PARAM.PFD

Description

This files are responsible to prevent tampering of other files of the same folder, the only purpose is the security of the folder contents.

Contains the cryptographic signatures of other files of the same folder (not all, but the ones that developers decided to be important enought to be secured).

Its used in several folders of the PS3, usually to controll data related with the user profile e.g:

/dev_hdd0/home/0000000*/savedata/SAVEDATA_DIRECTORY/PARAM.PFD <---- in all save game folders
/dev_hdd0/home/0000000*/trophy/NPCOMMID/PARAM.PFD <--- in all trohpye folders
/dev_hdd0/home/0000000*/trophy/_TROPSYS_/PARAM.PFD <---- in the trophy "index" folder
/Maybe other places ?

Usually PARAM.SFO is one of the files listed (and supervised) by PARAM.PFD but this can vary.

e.g. the folder _TROPSYS_ only contains 2 files: TROPSYS.DAT & PARAM.PFD (the first is the trophy index for this user, and the second supervises it).

Every time one of the files listed inside the .PFD is updated... the .PFD itself needs to be updated to store the new "signatures" of the other updated files e.g:

When you save a game there is some text that changes inside PARAM.SFO with your current "gameplay time", "character level", etc...
When you unlock a new trophye the file TROPUSR.DAT is updated with this new data
When you install new trophyes (with a "trophy installer" in .TRP format) from a new game the TROPSYS.DAT is updated to index the new trophy installation
... In all this cases, there is a PARAM.PFD that has been updated to store the new "signatures" of the other updated files

More info on other related file formats: PARAM.SFO Game Saves Trophy files.

Structure

Total file size = 32768 bytes

The file can be divided in 5 chunks: header 120 bytes, X table 456 bytes, file table 31008 bytes, Y table 1140 bytes, and padding 44 bytes, for a total of 120+456+31008+1140+44 = 32768 bytes (0x8000 in hexadecimal)

The size is fixed because the number of entries in both "X table" & "Y table" is 57 (when the entry is not used his position is reserved and marked as "not-used"). In the same way... the "files table" has the space reserved for a maximun of 114 entries (usually most of them are not used and are filled with zeroes)

As result the file contains the maximun number of possible entries

Note that entries in the "file table" (114) is exactly the double than the entries in "X table" (57) & "Y table" (57)

Header

From 0x0 to 0x60
Size = 120 bytes

The end of the header is not clear at first sight, but it can be deduced by counting the number of entries in "X table" & "Y table" and comparing the positions used in both

e.g. look for one used entry in the "Y table" and count his position in this table... then look for other used entry in the "X" (both tables matches in the used entries). Then count towards behind to find the first entry (the start of the first entry is the start of the "X table")

So the start of the "X table" is the end offset of the header

Offset	Size	Value	Description
0x00	08 bytes	PFDB	Magic value in ASCII
0x08	08 bytes	0000000000000003	constant
0x10	10 bytes	random bytes	key to en/decrypt
0x20	14 bytes	hmac sha1	encrypted, decrypted by vtrm
0x34	14 bytes	hmac sha1	encrypted, decrypted by vtrm
0x48	14 bytes	hmac sha1	encrypted, decrypted by vtrm
0x5C	04 bytes		padding, encrypted by vtrm

Offset	Size	Value	Description
0x60	08 bytes	0000000000000039	Number of reserved entries in the "X table" & "Y table" (57 in decimal)
0x68	08 bytes	0000000000000072	Number of reserved entries in the "file table" (114 in decimal)
0x70	08 bytes	00000000000000**	Number of used entries in the "file table" (it can vary from 1 to 114 in decimal)

X table

From 0x78 to 0x240
Size = 456 bytes
Entry lenght = 8 bytes
Number of entries = 57

This table has a total of 57 entries, with 8 bytes each, for a total of 456 bytes (57*8=456)

Each entry can contain a "file index" that points to the "file table" from 0 to 71 in hexadecimal (or 1 to 114 in decimal)

If the "file index" is 72 it means that is pointing out of the "file table" (not used)

The table does not fills with entries from top to bottom, usually the first entries are not used (marked as 72) followed by entries with whatever number from 0 to 71, and mixed with more 72's entries

Used entries has a number asigned by his position in the table e.g:

Example of a "non-standard" "X table" with a "entry lenght" of 1 byte, and 8 entries
72 01 72 72 00 72 72 02

This "X table" can be translated as:
-File index nº 1 (00) assigned as position 5
-File index nº 2 (01) assigned as position 2
-File index nº 3 (02) assigned as position 8

Some possible entries	Description
0000000000000000	File index nº 1
0000000000000001	File index nº 2
0000000000000002	File index nº 3
0000000000000071	File index nº 114
0000000000000072	Out of file table (not used)

File table

From 0x240 to 0x7B60
Size = 31008 bytes
Entry lenght = 272 bytes
Number of entries = variable from 1 to 114

This table has a total of 114 entries, with 272 bytes each, for a total of 31008 bytes (114*272=31008)

The used entries fills the table from top to bottom, not-used entries are placed at the end of the table filled with zeroes

The first 8 bytes of each entry (file index) works in the same way than the entries in the "X table", usually not used (72) and when used are randmonly placed asigning an "file index" to the entry

This "file index" does not matches with the position of the entry in the "file table" itself... so seems that this "indexes files" are ???virtually reordered???

In fact, for a theoricall file with alll entries used, half of the "file index" numbers will be spreaded between the "X table" (can only contain 57) ant the first 8 bytes of some entries in the "file table"

Example of a "non-standard" "file table" with only the first 8 bytes of each entry, and 4 entries
0000000000000072 0000000000000001 0000000000000000 0000000000000002

The "file index" of each entry of this "file table" can be translated as:
-File index nº 1 (00) assigned as position 3
-File index nº 2 (01) assigned as position 2
-File index nº 3 (02) assigned as position 4

Size	Value	Description
008 bytes	00000000000000**	File index (value 72 = Not indexed)
064 bytes	EXAMPLE.WTF	Name of the file included the point and the extension in ASCII (Null-terminated)
008 bytes	:Wtf-o0-	In ASCII, Usually zeroed (crysis2 savedata contains the string "CELL" and heavenly sword contains "s:Music") ???Unknown???
188 bytes	????????...	Certificate for the file. When the file is PARAM.SFO then the certificate is bigger in size and uses imput data from the attribute "PARAMS" and/or "ACCOUNT_ID" inside PARAM.SFO. Method unknown (Null-terminated)
004 bytes	1A2B3C4D	Size of the file in bytes

Y table

From 0x7B60 to 0x7FD4
Size = 1140 bytes
Entry lenght = 20 bytes (some kind of sha-1 hash??)
Number of entries = 57

This table has a total of 57 entries, with 20 bytes each, for a total of 1140 bytes (57*20=1140)

All the entries contains the same "20 bytes string", only the used ones has a different "20 bytes string" in the other "X table" (in a theoricall PARAM.PFD with no files listed, the "Y table" would have all his 57 entries with the same string)

Is directly related with the "X table", both matches in the total number of entries (57) and wich ones are used (e.g. when the "index table" has a entry in position 12... the "weird table" has position 12 used)

The "Y table" has a repeating pattern so an entry for each potential file with the blank entry (I.E. no file) being represented by the repeating byte pattern.

Padding

Size = 44 bytes

Cryptography

Unknown by now, but some questions rises...

Why the files are listed in this order and not in other in the "files table" ?

Because are not listed alphabetically, neither by size

Indexes files (in the "index table") seems to have different number for every one, never repeats, but there is not direct relationship between the number of entries in "index table" & "weird table" (both are fixed to 57) and the numer of files listed in the "file table" (114)... the most logicall explain if that this 114 files can be linked to both tables (57 each)... but in fact the only table that stores crypto is the "weird table" (limited to 57)... so what trick they used ? hmmmm

What are this index in the "index table" and in the "file table" itself?, his positions seems to be random, seems like an old school "lucas arts games" anticheat card where you pick 2 values and by mixing them you get the unlock code :D

But here what is random is the positions, and index numbers of the entries in the "index table", and the indexed files in the "file table" ??? 2 index ???

Source: http://ps3dev.wikispaces.com/PARAM.PFD