[PATCH]: ufs: fix symlink creation on ufs2

[PATCH]: ufs: fix symlink creation on ufs2

[Evgeniy Dushistov]

If create symlink on UFS2 filesystem under Linux,
it looks wrong under other OSes, because of max symlink length
field was not initialized properly, and data blocks were not
used to save short symlink names.

Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru>
Cc: Steven <stevenaaus@yahoo.com>

---

diff --git a/fs/ufs/super.c b/fs/ufs/super.c
index 0072cb3..14605c0 100644
--- a/fs/ufs/super.c
+++ b/fs/ufs/super.c
@@ -131,6 +131,8 @@ static void ufs_print_super_stuff(struct super_block *sb,
 		printk(KERN_INFO"  cs_nffree(Num of free frags): %llu\n",
 		       (unsigned long long)
 		       fs64_to_cpu(sb, usb3->fs_un1.fs_u2.cs_nffree));
+		printk(KERN_INFO"  fs_maxsymlinklen: %u\n",
+		       usb3->fs_un2.fs_44.fs_maxsymlinklen);
 	} else {
 		printk(" sblkno:      %u\n", fs32_to_cpu(sb, usb1->fs_sblkno));
 		printk(" cblkno:      %u\n", fs32_to_cpu(sb, usb1->fs_cblkno));
@@ -1060,8 +1062,8 @@ magic_found:
 	uspi->s_bpf = uspi->s_fsize << 3;
 	uspi->s_bpfshift = uspi->s_fshift + 3;
 	uspi->s_bpfmask = uspi->s_bpf - 1;
-	if ((sbi->s_mount_opt & UFS_MOUNT_UFSTYPE) ==
-	    UFS_MOUNT_UFSTYPE_44BSD)
+	if ((sbi->s_mount_opt & UFS_MOUNT_UFSTYPE) == UFS_MOUNT_UFSTYPE_44BSD ||
+	    (sbi->s_mount_opt & UFS_MOUNT_UFSTYPE) == UFS_MOUNT_UFSTYPE_UFS2)
 		uspi->s_maxsymlinklen =
 		    fs32_to_cpu(sb, usb3->fs_un2.fs_44.fs_maxsymlinklen);
 
-- 
/Evgeniy

Re: [PATCH]: ufs: fix symlink creation on ufs2

[Andrew Morton]

On Wed, 2 Jan 2008 23:47:04 +0300 Evgeniy Dushistov <dushistov@mail.ru> wrote:

> If create symlink on UFS2 filesystem under Linux,
> it looks wrong under other OSes, because of max symlink length
> field was not initialized properly, and data blocks were not
> used to save short symlink names.
> 
> Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru>
> Cc: Steven <stevenaaus@yahoo.com>
> 
> ---
> 
> diff --git a/fs/ufs/super.c b/fs/ufs/super.c
> index 0072cb3..14605c0 100644
> --- a/fs/ufs/super.c
> +++ b/fs/ufs/super.c
> @@ -131,6 +131,8 @@ static void ufs_print_super_stuff(struct super_block *sb,
>  		printk(KERN_INFO"  cs_nffree(Num of free frags): %llu\n",
>  		       (unsigned long long)
>  		       fs64_to_cpu(sb, usb3->fs_un1.fs_u2.cs_nffree));
> +		printk(KERN_INFO"  fs_maxsymlinklen: %u\n",
> +		       usb3->fs_un2.fs_44.fs_maxsymlinklen);

I assume this needs conversion via fs32_to_cpu() before we print it.

>  	} else {
>  		printk(" sblkno:      %u\n", fs32_to_cpu(sb, usb1->fs_sblkno));
>  		printk(" cblkno:      %u\n", fs32_to_cpu(sb, usb1->fs_cblkno));
> @@ -1060,8 +1062,8 @@ magic_found:
>  	uspi->s_bpf = uspi->s_fsize << 3;
>  	uspi->s_bpfshift = uspi->s_fshift + 3;
>  	uspi->s_bpfmask = uspi->s_bpf - 1;
> -	if ((sbi->s_mount_opt & UFS_MOUNT_UFSTYPE) ==
> -	    UFS_MOUNT_UFSTYPE_44BSD)
> +	if ((sbi->s_mount_opt & UFS_MOUNT_UFSTYPE) == UFS_MOUNT_UFSTYPE_44BSD ||
> +	    (sbi->s_mount_opt & UFS_MOUNT_UFSTYPE) == UFS_MOUNT_UFSTYPE_UFS2)
>  		uspi->s_maxsymlinklen =
>  		    fs32_to_cpu(sb, usb3->fs_un2.fs_44.fs_maxsymlinklen);
>  

Re: [PATCH]: ufs: fix symlink creation on ufs2

[Christoph Hellwig]

On Wed, Jan 09, 2008 at 11:51:48PM -0800, Andrew Morton wrote:
> > +++ b/fs/ufs/super.c
> > @@ -131,6 +131,8 @@ static void ufs_print_super_stuff(struct super_block *sb,
> >  		printk(KERN_INFO"  cs_nffree(Num of free frags): %llu\n",
> >  		       (unsigned long long)
> >  		       fs64_to_cpu(sb, usb3->fs_un1.fs_u2.cs_nffree));
> > +		printk(KERN_INFO"  fs_maxsymlinklen: %u\n",
> > +		       usb3->fs_un2.fs_44.fs_maxsymlinklen);
> 
> I assume this needs conversion via fs32_to_cpu() before we print it.

Yes, it does.  I wish people would run sparse before submitting patches :P

Re: [PATCH]: ufs: fix symlink creation on ufs2

[Al Viro]

On Thu, Jan 10, 2008 at 08:40:41AM +0000, Christoph Hellwig wrote:
> On Wed, Jan 09, 2008 at 11:51:48PM -0800, Andrew Morton wrote:
> > > +++ b/fs/ufs/super.c
> > > @@ -131,6 +131,8 @@ static void ufs_print_super_stuff(struct super_block *sb,
> > >  		printk(KERN_INFO"  cs_nffree(Num of free frags): %llu\n",
> > >  		       (unsigned long long)
> > >  		       fs64_to_cpu(sb, usb3->fs_un1.fs_u2.cs_nffree));
> > > +		printk(KERN_INFO"  fs_maxsymlinklen: %u\n",
> > > +		       usb3->fs_un2.fs_44.fs_maxsymlinklen);
> > 
> > I assume this needs conversion via fs32_to_cpu() before we print it.
> 
> Yes, it does.  I wish people would run sparse before submitting patches :P

sparse wouldn't have helped it in the current state.  The thing is, it
doesn't understand printf-style formats.  We probably ought to implement
that - would have a benefit of being able to get rid of gcc stupidity
in that area and actually define conversion specifiers of our own for
things like u64, dma_addr_t, etc. without having gcc scream at us...