Re: [PATCH v2 2/3] exportfs: allow fs to disable CAP_DAC_READ_SEARCH check

[kernel test robot <lkp@intel.com>]

Hi Erin,

kernel test robot noticed the following build warnings:

[auto build test WARNING on 14b6320953a3f856a3f93bf9a0e423395baa593d]

url:    https://github.com/intel-lab-lkp/linux/commits/Erin-Shepherd/pseudofs-add-support-for-export_ops/20241114-020539
base:   14b6320953a3f856a3f93bf9a0e423395baa593d
patch link:    https://lore.kernel.org/r/20241113-pidfs_fh-v2-2-9a4d28155a37%40e43.eu
patch subject: [PATCH v2 2/3] exportfs: allow fs to disable CAP_DAC_READ_SEARCH check
config: openrisc-defconfig (https://download.01.org/0day-ci/archive/20241114/202411140905.a0ntnQQG-lkp@intel.com/config)
compiler: or1k-linux-gcc (GCC) 14.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20241114/202411140905.a0ntnQQG-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202411140905.a0ntnQQG-lkp@intel.com/

All warnings (new ones prefixed by >>):

   fs/fhandle.c: In function 'may_decode_fh':
>> fs/fhandle.c:242:41: warning: initialization discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers]
     242 |         struct export_operations *nop = root->mnt->mnt_sb->s_export_op;
         |                                         ^~~~


vim +/const +242 fs/fhandle.c

   237	
   238	static inline bool may_decode_fh(struct handle_to_path_ctx *ctx,
   239					 unsigned int o_flags)
   240	{
   241		struct path *root = &ctx->root;
 > 242		struct export_operations *nop = root->mnt->mnt_sb->s_export_op;
   243	
   244		if (nop && nop->flags & EXPORT_OP_UNRESTRICTED_OPEN)
   245			return true;
   246	
   247		if (capable(CAP_DAC_READ_SEARCH))
   248			return true;
   249	
   250		/*
   251		 * Allow relaxed permissions of file handles if the caller has the
   252		 * ability to mount the filesystem or create a bind-mount of the
   253		 * provided @mountdirfd.
   254		 *
   255		 * In both cases the caller may be able to get an unobstructed way to
   256		 * the encoded file handle. If the caller is only able to create a
   257		 * bind-mount we need to verify that there are no locked mounts on top
   258		 * of it that could prevent us from getting to the encoded file.
   259		 *
   260		 * In principle, locked mounts can prevent the caller from mounting the
   261		 * filesystem but that only applies to procfs and sysfs neither of which
   262		 * support decoding file handles.
   263		 *
   264		 * Restrict to O_DIRECTORY to provide a deterministic API that avoids a
   265		 * confusing api in the face of disconnected non-dir dentries.
   266		 *
   267		 * There's only one dentry for each directory inode (VFS rule)...
   268		 */
   269		if (!(o_flags & O_DIRECTORY))
   270			return false;
   271	
   272		if (ns_capable(root->mnt->mnt_sb->s_user_ns, CAP_SYS_ADMIN))
   273			ctx->flags = HANDLE_CHECK_PERMS;
   274		else if (is_mounted(root->mnt) &&
   275			 ns_capable(real_mount(root->mnt)->mnt_ns->user_ns,
   276				    CAP_SYS_ADMIN) &&
   277			 !has_locked_children(real_mount(root->mnt), root->dentry))
   278			ctx->flags = HANDLE_CHECK_PERMS | HANDLE_CHECK_SUBTREE;
   279		else
   280			return false;
   281	
   282		/* Are we able to override DAC permissions? */
   283		if (!ns_capable(current_user_ns(), CAP_DAC_READ_SEARCH))
   284			return false;
   285	
   286		ctx->fh_flags = EXPORT_FH_DIR_ONLY;
   287		return true;
   288	}
   289	

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