From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============1938307108353976146==" MIME-Version: 1.0 From: kernel test robot Subject: [peterz-queue:sched/core-sched 16/29] kernel/sched/core.c:4049:6: sparse: sparse: context imbalance in 'wake_up_new_task' - wrong count at exit Date: Fri, 16 Apr 2021 12:45:59 +0800 Message-ID: <202104161244.Ge2IZFXI-lkp@intel.com> List-Id: To: kbuild@lists.01.org --===============1938307108353976146== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable CC: kbuild-all(a)lists.01.org TO: Peter Zijlstra tree: https://git.kernel.org/pub/scm/linux/kernel/git/peterz/queue.git sc= hed/core-sched head: aa17e873016be2a89e8cb3c0fe044801874b3ea5 commit: d264c7c1defc2280317a692e046aad5e7676d021 [16/29] sched: Use raw_spi= n_rq_*lock*() helpers :::::: branch date: 12 hours ago :::::: commit date: 12 hours ago config: h8300-randconfig-s031-20210415 (attached as .config) compiler: h8300-linux-gcc (GCC) 9.3.0 reproduce: wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/= make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross # apt-get install sparse # sparse version: v0.6.3-280-g2cd6d34e-dirty # https://git.kernel.org/pub/scm/linux/kernel/git/peterz/queue.git/= commit/?id=3Dd264c7c1defc2280317a692e046aad5e7676d021 git remote add peterz-queue https://git.kernel.org/pub/scm/linux/ke= rnel/git/peterz/queue.git git fetch --no-tags peterz-queue sched/core-sched git checkout d264c7c1defc2280317a692e046aad5e7676d021 # save the attached .config to linux build tree COMPILER_INSTALL_PATH=3D$HOME/0day COMPILER=3Dgcc-9.3.0 make.cross = C=3D1 CF=3D'-fdiagnostic-prefix -D__CHECK_ENDIAN__' W=3D1 ARCH=3Dh8300 = If you fix the issue, kindly add following tag as appropriate Reported-by: kernel test robot sparse warnings: (new ones prefixed by >>) kernel/sched/core.c:816:38: sparse: sparse: incorrect type in initialize= r (different address spaces) @@ expected struct task_struct *curr @@ = got struct task_struct [noderef] __rcu *curr @@ kernel/sched/core.c:816:38: sparse: expected struct task_struct *curr kernel/sched/core.c:816:38: sparse: got struct task_struct [noderef]= __rcu *curr kernel/sched/core.c:1936:33: sparse: sparse: incorrect type in argument = 1 (different address spaces) @@ expected struct task_struct *p @@ g= ot struct task_struct [noderef] __rcu *curr @@ kernel/sched/core.c:1936:33: sparse: expected struct task_struct *p kernel/sched/core.c:1936:33: sparse: got struct task_struct [noderef= ] __rcu *curr kernel/sched/core.c:1936:68: sparse: sparse: incorrect type in argument = 1 (different address spaces) @@ expected struct task_struct *tsk @@ = got struct task_struct [noderef] __rcu *curr @@ kernel/sched/core.c:1936:68: sparse: expected struct task_struct *tsk kernel/sched/core.c:1936:68: sparse: got struct task_struct [noderef= ] __rcu *curr kernel/sched/core.c:4759:38: sparse: sparse: incorrect type in initializ= er (different address spaces) @@ expected struct task_struct *curr @@ = got struct task_struct [noderef] __rcu *curr @@ kernel/sched/core.c:4759:38: sparse: expected struct task_struct *cu= rr kernel/sched/core.c:4759:38: sparse: got struct task_struct [noderef= ] __rcu *curr kernel/sched/core.c:5641:14: sparse: sparse: incorrect type in assignmen= t (different address spaces) @@ expected struct task_struct *prev @@ = got struct task_struct [noderef] __rcu *curr @@ kernel/sched/core.c:5641:14: sparse: expected struct task_struct *pr= ev kernel/sched/core.c:5641:14: sparse: got struct task_struct [noderef= ] __rcu *curr kernel/sched/core.c:6300:17: sparse: sparse: incompatible types in compa= rison expression (different address spaces): kernel/sched/core.c:6300:17: sparse: struct task_struct * kernel/sched/core.c:6300:17: sparse: struct task_struct [noderef] __r= cu * kernel/sched/core.c:6507:22: sparse: sparse: incompatible types in compa= rison expression (different address spaces): kernel/sched/core.c:6507:22: sparse: struct task_struct [noderef] __r= cu * kernel/sched/core.c:6507:22: sparse: struct task_struct * kernel/sched/core.c:10192:25: sparse: sparse: incorrect type in argument= 1 (different address spaces) @@ expected struct task_struct *p @@ = got struct task_struct [noderef] __rcu *curr @@ kernel/sched/core.c:10192:25: sparse: expected struct task_struct *p kernel/sched/core.c:10192:25: sparse: got struct task_struct [nodere= f] __rcu *curr kernel/sched/core.c:377:6: sparse: sparse: context imbalance in 'raw_spi= n_rq_lock_nested' - wrong count at exit kernel/sched/core.c:382:6: sparse: sparse: context imbalance in 'raw_spi= n_rq_trylock' - wrong count at exit kernel/sched/core.c:387:6: sparse: sparse: context imbalance in 'raw_spi= n_rq_unlock' - unexpected unlock kernel/sched/core.c:406:36: sparse: sparse: context imbalance in '__task= _rq_lock' - wrong count at exit kernel/sched/core.c:447:36: sparse: sparse: context imbalance in 'task_r= q_lock' - wrong count at exit kernel/sched/core.c:1927:33: sparse: sparse: dereference of noderef expr= ession kernel/sched/core.c:1928:19: sparse: sparse: dereference of noderef expr= ession kernel/sched/core.c:1929:37: sparse: sparse: dereference of noderef expr= ession kernel/sched/core.c:3261:25: sparse: sparse: context imbalance in 'ttwu_= runnable' - wrong count at exit kernel/sched/core.c:3746:9: sparse: sparse: context imbalance in 'try_in= voke_on_locked_down_task' - wrong count at exit >> kernel/sched/core.c:4049:6: sparse: sparse: context imbalance in 'wake_u= p_new_task' - wrong count at exit kernel/sched/core.c:4466:30: sparse: sparse: context imbalance in 'finis= h_task_switch' - wrong count at exit kernel/sched/core.c: note: in included file: kernel/sched/sched.h:1886:25: sparse: sparse: incompatible types in comp= arison expression (different address spaces): kernel/sched/sched.h:1886:25: sparse: struct task_struct [noderef] __= rcu * kernel/sched/sched.h:1886:25: sparse: struct task_struct * kernel/sched/sched.h:2044:9: sparse: sparse: incompatible types in compa= rison expression (different address spaces): kernel/sched/sched.h:2044:9: sparse: struct task_struct [noderef] __r= cu * kernel/sched/sched.h:2044:9: sparse: struct task_struct * kernel/sched/sched.h:2044:9: sparse: sparse: incompatible types in compa= rison expression (different address spaces): kernel/sched/sched.h:2044:9: sparse: struct task_struct [noderef] __r= cu * kernel/sched/sched.h:2044:9: sparse: struct task_struct * kernel/sched/sched.h:1886:25: sparse: sparse: incompatible types in comp= arison expression (different address spaces): kernel/sched/sched.h:1886:25: sparse: struct task_struct [noderef] __= rcu * kernel/sched/sched.h:1886:25: sparse: struct task_struct * kernel/sched/sched.h:2044:9: sparse: sparse: incompatible types in compa= rison expression (different address spaces): kernel/sched/sched.h:2044:9: sparse: struct task_struct [noderef] __r= cu * kernel/sched/sched.h:2044:9: sparse: struct task_struct * kernel/sched/sched.h:1886:25: sparse: sparse: incompatible types in comp= arison expression (different address spaces): kernel/sched/sched.h:1886:25: sparse: struct task_struct [noderef] __= rcu * kernel/sched/sched.h:1886:25: sparse: struct task_struct * kernel/sched/sched.h:2044:9: sparse: sparse: incompatible types in compa= rison expression (different address spaces): kernel/sched/sched.h:2044:9: sparse: struct task_struct [noderef] __r= cu * kernel/sched/sched.h:2044:9: sparse: struct task_struct * kernel/sched/sched.h:1886:25: sparse: sparse: incompatible types in comp= arison expression (different address spaces): kernel/sched/sched.h:1886:25: sparse: struct task_struct [noderef] __= rcu * kernel/sched/sched.h:1886:25: sparse: struct task_struct * kernel/sched/sched.h:2044:9: sparse: sparse: incompatible types in compa= rison expression (different address spaces): kernel/sched/sched.h:2044:9: sparse: struct task_struct [noderef] __r= cu * kernel/sched/sched.h:2044:9: sparse: struct task_struct * kernel/sched/sched.h:1886:25: sparse: sparse: incompatible types in comp= arison expression (different address spaces): kernel/sched/sched.h:1886:25: sparse: struct task_struct [noderef] __= rcu * kernel/sched/sched.h:1886:25: sparse: struct task_struct * vim +/wake_up_new_task +4049 kernel/sched/core.c 332ac17ef5bfcf kernel/sched/core.c Dario Faggioli 2013-11-07 40= 41 = ^1da177e4c3f41 kernel/sched.c Linus Torvalds 2005-04-16 40= 42 /* ^1da177e4c3f41 kernel/sched.c Linus Torvalds 2005-04-16 40= 43 * wake_up_new_task - wake up a newly created task for the first time. ^1da177e4c3f41 kernel/sched.c Linus Torvalds 2005-04-16 40= 44 * ^1da177e4c3f41 kernel/sched.c Linus Torvalds 2005-04-16 40= 45 * This function will do some initial scheduler statistics housekeeping ^1da177e4c3f41 kernel/sched.c Linus Torvalds 2005-04-16 40= 46 * that must be done for every newly created context, then puts the task ^1da177e4c3f41 kernel/sched.c Linus Torvalds 2005-04-16 40= 47 * on the runqueue and wakes it. ^1da177e4c3f41 kernel/sched.c Linus Torvalds 2005-04-16 40= 48 */ 3e51e3edfd81bf kernel/sched.c Samir Bellabes 2011-05-11 @40= 49 void wake_up_new_task(struct task_struct *p) ^1da177e4c3f41 kernel/sched.c Linus Torvalds 2005-04-16 40= 50 { eb58075149b7f0 kernel/sched/core.c Peter Zijlstra 2015-07-31 40= 51 struct rq_flags rf; dd41f596cda0d7 kernel/sched.c Ingo Molnar 2007-07-09 40= 52 struct rq *rq; fabf318e5e4bda kernel/sched.c Peter Zijlstra 2010-01-21 40= 53 = eb58075149b7f0 kernel/sched/core.c Peter Zijlstra 2015-07-31 40= 54 raw_spin_lock_irqsave(&p->pi_lock, rf.flags); 7dc603c9028ea5 kernel/sched/core.c Peter Zijlstra 2016-06-16 40= 55 p->state =3D TASK_RUNNING; fabf318e5e4bda kernel/sched.c Peter Zijlstra 2010-01-21 40= 56 #ifdef CONFIG_SMP fabf318e5e4bda kernel/sched.c Peter Zijlstra 2010-01-21 40= 57 /* fabf318e5e4bda kernel/sched.c Peter Zijlstra 2010-01-21 40= 58 * Fork balancing, do it here and not earlier because: 3bd3706251ee8a kernel/sched/core.c Sebastian Andrzej Siewior 2019-04-23 40= 59 * - cpus_ptr can change in the fork path d1ccc66df8bfe3 kernel/sched/core.c Ingo Molnar 2017-02-01 40= 60 * - any previously selected CPU might disappear through hotplug e210bffd39d01b kernel/sched/core.c Peter Zijlstra 2016-06-16 40= 61 * e210bffd39d01b kernel/sched/core.c Peter Zijlstra 2016-06-16 40= 62 * Use __set_task_cpu() to avoid calling sched_class::migrate_task_rq, e210bffd39d01b kernel/sched/core.c Peter Zijlstra 2016-06-16 40= 63 * as we're not fully set-up yet. fabf318e5e4bda kernel/sched.c Peter Zijlstra 2010-01-21 40= 64 */ 32e839dda3ba57 kernel/sched/core.c Mel Gorman 2018-01-30 40= 65 p->recent_used_cpu =3D task_cpu(p); ce3614daabea8a kernel/sched/core.c Mathieu Desnoyers 2020-07-06 40= 66 rseq_migrate(p); 3aef1551e94286 kernel/sched/core.c Valentin Schneider 2020-11-02 40= 67 __set_task_cpu(p, select_task_rq(p, task_cpu(p), WF_FORK)); 0017d735092844 kernel/sched.c Peter Zijlstra 2010-03-24 40= 68 #endif b7fa30c9cc48c4 kernel/sched/core.c Peter Zijlstra 2016-06-09 40= 69 rq =3D __task_rq_lock(p, &rf); 4126bad6717336 kernel/sched/core.c Peter Zijlstra 2016-10-03 40= 70 update_rq_clock(rq); d0fe0b9c45c144 kernel/sched/core.c Dietmar Eggemann 2019-01-22 40= 71 post_init_entity_util_avg(p); 0017d735092844 kernel/sched.c Peter Zijlstra 2010-03-24 40= 72 = 7a57f32a4d5c80 kernel/sched/core.c Peter Zijlstra 2017-02-21 40= 73 activate_task(rq, p, ENQUEUE_NOCLOCK); fbd705a0c61845 kernel/sched/core.c Peter Zijlstra 2015-06-09 40= 74 trace_sched_wakeup_new(p); a7558e01056f51 kernel/sched.c Peter Zijlstra 2009-09-14 40= 75 check_preempt_curr(rq, p, WF_FORK); 9a897c5a6701bc kernel/sched.c Steven Rostedt 2008-01-25 40= 76 #ifdef CONFIG_SMP 0aaafaabfcba8a kernel/sched/core.c Peter Zijlstra 2015-10-23 40= 77 if (p->sched_class->task_woken) { 0aaafaabfcba8a kernel/sched/core.c Peter Zijlstra 2015-10-23 40= 78 /* b19a888c1e9bdf kernel/sched/core.c Tal Zussman 2020-11-12 40= 79 * Nothing relies on rq->lock after this, so it's fine to 0aaafaabfcba8a kernel/sched/core.c Peter Zijlstra 2015-10-23 40= 80 * drop it. 0aaafaabfcba8a kernel/sched/core.c Peter Zijlstra 2015-10-23 40= 81 */ d8ac897137a230 kernel/sched/core.c Matt Fleming 2016-09-21 40= 82 rq_unpin_lock(rq, &rf); efbbd05a595343 kernel/sched.c Peter Zijlstra 2009-12-16 40= 83 p->sched_class->task_woken(rq, p); d8ac897137a230 kernel/sched/core.c Matt Fleming 2016-09-21 40= 84 rq_repin_lock(rq, &rf); 0aaafaabfcba8a kernel/sched/core.c Peter Zijlstra 2015-10-23 40= 85 } 9a897c5a6701bc kernel/sched.c Steven Rostedt 2008-01-25 40= 86 #endif eb58075149b7f0 kernel/sched/core.c Peter Zijlstra 2015-07-31 40= 87 task_rq_unlock(rq, p, &rf); ^1da177e4c3f41 kernel/sched.c Linus Torvalds 2005-04-16 40= 88 } ^1da177e4c3f41 kernel/sched.c Linus Torvalds 2005-04-16 40= 89 = :::::: The code at line 4049 was first introduced by commit :::::: 3e51e3edfd81bfd9853ad7de91167e4ce33d0fe7 sched: Remove unused parame= ters from sched_fork() and wake_up_new_task() :::::: TO: Samir Bellabes :::::: CC: Ingo Molnar --- 0-DAY CI Kernel Test Service, Intel Corporation https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org --===============1938307108353976146== Content-Type: application/gzip MIME-Version: 1.0 Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="config.gz" H4sICLQPeWAAAy5jb25maWcAnDxdb+O2su/9FUL70gJnW0v+DA72gaIom2tJ1IqUreRFcBPvNmg2 WdhJb/vv75CULFKinOLcA9zGM8PhcDicL1L70w8/eejt9eXb4fXx/vD09I/39fh8PB1ejw/el8en 43+9iHkZEx6JqPgViJPH57e/f/tjNZ1MvPmvfvDr5MPpfultj6fn45OHX56/PH59g/GPL88//PQD ZllM1zXG9Y4UnLKsFqQSH39U4z88SV4fvt7fez+vMf7Fu/l1+uvkR2MQ5TUgPv7TgtYdo483E2Bx oU1Qtr6gLuAkkizCOOpYAKglC6azjkNiICaGCBvEa8TTes0E67gYCJolNCMdihaf6z0rtgABBfzk rZU+n7zz8fXte6eSsGBbktWgEZ7mxuiMippkuxoVIBNNqfg4DYBLOy9Lc5oQ0CIX3uPZe355lYwv i2AYJe0qfvyxG2cialQK5hgclhR0wFEi5NAGuEE7Um9JkZGkXt9RQ1ITk9wZurGpLyJ0pI65IxKj MhFq+cbsLXjDuMhQSj7++PPzy/PxF2Nl/JbvaI4dPPdI4E39uSQlMeUoOUloaNKrbYJt885vv5// Ob8ev3XbtCYZKShWu8o3bG/YooGh2SeChVS6E403puIkJGIpopkN4zR1D49IWK5jrhZxfH7wXr70 pO0PwrDdW7IjmeCtFYrHb8fT2bVCQfEWzJDA6kQ3f8bqzZ00t1Qt6qI9AOYwB4uoS+V6FI0S0uNk mA1db+qCcJg3BXs0FzWQsR2TF4SkuQBWmbWVLXzHkjITqLg1RepTDXYc5+Vv4nD+03uFeb0DyHB+ PbyevcP9/cvb8+vj89eekmBAjTBmMBfN1qYgIY9gGoYJ55JCuOXg1IY3y/4Xcih5C1x63LV/2W0N OFMe+FmTCjbK5SS4JjaH90CIb7ni0RhUHyUKhMllzmYZtngXw97qPwxT3172hWETvCEo0hZxcVrS KcVw7mgsPvrLzhxoJrbgqWLSp5lqVfH7P44Pb0/Hk/fleHh9Ox3PCtxI6sAaHnZdsDLnLuMGB8Rz WDm33IngdcadGw6ephjD5TQaQ+ENwducwRrlORGsIE4yDnSR8uVKYtdG3/KYg/eAA4CRIEYQ7GPq XWCZD0mQ+yyFyRaG7ZSfLiI3CWOi1n87ZILAyXI4+vSO1DErpDOB/6Qow9bJ7pNx+MO9QiySXkwq aeQvOliYxybn0WPRG5ZC3KFyB/tak3OixJg03qDM8ng547Tq/JtltYZcpXEoSBKDzgqDSYg4LL20 Jioheer9BEMyuOTMpOd0naHETH6UTCZABQkTgKgRySmry0K7uhYd7SjI1ejAWF1K0hAVBTX1tZUk tykfQvTqpJkJuiPWXhn67UwKwGCqCUORY+O2OLUSDZCERBGJBg6/yVHz4+nLy+nb4fn+6JG/js/g ahE4BiydLcQf01P8yxGtJLtUK7RWMcLafZ6UoQ7kRhIJyRwSkAluTfF5gkKXqQODPhlovViTNkty ewlJFkMETCgHfwKWytJ/QbhBRQS+333C+aaMY0hCcwSTw1ZBbgleaiT2sphCfrx2Rj47Nb4cYlkb GFqCqBvKbc0iihwJ1mZPIKWwUxfKclaIOkX5kB7z0sizIDXxuyIgK+R0/KNvTq7k2RhD4Pfipvtd oFTntm2ulZ9e7o/n88vJe/3nu47oVhgy11kjAtxWTuVpgs0qRZXDIDR2izISwv+sHE0JLPPJEba8 JhHj22CxnI1S9EZbk8pKAfKGOhKhkayzOOZEQPF02d9rirBqo8Pp/o/H1+O9RH14OH6H8XDKvJfv sow8d7kO8K9j41ShAjL8aRBCzQSz14YRJIK16WLroVhUJpB3gk9X7lZ6H8M7rwUKwaYTOMDg1wIr oqhpoN7bWCFKn149u3SsrmgH5gPmSOKYYiodQxxbyYPM002XwQc+a43Z7sPvhzOU439qZ/T99AKF uc5OL4wkWVNhuU/aNTb94/jOblzsHg6YDEHE2BDl13kq/fekk65RvavmlBowhkN6wTGnsDlQt3Fh Y2TiEfK1Ewgl3RAOAZesCypur6Bq4U+sWNMQ3LGe+zPwOI1kzQ8OsLASBInbh6LPDkB1+nmEl4oJ llErJUBOxnKU2FDdagBPiIvb3C43neg6ho0IEb70IvLD6fVR7qIn4EAaBwtWIqgaAhFeJmORuQgE qUnW0ThWgiDhueA7qRiPLXDHMaVrdJ0j1HPUxTNF2M0z5eDU3Dy7Axel71DwNb0qFwTaorfYbmw5 oqWuCYIg132HhsQjElg9j8XqHSLDSF1UrYvuWYRpT+nnekdhMGvNh7KudjKMB+go0+VIBDWc3Q4z kNvb0DwtLTiMP1utDWuSiznwzDe7ZGppPIcgVGbKi9Li8wBfgDQN/hrOOXYPzoGMDTaRzWilIPL3 8f7t9fD701E1Tz2VNb4aqgppFqcCfFxBc8O7tQGmxceQGBtn+x2g7BXuctk1zFU/USDLpxqELIGK MdmjWy6zllEay4Fp1J3Eueu9ZkmQMsJm9slsopRyo+SXBU9UNsl7s/tjGlTqTY/fXk7/eOnh+fD1 +M2ZIcj5oRYxqiS5poxFRJYodkbI8wRCdy7ULqq870b9n5G+ygy3IDKZ7OWwbcYP56OGVCMsDQee sTQtVYlMUVKDu4CyoJK9oS6vzAjoCgpRlSFsUyuvSAg4YARHwzHfXc6YVRrdhaUrSt1NY72P7UJI IWcC0xBWArIu80Ev+bIX4+ruGGRk2FiLjn89QqUUnR7/svxEjjGUFYZQOMUUWQ09BalloVdjOsyG cvzh/nB68H4/PT58VfGrSyMf75sZPXYxiq4Vo5OsDUlAE04zhpgv0jx2t2RAb1mEEkgJHLqGylQx j2mR7uEQ6EZ66xLix9O3/zucjt7Ty+HheDIsda/WaXpEUokCXfhYDfwLte4qXVlIRykz1ALMzrm5 fbkurgjqub3KgIzj2a40rT9DiN2W8pJCWPmZhkUFlPOyqdPDXmovsEFgTbGZMxZkbR1L/bumAR7A eEJT66i1cLML0sD2/gCUpmZ3o53HdP4dvxrtUsONRBC2GxcXQvWrdq2zHUDGkHYR3Rkdr3WHNqps JHw7ew/q0FhGm7JKEJc3TTfU9jgN4NJguMxqcr5sRsaNofIXpJsFNZNNBUxlj7VFdB5D0dMibnBO G1REZVhdo0mFy3FFwth1ZnXuWFxLIxu5+gKsdPWiIMRkUBNUJLdu1JaFnyxAdJtBYmoJAPlwJM+Q BbNMhslSEta5A9OwIo9GsGRnz8rghCTIKEkgQZMBxvKDGlSjarVa3izcrZWGxg9Ws9HaYjesfLIy SeQPIw5HBUuHhAljuRsKcTzTBevH1bByUtUHk3TDyFCEkffweJbxHbz48f7wdj568iIAKiDv5eRR GU/0kCcoQI8PRurUsLcTlw7YCOQvXDjZP/44m9wsrDXX+VbgaBf1VNGCm5POYZFO9F5F6W6wTpYb 3ar1ZruUePzt+/eX02u3EgntdXsVSLWccySsNoPCbPapsxhRyBiF4Ft5j1mMewCop9ZEDFhrMFgT 52JTlE5TMwn7++okirHTB1rq0Bnd4/necH1tuCUZZwWvE8qnyW4S2MVoNA/mVR3lzJ2PQshKb+UR degL1HQzDfhsYsQG8NsJ4yXEbXmIae96B+URv1lNApS42heUJ8HNZDI1CgQFCYz+ZbsYAZj53Oo1 tKhw4y+XE1fTtyFQUtxMKqvLl+LFdB64fCj3F6vA6IWB74N11QTn0+ZmxhDPOk+VbNaC245iYthP vstRZvpFHDQeS1c9BE5w6p0vdt6VsgoDtXzgclEddm717zU4IWuEb8eHpaharJZzYyM1/GaKq4UD WlWzIZhGol7dbHLCqwGOEH8ymVnFib1QfbF+/Ptw9ujz+fX09k1dDZz/gLTqwXs9HZ7Pks57enw+ Sq93//hd/mleMPwPo12HxE6ALIyVRqFEEMgwIZ3NjWhP8IaZy7QOpb5PxZw2EGOb2+lkwy5lhhct EIU6EwKueQECVPYvmTb1IE0kaA1LTdvMp1vIP4Mi/vyP93r4fvyPh6MPsB2/DEMEtxwG3hQa6ixM 2yGFc4jzCrNFYstRqwVcnInrMEsC+FsWE3YVpjAJW68HFyUmAccoqxG/zfAgripFidZkzr29Udns cDdqLl8wjcATGsJ/HAj5FMd+sqRRRX7h1V2098T6wV7vXjXbTUVQhVE3yOqydUyLZcw3OBqoUIPr HByevFQeVyUQyqbxFfZNFjOYQmXY44x1CjDGt12yo/LVzDfOsOk6fUaEEsidX7vS6yY+28mHwOAI e514CYtpQsyCScJy+xzLbEA2mN3ZxfCR1fP3t9dRR0KzvDSKRvUTwkDE+7A4lpl2YqXlGsNVS35r 1ZMakyLZt20wSpjyfDw9yTupR3mL++XQq76aYQzqbTjtruCvCD6xW51m9waS3bVRZKcLNUMrY10T PWBLbkOmuyfGizoNg5TI9RjLQOfz+WplFOs25saFEdswcsA/C38yn4wglhOneJ9F4C9c6c2FQvaM wQ/TYrGaO3gnWy3MkPU6d542C1/LVxHEPV5gtJj5i2ssgGQ181fO4dqoro1O0tU0mDoHS9R0em0w pDjL6dy1PSnmboHywg9899VDS5ORvXBWExcKlkMcAxfBHTNzlPLSfv3WaZslUUz5Rl9purx3x0aw PdqbZXCHKjO37dHPfBFUzpkZnGtXimns4hQMvXJwFWlQC1biDUDcNrJPZpPp5LpSK3lerpNglPt+ ddVaQpy6Nlts61z3JVze6Zpr4vLhojmuhdUoQ5BxOMZ2FFNjEzpohJ1Q6oBiFhbIAV/HwdYFLuwX wxYCTPuasOuSwiFPzSesF5y86y0QdqE4jcieZlYX9oIUqXOtFHILTEYRdTANHMi9fB3FXNOkaE2S xHza0oknnxuyIhxDhdZDtA4nbyvcS9rTCH44MHcbkm1K125F4Y17V1BK8MjlYzdhWYRsXaC4uk6H +Hzi+9d2WAZaqxttqDbZwgZD+PEd2JxXOYqaCmk4cYeGtOK6jDGnaOF6m6WPm3pAYhiZ/t20QEFI zNJZPytRjofjghBj+w2grIjke0Vq+mITv1rl6WoxqdxYFPHlyix7beRytVxewd1cw9mFpQNvlaQ2 3vJkFqrwJ4EvWbtyV5NQpJBGp5UYmaKEqE8rTAs3PiwDf+JPryCDkcXLp/wsIzXF2WpqZwUW2e0K ixT5M3fYGJKuff/fkArBc5W9v6MfTTnrJ/oOCr2NV6abyZ18V7QI3UyczSiL6DZDYNBugTYozfmG jslLiKBjgpI1SpDbwQzJrt1JWNQVnk4mrrTVpIrLT1Tw0i3zmrGIVmNSbyD2EFdQM4loQsEaR3nw Bb9dLlx+05KjzO7G1LoVceAHI26AWGHJxoxso/Jz9X41Uc7YKbQm6R1yJyUkwL6/mry3PsiG5xPz zaiFTLnvz8ZkATcSI16n1Jk+WpTqh3sOmlaLMqkFHz1LNCOVXaq4xdku/fdOESTnqbrqcG9MBPWx mFeTxZgo6u9Cvpl9ZyL1955mo4xojdLpdF7Jhb+7NO2P35lyH4nVsqqueaU9lE3++4c9xf50uZq+ S6f+plCh/gtSjpXzcFWcPbpgMqla5zvGC2jeszlNNXfvtEYur8+wrOm78ubYbJeYmCKtxUjiwWlC UDSG49d2kAsfcuR3hOIije0uqYUtixjy32k/NrlIq9ViPhtRYc4X88ly1L3eEbEIAleNblG1BYE7 irKEhgWtd/H8/QhfsE3aJCDvzQrV8Lwayfvu1OMjA9mUi9bzKw1rE8iaZboAHmLHkJDo+bPBJBpq p4cWxkoMG4xK58AMez5WY0NIpMyeU9NCm1byJboQ5uPQttdXLZeLmynkFLmgAzkAvbq5WY5hteOo 830xwj1Fq9lQHtVqCiGi229DDGQE5VJEXC7QINpRq2bWGCyP6LhE20p8uukDC7IuE/lhxsg6CyLK cZbqZAT+apwCVXkAlpGT7WCsbphcYd4QtGvtKatU/xlVU47hJC2msENp2ecMuNV8OSiz8n3abc0A MyJGsV1N5nIRYPrX96xg8uNTefnDouEUEVoGq0mzC3w4j06d9REbnUcSLabuc6gjYm2/RW71gdy3 9e2JrJLpzNWS0njwMcHixqEbnKKR5LgRt9gF0ql0ix6iF/MrOlEE7RF1Xr3rPZKfZ3D7aFxYFSmd DYol1XHfHE4P6vEb/Y15/SszGeE7gdVP+f+bG28LnKPCalVqKPj7nAd9aIH2fVBzHesgBpB8DTIY UGAXNcpdE+reN7e+7CzH8hfZzLFX2ELqjM/nKwc8se7GXSq9POl2Xfvou5Y/DqfD/evxNHwEIszP NnaGYPAfzhL11i7j8gtC+G1StgQdbLMfwoCuA8v3yZH1oWOZ0eoGvJ+4NT9IVk8SRoHNC6Rgfnll lEQQcdU3uvJJ5sf2Wvt4ejw8eQ/9m56mU6RerGHTazaIVdB/PnIBm9/yMvXth/Pu1BjgL+bzCap3 CECZmeSZRLFsnW7duIFGLXGsq31z1MhMWVGXqAA5Zi5sIT+5T8mFxKkCUgmSRc6PdkwyxHP5+Hon eY1qc+++4jVFEsFq5S5DGjIWO2+i9cOwl+cPkg1AlC2oFw7DlxWaEeQrU38yGShOwyvHIuTaEipG PhzXNHxTc+xKnhu83ToygMbOD3jSmO6uz4pxVrlD0oXCX1C+rK7qNsTpYuq8S2kIGtf6SaB1s9FX 8VfWNEJZh7fyzdy7ElybXfGDXZTudGj5JlGIyqiA0/3R9+dB972qg3LsWDYvu3LulshGj3KBAOSC XaOXR1sv0B/oNuZJneRyxnE9KhqaxQmpnJL38KOiwC9SgaurI7qmUJSxwrHZQ6KW37iAPLev5Q2w a3CfTp7Tvga6px92qOg7ACyKRMX5wWoz/c4o6j0ZyOpNlLia65erWyvsmtDmJfVAveo9sTWq+edB WCnMZFhDOc2G8UR9w1kOI4N6TSyXCOzt5ET9Ywp2oZXkV7Wd5733GBdM82x3fJ9pntJa/yMPZnIv odLBQ24uUB8u30vqa3Anhoui90/IKKT+NkJfWMa9LwlMOvMJjgaA5x1wU/8AUuS85NWisD0pWBxb vMKBEFYepf8FAXOqC1D/6xKUpWTku/MLYYhmU1dPt6O4fHc4wGAwCDNR6zAVzTdEFXLNt0LyjZV3 P55hyqd06uGJXYHIj5RSlNUzd4HToWdmSMZFMKvMhHh0/nYI6Fl/L2B+g/T/jF1Jd9w4kv4rOs2b OfQ09+XQBybJzGSJm5LMTMoXPpWsKvu1yvKT5W73v58IECSxBJhzkC3FFwACIJYIABG4N7UemA/c vYbaF0rhpzV9lpby5GdJik7fpJzoGymUk1SBPKYnnzww4SyFkzIWYVIQELYJREMwvRd1LqrEIlqf L02vgheoNp6eDo96oq533U+t45kR5XRTRZUmgMWzfIQJjJzCdRNH/KjTFzqdu55de5w8t/R7c05K XJeTdtigHdhtEWiqRiZPgQoUGgvwcZGJ1XmYB0/18/Xj6/fXl18gNhaefvn6nbrgzb7baTfZmpBp Web1gZ6AeQmmK1srPImhpSv71HMt6qrWzNGmSex7NpV4gn5tJD7lB601xqoc0pb7w86Xo7daRi54 ctIzhdtDjo57sy0fOXn98+3968eXv35I3xk0vEOzK3q1akhu070h9wlNROmVMpZyF+sdHcbW77x2 PhbV6+53dCebdJG7//7r7cfH63/uXv76/eXz55fPd3/nXH8D2+YZWuV/1I6Sooua+vkljizHmEXM iZIynQTOvMovjvzBZD1opoxzYEGM0CfrfMhyn1fwgQ2FNMolOKRBgy6yycjp3h1kSldUfZ6qRU6q tjbG818wR3wDZQ94/g4dA5r66fPTdzZxaHdCnZRyG0NynzQdKB66ydl8fJm6LM9c+JaiR4KxN0gV KxPRq20hcdcJCkG3PXTfk7Ep+oVsaq507MAUfRo0ksjECHQpPVd2lIAlq0o6SU1lNLYsT3s1bXFX Pf3Ar5C+fft4f3vFuHGZ0G5CuskOo5V9hIfpLj5MkQXpSYwgjPJdIio4jHjuURkrpQuKbMGF5aJW PU4FfB1QRhYwx0c0nuiTNORQDAyglFVojWXZytTJEtvpRGW5RHKTssiDhhJPTXovB81EapfaUdEF lqOQ2aaDWkA1FPSZNIIDBrY0lD0NT7mIT4/1Q9WOhwet+yRVJnUVYXXQt3NQrHWNRf72/e3j7fnt lfexHzIz/EiLNNL6Mg+cwVKagI9HqZJ83NGRiVaG7hEGAHpP1v2pKeV8V4dYIeeKvg10VMNbLoYX 4cfft3fPr2/P/1TXmvwbC/3QHh8xRitezq/zHqPpjkBidQGrtmrRT+bjDfJ7uYNZDebJzyyGCUye LNcf/yvOaHphgnBFjQYmZRBDp5c6MyfAetL16KTJo/P69hIrqtkrQ2VOUpwe5N48qUvSt11I48VW qFoUTkZll8OtVWObgjX89fT9O6zFbJX9rM9SLGXoDQPzhTdUm89jqmzTVKNQs2vSSnHppjW3x/8s mzIGxCqJK6mcw+Fk0BMZeiyvmSJI2RyK9KI10i4KunBQqF1SJX7mwNdvdmetaH0TU0Ifu1SK84vE a5rFrqcWo84kU9tW2bjnnmNynA3q0y0KGKO+/PoOQ0KaJqY8FwcPuSacbnCG5Sx1qzf+daSVIqHj WVoqRneoPdnpJApVcHfQknG6KiTBRLrIchiPffW8+7ZInUi9WSmoDUqzTiNpn91o7lPxqanV8bHL Qst3Io0KctvVVR3p06GvJrBRHZ36eBuFrtqhNM8GXnO82RIFFNmx9a7CgCigN91Xjtg8pDmu16l/ qIaIjmEw4bp/hTS22Om2OuCqKI6l80fiqy2rrPY1tfnQDuhIiHPHdu3YNvbsaUDY+oBIXTeK6Es/ U7ctuqaj9nOm6eOE94ddsZJEZVhtLl/fP37C6rc57SeHAxi5aqhOSWBQu85SOCYy4zkNi7DCirH/ 9u+v3FzQFJmrzXXsMescT/RLFxH7Kq0BK2S0F1eW7kAH1iakEqXtXp/+9SILylWmYy7adgu9k8J7 LGSsluUr0gtQRDS3xGG75sTUfofE4bi0SJF4bVBK4VomwDYBZgFdd0xPtKIt89EhTkUe36JGmMgR RgbRw8ggepRbngmxQ7Gny51iTsG2yFl8F/HwYSWOSeeGjjTniaix86pMLNK3dghEMJd96sTkRXuR q+oDV+wYIrbcCzPBTBBTjSb150bxExNxvnDKmW931WTi+dzELWPriRVukougsezu3LayhSzSdc9t iokFWhEaJksmXFh+uOaaZOm4S9AsFzaY+U0s9Ig+S2oVB1he9CVQaHYd5iAvh3D2wX3cA+4xgqZn BcIgmJMkaR/Fni9d4ZqxVL0dpuJXx7J9KiUOOdKjV2QQB6tEJ8RkdEend7tOr+xEXISqkjrh5A2J dg9OOAwDVRsOGdyOVK5j9kDIP6tznD5fjuR9ZykS6aCQ78/4QEpyNmyVz7mi+0NoeVsNzVmIpmOI I2pOMzJfv6wkX865dc2dbL54qed4Gnxb52f3bMXgODPABdABVHAd6Ub7jBhn0rUw1gs2ecreDXzq 4HFlSD07cEpdsizv2f4xa1Yv8AOyuuzWsY5A1/Fsn2hRBsQWVWGEHD/cEBY5Qtcnc/VNxYGpYCrO jw26qsgTkJdulpFZ7VwvJD74ZHaEes9hg2Ba1TxiXpgvIBBdrvctWS+ZCzv1MOH5G1Ke0862LIds hsmI22r1LI5jX/InOtV+H+BVacP0rSwq7M/xUkgXMyYi3w4/FnrotPrpA7RvSrlfgkRloWvTZozA 4tmU04nEIFmHK1Khb+ZmWuTwzYkpXVbmiI2JyfsCIocturIKQOx4RDCuJOvDwTYArgnwzIBtAALH AIQWXVeEqM67cBx7UgrQQ+kcuzQMtr/bUIz7pCZ2gJcs8MYkQe+H1qbKTOGfpMDgticyFo7C1or+ kzOYdYFDVgijm23Wh9+nlxa3GduHNhhGexqInP2BQnw39DsdOHREAbP3CF16D/bquU/6nMqu9O2o q0jAsbqKaosDaGHJRlMATvQ/fixa68ixOAa2S3SvYlclOSkCIG1OX8TkDLhfK8+AC9RHxKD9LfUI oWFyPdkOFVgPI3LD0k8ATXoEpSA5ERBbcHwTEFI15ZBBTVS56IhsCMZkt54gyrATOECLICYaBByb nHgZ5NzK1TG0hOcEJlkB2hqDzHGXmhQRcMjmRSSwAn9zAWNMdnybx/CkisgTU8qVwODaimUvY+SO pcASwCxF1j8I3NgAUB2fAT7R7xkQEwNoko/uZlXautvLeJ9K7pILue0cNwqoKp1CmKBcHYA5cBjI /lMFtLftyhDSqqjAQPlHCjA9HKpw66MDHBEDoYqoWQcsepJqKDgKb1Qo3upPABNdA6ikDLHvuMQn ZIBHrtgTtD322jQKXdLqFzk8enTXfTrtqxad6a2ohTXtYfhudxDkCTf1JOAII4toNARii2ieuk2r UPSoXau1j/xYare20q77qYmuFa6xGxJ2u74j1ogOFDxiOgYyNZ8A2f1FklPyQ2dVDlPXdl/MQYkB +3xDdOBwbMulCgAowH2jrYpXXeqFFVUbjlCdfcJ2LjXjdekRLVPivVKJw9ka/IzDDcjEfd+FPh1W bRWvCoJNxT1LbSfKIpN51YWRs71oMZ5wU5WHxo+oblLUiWOR1hUiRl+bhcV1DFHl1lUj3LIr+2OV UmtYX7U2NUgZnexfDKEMdIHBs8i+j8italStb4gEMbNcetshI1XNDNfIDUOXsCQQiGzCkEIgNgKO CSDbhyFb/RAYyjDye8IEmaCgpmWH4XPcG4oELD9S91AXHuUUd+1bPQZysq1xV6WqncAWDPn5AU5i L5kUncE3eGbK2fuINXpn8qOIMcOnTceqW9/4npnnsrWiGjos2AzjW0Ds6bj+VJDvxs6M82PWhwZj EefteC26nCpQZNyjgcwentgUQkzCHghh4eE2k5hzJxg35UUGvLPI/rlZJi0eZ8zyy/6UP2x9+bw6 T06/myWp15sWBuZI78x8dB7F2DUpxcIZxDOgVVIOzl43OkXztViAurkmj82ZPv5buCYXpCk8cl5j n6NuiizsGMWTXZmDjNfuvsAswPI/+AH+9enj+cvntz/v2vcXfIb67efH3eHtXy/v396ks/w5cXvK ec74TYmqygww3IkWUplq6cEJE1crh6Kn2MTRwDPVW9PAz7LXdl+X9jFF6e2afU98eoksFCnsZ087 zkRavo9FA4FrAqQU8g2QDYcwfkCp58kdQ3XgU1Gw6As6MgdloOTgtwi3RMmuVJ58j53Kcz7P2sgT bWV3oKrBwqDo5OkiFsajWmkVRhZ0bJnYdTt88LkrdpLPabeTWTr5XjxLxRzEWDx1MgOBQckrK5qN ZDMsU/nTTPIF1V2KrxGuuawnkABog4Bdbv7j57dn9l4gD+eg+SZU+0y50I8U6hya0Ts3JLWpGVQu V1Rs2m1936E3CFiypHei0DLFSWQsLOAQXntPZW/GFTyWKRlWGzmgefzYEm1FRtXv+rHs2FkwRVOD ZCFSoQMY7UPAKo9j3KX2WxdUPIfGHPk8Im1ICnRCBoZQOuQMBkQRgUtkY5P+gAgekj7Hi93KPjpr gtR2B7VxOVGvxgxQbdk6gRMbBDgWAVgDc2BlDoDZjE/hFKlUGaRC9vTNTMxriUwt0KYZRKZN4bQs iugTxEAO7jB1j8H2/JA24DlDGAYbo2NiMH6XCRYvjq5UcbdpoUaeTo1iOS7dQnZMvYrH5CJyiiOF 2IOVbum0WC8xr/eODXYFUWb+aVDi3rF5Syfh+iBThGsQ69CcY1LRofgXWJ5+Wf7aOTajpn7vR9QW DEPvI0tpFb4+ysQuTzWdk9ELLwyGrfmxq3wxmPJCIirQ3T9G0CWlaTrZDb6lT8CyEGBrG4tXLtkj TYo2qcQQRrxs3dijTfcJjsKI2jfgeZfVWc2xTcoqIS26tgtsS7xhMV23Fg9mhdCCYkHrtWypqIlO bgEvsHR/YpZ6vhEu5zYBPrkfJeQXEcJFASVybFsk1aGpsm+yhCi+YByDSZE84Z9VO12hmJHknBXS My0ABJZ3o/NdS9sJ3W2esnJ98n0GVrx6654R2U13mUYcQjIFQPUhEIh6682AtvilnReWjqe26LXy bcsx1x5gQ7jpCcapeBs2jSQAPXWBU/d+Vhq1aHPEFHR6ZvEtw0HsIqInlziF1sxCO1JVixnh7hvy RLykMmzOTlMZahH0ziKf6vaUysZEVV2HZutJ7/LSXtk/VJ90k2K+Gjravd81ROT8cKgG7IsBg2U1 ZS8dsa8MGKnjPEXO6c5VTuaOWz5sx2eTC1SegzT7SBDXm9ZLVzIYWHSHXdnQAonIGVHm4VaKjmW+ K+oiAjKbKToyq/6EPLMNsSmQ5lQkQbwrU5nPpsiNRjFdARZ6hmJRyEjg0OVPBsGtjG3HNiZ3DBOU wkQtGkLvTWrf9X2f7NmIRZGhUxm8D4Wgqsx8oDKekIsv3qNZ0aIrY9ciJQIocEI7oSWC5SggDT+B RV9rBBD0o5AUmCHkJ2b3cckRqbqiyQjd4uW0aJqgIAwoCC0ePzJBUeCRGTIoMHxcbqFstiXj8Q3d m1srNzooN19uczGD65Y0s/1lyiKyqGs+KpNDNyS3oeU1R8ZD8b6DDEWxqZ3S1ga99oZgre/ZtFht FPmxIWvADO6KItNDGBtsYYELrEebXr1lJjImt8zik3P1YrMaMiavIq0s6P3q+Yb0xhC/Istkft5i 20eDdaux2v35Ez6MeovtAlMreVlE4YnIWZJBMQ1dK4p8Srp2l59OjxgjQXopBuNM0E3H7OhNETWz WoBAMzRk23v0mxUiC7f6CaS6OGTFO6dqE4ucwRHqbBryqygMQlrS2XbflLUrD2BNWLRMkN4KDCsW gJFDRpdWeMKazgAMQ98OyKcCJCbFGJcxxzjyJqPbuTUsZkv+phSyv7WC2a5hjpxt+9vZO3RHnDDP oAluOnYrbLR3t8ZEqgqrTw+R+8UQZGPlUG1ECfHozic4gc9YatrUStetsNmiyrMiYXT0/ZOiSzLm Y+jKZw9TAs6sHY4c3p++f/n6/EMP+5KJzsTwBz42U4zZrqConULN2jE5D3q8NsTuq47HGNPp+90K rWffAO53GGNk+xAd+TAS3Qg1zsAEPFVqBCFZwlR8LBxpBwwoUyVG6SRsiXHx8u357fPL+93b+92X l9fv8BuGEfshtSSPXhda8vs2M9IVpcmVf2bBeEc92A6xIXqzxqc+2yGEjzBJzKqUnCoh5qSU/30D nSkhsxVTifW+HJRYkUiDtjR8lVOanPAk9ZhVSo9iSHkRX9VFcpvU7IyeiZp9/fH99ek/d+3Tt5dX 5QMwxjHZ9eMjjPdhsIIwIbIaExQQ7H7oZvLZosDSnbvxk2X1Y1/5rT/WPRhrMaUJr2l2TQ4WFqrW ThhnVMHI0V9sy76eq7EuA4onw6e4KwoxtAzGcmvFk9YVycsiS8b7zPV7W97JX3n2eTEU9XgPUo1F 5ewSw+6clOIR79TsH63QcryscILEtagDqDVNgbGE7+G/2BX3IwiGIo4iO6VlLeq6KTFmohXGn1LK vWTl/S0rwHYDCavckp/gWnnuj0mWdGPfSQ8HC3hRH7Kia/Fi1n1mxWEm3s4VvkyeZCh92d9DTkfX 9oIrXQOBE4Q6ZnZEngIKH3d61HYss1iK6ipkCeDOcv0Hi2xYhA+eH7oUWMNyUpeR5UXHUlTSBI7m kqDAbADYpAACSxCETkLXXOCKLfJd45W3Suq+wBiXyd7yw2vuk6I1ZVHlw1imGf5an6EbNyTfqejQ R/Y4Nj3u1cUGCZsuwx8YCL3jR+Hou+RDwWsC+DfpGgyafLkMtrW3XK+WtwJXXoMRcGOonZLHrIC5 4lQFoR1TijvJGzl0fz819a4ZTzsYFZlrEHTucF2Q2UFGaV0Ub+4eE7L7CSyB+5s1WGQ/lLgqUnaF RT5wMLNNastmPfMoSqwR/gRzI9+T5hGdLEm2JW32kB3Nkhf3zei518vePhjkA72qHcsH6I4nuxss ehdA4+8sN7yE2fVWNWZuz+3tMrfIAdYVPfQYGIhdH4b/HxbDEiMxRTG98yuwNzVGhxg8x0vu6Q0E ndkP/OSeCpy4svZZM/Yl9P1rdzT1/r4Fnsxyoh4mjO0m5KyeW/V5QjYO42jx/VQSPZ3LR65YhOP1 YTiQysql6EAZbgYc17ETxxQPzHxtDj1uaFvL91MnnKwDrrcpKpOYfHcqsgOpOiyIpHUV3z5e3v94 en65271//fzni6KApVnd6YZCeoSvj+dDqBy7yvifl1cg1VoY3Enbh7UAprWyjwNyT11nOg+a/oC6 04hvYFHncszGwbdyjkWLLhVZO+CW4yEfd5FvXdxxr63m9bVczCJjD0U9ve1r1yN3m6a2PiVZPrZd FMgWnQKSgTaQB6wK+CmiwFH6GBBjS36SfiY7LuUPMaGoSZK9oj8WNcafSgMX2tK25JNdxtF0x2KX TOfPYUBtjRBsij6loOGNQugdbJ2R9MdibLAa71tPHaF4z7oOfPi4kWbMYZI2s53OsmmvNGSa3rCH KS+ph8D1TMWLbKF0+iuhWWsAMFng+DLKAmxnl9BXNToB0A1jNnarY9ZGvqcYJquxJtWRkzEr0l40 TxpKW53S9nA2jctBsXmAsBdusOIWKpKPQ+T6YaYDaFE4cuwHEXI98nKHwOHJXWCGqgJWCfeBvhg/ M53yNmnpdxE4ByyJPl0ALpauT6WVVNu87tl2yfhwLk73y7X5/fvTXy93v//84w8MCq2+L7HfgY2Z oXv62mBAq5u+2D+KJOF3vtfCdl6kVCn87IuyPMH8rQFp0z5CqkQDwFA+5LuykJN0jx2dFwJkXgiI eS3tiFI1p7w41GNeZwXp8DiX2LSdlGmW70FXz7NRvsADCMb54htEtJcl8OC2AkrTF7LTif5hvszh zIngKZAR+kmwQO8G0e1svo4qpmLX4egUxa4aD0Pv+aJaCnQ9mA0Q+f0JiVblqMo1ldxvlh0ISQww pVz17gOfGsjeySq/e3r+5+vXP7983P3XHVh26tN8Qvug3ZeWSdfxZ2GIGmO0sxKfk5YYV9FX/L7P HN+lkFYO/7gCxnflV5YH9uB4mWd0Dvp70RpLkuFhokUJxiA5ssoKzpd0NzOnombNGDvTtxI6dwbS oQ4EJlhJ/BsC4CMjYjjnFRLubmoYFbNpxvgtUEqei+9Y4f9R9izNrdu8/hXPt2oXvdXb9qILWZZt NZKliLLjnI0nTXxyPE3s3MSZ6fl+/QVIPQgKcno3cQSAFMUHAJJ4pHxYuYZoNg9si60YpNQuWq85 VG0XxXZjTFJ0fDG7m/LbZB7nOqfpqq4FsVoI59PH+QWYRy1oFRPh1sp2GV7JIjXfZBmTPouAMVn0 JluLPyYWjy/zO/GH47dcoQyzeLZZAB/t18wg6zBmmD4rC/XAhRwtZrVFmfdVjTUTr8KbON/WFw5N 7pvrfdfUC2qWdqaET3t51ARyYM0joKt1kwENE6WbynFIaODe1UxTTOSbNU3dtp73RMkqmffvdFa6 dIaHLmBhVcIesVrptQK+DPmMkhusvT9XsMYu6ruKY/x2eMRUcFiAkWNYIvTw/G2gujAqNzvaZgna 6/E5JbQwpIwEbkAj4G7w5LfH6U2yprVEKzyGM2EJPN2bdUf5ZhmycdMSdAGIwjQ1K5JXcQbsvgAR LigQ+n2Zr0vlt9rqdQ2s9+kxXpwtzPbFaQwSZqB98TeVkJmUWMbZLCkHB3ZB491LWApKZj4QUgIJ tsk2TOcDmRcS9Da5lweewwT3HFNCzF2YVrrjpXpdfCcPXXsNvS+H7w2RIEEPv4E3JVVMX/NnSJKd I6i6S9ar0JhNN/FagJpn5HdGTBoNx1SVeNZHVmHW+Tbv1Qc7viurCFTgJMpgoIwPyaAPy37rsvB+ ATrRUG3ANuVk7BVLojJH19GhVuBJTxkbqwITRidyElD4ukooIC9JUl8EgY6AjswwCwk/1MCwMAY7 uYirML1fc2qIRGMmzchgmDWQ7H50eCvdeTSMq7HWizRcyxPQqI+4F60o676tA1/9NhSVQ18mwkR1 JSlSH00PlcFgeqlKg0mLVXE4kDlRYeMU82LGw0wC3lqkV5hIOZS6BRc2XpSEIuHu+2TdoDBUf+b3 +AIiMjW40Y/6qk62ubHOYTcYU51dglewzoeYbbXCRHlmxiYd2mPpmDb3bl8Il4LvkiTLTW60S9ZZ jyF8i8vc7FQdfT8HMdpfwiq+xX61mQ3J47QgSaw4+d6lZ+N0EJliDpeznsCzhcFuM58nJCOlWZNZ qHUVa1K6M7QbARvmVZRg2toK1D+16de/HikYI5kam2XaiVhxV4r4FsRuRuRMDVb7XXbCQoH9LM2j G+YN6P68N5N7YwEMsT1YW1TeFzRNn3J5zqLfxfx3rHO0On9cvkgChjUNxRlHnJhD13Xf34L2mJ4h ikB/yfVtSIcv0mqRcQgY6LAMBR0CipYM51qDJFU1tQfqn99FmVhF/AuYBGQ9mgX+0rugDpkl6SwO NwNzpecuBqANFEyCMk9Zbzkcy9teL6/ELQVk1Q3fnh2oBdwhltYlGfERbeFhRgLndQhMTF+uMUpL pjt8gbZZJdFNH9J642iJncTl+Pg3p/i3hTZrES5ijMC/YTN+ZRiCQK0a7ZWihfReNjzf68Lr+M6Q w/ikToE42F7qQixG6i5NUm4dPStRB1jDysCMthFohsu4zbiGyl0v/oAsxrlPS0S4di3Hn3JSTuGF GxAHIAXFiGau2TTMQa/75nRQf9J7cc/TxkCXlmV7NhucWRLEKWZ+dy1q+iBR0lGRWwwd1jFaiWdL Hgec6u4eLdSy+z05aBAvsSqrk9MrVcOH47lLqgHvG9Ue9PP1+p0AYJ83qqrxvjUQ56zB+20EueF3 03PC7pN8s9tqaO+orEUazjyUoHGCBP2UVT0kkemaVQMj2/GENfHNBtEjVglrbWmHXjGbOxOrP4pp 5fpT3m5azQ3lLzJUaxWFaFHdq7ZKI39qsy5xqlrGC6ZBDPrjtIvA59L/SiyeSwfT/mcmwrUXqWtP B1tUUzi7XZ8njb6f30d/vRxPf/9i/zoCtWhULmejekP6idmbOKVv9EunBpMMumo8cOPAbxIkvp+e xeiodAdjPvQ16KdpzBqZffRe15TVSEmH+S7eYo+NjBmgM/bMajrbfeM7lplr0/t4da/z8vDxQ+aF rM7vjz8MCWBM7mri08vjdoCq9+Pzc19qVCBsluQgWAer/PYDuBxE1Cqv+lO6xq9i0EhB0eH0HELI 7HsJPio2A5gwgl1WUt0PoGnkBoJq4l/JoZSddHy7YCLOj9FF9VQ3ZdeHy/fjC+bDfTyfvh+fR79g h14e3p8Pl/58bbuuDNcCb1G/+vwozEjqSYIsaOgvA4cHsOZcbPumjhLAtk12WTs3Zrhi+YXHriul tycztK3l8tmWVUSz4iKg0ZHaWhC4iqocFvBAHZinNV9FtJ4a2Ny5/ef98mj9h9ba248QrEyv1Fsj gBkdG4MCsrKwTLKuFip83EBTJQFom0ZjJZh4MejQ/SaJpV8CRc/Lrdy5/aE5CmDzmFXfkHPXcjwR e3XXUISzmf8t1k8OOkycf5ty8N1Ej/TUwntxrBrMXPRvbxkSNraqRhAQ990avrrPJn7ANB+DoU2J S02HqCP3cIjphPsCRvT2KYyIQg3GiGPTgoUfudwXJSK1Ha6EQjiDRRzm5TuA+32wDPbsML0mEUaU K4JzeYd3nYQbDYmYsNVmnl1NWJesmqAfHqBB3LrODVclk2yzT3PVZUwnGnAYa8YxQufJKdcMAXuS KZu6oqFYgALgMnO0hCVm83Bfzyym01PDpAYTZ7AVvL72yi2QsB6iGoHLTLsS/VeZwRZ+xgDnwAMm f2hpwSmDY2bFlFvWCPf6cMljmDZKOLMCEO4x9Us4wxwQPuW5STC1uWU/JXbO3Uh5agT7w4AMwbs2 DoqlsSwW1pdj857bTeGoGE+NjpBWfOt5ff7UDgxqnv9CAs0FbI1Zt1XSqDH7sTjnptG10uUuUObO Kpn6y8MFthqv12cNjJLD8WCAExNGHe7zsyCY+PtFmCU02SIl+EqkBRPOI0cjGDsTdtUiyvu6/vFk woZ60WsZkMiOZ/EehC1JzyaJJRnIcdGu+urGHlfh1VntTSpuzBDuMisX4T6jl2QiCxyPYQGzW2/C sYay8CNuieLUZFZ6z3m3m+W9SHkN7tv9+jbrpzk7n36DHc5X62v41LkVHxX8pwQFwxH4mDctOzDC T7aIJsqg2Vt19mezkXUwwdaKRBxOH+f368u0byM4x4igqN0LDtaeFXfGvh1u29P9lZ9BFvaNVQG4 j9dLYqyKsDa+0ypcr+OUNkLdPXVboRS2YCFMt+U8Y/0G72SyMUBqO4OFSPcxORlXca4SgOmG63Ug YjVv9vNCFWnfLa0AV1hony0zbp/ZUWjfcCdbY7iG19AOIGA7o8q1HRi9HA+nC5mdobhfR/tqtx/4 +iysNzK9Lt+XYTLXap9tFqPzG0bz0kNMY+0Lw5lW3Ek4/zqoZp/l27hnclzjjKhfNVTE6QIbKoxZ hbhVHBbG/XJjh05b3dQZbna180n3HnRmp/YAc88bTywmfUeNYRlpkmGXR0mCRqr8vXZlBzdsCD8o 4WhTuQhLvK9snaBbsHJAlMguhHoNLnM5GD4xl4jT+pJin8VChAMJXuse2M8wuwBve6CTcOfRGt64 Vmk+orNLXLCn6biy9yostNYVCKUn1gqC0VQ3bEu3Mha1iVbXScfH9/PH+ftltPr5dnj/bTt6/jx8 XIjxZBPD7gvSpn3LMr6fGSYIVbhMWHsLjZu21A1sXyRs/NEV2plGqXZJBg94PpHmucrqbBBiWHeY Ixr/UPO4rqQb0RbK2CRzVFciQ1Gqqaef9ms4Y4OtYUTiuzQVlIEcSHdDqdjrKkrieUPvH1ssJppH 8ZhGdzCwRhRhlkw4Fgag5W/c9HaoyDnXPwOFFvwu4zXbYibsmoYl8Yo0+Dbix4yJ66dhVTxGM7sF FZHbiISyXd3BzmVtWksowfVyfvx7JM6f71wkd3neTcxLFAQ430w3nZm40NVVVk4YmE1D3aJhGzoA wuKrAm/GShK2UVodYZLO8l3vW8rD6/lyeHs/P7KaY4z2Pngmyb6TKawqfXv9eGbrK0DLqVkjXyMp 2cpwtDi+S8rW5xM+8/R0d3w/aBqZQkBLfxE/Py6H11F+GkU/jm+/jj7wvuj78VG7i1ehRl5fzs8A FmeqNjcxRRi08gN5Pz88PZ5fhwqyeJWOeFf8vng/HD4eH14Oo9vze3I7VMlXpOrC4X+y3VAFPZxE xie8oRilx8tBYWefxxe8oWg7qW8akFSxfk+LjzLVCwAw9W1KrncUdjMrYxi35Fv8h9c16d+/XLb1 9vPhBbpxsJ9ZfDdnQKlsTyF2x5fj6Z+hijhsa2H2r2ZUq0VkTTKeVi9Vj1xemCZtj0z1otxb8/U8 zkJqXq+TFXG5yEuguJoHSFKiga8AYTtUVRte9quKQiGSbWx+D2PH1X38Pt7yt1bxroq6+7L4n8vj +TSYHUYRy3xBf4YR0Qpq1EKEIMV5RbcmGTCKqLGY9NP1fabq5tb8Wt1NJP8vaAYv2WuSolr7thkv iZKU1WQ6drn9UU0gMt/XzyRqMJoV0ivMDgHrA/66+tFEBvyeOLgUy3A/X2D6qES3l9QrxEzFyruF g+2jGQumW1cCN/fTGhaNmXqxkBF/s0gWkoqC68tKkP1cC9W/C8GW6ZHKtwpcgS2Jo22rgAh2lYPu fjW+q5yU7NrZWzhKUj0+Hl4O7+fXw8VYcyHsE+3AGQjB0WC5U8NwvktdPYVxDaBx1BsgCWkigfpV Uw1gqcxo5bMsdAaCiwLKs7i7kVkWwQKRl8baRlOH0lYTDI3GEjp67Mx56BJP9Cws50aMNAniHQol biDi8s1OzLluv9lFf2JAKz39R+Q6+qVNloVjTw8GXAOMCPc1sGfrGY6DYCBMfRZO+LiVgJn6vt3P e6HggyX0r9hFMHiElQIocNiUOCIKXRL4SlQ3sLlzKGAW1uFbG42MLgO1NE4PoKaNLufR0/H5eHl4 QQsLkCMXIkpCTN+xzELMZFiF+hwdW1O7JKtgbDsefaZGTgBxAm57iQjdHlg+94qySQAA4emBnOE5 sHrP+2SB4eAx+AZoXalRc0cwlAUAiMZDLYct8562fawvFHw2vs1IpQmQyUC6ZkBN2bskROiRqPF5 uqO1Tr1gqNZE7jGN5DEdXmWLuoaeTAZSz0QRxm619ypjSgPEo+m9kUQlXm/jNC8aJ9GBLM2rBNQE biGsdmOdAam8svS1mOPeG5NTBwlir2okZhr0iFlbAwwSbjl6fhSZgJ4ayyrYQBoHzEvPRqlAjKtf 3OOBS2CTr8Cs6g5rTIIYTzdMQMBU7yjpy4gx21SQaHNU1uFmKLR3nRLPKCDmUs/M8nnftLMlquRk syY2N2UapH6x3cA8YelR7xXYdmx30gNaE0Fy+za0E2HRuOo1IrBF4HBrWuKhLtvvlRLjXnhOgp64 HndEVSODidlqoWxnKVRlHSIzGcPTpJHne3pf1BlpMkqJ6WgAuiwIeLsIbIvWWZ/e7JrxbETFNbGg C47F+/l0gZ3pkyYtUCUrY5BRaczUqZWojxHeXmAzaMibiRuQVbjKIs88gWsPGtoKlGr34/B6hG1/ ffVF9b0qDUEJXdWuUhxPlRTxt7wm0bWiONB5uno2NScJIzpTFIkJYVLhLb35KTIxtozE09HcHcwl iA1LSnQhFMtC13xEIajHy/bbZLpjO63XSerC8PjUXBjCII6i8+vr+aRv9XkCfeAz0aVZdjpHclE0 5bRKdeVeFHU5w3utO0noVUE2B5XxWh5HBsbA1YNSR0VTsx4WwIOatrxa5FsBUXZ8kiUOn6kW4JP0 HfjsBcYzkee+P3XQPFd3/a2hBsAluyIEDRgUACpwvNLUcjTshDYJns2diR9MA3NjAtAxq6xKxIQU Hwe28eyZVQ0oWv54bNFPN/UqVzeBAkYyIeENixzDu+gBsoTn6QoriHvbyOGBGgAfED4LHFcXWCCl fZumBATIxGFFfFR4Y2okhqCpw70IWD+02po41MdCgX1/bJuwMdmY1bDAJiJQiQNA8Her15aAsooF DvD0+frahNjQGUUPV0dIOvzv5+H0+HMkfp4uPw4fx/+i98F8Ln4v0rQ5WFZn8cvD6fD+cDm//z4/ flzej3994iWvvvimymTSOMMfKKesln48fBx+S4Hs8DRKz+e30S/w3l9H39t2fWjt0t+1AP2TrGQA 1Epl/fb/b91dwKSrfULY0fPP9/PH4/ntAEPVl23ytMJibTcVjtg2NqCArjx54sFGNQznu1I4U1IF QDyfyMSlHfSeTRkpYYSlLHahcEBz1uk6GC2vwWlqumLjkmjPNYAVCMv7Mt+7sP8RPApN8a6g0UPF RFdL17HIhnt42JS0PTy8XH5oakoDfb+MyofLYZSdT8cLOfMOF7HnUVVBgXhej+ezls2eCdUoEk6U fbWG1Fur2vr5enw6Xn5q07FpVea4VHWeryr2ynOFWrtuwA4AxzCiWlXCcfgzulW1YdmrSMaWbmiK zw4ZnV7jFVMD7nFB96jXw8PH5/vh9QDa6id0BmEHuEo8utGrgQNLR+LGPlOAXa+zLDHWUcKso4RZ R7mYjEkIuBpirqEaSkrfZLuA7Ka3+yTKPGAH5Et1+EA+REJiHLEhDpZmYPWzKbI0vKJSr8ZUZMFc 7HqrtIaza7/BNe1qRNbwyOsV4LBR3xod2p2UK98eGRGMZdZ/zvfCZVdEON/gwYLOaFOX2J/DM/Aa YooWFnMxddmlLlFTqtOEYuw6A/mnZit7zKZqRoSuz0YZ1EGtpxHk8maqgOJ9dyN09vWNWoJgwBhk WThhYbGHEwoFHWNZ+k0G5sm2oc80Zt3uFUQKEk1PcUMxNDenhNlsOuk/RWg7JElNUVq+ruynVUnS G6dbGFIvIkZFwJSBlw/xa0Rpu4N1HlJb+ryoYAKQ4SigVdKBm2eRtq2HCMRnj4yDqG5cd+CEHhbS ZpuIAbOYKhKuZ/NSSeLG3AA2nV9BR/vU60WCWO9exIz16xQAeL6rdfVG+PbEIbfC22idDvS0Qumm zts4k+cppAIJG/N9s00Dm+Xr32CMYECI4kh5hDJhe3g+HS7qoJ6RrTeTqe7VKp/1M/gba0pO++r7 nCxcrlkge/sjEVTBCpcuiXieZZHrO3ruiprDyrK8+tS87xqa0a6ambHKIp8kfzcQ9FNMJPmcBllm rk3uUgjc3OYaWF443YdZuArhR/gu0TnYgVVDjul0314O/xg3lPJMZcMf4pAytfby+HI89SaOJuMY vCRovJNHv40+Lg+nJ9jxnQ70TG9VKlOsgVtYtKAuy01RNQQDUrtCR+M0z4uhiqSXJ1dJ+xl8Y2uJ ewItVvrMPJyeP1/g/7fzx1EmKWbksBQO3r7IeQPjf1Mb2aG9nS+gNhy7q+ZWpvuOzqPmwia+WnhO 4JlHCZ7uW6YA9HAhKjzL5i7EEGO75nUBckV+n4Dk1gCjr4rU6iX2NnY4xmezXQIjpSvRaVZMbYvf MNEiauv+fvhArYxVpmaFFVgZZ4w7ywqHHtris8nwJMxQU+fpChg8m/uoAMWN7k8Ki5NMSVRgr5It aWrTbZGCDKUVV0hjm5u6qo4GIHzzZkhChupUSPPeG6Aud8lV82UjzqUOZXVshSENr3xPn/GrwrEC reC3IgTNMegBaPUN0NDde7Oj075Px9MzO2mEO3X5G4V+uXoKnv85vuKOEfnB0xFZzyM7IaWiyIfq xrxdpbTz22/15T6zHRq5puCty8vFfDz29CsuUS703FViN3V1IQ3PJCY4kmv6Lmo71MFqm/puau1a wdd28dWvr81hP84vGIXkS3sCR0yN3bMjbMfM8Noayl6tVsmvw+sbnvsNsAjJ560QZFOcDUREqiJn yiqYwHkTTJMRl1ke5RuSkS1Ld1MrsMnxtYKxB8ZVVli6YYJ8Jjy9AunHThyJcDQHFjzBsSd+oA8S 1wsNPbEJhwclZ8mUA2BYZZjBJuJCCCIePacWVWYWq/uX7VnEy+BD7JU7Yqu7lDYNAHWAbaXWlLcy ySETj7C8RdNufWu8X+jxOsp4HYtQKOPwFSopZFb3atamTBFGN7CfZwNOxCKuWAtehZmVUSaqWX0N qneWwitvsyUfn1mRYKKSXmAdxYdW9yPx+deHtKrtuqKO3bwHdNccDVhntSHoWYR5IdchGjQ6tCSW qF0e91VeQj9WPLKusRtuDScSUAk5u0tCFKZ6oE5E4SxLst0ku8WWmbVnyS5Ou8/hJx3QFbtw70zW 2X4lEk4UEhrsAfNFeRSnOd4QlnMzAmrDFclYtBWj9XCkR+ybVwVZM1k048c+FL3xDk9P7+fjExFe 63mZJ3O2SQ25psqwsWT/r7IjWW5cx93nK1I5zVR1v44dZzv0QZZoW21t0ZI4uajciTtxdbaynXrp 9/UDkKJEkJAnc+jFAMSdIEhsMgqL8ZyAP1ue0DZHxr+vBXoxkA4o8pwL5DK7Pthtlnfy/LT3a1Ea VcIP9OsoU9Rw0rDPHQqzeHC20Egh49TT8oq0yn1hBKYnRTbYfYGR1NakwdQ1rJ6WXFjlFl2Uhktc C42LioFmZoTkFtp51upXancw20fcbEqfAJUnbJYD63JsB1oqaZQcT/OWvOi5xLaEjY0IfWjWyNAX I0d2brGx588W6XBfDW1KKrsfk1yIW52yiu1K07AMg0mpg5m7c8pacjElWQXSCQ/XJtsupPYmFQNN wrRopg/OizqxgyVOCq7nMvQktHfRPRUbl3k25GaF1kDTs4shf8g2+GIwYrPLI5pasiMEXbrI8wTT hpanxXWaZWRDhSnHV4oojG1nSQApV+ienNzy7cBXieK6JsKcJiV9HwDRo76svIBP99a5eoEAA+dL VlYk+3ZqekLjr9pXoaG6Gyh1plBa6/UTSFOSwxsy7JWHojyI8ZMCTUpJFDsAhTRiq1iUw9rkrA2g XnhlmbvgLC0wCZhPZAeNLIRf5XzgLyA5rqlU14C6Itn1o6n2lz1yyx59quzRZ8q2XMMlbF4lYVlr 1/AG82MckLMaf/fGQIaK47EPjIgwmVyEMGWAYwOK/ZAIo8K+KfnR0y1C0B8NTX6O8eAxnC4fQ33R 18bppKBLalzmVrM1hG97i4WhAUkX9+C0tx8tcV4lILHBpNyoWdlD3d9vhfcKmALuDO4qExPM563i CHSyRxipvnOzPdRj0PGMYTPK1heUQO/FvjLVIJmjqz6TQQzD5AdwLnqSNIUC15OPISwyuk054IgF mnH4NPi2KAOnr6rgnM0IZC2GdqthtAHKnxREhWquaRq1MBLogIxJtIlQC2IpWv/eEAp+yIsaZH+M fx6ykQYAj/NuBpZsQW74kQ41rkI4WRN0V0g85P78jAcqOEWP7lnipIcf33bP/bpBXVZpSWSyLIdt pcD1tZcn/KOOwlvsTwFLkIIM2CQu66uBDRhaX/mlMbdeVaaTYkSWroLR1VxhchqycXwAcQoyFZuB 0mIWJsz2OnHvL/7y7pEGdJwUkh/zlqmKWpEHX+Ha8S24CuQh7JzBYZFenJ4eWU35kUah4KPu38IX LNuogokuRbeDr1s9bafFt4lXfhML/Dsp+dZNJMcxH2rhOwK5sknwtw6MijltMw9E49HxGYcPUwzH UYjy++F6+3p+fnLxdXBortKOtConnFAom29JJT01vO9+nR+2V8XS4bES1HcCS2R+TeSsfSOoXjm2 q/f714Nf3MiiSz1ptwTgc0kZ0d2HftazMApywXGZucgTsxjr9qv+6fqq7/xuyzoBtFDhcjBGqoiN stIcA7RYe84LnIFsQDBcnCXHxCpASBbKg5pgMIpHN8iZ9T38VjlUTKFBOI2SoL7ZHdttsn77uRe7 v9W5QgKzFiDYFzNas4apE8XhGixVENIcoy0Wb4lxVmNWJ+tpwKKQ1zPe6oGjRK9WDCC294M+4aIl uFVGRe6XcPTv+07JEEyFt/sbhKLDvnJHmBPiaizjptzywyXisYCr2N5iJrk3jQUcgnLyVFnHLQNc WIslDjGbvHW0xH0i8Cxzluplshg55Cb2tK+wvKnHeP2SEAyOjS7QN3YeC4VOExueFSX175C/W646 xyAeGN68+I7Jto8MdtUSRnix1JIj93KpKGHqWyqnPlg2e5Ezvx99Phr2I3Hh9GP3VNk1Vg8GYdZu szUZ/8LL9eQzX5id+ww96S/3AT8AbR8Pn/55PXRK9d2knTYJhnvprwfYqDs1aeIuvHHkLE6E4R/k v4eHDE4uUrlXT0cMOvYWcDfzCrjTGF78cOhd8XurcvapgtTXILVzHL2aFM5zQJ46pWhY7+HUEuh3 FvdT9u3CJtJ3fLaE25AL7ZiYBofwo1sPnLiGBFriq0esKp6QnB0TtSHFnXGKNkJybhrKW5hhL+ak F9PfmHPWItoiGfQVfDrcUzAfOcQi4m0QLSLeiNEi4jySLJKLnn5cHJ/2YXon4uK4v+8XIy44Am2M aSKIGLj64Kqrz3vqGwx7mwIoa4ZkDES7eboGTn1t4od8w477yuufQk3Rt941/pSv8YwHX/R2jDMP IAQ9Yz6wds48Dc/r3K5GQqueKjBiKEgcZpZSDfYFJrKyS1OYpBRVzoXwaUny1CtDttibPIwiU5mu MVNPRHyFmNGRS5Cn8SG01QrQ1KKSKuSOVdJ5K++fxpVVPg/ZlKdIgdffrhNVEuLCdgB1gqGhovBW pefUEUk7ujCtry/NiyBRDCg/19Xd+wZtZJxAqnNxYxwH+AsO0MtKFI1gTCQhkRch3B9BaAbCHK4r 3JlaYgZKEVglN69qDhx+1cGsTqFs2T8LJV+0Qt9G6WOvDuAyKS0Tyjz0idi251Ffo8iNEzXKMy8P RAJtxDc2P81uai8CGRrfDs2iHTJWDQxt9iUFplVXWdWNDnBozNsy+374bftz/fLtfbvaPL/er74+ rp7eVptWGNLiejcEpttoVMQg173e/b5//fvly5/l8/LL0+vy/m398mW7/LWCBq7vv2A2kwdcDl9+ vv06VCtkvtq8rJ4OHpeb+5U0MetWyr+6ZHgH65c1+qKs/1lSr8cQtSHQKX8OU5aQVSNRGPALh9JI r8Oq2xQpqlhpIp5OFci3Q6P7u9H6a9tbQVe+SHN1YyJvApgJ83ujCvU3f952rwd3r5vVwevmQE1M NwaKGPo59YifvQkeunDhBSzQJS3mfpjNzGVkIdxPZiQargF0SXPzQaaDsYTGPclqeG9LvL7Gz7PM pZ5nmVsC3mBcUuC+3pQpt4H3flAHYeGNI2Er8Rqq6WQwPCcZeBpEUkU80K1J/kMOFd2VqpwJGqla PS6+/3xa3339vfpzcCcX28Nm+fb4x1ljeeE5lQXuRAvfZyoXfsCdSC02D5jSi3jIFAUs6EoMT04G F05XvPfdIxpL3y13q/sD8SL7g/bpf693jwfedvt6t5aoYLlbOh30/Zipburz6d70RzM4sbzhUZZG N+iCtI/WE9MQs0D0D0QhLsMrZkhnHrCpK80TxtK5Gxn11u3E2HeXhJnjW8PoDbCFcqdr2wy3mKh5 yqbQdMKZTDbIjGvigkY71ztS3FznbK4AvQVmetzdnYt5kcuKm1LUQF65plqY2a5nUEkwes3ROOBC dc6u8Sr23I0XrB9W251bWe4fD5lJRLBb34LltuPIm4uhO10K7jIeKLwcHAXhhNsAWMOe6eybgDgY MTCGLoTlLW0Y3U7ncUB8f/U2mXkDDjg8OWU6AIiTAWeB3OGPGd7DwFABOU6nTB3XmVWFOrzXb4/E 2LtlBNxqB2hdchZS7eSl1zQ1hYVw0kHq2fViAZcXl8H6HorcfR8V5QnHIgDORkRvTgThrq6J/Net oWGd7jiLPCMmtu2cuCuqvE7ZMWngXe/UhLw+v6F7xJpG3WnbLh+b+ztHTBQa2PnI3ZXEaqGDzdwF 3tgsKI+B5cv96/NB8v78c7XRsTv4lmIyx9rPclaDrnuTj6cyK4B7aiOGZV8Kw3EUifFLV+RBhAP8 EWIySIEm4dmNg8UKQO6e2ELu0/rnZgmC9ub1fbd+YfgwOpV7zAKTzuaKtWkr7300LE6txr2fKxIe 1Qol+0toyVg0t3kQrnksyGD4DD3YR9JVby8ak2yfmNJ1lRdrXOpe1jtjNbfFTRwLvDzL63Z5k5kR nTpkVo2jhqaoxr1kZRbzNIuTo4vaF3lzmxeOaWI294tzNEu5QiyWYVPoslt49y4B357pHCMN3uX/ GEXhlxRHtzIb8Xb98KI8Uu4eV3e/4bJomINLNbn5lJETfbWLLwx9RYMVizL3zE473zsUTWz1o4tT 460iTQIvv/mfjYENhdlyi/ITFHLX4/+w1Z1pySeGqHFg62MOmLvntM4uzenRsHoMVx5gdTn3EBeF ifDyWlojUOcfz7HWajDjEGQATNhCso3kgbnfYUHFAm5n8ZgkdlFvTR65x/lw6QBWSUBmxj6kaAUz AxaWVU2/orIh/DQf7YxtKTGwrcT4hg/ESUj41+aGxMuv+05LxMNQWVWfstlKcp8clr6ZuTocu9Kw b7xf2uIvTGWQxrTzDYpX4yJUmS9QOBog4BEVkS10q5i0BeU1zwjlSuZV0X06aKRm28frnSWYo1/c Itj+XS/M+HYNTDrsZC5tSLOBKaCXxxysnMHqdxAFsEm33LH/w1wnDRQnjzP0bXy4zDdSPfcCGFWR RimRZk0oPgif96CgTjMuX1Gkfggb9kpAd3IzeRYsODRoN92FEESzmsnw0H7kSd32TFCHMcQmaaIR GP+a+BIg3kN/rl6TYZlkzcv2MrdiGqlR6iqeRumY/mL2STvCZQo3MnPK/ei2Lj1ilxPmlyhScOri OAtJZCD4MQmMetIwkB4ncKkyDVtTGJCiymTCLwKlZsxIdv7BvaQ0KJr3RgJPP3qC/Ejs2QebTEni MjgmIqzPKdMD5p/YLaEkaMZTjz5O+ymgYXy4AYkdHH30xExuBivB3vY1HdCD4YcZAFmCQTofnH5Q rWrTFm5QC/TjS42jq1JZ1zHNg29K+AVwfeLsgUqSZEoPo9a13DrRqVJAi0cS+rZZv+x+K3fr59X2 wVUqSeviucyQRI5yBfYxzDlncuYr05M6SqcRHOxR+8p81ktxWYWi/N6agmgZ0CmhpQhuEg/2kpPu zwRbGRBBvB2nKP2KPAcqEjy4dyzaK+76afV1t35uRKitJL1T8I07cpMcKpAW2dIIy1BnwdxlwAvR FTJmbc6EF2AmizApSs98oVZdKpQfAFqCxl7pG+e0jZG112kSGUsJzgqEw5pUDcxS6RllWgibcHOE Pj0GcsTkTXt9p1desPr5/vCAepzwZbvbvD83aSb1dHvTUNrX5pddSwxgq0MSCb70f4fdy1HBdSE0 ZUEXh4/BFToFG1K+Gldq96Nhkttf49/8maHJUAUhKWN0KWNm1SowIXZUUmcpp2Y+DchZUI0LL6FV N9PxqQG2W4r2yCJyrlSNLq4tg2QNwm0I9xuMCc3q+iQBrJciTehdJarGTbWm7CDBqDUt7JWtsttI 3aFxPvpSEph7MA7MK4DCXqc5Xi1gVKXjEFy9ai8IGmHN1jh2vVQPxvjzIH192345wOCq729qWc+W Lw80fRIU7aNSM00z1mTSxKP3XiVIdk28VQMfTStMumn4A+6rXun8YbPdv+MOoxOk9aAMmg4sVjsX IlPToy5+qOzo1sy/t2/rF1SAQCue33erjxX8Z7W7++uvv/7T7dJrWN6VzB/GHTz/R4n2sgSJB8T/ KW9SK7kY7Jq6SgoQLkG8VBcS9+lfDuVvtR/ul7vlAW6EO7z2kqxeWKO8MrfLSK3ROvBKD48MjDEV Um313rLVQ6Nf8fNDEe028DChGDVFl6B6no7R18vp3iMGmCBVmEd7udrucNxx7fiYy2z5YETXkk6O XX+Vz6McCdN0onOFtGFiodrG4XBqbM2r2pawGf30qi5zmRWFvJ7lwCXxQo8f4s61E622YkBsr7O9 /XWsB5S881+U423+EFQBAA== --===============1938307108353976146==--