo.

Apparently GNU Hurd uses a unsigned 32-bit int for time_t, so they
have a 2106 problem.  They "have no plans for a 64-bit userspace, but
they have plans for a 64-bit kernel that can run 32-bit user space".
Comments from the the GNU Hurd folks on an IRC chat from 2013:

    <braunr> which overflows in 2106
    <braunr> and we already include funny comments that predict our successors,
      if any, will probably fail to deal with the problem until short before
      the overflow :>
    <azeem> luckily, no nuclear reactors are running the Hurd sofar

    https://www.gnu.org/software/hurd/open_issues/versioning.html

> > +	/*
> > +	 * Warn the user that filesystems with 128-byte inodes will
> > +	 * not work properly beyond 2038.  This can be suppressed via
> > +	 * a boolean in the mke2fs.conf file, and we will disable this
> > +	 * warning for ext2, ext3, and hurd file systems.
> 
> Um... the conffile changes only disable the warning for Hurd?

Oops, good catch, I'll fix up the comment.

> > +This boolean relation specifies wheather mke2fs will issue a warning
> > +when creating a file system with 128 byte inodes (and so therefore will
> > +not support dates after January 19th, 2038.  The default value is true,
> 
> Nit: need a closing parentheses after '2038' or no opening paren.

Thanks, fixed.

> > +except for file systems created for the GNU Hurd, which does not support
> > +inodes larger than 128 bytes.
> 
> I wonder if this statementt about Hurd this belongs in the conffile as a
> comment in the hurd section?

We currently don't have a Hurd section.  We probably could document
more about the magic that mke2fs does when you specify "-o hurd",
which probably should go in the mke2fs man page but I can't quite
bring myself to care.  Maybe some GNU Hurd folks can get interested to
do this?  :-)

						- Ted

[PATCH -v4] mke2fs: warn about missing y2038 support when formatting fresh ext4 fs

[Theodore Ts'o]

From: "Darrick J. Wong" <djwong@kernel.org>

Filesystems with 128-byte inodes do not support timestamps beyond the
year 2038.  Since we're now less than 16.5 years away from that point,
it's time to start warning users about this lack of support when they
format an ext4 filesystem with small inodes.

(Note that even for ext2 and ext3, we changed the default for
non-small file systems in 2008 in commit commit b1631cce648e ("Create
new filesystems with 256-byte inodes by default").)

So change the mke2fs.conf file to specify 256-byte inodes even for
small filesystems, and then add a warning to mke2fs itself if someone
is trying to make us format a file system with 128-byte inodes.  This
can be suppressed by setting the boolean option warn_y2038_dates in
the mke2fs.conf file to false, which we do in the case of GNU Hurd,
since it only supports 128 byte inodes as of this writing.

[ Patch reworked by tytso to only warn in the case of GNU Hurd, since
  the default for ext2/ext3 was changed for all but small file systems
  in 2008. ]

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 misc/mke2fs.c         | 11 +++++++++++
 misc/mke2fs.conf.5.in |  7 +++++++
 misc/mke2fs.conf.in   |  4 +---
 3 files changed, 19 insertions(+), 3 deletions(-)

diff --git a/misc/mke2fs.c b/misc/mke2fs.c
index 92003e11..881ffd31 100644
--- a/misc/mke2fs.c
+++ b/misc/mke2fs.c
@@ -2603,6 +2603,17 @@ profile_error:
 		exit(1);
 	}
 
+	/*
+	 * Warn the user that filesystems with 128-byte inodes will
+	 * not work properly beyond 2038.  This can be suppressed via
+	 * a boolean in the mke2fs.conf file, and we will disable this
+	 * warning for ext2, ext3, and hurd file systems.
+	 */
+	if (inode_size == EXT2_GOOD_OLD_INODE_SIZE &&
+	    get_bool_from_profile(fs_types, "warn_y2038_dates", 1))
+		printf(
+_("128-byte inodes cannot handle dates beyond 2038 and are deprecated\n"));
+
 	/* Make sure number of inodes specified will fit in 32 bits */
 	if (num_inodes == 0) {
 		unsigned long long n;
diff --git a/misc/mke2fs.conf.5.in b/misc/mke2fs.conf.5.in
index 08bb9488..62d0fdb5 100644
--- a/misc/mke2fs.conf.5.in
+++ b/misc/mke2fs.conf.5.in
@@ -505,6 +505,13 @@ This relation specifies the base file name for the huge files.
 This relation specifies the (zero-padded) width of the field for the
 huge file number.
 .TP
+.I warn_y2038_dates
+This boolean relation specifies wheather mke2fs will issue a warning
+when creating a file system with 128 byte inodes (and so therefore will
+not support dates after January 19th, 2038.  The default value is true,
+except for file systems created for the GNU Hurd, which does not support
+inodes larger than 128 bytes.
+.TP
 .I zero_hugefiles
 This boolean relation specifies whether or not zero blocks will be
 written to the hugefiles while
diff --git a/misc/mke2fs.conf.in b/misc/mke2fs.conf.in
index 01e35cf8..05680992 100644
--- a/misc/mke2fs.conf.in
+++ b/misc/mke2fs.conf.in
@@ -12,16 +12,13 @@
 	}
 	ext4 = {
 		features = has_journal,extent,huge_file,flex_bg,metadata_csum,64bit,dir_nlink,extra_isize
-		inode_size = 256
 	}
 	small = {
 		blocksize = 1024
-		inode_size = 128
 		inode_ratio = 4096
 	}
 	floppy = {
 		blocksize = 1024
-		inode_size = 128
 		inode_ratio = 8192
 	}
 	big = {
@@ -44,4 +41,5 @@
 	hurd = {
 	     blocksize = 4096
 	     inode_size = 128
+	     warn_y2038_dates = 0
 	}
-- 
2.31.0

[PATCH -v5] mke2fs: warn about missing y2038 support when formatting fresh ext4 fs

[Theodore Ts'o]

From: "Darrick J. Wong" <djwong@kernel.org>

Filesystems with 128-byte inodes do not support timestamps beyond the
year 2038.  Since we're now less than 16.5 years away from that point,
it's time to start warning users about this lack of support when they
format an ext4 filesystem with small inodes.

(Note that even for ext2 and ext3, we changed the default for
non-small file systems in 2008 in commit commit b1631cce648e ("Create
new filesystems with 256-byte inodes by default").)

So change the mke2fs.conf file to specify 256-byte inodes even for
small filesystems, and then add a warning to mke2fs itself if someone
is trying to make us format a file system with 128-byte inodes.  This
can be suppressed by setting the boolean option warn_y2038_dates in
the mke2fs.conf file to false, which we do in the case of GNU Hurd,
since it only supports 128 byte inodes as of this writing.

[ Patch reworked by tytso to only warn in the case of GNU Hurd, since
  the default for ext2/ext3 was changed for all but small file systems
  in 2008. ]

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 misc/mke2fs.c         | 11 +++++++++++
 misc/mke2fs.conf.5.in |  7 +++++++
 misc/mke2fs.conf.in   |  4 +---
 3 files changed, 19 insertions(+), 3 deletions(-)

diff --git a/misc/mke2fs.c b/misc/mke2fs.c
index 92003e11..a4573972 100644
--- a/misc/mke2fs.c
+++ b/misc/mke2fs.c
@@ -2603,6 +2603,17 @@ profile_error:
 		exit(1);
 	}
 
+	/*
+	 * Warn the user that filesystems with 128-byte inodes will
+	 * not work properly beyond 2038.  This can be suppressed via
+	 * a boolean in the mke2fs.conf file, and we will disable this
+	 * warning for file systems created for the GNU Hurd.
+	 */
+	if (inode_size == EXT2_GOOD_OLD_INODE_SIZE &&
+	    get_bool_from_profile(fs_types, "warn_y2038_dates", 1))
+		printf(
+_("128-byte inodes cannot handle dates beyond 2038 and are deprecated\n"));
+
 	/* Make sure number of inodes specified will fit in 32 bits */
 	if (num_inodes == 0) {
 		unsigned long long n;
diff --git a/misc/mke2fs.conf.5.in b/misc/mke2fs.conf.5.in
index 08bb9488..0b570303 100644
--- a/misc/mke2fs.conf.5.in
+++ b/misc/mke2fs.conf.5.in
@@ -505,6 +505,13 @@ This relation specifies the base file name for the huge files.
 This relation specifies the (zero-padded) width of the field for the
 huge file number.
 .TP
+.I warn_y2038_dates
+This boolean relation specifies wheather mke2fs will issue a warning
+when creating a file system with 128 byte inodes (and so therefore will
+not support dates after January 19th, 2038).  The default value is true,
+except for file systems created for the GNU Hurd since it only supports
+128-byte inodes.
+.TP
 .I zero_hugefiles
 This boolean relation specifies whether or not zero blocks will be
 written to the hugefiles while
diff --git a/misc/mke2fs.conf.in b/misc/mke2fs.conf.in
index 01e35cf8..05680992 100644
--- a/misc/mke2fs.conf.in
+++ b/misc/mke2fs.conf.in
@@ -12,16 +12,13 @@
 	}
 	ext4 = {
 		features = has_journal,extent,huge_file,flex_bg,metadata_csum,64bit,dir_nlink,extra_isize
-		inode_size = 256
 	}
 	small = {
 		blocksize = 1024
-		inode_size = 128
 		inode_ratio = 4096
 	}
 	floppy = {
 		blocksize = 1024
-		inode_size = 128
 		inode_ratio = 8192
 	}
 	big = {
@@ -44,4 +41,5 @@
 	hurd = {
 	     blocksize = 4096
 	     inode_size = 128
+	     warn_y2038_dates = 0
 	}
-- 
2.31.0

Re: [PATCH -v4] mke2fs: warn about missing y2038 support when formatting fresh ext4 fs

[Theodore Ts'o]

Oops, I sent the wrong commit; please ignore the -v4 version.

      	     	       	       - Ted

Re: [PATCH -v5] mke2fs: warn about missing y2038 support when formatting fresh ext4 fs

[Darrick J. Wong]

On Fri, Aug 13, 2021 at 03:35:18PM -0400, Theodore Ts'o wrote:
> From: "Darrick J. Wong" <djwong@kernel.org>
> 
> Filesystems with 128-byte inodes do not support timestamps beyond the
> year 2038.  Since we're now less than 16.5 years away from that point,
> it's time to start warning users about this lack of support when they
> format an ext4 filesystem with small inodes.
> 
> (Note that even for ext2 and ext3, we changed the default for
> non-small file systems in 2008 in commit commit b1631cce648e ("Create
> new filesystems with 256-byte inodes by default").)
> 
> So change the mke2fs.conf file to specify 256-byte inodes even for
> small filesystems, and then add a warning to mke2fs itself if someone
> is trying to make us format a file system with 128-byte inodes.  This
> can be suppressed by setting the boolean option warn_y2038_dates in
> the mke2fs.conf file to false, which we do in the case of GNU Hurd,
> since it only supports 128 byte inodes as of this writing.
> 
> [ Patch reworked by tytso to only warn in the case of GNU Hurd, since
>   the default for ext2/ext3 was changed for all but small file systems
>   in 2008. ]
> 
> Signed-off-by: Darrick J. Wong <djwong@kernel.org>
> Signed-off-by: Theodore Ts'o <tytso@mit.edu>

LGTM, insofar as I don't know how useful it is to review "my" own patch.
Reviewed-by: Darrick J. Wong <djwong@kernel.org>

--D

> ---
>  misc/mke2fs.c         | 11 +++++++++++
>  misc/mke2fs.conf.5.in |  7 +++++++
>  misc/mke2fs.conf.in   |  4 +---
>  3 files changed, 19 insertions(+), 3 deletions(-)
> 
> diff --git a/misc/mke2fs.c b/misc/mke2fs.c
> index 92003e11..a4573972 100644
> --- a/misc/mke2fs.c
> +++ b/misc/mke2fs.c
> @@ -2603,6 +2603,17 @@ profile_error:
>  		exit(1);
>  	}
>  
> +	/*
> +	 * Warn the user that filesystems with 128-byte inodes will
> +	 * not work properly beyond 2038.  This can be suppressed via
> +	 * a boolean in the mke2fs.conf file, and we will disable this
> +	 * warning for file systems created for the GNU Hurd.
> +	 */
> +	if (inode_size == EXT2_GOOD_OLD_INODE_SIZE &&
> +	    get_bool_from_profile(fs_types, "warn_y2038_dates", 1))
> +		printf(
> +_("128-byte inodes cannot handle dates beyond 2038 and are deprecated\n"));
> +
>  	/* Make sure number of inodes specified will fit in 32 bits */
>  	if (num_inodes == 0) {
>  		unsigned long long n;
> diff --git a/misc/mke2fs.conf.5.in b/misc/mke2fs.conf.5.in
> index 08bb9488..0b570303 100644
> --- a/misc/mke2fs.conf.5.in
> +++ b/misc/mke2fs.conf.5.in
> @@ -505,6 +505,13 @@ This relation specifies the base file name for the huge files.
>  This relation specifies the (zero-padded) width of the field for the
>  huge file number.
>  .TP
> +.I warn_y2038_dates
> +This boolean relation specifies wheather mke2fs will issue a warning
> +when creating a file system with 128 byte inodes (and so therefore will
> +not support dates after January 19th, 2038).  The default value is true,
> +except for file systems created for the GNU Hurd since it only supports
> +128-byte inodes.
> +.TP
>  .I zero_hugefiles
>  This boolean relation specifies whether or not zero blocks will be
>  written to the hugefiles while
> diff --git a/misc/mke2fs.conf.in b/misc/mke2fs.conf.in
> index 01e35cf8..05680992 100644
> --- a/misc/mke2fs.conf.in
> +++ b/misc/mke2fs.conf.in
> @@ -12,16 +12,13 @@
>  	}
>  	ext4 = {
>  		features = has_journal,extent,huge_file,flex_bg,metadata_csum,64bit,dir_nlink,extra_isize
> -		inode_size = 256
>  	}
>  	small = {
>  		blocksize = 1024
> -		inode_size = 128
>  		inode_ratio = 4096
>  	}
>  	floppy = {
>  		blocksize = 1024
> -		inode_size = 128
>  		inode_ratio = 8192
>  	}
>  	big = {
> @@ -44,4 +41,5 @@
>  	hurd = {
>  	     blocksize = 4096
>  	     inode_size = 128
> +	     warn_y2038_dates = 0
>  	}
> -- 
> 2.31.0
>