On 11/19/05, Pekka Enberg <[email protected]> wrote:
> Hi Phillip,
>
> On 11/19/05, Phillip Hellewell <[email protected]> wrote:
> > +struct ecryptfs_session_key {
> > +#define ECRYPTFS_USERSPACE_SHOULD_TRY_TO_DECRYPT 0x01
> > +#define ECRYPTFS_USERSPACE_SHOULD_TRY_TO_ENCRYPT 0x02
> > +#define ECRYPTFS_CONTAINS_DECRYPTED_KEY 0x04
> > +#define ECRYPTFS_CONTAINS_ENCRYPTED_KEY 0x08
> > + int32_t flags;
> > + int32_t encrypted_key_size;
> > + int32_t decrypted_key_size;
> > + uint8_t decrypted_key[ECRYPTFS_MAX_KEY_BYTES];
> > + uint8_t encrypted_key[ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES];
>
> s32 and u8 are preferred in the kernel.
Thanks for that, lots of little things like this were undoubtably
missed, we use u/s# in other areas. Changed and will be resent after
all comments are in.
>
> > +#define OBSERVE_ASSERTS 1
> > +#ifdef OBSERVE_ASSERTS
> > +#define ASSERT(EX) \
> > +do { \
> > + if (unlikely(!(EX))) { \
> > + printk(KERN_CRIT "ASSERTION FAILED: %s at %s:%d (%s)\n", #EX, \
> > + __FILE__, __LINE__, __FUNCTION__); \
> > + BUG(); \
> > + } \
> > +} while (0)
> > +#else
> > +#define ASSERT(EX) ;
> > +#endif /* OBSERVE_ASSERTS */
>
> Any reason why you can't just use BUG and BUG_ON()?
No reason? I think we've had this comment before, not sure if we have
had a decent reason though... let me see if I can find out why we do
it this way.
> > +
> > +/**
> > + * Halcrow: What does the kernel VFS do to ensure that there is no
> > + * contention for file->private_data?
> > + */
>
> Please elaborate?
I believe this is an old and lingering comment. While I don't know who
originated, it sounds like a question to Michael Halcrow regarding
locking of a struct file's private data... It will be removed for next
submission.
>
> > +#define ECRYPTFS_FILE_TO_PRIVATE(file) ((struct ecryptfs_file_info *) \
> > + ((file)->private_data))
> > +#define ECRYPTFS_FILE_TO_PRIVATE_SM(file) ((file)->private_data)
> > +#define ECRYPTFS_FILE_TO_LOWER(file) \
> > + ((ECRYPTFS_FILE_TO_PRIVATE(file))->wfi_file)
> > +#define ECRYPTFS_INODE_TO_PRIVATE(ino) ((struct ecryptfs_inode_info *) \
> > + (ino)->u.generic_ip)
> > +#define ECRYPTFS_INODE_TO_PRIVATE_SM(ino) ((ino)->u.generic_ip)
> > +#define ECRYPTFS_INODE_TO_LOWER(ino) (ECRYPTFS_INODE_TO_PRIVATE(ino)->wii_inode)
> > +#define ECRYPTFS_SUPERBLOCK_TO_PRIVATE(super) ((struct ecryptfs_sb_info *) \
> > + (super)->s_fs_info)
> > +#define ECRYPTFS_SUPERBLOCK_TO_PRIVATE_SM(super) ((super)->s_fs_info)
> > +#define ECRYPTFS_SUPERBLOCK_TO_LOWER(super) \
> > + (ECRYPTFS_SUPERBLOCK_TO_PRIVATE(super)->wsi_sb)
> > +#define ECRYPTFS_DENTRY_TO_PRIVATE_SM(dentry) ((dentry)->d_fsdata)
> > +#define ECRYPTFS_DENTRY_TO_PRIVATE(dentry) ((struct ecryptfs_dentry_info *) \
> > + (dentry)->d_fsdata)
> > +#define ECRYPTFS_DENTRY_TO_LOWER(dentry) \
> > + (ECRYPTFS_DENTRY_TO_PRIVATE(dentry)->wdi_dentry)
>
> These wrappers seem rather pointless and obfuscating...
I find them make the code a bit more clear. Then again, I understand
what they do and use them a lot, so I am clearly jaded. You are the
first person to comment on these wrappers (all previous comments where
regarding some debug wrappers we had in when we first submitted, I
like those ones too, but I guess my opinion doesn't carry enough
weight here ;P) There is no functional reason why these can't be
removed, but like I said, I think they make the code a lot easier to
read once you know all they do is sugar-coat a potentially long, and
potentially confusing, chain of refences, and turn that into some
phrase-like macro.
>
> > +int virt_to_scatterlist(const void *addr, int size, struct scatterlist *sg,
> > + int sg_size);
>
> Doesn't seem ecryptfs specific, why is it here?
Ah ha, that should read ecryptfs_virt_to_scatterlist. Again, will be
changed for next submission. While reviewing the code, there seem to
be a few more like this. I'll make sure they get updated.
>
> Pekka
> -
> To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
--
Michael C. Thompson <[email protected]>
Software-Engineer, IBM LTC Security
-
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]