Title: iPhone target analysis and exploitation with unique device identifiers

Release Date: 2015-01-17

Document Date: 2010-11-12

Description: This GCHQ research paper dated 12 November 2010 discuss some of the agency’s attacks against iPhone handsets: see the Der Spiegel article The Digital Arms Race: NSA Preps America for Future Battle, 17 January 2015.


12th November 2010

MHE Team (MHETeam-GCHQ-dl)


Benhall Records Centre

iPhone target analysis and exploitation with unique

device identifiers

This paper describes standard analysis techniques
that have been used to both discover iPhone
target end point machines and implant target
iPhones directly using the QUANTUM system. It
shows that the iPhone Unique Device Identifier
(UDID) can be used for target tracking and can be
used to correlate with end point machines and
target phone. It highlights the exploits currently
available and the CNE process to enable further



Page 1 of 11



12th November 2010


[a] iPhone applications and privacy issues: An analysis of Application
Transmission of iPhone Unique Device Identifiers (UDIDs),|

October 1 2010

[b] CROWN PRINCE - Technimjefondentifying Apple UDIDs in HTTP traffic

- B/7844/5001/1/105J

[c] Strategic Framework Task 4138585
October 2009 Issue 2


- Report No: 72/09/R/416/C, Roke

[d] TheGoodpeneti^

[e] Current SEPP targets

[f] iPhone target list


1. Every Apple iPhone, iPad and iPod touch has a unique hardware
identifier called the Apple UDID. The UDID is a 40 character hex string
(160 bits) that seems to be a SHA-1 hash of the IMEI, serial number and
the Bluetooth and WiFi MAC addresses. The UDID is available to
developers of applications for these devices, and it is used to identify a
given device. As highlighted in [a], the UDID is seen in multiple apps and
can be used to allow target tracking or be used to correlate with other
personal identifiers.

2. The Mobile Theme has invested a large amount of research into iPhone
apps and metadata analysisover the last year accumulating with a
detailed report done by^^|[c] in October 2009 and 29 SEM rules
created by ICTR-MCT [b]. These rules have used to extract iPhone
metadata for a number of apps and in particular the Unique Device
Identifier (UDID) from any carrier being processed using DEBIT CARDS.
Further TDI rules are being developed by GTE that will in the future
extract UDID events from carriers processed through the MVR system.
The resulting events have then been used to populate both research and
corporate QFDs (Query Focused Datasets) such as MUTANT BROTH
and AUTOASSOC and will eventually form the basis of mobile
correlations in HARD ASSOC.

3. Initially, an exploit was developed by the Joint CNE/TECA Mobile
Exploitation Team for iPhone that was to be delivered to the target
phone when syncing with an exploited end point machine. This was
successful for a BROKER target and resulted in the extraction of SMS,
call logs and contact details. After this initial trial, CNE and SD undertook
work to discover other single end points seen syncing with iPhones.

4. At HANDEX 2010 (handset exploitation workshop) in August, various
aspects of the iPhone OS were investigated for potential vulnerabilities.

Page 2 of 11



12th November 2010

Using an open source PDF vulnerability when using the Safari browser,
Joint CNE/TECA Mobile Exploitation Team were able to develop an
exploit to deliver a WARRIORPRIDE implant to a target test phone.
Further, investigation, liaison and testing with the CNE QUANTUM team
resulted in approval for the implant to be deployed against QUANTUM
iPhone targets.


CNE Endpoint

5. As part of the SD Mobile Exploitation theme to identify further end point
machines that had been seen syncing with iPhones, a survey was
undertaken by CNE and TAO to scan all target end point machines for
the appropriate iPhone registry keys. Scanning of all CNE stored single
end point (SEPs) registry keys on particular process IDs resulted in 9
CNE endpoints seen this year sync’d with iPhones. The resulting Unique
Device Identifiers (UDID) were extracted from the registry keys and ran
in MUTANT BROTH and AUTOASSOC, resulting in 6 correlations with
either iPhone Safari user agents or the iPhone Mail app seen in passive

6. A CNE end point operation against SO LINE

EPI^ONHtad resulted in access to a windows end point machine
As can be the

UDID has been seen with the Admob SEM rule type and with the Apple-
IMEI-URI TDI type. Admob is the largest mobile advertising network
allowing games publishers to embed adverts and therefore receive
revenue from a number of different brands. The target iPhone OS is 3_0
as shown in the User Agent profile in figure 2.

TDI type

TDI value

0 EHP_AdmotHsu-Uft)

[7] EihUTÛ_Apple-imBi-Um
[^1 EHP^Admob-X-Admcb-Iiu

Figure 1 - MUTANT BROTH Matching Identifiers



£vent Count

Mozilla/5.0 (iPhone; U; CPU iPhone OS 3_0 like Ma c OS X; Hw iPhonel.2; emus] AppleWebKit/525.18,1 3 <18 %0

(KHTML, like Gecko) (AdM*b-ISDK-2a(J9l)Z(J3)

iPhone Mail (7A.341) 11(68%)

MQZIIfa/S.D (iPhone; U; CPU IPhone OS 3_Q like Mat OS X; MW iPhonel.Z; en_us) AppleWebKit/SZS.Lö.L till %)

(KHTML, like Gecko)
Figure 2 -|

iPhone UDID User agent profile

Page 3 of 11



12th November 2010


The target UDID can be used track the iPhone seen with ASBOLINE
EPILSON end point machine. In this paiTiculaj^ase^he^arge^pID has
been seen 16 times, the last time off on the

24/10/10 using the inbuilt iPhone Mail client to access his yahoo
account. The user agent for this is shown in Figure 3. In this case the
EAUTO_Apple-imei-URI TDI rule was used to extract the specific UDID

TDI-Scope Auto Route________

User-Agent iPhone Mail (7A341)
Labelled-Ro ute

Figure 3 - iPhone Mail client user agent

Bearer Siged Pddg Ssdg Other Info


TDI-Scope Machine Ruute

horro at-Tran Storni SEM->TD!

User-ftoent Mozilla/5.0 (iPhone; U; CPU
iPhone OS 3_0 like Mac OS X; HW
iP h o neljï ; e n_uí) App leWebKitf 5 25.13.1
(KhlTML, like Gecko) (AdMob-
iSDK'20090609) AS-IP-Src NO_MATCH

Figure 4 - iPhone Admob user agent profile

8. The UDID for all 6 targets were run through AUTOASSOC. The result for

is shown in Figure 5. As can be seen there is a clear
correlation with the^^^^^^^Hyahoo-Y-cookie.

Known TE>1



EXP Admob-X-Adrrtob-lsu:

Putative TÙ1

Yahoo - Y- Cookie :

Yahoo - Y- Cookie :


Yahoo - Y- Cookie :

Yahoo - Y- Cookie :

Yahoo - Y- Cookie :

Yahoo - Y- Cookie :

Yahoo - Y- Cookie :

Yahoo - Y- Cookie :

EXP Admob-X-Admob-Isu,


CORINTH tasking ? Score Events Most recent
- 1.000 13 24/10/10
- □ .SOD 1 23/00/10
- o.soo 1 14/11/09
- o.soo 1 14/11/09
- o.soo 1 19/06/10
- o.soo 1 10/12/09
- 0.500 1 19/00/10
- o.soo 1 23/08/10
- o.soo 1 10/12/09
- o.soo 1 19/06/10
- o.soo 1 15/11/09
- o.soo 1 19/09/10
- o.soo 1 19/09/10


9. The IP address, identifier type, bearer and user agent type for all 6

targets was extracted and formed the basis of further target development
work. Running the resulting 6 UDIDs through AUTOASSOC, resulted in
two correlations with a high enough score. These were
and and confirmed to be correct after discussing with

Page 4 of 11



12th November 2010

Checking BROADOAK tasking revealed that both targets
had known associated iPhone IMEIs already tasked.

10. A recent rescan of all currently active CNE SEPs resulted in only 5 of the
9 identified CNE machines actually being available for exploit. The other
4 implants having been removed. Of these five UDIDs, four returned with
correlated UDIDs that had recently been seen in passive collection.
These are contained in the iPhone target list [f],

11. Analysis of all TAO SEPs resulted in 116 UDIDs being identified. Of
these UDIDs, 15 were correlated with iPhone user agents and the
resulting identifier type, project name, case notation and IP noted. A full
list is available in the iPhone target list [f]. Of these four had the Cydia
user agent as shown in Figure 6 indicating the target had jailbroken their
phone. All four end points were located in^^^^|.

i |][-Hrrcpp machine fficuitp


Format -Transform 3EM'»TD1
llsor-Agent Mozilla/5.0 ReckApp,J2.6C .L
(iPhone: U; CPU like Ctdia/l.D.3172-66)

fippls WehKiVSlD .13 (KHTT'IL, like Gieko)
Version/J .0 Mobils/7Dll 5afari/52S.16
AS-lP-Src AS-IP-Ostl


Euent-CSL EEQ7CB Stream-C
E F0 7CS0 0 0 000000 0 Q 300000 0 0 000000 0 Z 0

Figure 6 - iPhone jailbroken user agent in MUTANT BROTH

12. The same 15 TAO UDIDs were run through AUTOASSOC and resulted

inthreeaoodcoirela^ yahoo sel i li n

were in

turn confirmed to be correct correlations with TAO target end points and
two showed associated target iPhone IMEIs. Further, analysis of the
Yahoo mail used via the Safari browser clearly showed the resulting
UDID was transmitted in traffic.

13. One of these TAO end point machines, SOLARSHOCK116

lranian)Jiasbeense^ with iPhone


th ^si n
AUTOASSOC. The UDID was last seen on 23/10/10 at 03:46:36z on

using the EAUTO_Apple-imei-URI TDI.

Jtiraiivn 75?


tmei war;

Affsfiue 70/


i.nnn 33 ?3/srvin FI
i.OOc« Si tt/Wi 0 FI
U.DJC 2 M/oi/w El
0.WQ 1 E3/06AU 0
0.509 t ? VC 6/10 0
U.bULT 1 £■0/12/09 0
0.500 1 21/2 0/10 0
n.sna | w/ee/w 0

Figure 7- SOLARSHOCK116 UDID AUTOASSOC correlations

Page 5 of 11



12th November 2010


14. After extensive testing of the QUANTUM tipping and redirection to the
SHORTSHEET exploitation server by the Joint CNE/TECA Mobile
Exploitation Team, further target development work was needed to
identify iPhone targets recently seen active on the appropriate user
agents. Bulk extraction of targets over a three-week period from
BLACKHOLE by ICTR resulted in a large number of iPhone targets and
further queries in Xkeyscore resulted in others. Others were passed from
discussions with various IPTs and two were also passed from TAO

15. In total 44 selectors were checked to verify that the correct user agent
was present. Of these 44 selectors, 41 were yahoo selectors and 3 gmail
selectors. Of these 26 were seen with a valid iPhone Safari user agent
as shown in Figure 8. A summary of OS versions seen with target
phones is shown in Table 1 with OS highlighted in red currently
exploited. These are 3_1_2, 3_1_3 and 4_0_1. In all 26 cases, the target
analysts were contacted with details regarding their targets use of an
exploitable iPhone.

iPhone OS Number
3_0 4
3_1 2
3_1_2 6
3_1_3 15
4_0 3
4_0_1 5
4_0_2 1

Table 1 - iPhone target OS summary

24. One particular case was target,with yahoo

selectorwas seen active on a iPhone OS
3_1_2, as shown in Figure 8. The resulting Yahoo-B cookie is
^^^^^^^^■ancha^arUi^eenthetarget has been active off

Running the Yahoo-

B cookie through MUTANT BROTH resulted in 171 events primarily on
case notations GWUKG005, GWVCB003 and IRUKC036. The resulting
information was then forwarded to the analysts in the^^^|team for
tasking by the standard CNE process as outlined in the Good
Penetration Guide [d].

Page 6 of 11



12th November 2010

TDÏ-5 Source CPC_IDEBTTCAR.O Formater

Ser-Agent Masilla/S.ü (iPhone; U; CPU iPhone OS 3_1_2 like Mac QS X; en-us)

Safah/528.i6 j

. TCI Yahoo C Cuij Y jI'iüli Y Cuukic

FaWebKit/E28.10 ÎKHTML,Tf^jecka^7ersiü'R/4.û Mobile/7Dtl
I Event-C$1 ËF07C6 stream-CSL EFÜ7C9

TDT Scope Machine Route Sauren CPC_DEBITCARD Format Transform ÏEM--*TD: Yahoo B CookieYahoo Y Cookie

Iser-Agent Mcmlla/5.0 (iPhone; U; CPU iPhone OS 3_1_2 like Mec OS X; en-us) Ap3ÎeWebKrt/528.lS (KHTML, like Gecko] Versicm/4.0 Mobile/7Dll
Safari/520.16 AS IP-Src NO.MATCH A5 1P~Dst ^^B Labelled Route Event-CSL EF07C6 Stream-CSL EFD7C9

TOI Scope Machina Route IJ see Agent Mo^illa/SG IPhone; IJ; CPU iPhone OS3]l2likelMac OS X; en-us) AppleWebKitfEZa.JLi (K HTML, like

Gecko) Version/d 0 Mobile/7D13 Safan/szg.Lo Yahoo-V-CookleBI^B Yahoo-H-L'ook le^^^B^^^^Yahoo-T-Cook e

af=OXdBGjFOJnR£pTeyOOOOMDHwfyjUmcHM9üVRkrXF2WUtWai540USUWndm2IM2yOt Yafioa-Y-Cockie-Full v=i&n=abqikÖqdvtMn&l=c0j]Ö4S/o&
p—fin2nniOOCOO1300QOOO6ir^39&lg^en-US6;iritl-uS&np^l Cvcnt-security-lubcl 10007F Strcâinfï-ScCsirity-labeî 40QO23ECFF

TOT-Scope User P.Tute^^^^^MlUei-Agent Mç-?illa/5 Q (iPhone; Ut CPU iPhone OS 3_1_2 üke Mac OS Wj en-us) AppleWebKi^Sâ.ie EKHTML like
Gecko) version/4,0 Mobile/7Dii safari/5ZB.i6 Yahûo-Y-cookie^^^Hrahoo-B-cookæ^^^^^^^^|yab@o-T-cookne

Figure 8 -Valid iPhone Safari user agent in MB (



The target with target selector

^^^^^^^^^^Bwas seen active on different iPhone OS
and more recently on an iPad. Three other targets were seen active on
iPads and two others on iPods but with no other associated iPhone
device. Of the 44 targets seen, 16 were seen using the iPhone Mail
client that comes by default with all OS. A total of 7 targets were seen
using the Yahoo Mobile Messenger app.

26. For all target selectors seen with valid Safari user-agents, further

MUTANT BROTH, AUTOASSOC and MARINA queries were performed
to discover any other possible OS versions seen with the selector.
MARINA profile queries were performed resulting in the OS version
being returned within the MachinelD field. The resulting user-agents,
time/date, bearer, IP and any other associated selectors are shown in
the iPhone target list [f]. Associated selectors are either from MUTANT
BROTH, AUTOASSOC or seen directly as stated in BROADOAK. All
Yahoo B-cookies as shown in Figure 8 were run in MUTANT BROTH to
confirm their uniqueness with the iPhone and target yahoo selectors.


27. The QUANTUM redirect and PDF Safari browser exploit was developed
to work against 3_1_2, 3_1_3 and 4_0_1 Safari OS versions. Of the 26
targets seen with valid exploitable OS version, 5 were added to the
QUANTUM system for targetting.



Once notified, GCHQ IPTs either added the target to existing ONE
section 7 warrants as defined in [e] or developed targetting aids as
defined in [d] and wrote the appropriate warrant. The resulting section 7
warrants were approved by IPT team leaders and signed off by the ONE
Legal Team. Three NSA targets were discovered and were added by
ONE Legal Team to the Partner Agreement Forms to allow exploitation.


and B-
was not seen

yahoo selector

cookie was put on cover on QUANTUM,
active recently on a iPhone Safari browser to access his yahoo account,
preferring instead to use the inbuilt iPhone Mail client or his iPad. Three
QUANTUM attempts resulted in no redirect to SHORTSHEET server for

Page 7 of 11



12th November 2010

and after further analysis it was discovered that this selector was
not^^^|but an associate. It was removed from tasking.



Target analysis and warrant was completed for a further five targets and
a successful QUANTUM redirect and the PDF exploit was delivered for

PFVT658) on the 30th of October. The resulting
WARRIORPRIDE install was also performed and beaconed on the 2nd of
November. The target phone was shown to be jailbroken and on the 3rd
of November content was successfully extracted from the phone and
was available in Looking Glass. This is highlighted in Appendix B with
the resulting iPhone directory structure presented how it appears in
Looking Glass. The WARRIORPRIDE exploit has resulted in extraction
of the target's address book, sms, call logs, notes, WLAN logs,
bookmarks, map query history, Safari browsing history and some
images. Detailed analysis of extracted files will be covered in a further

A successful redirect to

in October was

performed but due to what is believed to be Javascript being disabled on
the phone the firmware type of the phone could not be confirmed to
enable the first stage implant. This initial survey of the firmware type has
now been removed after discussions within ONE but the target has not
been seen recently on his iPhone for exploitation. The two other targets
tasked for QUANTUM have failed to be seen recently in collect.

32. Currently, there are four CNE Single End Point machines that have been
identified with recently sync’d iPhone targets. The most recent being an
OVERLIT target seen on 29/09/10. These targets are being monitored
by CNE and will have the SLIDE exploit installed to allow implant of
WARRIORPRIDE when the iPhone is sync’d with the existing SEP


33. With the analysis of the UDIDs on target machines and correlation in
passive collection with known target yahoo selectors, the UDID can be
used to correlate iPhone handset to end point sync machine and tasked
yahoo selectors. The UDID can be used for realtime tracking of target
iPhones and could in theory be used as a selector for QUANTUM events
where other traditional selectors (yahoo-Y/B cookie etc) are not present.
Of course an exploit for the application would have to be written which is
not trivial.

34. It is not possible at this time to take the UDID and reverse engineer the
SHA-1 HASH to discover the IMEI, MAC addresses and serial number.

35. Further work is ongoing to identify targets of interest that are suitable for
the QUANTUM exploitation. Development and monitoring of current
identifiedtaraets is still being done. Discovery of an associated IMEI for

will help in firmware identification and exploitation of this


Page 8 of 11



12th November 2010

36. CNE are now conducting monthly surveys of all target machines registry
keys for iPhone UDIDs. A similar tipping mechanism for all BROADOAK
hits needs to be completed using a XKS workflow. Hopefully, other
targets will be added for QUANTUM exploitation or discovered on new
CNE SEPs in the future and a successful exploit of a target phone


37. Analysis of three PRESTON accesses has resulted in the identification
of a number of iPhone targets. Development work against at least three
target sets is needed with extraction of appropriate UDIDs and ¡Tunes
XDSID values. Suitable section 5 warrants need to be in place to pursue
these end point machines, once CNE have gained access and the
resulting target iPhones have been discovered.

38. With further work being undertaken by BSS and TECA on the
WHIPSAW redirect and exploitation server, it should be possible in the
coming months to implant directly the target iPhone. However, the
WHIPSAW exploit is only available via the tasked ADSL line.

39. An automatic implantation of SLIDE on to an iPhone is needed.

Currently this is manual process requiring a CNE operator to be
connected to the endpoint machine whilst the target is syncing the end
point machine and iPhone.

40. A larger number of iPhone TDIs need to be written to allow further
events to be populated into the QFDs allowing correlations with other
target selectors. This will also enable further real-time tracking of target

Page 9 of 11



12th November 2010




12th November 2010


Looking Glass iPhone directory structure

\lib'ary\preferences\systenconfigiration [see
Nellie d -, mB i Si« File Type Path Dale Mudified
© IaI /vai /mobiletmcdia/dcimtlOOappic/img 041C.jpc autowake, plist FT- 034.00 ... \librorr\praferences\,. 01-Jan-1D00 00;...
®--[Sl /var/mobilein-iedia/dcimilOOapple/img 0415. jpe nntwnik. ¡Hnnhifirah nn. plkh L 20.22KE PLI5T File ,,, \librarv\praferences\,, Cß-Nov-2010 17.,,
â-[Â|\ 0 rrm.^pple lAiif pliih L 1R.I11KF PI ET File ... \lihrarv\pr=fi=n=nrps\. in-Nnv-7mni7...
!--& bin 1=1 netwcrkin:erf aces, plist L r 1.35KB ... \librarv\Dreferences\.. 01-Jan-1300 00:...
i—& dev |i| preference s.plist L 16.16KE PLI5T File ... \librarv\preferences\.. 03-No v-2010 17...

3> library



? private

-l(2?> lihiwy
©■■& apcleotp
&■& appleusbdevice

carrier bundles
©■& coreservices
GJ £> datadwsmigrators
©■& extensions
ra filfisy sl-e.ms
¿1 & fonts
SMS’ frametAorks
S-& internet plug-ins
III keyboa'ddictioriaries

l±l& launchcaemons
©■■& lockdown

Cp manogedccnfigjrotonbjnde;

¿1 mediacapture

©■& privatefrarreworks
©■■& puitlishhgbundlas
©■■& seerchbundes
©■& sprhgbsardplugins
©■■& sys:emconliguration
©■& texilnput
GJ £> videodecoders
© videoencoders

& bin

returned files

fe3 IS, lid, 11.] '¿Z I

File Searcher

default |v] defaut: |v default: v HI default:
after llfl:/20C0 § after Il/:lj2000 g| after U/11,'2C0C after
upfr nrrpnin np-n 11/ 1,701 fl §j| up to n/it;?nr g| iptn 111/1

default: |w| I I Ö
mh 0 Specify Path: 530file(s) found
may 500715112547405151 I I

i^Files Relieved

result reached (500),

a lit 1 1 1 1 File Typs 1 Fat. Date Mud Tied i Date C Bated Lil kToytl | 5IoLjs i DaltReusived
Id cdcressbûok. sql tedb IB v 236,0KB ... \\arivate\va'\mobile\... Jl-Jc -1800 00 ... Jl-Jan-18300C:.., RECEIVED 02N1/10 13:0E
cd cress book, sql tedb L 230.01® ... \privotemob Ic'Ji... 31-Ja -1000 00 ... 31-Jan-1030OC:... RCCCIV3D 03/11/10 L 0:0C
» ans. da ■ 1 8S.CKB ... \\arivate\va'\mobile\... 31-Je -1800 00 ... 31-Jan-1830OC:... RECEIVED 02/11/10 13:06
m sms. do ■UE3 ... \private\var,mûble\)i... 31-Ja -isnn nn ... 31-Jan-lSJOOC:... P.ErEIV-D 03/11/10 L6:3E
1 ujlii.cppe._uiilli ue iter .pli:L L 134,06 ... \\va\."ubilediLia-y\,,, 31-Jc -1300 00 ... 31-Jai 1-1830OC;,.. RECEIVED 02111/10 13;0E
in cam. epp e. romrr center. plist L ? 134,OB ... \var^noble\ibrary\p.. Jl-Jc -1800 00 ... 31-Jan-1830OC:,,. RECEIVED O3HI/IO 16:36
Ë com. app=. .-•.■ili. plist L v 14.5KB ... \\ibiarv\p,ro:eiorcos... 31-3a -1800 00 ... 31-Jan-1830 OC:,,. RECEIVED 03/11/10 17:12
a com.eppe.Ailfl, plist L •: 15.01KB PLIST File ... \llarary\prefarences\... 33-No t-zoi: it . 33-No V-2C10 17,, RECEIVED 03/11/10 17:11
m oarn. epp e. nebilesafari. plist L 1,23KB ... \\va'\mobiletlibra'y\... 31 -Je -1800 00 ... 31-Jan-1830OC:... RECEIVED 02H1/10 13:06
a com. epp c. ncbilesùfcrï. plist L ? 1.22KB ... \u^moblc\ibfory\p.. 31 3c 1SOO 00 ... 31 Jon 1S30 OC:,.. RECEIVED 02111/10 16:36
a com. epp e. youtube. p ist L Ibb.UB ... YVa'\mobiletlibra'y\... Jl-Jc -180U UU ... Ji-Jan-iBJUUL:,,. KECEiVoD U2H1/1U u:ufc
ë cam. epp e. youtube. p ist L 166,OB ... \varrrnoble\ibrary\p.. Jl-Jc -1800 00 ... 31-Jan-1830 OC:,,. RECEIVED 02f 11/10 16:36
a com. epp e. accountsettirgs. pi st L 2.27KB ... W'o •\mobilodibro y\... 313c -1000 00 ... 31-J,in-1030 OC:.., RECEIVED 03/11/10 13:06
i com. epp e. accountsettiras. pi st L V’ 2,27KB ... \var^moble\ ibrary\p.. Jl-Jc -1800 00 ... 31-Jan-1830OC:... RECEIVED 02/11/10 16:36
is com. appe. nips, plist L ? 1.12KB ... \Va'\mobfediba-y\... 31-Js -1SOO 00 . .. 31-Jan-lS30 OC:... RECEIVED 02'H/IO 13:06
a uuiii.cppe. ii^js.plbl L 1,12KB ... \vary 1 iuble\ibraiy\p,, 31-3c -1300 00 ... 31-Jai 1-1630OC;,.. RECEIVED 02,111/10 16:3E
m com. epp e.areferences.datet me, plist L ? 100,OB ... a '\m o bilei, 1 i tra' y V.. Jl-Jc -1800 00 ... 31-Jan-1830 OC:,,. RECEIVED 0^111/10 13:06
Ë com. app e. :r*f stances. da tot me. plist L 100.OB ... \var^moblo\ ibiary\p.. 31-3a -1800 00 ... 31-Jan-1830 OC:,,. RECEIVED O3/II/IO 16:36
a com. epp e. weather. p 1st L •: 1.0KB ... \var^mobIe\ ibrary\p.. 31-Je -1600 00 ... Jl-Jan-lCOOOC:... RECEIVED 02/11/10 16:36
a com. aap e. ncbilephone .settings. p ist L 88.C6 ... ttva'\mobiledibra-y\,,, Jl-Jc -1800 00 ... 31-Jan-1830OC:... RECEIVED 02H1/10 13:06
a com, epp c, «cot hei, p ist L ? 1.0KB ... Wo'\mobiloditro-y\... 31 3c 1SOO 00 ... 31 Jon 1SDOOC:,,. RECEIVED 02'H/IO 13:06
in com. epp e. ncbilephone, settings, p ist ■ r 88.Lt! ... \varimoble\ibraryYp,, Jl-Jc -180U UU ... Ji-Jan-iBJUUU,,. KECEiVoD U2H1/1U ib:3fc
a lexiana ja jp-rlynamr-fRxt.raf L n.OR ... \\va-\mnhilft(libra'yY... 31-V -i8on nn ... 31-lan-1830OC:... RFCFTV-D 07/11/10 13:06
a cr-dynanic-text det L 2.41KB ... W'o •\mobilodibro y\... 31-Jc -lOOO 00 ... 31-Jin-1030 OC;, RECEIVED 03/11/10 13:06
m er-dynanic-text det ■ -, 2.41KB ... \varimoble\ibraryYk... Jl-Jc -1800 00 ... 31-Jan-1830OC:... RECEIVED 02/11/10 16:36
n cyramic-teyt.dat L ? 13.501® ... \Va'\mobfediba-y\... 31-3a -1SOO 00 ... 31-Jan-lSJOOC:... RECEIVED O2III/IO 13:06
a uyi aniiL-Lcx.L.dd. L 10.931® ... \vary 1 iuble\ibraiy\k... 31-3c -1300 00 ... 31-Jai 1-1830OC;,.. RECEIVED 02111/10 16:36
M fe il-dynanic-text.det L ? 170,OB ... a '\m o bilei, 1 i tra' y V.. Jl-Jc -1800 00 ... 31-Jan-1830 OC:,,. RECEIVED O3III/IO 13:06
a rotas dt L 32.CKB ... \\ya'\mobilodibra'y\... 31-3a -1800 00 ... 31-Jan-1830 OC:,,. RECEIVED O3/II/IO 13:06
a fejl-dynanlc-text.det L 170.OB ... \vargnobIe\lbrary\k,,, 31-Je -1600 00 ... 31-Jan-1630 OC:, RECEIVED 02/11/10 16:36
m rotas db □ V 32.CKB ... \vanmoble\ibrary\n.. Jl-Jc -1800 00 ... 31-Jan-1830OC:... RECEIVED 02H1/10 16:36
a tocknore, plist IB T 4.42KB ... Wo'\mobiloditro-y\... J1 Jc 1SOO 00 ... 31 Jon 1830 OC:,,. RECEIVED O2III/IO 13:06

Page 11 of 11



Click to send permalink to address bar, or right-click to copy permalink.

Un-highlight all Un-highlight selectionu Highlight selectionh