Hello,
I notice that on versions 2.6.15 and later, the function
void d_delete(struct dentry * dentry)
Implements a mechanism for notifying the deletion of the d_entry.
The implementation assumes that the d_entry has a non-null d_inode pointer
and that seems to be terribly wrong. In other words, how can a
"negative d_entry"
be deleted? Implementation of rename may use dummy dentries (that have d_inode
set to NULL) and that is a problem as well.
Please CC me if you have any comments/feedback on this subject.
Thanks!
-- [email protected]
PS. For more specific details on the code I'm referring to, see below:
File .../fs/dcache.c
void d_delete(struct dentry * dentry)
{
int isdir = 0;
...
spin_lock(&dcache_lock);
spin_lock(&dentry->d_lock);
isdir = S_ISDIR(dentry->d_inode->i_mode); <======= MUST CHECK d_inode!!!!
if (atomic_read(&dentry->d_count) == 1) {
...
}
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[Index of Archives]
[Kernel Newbies]
[Netfilter]
[Bugtraq]
[Photo]
[Stuff]
[Gimp]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Video 4 Linux]
[Linux for the blind]
[Linux Resources]