Add a section on advantages of unit testing, how to write unit tests, KUnit features and Prerequisites.

Signed-off-by: Harinder Singh sharinder@google.com --- Documentation/dev-tools/kunit/index.rst | 166 +++++++++++++----------- 1 file changed, 88 insertions(+), 78 deletions(-)

diff --git a/Documentation/dev-tools/kunit/index.rst b/Documentation/dev-tools/kunit/index.rst index cacb35ec658d..ebf4bffaa1ca 100644 --- a/Documentation/dev-tools/kunit/index.rst +++ b/Documentation/dev-tools/kunit/index.rst @@ -1,11 +1,12 @@ .. SPDX-License-Identifier: GPL-2.0

-========================================= -KUnit - Unit Testing for the Linux Kernel -========================================= +================================= +KUnit - Linux Kernel Unit Testing +=================================

.. toctree:: :maxdepth: 2 + :caption: Contents:

start usage @@ -16,82 +17,91 @@ KUnit - Unit Testing for the Linux Kernel tips running_tips

-What is KUnit? -============== - -KUnit is a lightweight unit testing and mocking framework for the Linux kernel. - -KUnit is heavily inspired by JUnit, Python's unittest.mock, and -Googletest/Googlemock for C++. KUnit provides facilities for defining unit test -cases, grouping related test cases into test suites, providing common -infrastructure for running tests, and much more. - -KUnit consists of a kernel component, which provides a set of macros for easily -writing unit tests. Tests written against KUnit will run on kernel boot if -built-in, or when loaded if built as a module. These tests write out results to -the kernel log in `TAP https://testanything.org/`_ format. - -To make running these tests (and reading the results) easier, KUnit offers -:doc:`kunit_tool <kunit-tool>`, which builds a `User Mode Linux -http://user-mode-linux.sourceforge.net`_ kernel, runs it, and parses the test -results. This provides a quick way of running KUnit tests during development, -without requiring a virtual machine or separate hardware. - -Get started now: Documentation/dev-tools/kunit/start.rst - -Why KUnit? -========== - -A unit test is supposed to test a single unit of code in isolation, hence the -name. A unit test should be the finest granularity of testing and as such should -allow all possible code paths to be tested in the code under test; this is only -possible if the code under test is very small and does not have any external -dependencies outside of the test's control like hardware. - -KUnit provides a common framework for unit tests within the kernel. - -KUnit tests can be run on most architectures, and most tests are architecture -independent. All built-in KUnit tests run on kernel startup. Alternatively, -KUnit and KUnit tests can be built as modules and tests will run when the test -module is loaded. - -.. note:: - - KUnit can also run tests without needing a virtual machine or actual - hardware under User Mode Linux. User Mode Linux is a Linux architecture, - like ARM or x86, which compiles the kernel as a Linux executable. KUnit - can be used with UML either by building with ``ARCH=um`` (like any other - architecture), or by using :doc:`kunit_tool <kunit-tool>`. - -KUnit is fast. Excluding build time, from invocation to completion KUnit can run -several dozen tests in only 10 to 20 seconds; this might not sound like a big -deal to some people, but having such fast and easy to run tests fundamentally -changes the way you go about testing and even writing code in the first place. -Linus himself said in his `git talk at Google -https://gist.github.com/lorn/1272686/revisions#diff-53c65572127855f1b003db4064a94573R874`_: - - "... a lot of people seem to think that performance is about doing the - same thing, just doing it faster, and that is not true. That is not what - performance is all about. If you can do something really fast, really - well, people will start using it differently." - -In this context Linus was talking about branching and merging, -but this point also applies to testing. If your tests are slow, unreliable, are -difficult to write, and require a special setup or special hardware to run, -then you wait a lot longer to write tests, and you wait a lot longer to run -tests; this means that tests are likely to break, unlikely to test a lot of -things, and are unlikely to be rerun once they pass. If your tests are really -fast, you run them all the time, every time you make a change, and every time -someone sends you some code. Why trust that someone ran all their tests -correctly on every change when you can just run them yourself in less time than -it takes to read their test log? +This section details the kernel unit testing framework. + +Introduction +============ + +KUnit (Kernel unit testing framework) provides a common framework for +unit tests within the Linux kernel. Using KUnit, you can define groups +of test cases called test suites. The tests either run on kernel boot +if built-in, or load as a module. KUnit automatically flags and reports +failed test cases in the kernel log. The test results appear in `TAP +(Test Anything Protocol) format https://testanything.org/`_. It is inspired by +JUnit, Python’s unittest.mock, and GoogleTest/GoogleMock (C++ unit testing +framework). + +KUnit tests are part of the kernel, written in the C (programming) +language, and test parts of the Kernel implementation (example: a C +language function). Excluding build time, from invocation to +completion, KUnit can run around 100 tests in less than 10 seconds. +KUnit can test any kernel component, for example: file system, system +calls, memory management, device drivers and so on. + +KUnit follows the white-box testing approach. The test has access to +internal system functionality. KUnit runs in kernel space and is not +restricted to things exposed to user-space. + +In addition, KUnit has kunit_tool, a script (``tools/testing/kunit/kunit.py``) +that configures the Linux kernel, runs KUnit tests under QEMU or UML (`User Mode +Linux http://user-mode-linux.sourceforge.net/`_), parses the test results and +displays them in a user friendly manner. + +Features +-------- + +- Provides a framework for writing unit tests. +- Runs tests on any kernel architecture. +- Runs a test in milliseconds. + +Prerequisites +------------- + +- Any Linux kernel compatible hardware. +- For Kernel under test, Linux kernel version 5.5 or greater. + +Unit Testing +============ + +A unit test tests a single unit of code in isolation. A unit test is the finest +granularity of testing and allows all possible code paths to be tested in the +code under test. This is possible if the code under test is small and does not +have any external dependencies outside of the test's control like hardware. + + +Write Unit Tests +---------------- + +To write good unit tests, there is a simple but powerful pattern: +Arrange-Act-Assert. This is a great way to structure test cases and +defines an order of operations. + +- Arrange inputs and targets: At the start of the test, arrange the data + that allows a function to work. Example: initialize a statement or + object. +- Act on the target behavior: Call your function/code under test. +- Assert expected outcome: Verify the result (or resulting state) as expected + or not. + +Unit Testing Advantages +----------------------- + +- Increases testing speed and development in the long run. +- Detects bugs at initial stage and therefore decreases bug fix cost + compared to acceptance testing. +- Improves code quality. +- Encourages writing testable code.

How do I use it? ================

-* Documentation/dev-tools/kunit/start.rst - for new users of KUnit -* Documentation/dev-tools/kunit/tips.rst - for short examples of best practices -* Documentation/dev-tools/kunit/usage.rst - for a more detailed explanation of KUnit features -* Documentation/dev-tools/kunit/api/index.rst - for the list of KUnit APIs used for testing -* Documentation/dev-tools/kunit/kunit-tool.rst - for more information on the kunit_tool helper script -* Documentation/dev-tools/kunit/faq.rst - for answers to some common questions about KUnit +* Documentation/dev-tools/kunit/start.rst - for KUnit new users. +* Documentation/dev-tools/kunit/usage.rst - KUnit features. +* Documentation/dev-tools/kunit/tips.rst - best practices with + examples. +* Documentation/dev-tools/kunit/api/index.rst - KUnit APIs + used for testing. +* Documentation/dev-tools/kunit/kunit-tool.rst - kunit_tool helper + script. +* Documentation/dev-tools/kunit/faq.rst - KUnit common questions and + answers.

Describe the components of KUnit and how the kernel mode parts interact with kunit_tool.

Signed-off-by: Harinder Singh sharinder@google.com --- .../dev-tools/kunit/architecture.rst | 206 ++++++++++++++++++ Documentation/dev-tools/kunit/index.rst | 2 + .../kunit/kunit_suitememorydiagram.png | Bin 0 -> 24174 bytes Documentation/dev-tools/kunit/start.rst | 1 + 4 files changed, 209 insertions(+) create mode 100644 Documentation/dev-tools/kunit/architecture.rst create mode 100644 Documentation/dev-tools/kunit/kunit_suitememorydiagram.png

diff --git a/Documentation/dev-tools/kunit/architecture.rst b/Documentation/dev-tools/kunit/architecture.rst new file mode 100644 index 000000000000..bb0fb3e3ed01 --- /dev/null +++ b/Documentation/dev-tools/kunit/architecture.rst @@ -0,0 +1,206 @@ +.. SPDX-License-Identifier: GPL-2.0 + +================== +KUnit Architecture +================== + +The KUnit architecture can be divided into two parts: + +- Kernel testing library +- kunit_tool (Command line test harness) + +In-Kernel Testing Framework +=========================== + +The kernel testing library supports KUnit tests written in C using +KUnit. KUnit tests are kernel code. KUnit does several things: + +- Organizes tests +- Reports test results +- Provides test utilities + +Test Cases +---------- + +The fundamental unit in KUnit is the test case. The KUnit test cases are +grouped into KUnit suites. A KUnit test case is a function with type +signature ``void (*)(struct kunit *test)``. +These test case functions are wrapped in a struct called +``struct kunit_case``. For code, see: +https://elixir.bootlin.com/linux/latest/source/include/kunit/test.h#L145 + +It includes: + +- ``run_case``: the function implementing the actual test case. +- ``name``: the test case name. +- ``generate_params``: the parameterized tests generator function. This + is optional for non-parameterized tests. + +Each KUnit test case gets a ``struct kunit`` context +object passed to it that tracks a running test. The KUnit assertion +macros and other KUnit utilities use the ``struct kunit`` context +object. As an exception, there are two fields: + +- ``->priv``: The setup functions can use it to store arbitrary test + user data. + +- ``->param_value``: It contains the parameter value which can be + retrieved in the parameterized tests. + +Test Suites +----------- + +A KUnit suite includes a collection of test cases. The KUnit suites +are represented by the ``struct kunit_suite``. For example: + +.. code-block:: c + + static struct kunit_case example_test_cases[] = { + KUNIT_CASE(example_test_foo), + KUNIT_CASE(example_test_bar), + KUNIT_CASE(example_test_baz), + {} + }; + + static struct kunit_suite example_test_suite = { + .name = "example", + .init = example_test_init, + .exit = example_test_exit, + .test_cases = example_test_cases, + }; + kunit_test_suite(example_test_suite); + +In the above example, the test suite ``example_test_suite``, runs the +test cases ``example_test_foo``, ``example_test_bar``, and +``example_test_baz``. Before running the test, the ``example_test_init`` +is called and after running the test, ``example_test_exit`` is called. +The ``kunit_test_suite(example_test_suite)`` registers the test suite +with the KUnit test framework. + +Executor +-------- + +The KUnit executor can list and run built-in KUnit tests on boot. +The Test suites are stored in a linker section +called ``.kunit_test_suites``. For code, see: +https://elixir.bootlin.com/linux/v5.12/source/include/asm-generic/vmlinux.ld.... +The linker section consists of an array of pointers to +``struct kunit_suite``, and is populated by the ``kunit_test_suites()`` +macro. To run all tests compiled into the kernel, the KUnit executor +iterates over the linker section array. + +.. kernel-figure:: kunit_suitememorydiagram.png + :alt: KUnit Suite Memory + + KUnit Suite Memory Diagram + +On the kernel boot, the KUnit executor uses the start and end addresses +of this section to iterate over and run all tests. For code, see: +https://elixir.bootlin.com/linux/latest/source/lib/kunit/executor.c + +When built as a module, the ``kunit_test_suites()`` macro defines a +``module_init()`` function, which runs all the tests in the compilation +unit instead of utilizing the executor. + +In KUnit tests, some error classes do not affect other tests +or parts of the kernel, each KUnit case executes in a separate thread +context. For code, see: +https://elixir.bootlin.com/linux/latest/source/lib/kunit/try-catch.c#L58 + +Assertion Macros +---------------- + +KUnit tests verify state using expectations/assertions. +All expectations/assertions are formatted as: +``KUNIT_{EXPECT|ASSERT}_<op>[_MSG](kunit, property[, message])`` + +- ``{EXPECT|ASSERT}`` determines whether the check is an assertion or an + expectation. + + - For an expectation, if the check fails, marks the test as failed + and logs the failure. + + - An assertion, on failure, causes the test case to terminate + immediately. + + - Assertions call function: + ``void __noreturn kunit_abort(struct kunit *)``. + + - ``kunit_abort`` calls function: + ``void __noreturn kunit_try_catch_throw(struct kunit_try_catch *try_catch)``. + + - ``kunit_try_catch_throw`` calls function: + ``void complete_and_exit(struct completion *, long) __noreturn;`` + and terminates the special thread context. + +- ``<op>`` denotes a check with options: ``TRUE`` (supplied property + has the boolean value “true”), ``EQ`` (two supplied properties are + equal), ``NOT_ERR_OR_NULL`` (supplied pointer is not null and does not + contain an “err” value). + +- ``[_MSG]`` prints a custom message on failure. + +Test Result Reporting +--------------------- +KUnit prints test results in KTAP format. KTAP is based on TAP14, see: +https://github.com/isaacs/testanything.github.io/blob/tap14/tap-version-14-s.... +KTAP (yet to be standardized format) works with KUnit and Kselftest. +The KUnit executor prints KTAP results to dmesg, and debugfs +(if configured). + +Parameterized Tests +------------------- + +Each KUnit parameterized test is associated with a collection of +parameters. The test is invoked multiple times, once for each parameter +value and the parameter is stored in the ``param_value`` field. +The test case includes a ``KUNIT_CASE_PARAM()`` macro that accepts a +generator function. +The generator function returns the next parameter given to the +previous parameter in parameterized tests. It also provides a macro to +generate common-case generators based on arrays. + +For code, see: +https://elixir.bootlin.com/linux/v5.12/source/include/kunit/test.h#L1783 + + + + +kunit_tool (Command Line Test Harness) +====================================== + +kunit_tool is a Python script ``(tools/testing/kunit/kunit.py)`` +that can be used to configure, build, exec, parse and run (runs other +commands in order) test results. You can either run KUnit tests using +kunit_tool or can include KUnit in kernel and parse manually. + +- ``configure`` command generates the kernel ``.config`` from a + ``.kunitconfig`` file (and any architecture-specific options). + For some architectures, additional config options are specified in the + ``qemu_config`` Python script + (For example: ``tools/testing/kunit/qemu_configs/powerpc.py``). + It parses both the existing ``.config`` and the ``.kunitconfig`` files + and ensures that ``.config`` is a superset of ``.kunitconfig``. + If this is not the case, it will combine the two and run + ``make olddefconfig`` to regenerate the ``.config`` file. It then + verifies that ``.config`` is now a superset. This checks if all + Kconfig dependencies are correctly specified in ``.kunitconfig``. + ``kunit_config.py`` includes the parsing Kconfigs code. The code which + runs ``make olddefconfig`` is a part of ``kunit_kernel.py``. You can + invoke this command via: ``./tools/testing/kunit/kunit.py config`` and + generate a ``.config`` file. +- ``build`` runs ``make`` on the kernel tree with required options + (depends on the architecture and some options, for example: build_dir) + and reports any errors. + To build a KUnit kernel from the current ``.config``, you can use the + ``build`` argument: ``./tools/testing/kunit/kunit.py build``. +- ``exec`` command executes kernel results either directly (using + User-mode Linux configuration), or via an emulator such + as QEMU. It reads results from the log via standard + output (stdout), and passes them to ``parse`` to be parsed. + If you already have built a kernel with built-in KUnit tests, + you can run the kernel and display the test results with the ``exec`` + argument: ``./tools/testing/kunit/kunit.py exec``. +- ``parse`` extracts the KTAP output from a kernel log, parses + the test results, and prints a summary. For failed tests, any + diagnostic output will be included. diff --git a/Documentation/dev-tools/kunit/index.rst b/Documentation/dev-tools/kunit/index.rst index ebf4bffaa1ca..75e4ae85adbb 100644 --- a/Documentation/dev-tools/kunit/index.rst +++ b/Documentation/dev-tools/kunit/index.rst @@ -9,6 +9,7 @@ KUnit - Linux Kernel Unit Testing :caption: Contents:

start + architecture usage kunit-tool api/index @@ -96,6 +97,7 @@ How do I use it? ================

* Documentation/dev-tools/kunit/start.rst - for KUnit new users. +* Documentation/dev-tools/kunit/architecture.rst - KUnit architecture. * Documentation/dev-tools/kunit/usage.rst - KUnit features. * Documentation/dev-tools/kunit/tips.rst - best practices with examples. diff --git a/Documentation/dev-tools/kunit/kunit_suitememorydiagram.png b/Documentation/dev-tools/kunit/kunit_suitememorydiagram.png new file mode 100644 index 0000000000000000000000000000000000000000..a1aa7c3b0f63edfea83eb1cef3e2257b47b5ca7b GIT binary patch literal 24174 zcmd43cT`m0vNhTS0f_=48AJt<jDR2@D4-}ArO82Z&N*Ws3#dp=0+N~}IR^<MAR;0( zIR_P*C?Jx)YJTUQd+s;h81IeuzCRuVk>2dRd+oJqRn3|;tHae)<jIH`h!F?`*@OFX z4-p7lCj<hgiHHD>#6>A;z<+0*?(4WB5F}LCe>mSJ0}c_0i--qucQm}lRwo1P^!EJu z5BKOp>~Gs=;83s#d)%?ui?1j9nk}4eM@sR4Fq1$(A!}VE%17|d)mM(+!eTVI9oB_7 z9?F(t+B96WCsaH*ZfPV>aJoN#H0>5IpqR?~So2&gZiu$>!!(?qpNmMs$Id!Ev6OcS zb?{1GmJED3?MQy@;ky8}OF4Ta0(MCYH$$gobsPhmTh8~E-<E@~;`y2peC#izPS~Fv z2>Jhh@&EdeM89Lj*)ZK{Hf3M@RwlhHW4{{GdHWl9l!C341Gx?2xBBjlm|cx#wY+>z zebAVeixR%@n6N+L1nk#8FaCdeXux4_+zvM@ZN6i3-VXoLB?7TaHk9jl8)ftiW6g4M z%CcoVk;T}_{VpR$5FG|zIHy19X0X5i>&1T_vSKg6lb3UUbbgT%5u@Srxx8rIVaYWt z%qlK0rjDG6vY4&apSIN#h0kv;lp!l8H&jLv_uqG72sfPxv*MkaYFj=>sZPTLw}Q7e z83|VS-`De>#r+>k{LhQ;TeUCw2gsf753=A0bj4fIP4l*UrS#p#8b#$#L$A%dH`+KP zP7b%>aZQ3QXMG$Hj*7Z1S4rYTdb(==9~a7=dM#L?^33_WH_tq7!IZ9XA9K|Z2H8f1 zvi|zBl^|`-O^vk+b&T558K-<6w_9euJzCX29`52j4!V3gp7NKwFM0DP31_yI-IDe; zRZi^)GyL9zxb=g{eOKcZE-S1p!^3J=K0e?ej;jidc!F`Yqq&5=+@_?}LpsP1F=b+9 z`uH+-J+nB-?XTkaIfphY)2<6+&2d15$iav|q&iz-1LIFTLygcuoKOT6HqM`JoMNVl zkTO@4eG6Ym<zD0y40Uh$an?6#3hS}UjR*gnabcpF{9^Qy?v!b<F~(o-af6M{a93l+ z>7XP8YrSzAUkvMP$^NON`#weJ)K^|4yv~7ri~$ECM}9mn+}5cJbqh*@?|+ioonF(x zTRTZHql`$uaMs=$VY*7ijve?|@oX_kYm8dGpkKxQ`Bpsy6{-E{>7<T9(My=?M~GIv zc=Jy0no~QZn@W~w^e4V(p)(xgm3Mk5zpj1!dXOkjuXt|*m=DZaAM4aHPHPQMZ}OtS zrPImjZyP^C1dKeN#fS$HUFAG?IzI{z5GR?VZ7JqDMvKL4m$ERw_3UtM?nPwzk@VG2 z`iIO$g@wTe;0modz&Z4eH5fKP<avAk^Pc7FI3t5*=--x1+xJf2ljgi=-nv71Z~YgM zseewXcf2tx;%s^5biq*X#CD76ukyY7_YF9c9akCK87V16m|-jRy`xX~K{s!osp)=y zu(?-r#NUCQU|j-MahOxZ@Q(?+yL_s*0`Hnc;Kuqfv(#Q)e;K>S6Qh&A=s$Tz$5aTq z85nA{QvTV}pVU(qY}{4-k5;}4N0F70(1kqN?w9Tl8DWR5og7?qw)9%>xXkIGhx~Kk zd=^KNQ|m*%<?Inn`Tb4%$v+mwrEiCOI(;_0du3;>^3={R7Di;@k5XZ+Qo`Vo<_u5o zUy>;XT~n_}oWwIuNcA_;#*ydLvz_QEZ3uiv@GJ4!#`kbY5o#k0*h?Bd(2l|}(y?Pn zu%wg6v2f~pa0^z?&fncH!Srh3q)uL2x4(Enwzi>xt&35SL-1Ur$U?pAzGNyN?YAUS zrNQUuWm&$A>2K9~!lYjQ&F)RrH7%^M^<NzMP!Tj@V{5E_dlfYKozm^ocpH|T{6;wE z%6NKBdNpn^#q>%CPPE=$bQ62oYhj)^{%eB+cjBI*Z^!&BVF=3POmy}5*&8@jPF@GD zdR`5cgYDz*N_rKB-<7ogbcvJXXl4oPJV!)8NklN~R;Pe6%EA?mS2B5s7A#m1;r6TZ zGi2%g>C%l#YwFQFt~%<^B3|0rUohRBk7AGA(r-FC+FHl^a8YTnXq3;i`<-7DdbvHz z_+nc3<BJz!(QN7HsR8vJe=mQjeQmysuG>7CF7{YEuJZ5Tc*`)0IfGd4M8})os)%hL z&oX|@`Kom!N6|@~$^BNK07K_o8!lp|uY9d_>OTIg8WDWQQd1YQtSB3VS|@)puSiTv zZhKc^$`OY;7#nnT-SRB$y8LVquCmC@6%d!=i@6!Az;;VCv@`fjVs7}r<|Z=6KA9tg z>MUv3JX~&t7AEa5+3=C1i|#UK7f9$&kaUE1eOHp8@e|j;Ab8oL0)_&A9AFIJNQvSy zq2*FH=dFtx4xjJxUasb7Y$BQ&uWl)aHO?4!x^}`EbTU|8ByGF!z@?6<ZF2I*%FjCu zxkEb)rCoTAWh7r$Htq5yC=s7?*x_yIUfffxzlZi?8k7@{{Qys`hNQuuB7&x4Qj$oc z-Q95z_hKKUk_Ve@6n0gayk%-|E0~+#<ujlhhVduno<$;Y5b_fa0j}zkVAW&vwdZ?> ztJed0``r-73HE<>T5!Fp2OOmm`m+rNH8lpRdmanx=70YwJ|-a!tBqzR>dRqA<QU|{ z)|0zhpT%imeKx=ehKT#&BAj8Pas`G1lP#c3jaNnoHX1?|9??%TiGl@RNuRiV<vp%^ zrO*UkP-)l68{V?+fb$1W?2J;-h~FX~7LQsul~S!ZOS_(xMbge<R(W3NkOXTIWQ%`{ zuovziyO*skOp2WWWe&kVs};7*1n$#Ph}O^FH9Bof9Irfu#m=?P+Doz&cynLFuVoP< z^r+|tIXe}dB8T(mbtB3tm|CJx87J{HSbv#|pU5PzQ5<pTi&3<DUd$GMmXhsDIghRU zo>e|7qB))A1xGBLUOw)~5%(00i8lzdMXLT_2<v=rMVFYG_%gHir(2JJhid?fVN9My z@K<4FF=q0c<PI(-xJ%!t+EgW<+}5I9^Av;8eA=$G)$Lp9UEJRI>j!p0NgllMwpJ<> zFEWIUJD|?5`D@54qVN6(zjTvx1~L<M85`8X!-hX`{UHcy;W>_bSbbq4`aF5?%id)Y zgBJnPZJFnM4h{xjhwP1rHC$$)uy|OIY@Ayv_R{F{41N#{9UfrTOaBXQ@yDM|Zw&@v z;a+_)TX@KUVP+0a%23H$&E2V=bP4(OdDX7F{OVBjExM?lwZW<505EVhBr+PutzMt# z&r$Hkk8PSSE1S2GLnwm}CNG#QGarf&Cva!kA0fDUL~#z^pXJBk^_H{P#Cb@*vyMR| z9vpw?K2)<WcO=XpKw$AlLyzU3OF_qT31@rD)9thRgLm2@`DOIvYqN$;=?K0^pGbmN zuW=negGD-(IIyL&IU~-@&?D6Essp;sDt0+C?KT^_vp1`jZ*j&(C~VU>PK-HPq>sS< zxszv&zHLeB4{>&g|Cel<&Uvb+zp;!kIxfiKY$@Ye@#EHSpsQ?2Yh+RFref^N9Ts<& z8-D^gE9^$7|NC#-5m|><Z=5|vt0ICB>^=Q%q(8!WTaFCNS3ez&H``hZS425t#uKNp z`vP0mJ>xUcR<w*a199*sml<3dx#trEU5;gBOAZ{4$)l_5Hm437YAyNZm6jh&^i5mG zpWDkd*D2LWol}(~>mX*Lv`=>Bm$W3&X|Wmc-jyJq`HS(UUrJ_gq@=W`(VPh*LI@P9 zf`=`VipsWMd)?K6zP!vovl@QYdAB20Y;1|IARwH0B5t!caVSr)P!(OYG1)i$s!M!q zDIO`|8&`a2dyvuZmX41|{V92kwSslM=a2c6{wcOfTdv+@Gd%jhnchOz70~shKK(?a zl2*FOy@_rkSAiw4a#_$VI69l#jj5gbvdGi<^J__pwE?~?{6_v{6@P`N9!%yRj2X3T z#vdNNJrKBO5{&pWsE6L%8bWWxZnBzcW-KXlWHARD<jXZoT-`S^p80$;^_Mr3{TQBm z(-K@kUQa=Lyn6G+_(%R5Ze+cRy>u{%sz=sF570k!CxsHQGv{?|!sPqC$?6enYx}WE zqv#1=y-b(q;?dEf^zmisg+??M=1TNU9x^kzv3@L)3YPa}8Q<ELrr}*B<gYgvuOxI@ z3$`qipR6glNQB^*qHd2gf8tu>gBjA=XS<3;ylL2>n9$DuIG&G%Ipo$vt<gMFzbrfP zXh3S>hE&I={RAiT!|Ulh1QOn4OAq_zS0`&|5gYK7ze+gP%qFN~t(NEL#4fe5xbv&* zNJewdOUE1GwMrB)!m5{AE9qMQKrtV7ybJSPDknAAyGi~@V7Efan5unMHP<%P_AESI zu(Q*$QQIw7de3c7r`+u;hRQFtIH<9y5=PP>FM1vfX(}a<mf}ESfJQu;HY0r2HZ(d* z`QTe}jOSN>y7hMs-&2GCTJ_H(f=Zdixh(6`VVdE-VtZ|_wbujc%<kFM9Yy0=4YIV3 zQmf82?Rz@&9j8i8C4Gsl?cY2zs;nqi`G`~B_I33RUOwk^hvhhhNGt_UuTMU5Z1ZLA zE!@Xz2aifOWUIQHnf(|Lrzu}G0i_5@^pne3AwG`BlPF{@ynze-xl<IE#o_$bUz?8% z&{N7Y?AAI3h}i%BGdKDlNz*?s{x1%BkUQkutrV&aSl$bul|KCFUERTPkT!i_a5(OG zJalHp9tLund;YDNKICSRbV(#D>ps>zvAuzNnp()3;mpLyo+R-OBq9>b70;a}pvB=? zoQ8-PgcD9}=^tB)ApZDUJ6Ge)gGWqGljX`=nV9oI(l|3`HulAhc2la-mtW@U*x6?l zyY~v5homqojb<VQCsjw*YW#iS;GmsX7O%_PpG`8>XYhde-cR*tgVyo>^bnA5``SIo zN{hiBz1cQigYcthopNtHL?ga+8cAE&Qel%~_4)R3G|pj-d*g3BM4+k8FhcLu=|olY zHF~2dr=)NN>;T4(JpE^EJBA;}>Xicl8a}weII#AyNiE!??A11{z*qkgwyuWj0rcP? zgcs`9m)2*jQW)3!S|WXyu77b$@0dJHi?=<QtF5V0(*8pZnEG_y#KB1D(&?^+6d8%E zN**v1Ivwpw!qznPwytAu*BcNW-k4a4H$U<gYM~z#p557ph$D(gj10jk?bq5Yd!kJ` zySAn*k#OW?+I@w<q%5TJ5no2=;G&Fp?5J7_06V98eX<4>>{h$*f5MerAbw~muG-VJ zihb6)YvT>I_t$Ax{QepsURj)wCMhbWFign>jdQhcig8@!Kp<Ehw+`OBiZcO519Zk{ zQ(;?h-sdtU1hqrL7qBoM+2bEWy><w*v91Ol|Ci74qXEOi-@`j5ttnjIhXd(dUt!mk ze0A1&Rmj4W=hUy?wllu%wg1Z&+{LEKuX~f|<G)cg3+sfFp@LN0_^laJ_0b8X%^?{t zPUtFp9`QXqxN7UQW~39p<6NQdB_8_(z5Hx&@#)|q;7z=sqBq_EXqh$}A<_c0eE~>V zY`0W-P!7=jLgG&)0=%-sBJE;}hiG$l+lOeybjK5)hiC}&Pv8RY5;qRKnF`_w?d#zW z47%Qxfd60I^2UF9$D2o;tU_zFr=|L&uCcEt#?rBr=E<nAb4&Y&z*g92R-%8eoN$w- z-VA+oD<)gn_iYIp>p}1A8Lz7E&AnX~aA0g1`6|N0&uhAHcxKhp4+rhoKW`kkTp#V) zum90^dwmXO=Dqx6<P8jt<m&nxWFyDCLAxt;pWCc#IDR7wh07=pLAE+zGr)a6>*{1$ zQ&OIx=W!5#L-cP%mMJFK5WK(Je%r@Eg{8t^PykYo2&714$dgPtT+;-;4?6O11(m5O zY#87^xB=23R5^X)jNuoVh}oEkyCd{P9Dj)`4a)JQbbkJRp2MA4s6gY$xVETI4xz8F z4fyZ;y{<0v+Z<RB@Du#!So2|P1lO3RH`iEb$)hfal=I+2=df{f{EyS*opT67KMsxt zHOn1qo|InYp{35XL7$rLzi*RJj=Nul<CU^RsqXNl1fx(i8J?flOIN(m53k`L1_10x z?yBrvbRqKe%r4C%#3XC(?d;tnzV%IBGM&tGFj-NNF09k+E9v+BbnY{~ZXC~=@8hFT zS%AMb4xB5rJGs55_x&<ocM&sT-MtFf5M={h*VZ?ep^}fjicw^@Rqq~{(Q*Yv%nw8O z6^W_vG%^-BHLsPL<;7Cn$=dOORa-BI>QO#Jje3CdVg`KA8ybw0mp;mT6>q?UtZ+<( z%+cXdoQ(hjug0$kgNi4QohF}r)#vj2h?#7Nt>-(L`m(Z(omGXaSdYM035yCwG2ini zW)Y3pQ29l$megy@5Ua^hZ$l_8cx^7dj;FzC-IeL3@v3E#%--@pk*wfr&WhBFSKIuv z={kvNKL5nA8+x$~;6PN9F>`6;gdh0~WMIXdB%|-PYio&S+)B(x)t7|h8O9Kr((<?I z>!N(f7|9ZLo{J3x6C<)|J{AwTlKpD-M+7AmTw5sD*)M!3>Hr24O??Ht@K|TP3Ue2P zek?3t+4al<q=Jp`XmaMeDZ4o$tm_cju_)yx$qv{m4#G&nP2-E(t)XqP-F#o?wcuq4 z>@M9eMqFUqf6cuWRB<)E18dm&hc~?&h0?A%XUvm$agCvSa7uo1IT>cCMh{WZT+AEP z5h?;Bw(Z<Bj_oWnZ4+!^6)HIIvM*`z@EXj%PB%*93752kk!lR7O4f=yWD3^FTdB6> ziSEx!z+FChd@k3zhfhCm@LKyw-@Z2@Ux!UtnHq)aadyhJzLe14zOr3Rmwi>Ha&UI! zw@%A<@mgxU0X`2H??T2N4E$)<uiJFDt1|-=sD;g<0T*n($9NT<5g2f<zoc=&vRqTV z`AE0d$Oz-5ywI$9JOR!wSkS05DQ#Bme)NG2S{VQ7hX`agqyrul)wa6biB(RV&r`Oy zT{M>-3rBb;b3ZRL(ZbkJ(?eDPCiPB;9!&kYEZgthz22_jv10Fz`8$q*(vUk9s+j$V zl>0h99kUxvzJGk+<MS4Uvdb$oqXKL`w~|IK{~#s&236NKB*ri9iJrf1ZTS$W2JWxB z{s@}I6fD`4u=~VZhoVYBg5_JKJmRx|;IFvchU&eg>0;gB@pZ9_2{_i8)8tZQh-STz z&0UUwHS`w6I<QuMBMAZniXi;H9i}28uqBYRQ32w)i??0vh6^bm=lo2FVhI|(VY&~* zV+-mpqz=x$><zLD9n5^W>ab|i+&}?4Zs5Yh^PRn#R?H?%4Hj31QQ4E>v2<Z!)x()0 zQYTy~FR;wkukEubLCJT7_~sO`AsEXy{_`VBC|;?@edCwNxkzLh=>V!x%BWj~X?D$_ z?inyJ+&c9+P1&+iOKx=n7J92*JV|c?S&0aw$2p4HT*Gmq^DkfrnH5v#RU%9mWdbQ; z_Wb;(MHg*-?c3mKYHFid)JE1m3-`Td=IAi2{Hc_?F*Y|{h>FNeFnoQ`<5H>xdjJ?s zW`_X*NanOOf?W0+4}J4`**$NZL$Z`{I(t`jHD4_lu{Y*P>HuovA(~tBdk?qzE~d$) zF11J(;!a(p3|~g6ZqwFZzb%nV=ep;IYLCmz%HM6Vqe6E5^)SoSM&r|3-s8x^5J<J} z_xE)h2t@B6mA|gGqEkawnax_vE*(lXuJwE9CZ#4~`4*af<7u4dFl(r~PL7K<yw?-W zf%r(<3ibaJeidwEqSOt7*wg!b8S>x>z}hn9vebALas>V-yW;%&-X)j)XaxyjrC;I~ z$nbvF-6Xx2?UHuGIepfPDUyT)AvJDjE1bk*mloFtVBnx2!nX8+9hX5MFhgpWo`7jf zag-?I_n8K?=wHeK`U~&?FgY*3m<c?Z2vDk={Wr5?h;FW_^J77wE^t}bF(Hb6aL{Cq z>+8P0^9X-Y;hU?g{(3Y{w3U0jdCn-YTyeXwGHi!bzLLlrHwB-7TjF6LfO(%N#-`Nd z!3w+bn3bP62>kkbv;1?wx-4vY(|@w5i&tQq*=vWyTa@b6<kPtc!x;HBZ>}$wpT-i| z0##skf;R26??Dx~9c7U`v~%WQsR+G15cPA%Yhl<TOTi*J*KV-+zMSS)h=dLn*G1aL z)h@A>t}KV(0NpXF?DKg&;e$XzFIz%y_Zl*<UEO(IOE;WkpQ`w`I*H-7KVoCJ^W;kd zx^*{)gWRdJ^^uf?0{#WcEX2&-9;UD^)T(v83zkAOaxw>nG7(@Xu5&7}POfd~spUNo zc|#pn;l}Z3q+@D+)F}lXUSBcC?5?v&T#0g}wKp1LgSJSvFs2w)yHC)u)su-r-KUj> zg%;B!ZM<zvF{ZAf3uL}-o+i_Y2P@)mI^$S%>J~w*!Gp;f3WOP2-&+8}JZH%BKto+! z5v8+DuZVmuzmnnfUZN*urxL+SMWJld7cN>jSKBg#%sV$0tgzG6?u67hd5VH_N>#ZE zi;Cr-hH`}-J)-F<jRVRPFozYoGI>a0;8-OT@1*g=UJqo6GSz2F_x%xGfA2cjG>t~2 zjJEX0TWuk<R(<>)p@Dq?PUieaKVF_s4t3q(XJJ`6+zmX|kwOJ4TOwJu1BO+Fnja*V z_DtRUnEYQ6flUH^Rqk7vtj`yETPuMn??$YGeJ$?Qk8gMGUQ*k`=%J}NvXlu;1eH{j zTJ?2h!2B1g;zMh03xs{J($Jh*#TokP^l8(Ub$PqoR!d{zFV=N(!Q!So^q(LD4mw|f z?o!t9C7IyFqnjR^CTuU7HVKpl-#lBuO!^LiGxnNvxQXX#;h{<J@*lb<B9oJenC4^s zzTXcFc0F|~e1>SlKUIB|LsZ4I>rn`j_pfYcf=h~S(pz!v^mSPO!TCAFU7TAJEmyOw zc1gW_N>4oYn$m;->AokhP%(_;!A|1mU&qpY&dEUyNqmvv&C!nU)O?FM_!raD&9_3K zYk!a>o4iXDRJX3$)J8GH!L2!i%@8q|KcDsDbL>7Wchy5#Fh$vRZfD=AK3bqrXF2h^ zyPsXk%Wpj@Ow!SMA!V8C_TUQ@{5zVihKKzUdUIdcEc9&I$25PX`938Ac{|&W7!Rm^ zjbk!78dy9ss3?bkiABFRjWkzVMHN4*(JRB1*Z39q{_JUrY2{wRQ6I0$+xz#d<JnWn zzSu#P+JMB<=;CRWL)<SXyd-!sPmWiC)JGLPveXSeHfo{p7GM1Meaq|YBs1h4ChlZo zdw>00T802+*9HD~S5i7MdRx-6=p#A)?z~;RO;_r7cQJ#KkckZ4y|;~hv}AaO8IH2c ztco1c<IUU|q3GqpIz~b~CUu&KtW=w<-hC@kkBsQlmY^qVu4Ff-fUp6!Ba;p7FB(WF zKR)5*cx*5sl;8}gX>DnJBgKZ6`q`8%FRm>gFOrP>Qf*k%9WbRA(J=|n>u0?h=Qwt` zWj;-$>KnE<?<Q-{_cd1`Vd;4O>Y3h*jw&ImL3S0n2WmDD#*O(2l?*gXgbdCGq31vR z_#+0j5^-khEL-o$8q@hNO&d%hSMNozt!R-+^apg|=(pU8dA8QBZK!MEmHnEy|6NIT zQ#RJp*-3hEU&b+&_Sa<TS|)P<x^}LwHHg=?y8)~)Kh>*D6^)rjDOP|4!Dk!{N?!M* z=C=knG`#9vhU!E&v7XVDhvn0GN2yQ);A2^Jt1}zQKNN%1M$*w&zAGX^C#SU2r0x_N zX*O2du$6+Dt_|8asve-{`r)+30B8M!ZFrzXdxvBW0u#Y~A4)IG`eXp<!7bW#R|s%s z&Z>IOF&%T=xz;_Vjvnr@IME*09b@&o^5V_OR(DeX^8X*E(E{=?GP%}svE`2q{!0sR zM_HCH<t}(raL?SPb9-AhX@79|JX9bBs?#S2Ry@EkhxkqJqQL$~Ml9hh9;;JkeYnu^ zeBz#h#t3deOA!w*!#%c_=Svrr4fs<<vDwX=3EjeP&Or3PTV8r!`(CYrw?nD~BH*XR zZ8~PwQm@Ynr8E)>-y1rS{}EG7?rJUCU=6d0g#z5B&Ijs)Z$7tJyvZr$EFWI`{O)17 zvfRb>C^-mO@NKs9{ge|XzEk71s(UbTgAr<N<1*@R4hAIWrwykud-MOI3PX%oKY*Du zYISXgUcL8qcahL(w<GWa5N~-|{#%>CzQxki0->pOSxjX%z3w?ADUe{Dvs9G~2eSeO zDt~=x(u)K#__-ZMN8^_HAkr7WI5K-~x3sk;xsFsg@&m!QA63&ErsndlHp7zlq_I+? z9218^iJ-+Ln2A69bV88Jvk@)0i1cRzoETQH`{f{AT37iTp!KtX9}!hwSI0z(oz@t> zZEwT0=gDTadE4SkH9){fRas7MiyE=K@pUK|V8{EY_(2`Ct!-7E7r%*U2teQgGb*8q ziY=z>?kg&xG9v3~gR`7Jb?rrpS_eF=<D1l^#xO%Qsu66<8vXO`p}FuxVn3u_-1=q6 z^qc9U{X8vWeaK>t`3u<PN74LPCKnQ8eY7*APuTq}Ikrs)iN^qB8lgvN<q`6g41>Ze z#BwQjCTmLPZW3k>$mHaYEYcGZh)NPDX|xV+PiY|ICbRooVi2F!9v?W6omN)zQ}NmA z<F$Mpi-EqV?(OENi=LefxD0ONEvA>^v1Z0tJn)|wS#-fXmY}GTmB&)Wwf26sYyVK7 zu?G1--^TZu2Xv_$7@HS9ji{^C*h$UWx*>%8lg?BuDB)i9)|7&4#oH08Ho-V2$P(j1 zOP7dW=6or<`|POVrdMIBGWKZUQTHiQ^cRFYud4bRV?>7IQ^oxPa~<h1FqeNgRd6F8 ziwwPGxBn}X#d1rA&3_pX8ca1F{rV{H)TV=7Tm<$|$#i-`OGgKB)=+?#biOrNN}RQH zhIKM5Atsr?(b-YedFqL_5_vQ2Bx-l9%O+Fyd1aCYG3>u&d!BSOCIp4@s?d%i#(&1H znDm|ay8o?P5M=;;>D+9*DNTXcN=$L5usLHmV3?&SnbzI<(}m{a*$aj7S+@0!k#&bq zn`GH@RPT%fuMr`n>_m2gm^qm#<{Zgh>nup`ayCEg+OkD>U3JIX`)uAATVGZ(KUJ8$ z#m;dTytX;*ntwFj^%qAxkA57d9(B7#x$@G~y8#{Ka?Cy<L{jkx82FCDaP=uOuIHI^ zGds5hYE1~tyc5TGOn4F#U)?NNf%0LcErSL!7(M?A{Y|UrN;sbk<y$W1yalHJsO^-Q z@JvHozb@pyF=}bc@~B$TStl-OROhs`!(`gMP>&AR!?0*bNFyK-g`-e2Gj}5K^fVP< zh+Cgt@#xiQ1Os=`T%n916e@U=8qCG<^0{ngd|EIHBKx}4SUp8*RwMaa-xPjR`rDt% zi$>GFm4pgoC=ssH89wI9rPw-$r>m!cq~KFv6`MVNaGr*bA5V)uxd#&)qykBl(3pEj zw=3NeTmJ-0RlhyhK+(K!*q89yOiEn6VGktq#>!>vycS9ETINL90{QsCAGQk=NW6=) z#*va1(uT#`=q=35AY!J$oX*kl+g>ZSt~9m$A9y10t(x}6&b*)D@r)yWX@BSk%M9?l z3CMR(w2LxR+8-5wu@LBO?oh_aPsQGoa6S}sp#K=sTth!&#oZ44EPY7L(pjetL>EG< zM_3sMvlj;S8V@_icw+0Lp=b(B1=s!o#D2+qYZ2P*oG7T%5lL_21KKo0S3q*NUr&R} zlt7tTOzPODmT~++g~YLM4Vd-9k~1~a&WO_K8HS$klD~As%`SKNm@W3;F^fbG&T2cy z=#~AgTSIta<qI))+8dj6PR1RxHg;tZzBQZIJ<Jx1X|OyTFpPFYm$EE~oDj|1iv({H zJR6LT=cVV2ju(v(kLB^j^cL+U+Q$tLMwVWuG6cjXt&tgtL;1*d|NMH5mFp}okz=Xp z!BrkJL~1q9ifKSSsE+bxu3!2dHglthv*+qK(^NdpVUD?uVTkA@)H^>@kA|(jnO<H7 zf<~zfeLXv#Bz^aR5g)+PE=eS1#!a;<2X)UOH#ECdqp;dF$^)2yCSgLm_!hc%WxJbQ z1(ik{G~s^-_A3?$kqDt(Dd=Xb^3K#Ozi%Shu|T-##Jk6P28W?2Qxee)4;m&ROKZOW z?tOGhe<)B70KBZ$e`xRR09WBs;eNo9rV^U{p0VwZ*Mr;i%&ZA3b;!Z6CMnvQ_X(W0 zM6X&dX-cf$<_Q_tbfzBzEWY&ZTZrdSzB{~U{*dhWQAAeb#FY=2%LrB4LI>+9=B!Cm z&R3iT_re9tlQ|$?)1MduA%cRAcc&HStC#&c_MuPQI4;vZp<$CfOdB~y9j0bQyy;yA znw6~|3jf_b(fO^wWrIsm`c+@_2_QaHkj>9nJw(sCH!P}P5z>~v$<@%U`LHzNj!PdR zL)5p9d&gxZrvpAi<N<OBsBt9iNnCH)`3%$2cX5H3o&9rbYa!_6*vRd8X~(MV2$2Ig z?eC>mZRE648GcP8y;a|i5Ep6LJ?izZBZgfm8l|RlwM4N7EHnN-i_;o$f<MDJ&NLx) z#_2;j4{goIOJiw}oVzeyd;5MUc0`8ju!Gv=<TK=`O27U_oP!LTCcsKn>9Mo1@sMki z4yo$y?d2UWZ212}DLq0<dqF@*Bp8GZGBri*Pd-?khhn|bV8>TQeq%8fex1M~NL_oQ z!Xi>fnr^$XMMuZXf0&bN10M{IWCBk^W0{M&OIR6}5197C5^sxBaRlTkJbXMC6Z&bV z+B@)5$q7T|H<9Qf#<0$>%^~g>&^B;Pd*MOq2J4?#W`Bm@@z8d1ZaCyrFaT+O&7U3R z7a{lex|NRFlWkNT;J_q@zz#3*eVLwO`{QYUl@qfVO0?1&hY4@y%#8Kx-oO9SL4{g{ z{(YmY##*+%2pCqB>H`oUpvM3?NkV_G`!H~#pa8*gxb^i5D~=Y|p|V{rv6DD$RMpFu z&_ADO-x*Na_2?U!Z8FEV&Q2dB@@@m_(#+nx1;=IC4}Eiq#arkSD5THK`!$>KBO3rY z0MbadO&+QkzwHyi0&&TW6Ca7R4bg1?sPT1e)NZo$CNnV&Dx6$C(^ev(@Sv({cufxz z5LSK8w`*88Ms?zuMJq8AlFY7?Ykc~AgS=zb*SqzTvluj@B>R2txGbq-WxIbHN50Y9 zLdEF2Ldw)QEk>FPt3$cd(*PenF6}NCRg*lkPA=3cjbJPtZxFfaZg6}H!L;&AmlA*F zEZ6RQr1BbGfHaWuH($0iWt+Ai_(beFm$+7VQs$<7u-TJEA|Y4(#AoMAdjQyx9%LDy zWjY^p1Hp%&pb0$Dk2SC$xII>0gfk_03AK<o&ELg&<}e^1^;hSIpz>R~d&hNuo{g6n z3@V_c3dw@BknTgHVgF)p3s85Qe&i$o=`Hx?{^Q&)#HmElbARge*D3^^uo*ukXlkY` z(_c+ytemyqefzq18O<~&x9BRIROCapD==<VwH6AYaSyBN(Y~%DqKzhACw2NB_7a4H zM6LdvMkRCWMt0v6k41v^aNgm(aK833c^A<t&Cv2#QTV-o&7F8<Bk6&tdMi3H15)z2 zb&#O6{khTJ3bs+2m$y&_>X+xS-E9km%6jfpP^NfgT8ug)DY$0yKa?9Vbx@yySPjAA z`PA`oo6AL-j}b@|70Y0yjOg23n)vIRUIIWE997ve0192G8fwcWx*@u}JiFtiFjYY( z{l}t=Tl^ep8>IexA9Z`JVSn8G!lS<$(iE@Iy@uK0>na;YC*@wVe5qN_ng>a~zZC2! zkes;aXA%_;t>NJ5=QxsPP=K)^38f--v^zpY&XJmIMTy!R^BaTE>RjZ|YJ!n4DNwaF z3sY9ew((Fz^LN~Yzv_aOP5mVUBCa^CF(%Ug%*uybH;%?&2MEL~#lP52_M0ag^(XMt zEH2j?g#r%NQ0ASg_`Y9u5cRQjqag2^2Vx<iHzI7FpMd}}z(u<uY}ma>$3q`Oh>(|L zfj6bKh`I1=Im`GH&Q8~10R1J5=`XoI=kD|dq;8>Jqfj7pS^TWxTcw0V0-slYX{f~7 zrTaDTC#x1kyA`@f<jr9Ls6^~S(eS!i$@0{&c<DtBCJ4`9nHo*3h=q4lf&+JyxC-3C z9kMpE^j9mJsqyscqDd&Mlil5Ye8yXJ?j|bs_u1<PrF?e>xz0G%OoW7>-+H{UpC>im z!~vrD+w+s*0JS24Mv9rPqVi(UGD{H5l(tCk5Hp&_`b+!~MjYnpg*e;CHnX;bw@tO~ zh_%q1Bw5D?_UnZc$=PCZ^-BR<gm9_TZ+Uo_SlHN1&VRQu-Fy5Hz3D8)-(na^-4!z* zZ_LS9m!6x_adFs>a%CXwPVnN&kFYA69<<LjEqj9ueW@5H&%KB$qi;z{Yzs8~IOm7M zC~arMY;6SlzLRZJ5>uahGk4{AEND9Er?d-3?_wI_ZKlTU{w`t3;_uZQ+Sw#DzzM$_ z*FqwH@BDrYyUR<c`gV@5lThpj%5j(LdYBsB%mJ6tBA&(gz=}7eJ0>+_^-^<;{0Tsq z0yC*YPQ&fCytV&Y!~|9vLPUItTvK~MZ6IZ9WanrxKkZeua~O`D=)40F+IJ3y;CO#e zOvIB~d5Sa9^bawHf7joA%HT->NJ$sX$I`&j8eYl$k+6teY!tntj;wlj5-7%!V0x}2 zXi%ZTMD8W^{Iw2u;lTha-XD>O=&bjM_a&S?x9+NcBXz4(oX}d{p0a#IsV*YnbHQf+ z_wv@A*sQFp)uZ|G?*@}dOK7zI@EI}J-}NI5=9*wRAAc4n*;Z*{SxwvM3*HPP1pA1# z)j?sg8_^!=J`HO0z^kK}n_fEZYuFsT$OcIelaMG=jFv%?cv|--od+s;k2Ep5;pRH5 zdhzbV<wZiB&XyE9@i_=F+!;LthnhCiUN#4@n><b<VkRSEsLuGV`r2|04+gQrH<?{g z5N6dySgT-=EI+Z<cI?!NtOEt+f{RO|9$$u$x(@YZfr<MXRQVtSDHeNTT=h5s@E3o? zhqc1aCb>A1`~$`F7y6Ifsr6BW!3Gy!%#GBQZtErJzRKd<ImX+f>}s)3FEeu)Ji)AC z)=O&#Z|{d}Fq%uJTPJV)^(*ANK0%twr`6rRU}KLgz1#kh`a(fBZtBRei9u;q@y=_Y zx8s!pZU_xkm&9lv+=bc<yzUt-z!rW=sSOHL5=d}bTHX`A?GJ^5&CT?ZP7c8$k&&aj zWhTp2JRpTo6a4D`zAvGgFt}%e8#KOIoIL_9!3CXOD}GaYcBno8Vca#pEQ9Dmc|Tf< zx_ZGTgJMz}gVW@s%VJnm^^nT3_xKhL9~fDc8?fg;qU02$Di>Mjwz>d2FQA)V-S7Tu z4?m4>Sl{5zAUuCTs`BK>Ns`YpTb!Bv6f5+}Ad&mxf+X(F#+gf(o=Wh1J$gs?g9Uek z5W)9WEp1#SgK$IN8Mk9))iB3GYl3XiidpsOjo!!N82{4V+X6&`!X(_3KYWbkgVr3N zdH@o{g>Pze(p$Ud9v)YDm55o?HZWpos35&4>trJQXInQ-l6aDqM7>9gwp_;(dZ_Nc zwo$agdAZCPNQM|y$ztX@(+%W+Y3GsUxCmL(t4|TtJ&W7H2FyKEmisp~^~MS{N2t5_ z108p_BYbr*jd4<sOvHw4N+~lkiiyT1KBo^eEJ;=K*tVAQl=6PIUmmIJ$N6zD(>Lyp zLlk7~|59l*{OoG@Z<*Mqs+fPcLPXHN1Z5SXd%f;b{2x`9#`@kp3}zE@b%jb%kaaSU zK>W|cu!`3EC2t**bM+Z)hLnFQKU%w1U;92tU&?5N#l{2|i>B*OxK5H-p&b3PpfIuQ ziFpMVvi+ahO5Gsb2$}E|mzesXktL>)U^5vg|MFDLhFms4ku?3jAZcvhtR+ocv}dm` z{|76w*Njuy+e?EGx7x(IRJTKgtj_KN?5%WFCdKpi9q~3pU$|v-eFB;{4(BWtDj!s# z!{)wf@|1jTQnQvdCT~RqJ(;Vm6ykiLHBRmuac51ME8_Kgo|Vu>Sv|*=c?6aaFfLJ_ zB|`v9naV46j2bO;41D(2B0yC;e&=$)>tRRXq>i6z`AaI|LmVY>___PufFcGvn<RVD z)WtMcE|6uHd}SL+=%Xt26i-{R3AHVF!-RXKnjBB9922W8(`(49D@<A^TIki+x4$oA z#l8Dxhh|U6K==hsp+wkAed|zSN`fv(1Xq_)D8Pp<-Tcmzh~cW{G_hu{m^t0mS=v8# zffHN1T+*c@1WJuvgQwNfy6v`mC2=%jHn{?O*(S&}qH52gi2%fjvih67NCQocJggii z-HS@=BY$cB=h2o6fh>?z0|yLhzK_5QbEkU(3ygq#xI3Aq8CHL>46EL;pKqr@R&SMd zW|Y1#0;OA&j=!$d?wpXIx8%V}4O{JGDW?L%j(eJ64bKXxbodH2KrK*kmXz;CN0X7I z(mp}7))!t=2{mm;zkg-h($eEj!A=z%@upvEOxforRVsPt#0SoaNwPs<<75jqrLlTa z4Cc|E2+dW(u?F_5?#!TH)j;yv2^$r;+=L>xvfj!j_ay&76n8YQ%!<u2jlssQL88*& z#2Z`lpll0Lm!EUmb$K`wlOGlLiGy_RsyBBwJj?){($`%7voX7pn&|EIrL|Lzj0V#V z+z9X}iY$RgAN98OfLALd=+dIjJ{7Qm;t(W*bT^Goc5lOlA)v$oS)!?z2|mo=yd~7@ z7Yj>g^J;*R`L<Q<R(j#;C|!Ojm;YpWQR&vBT0s-i-%#rt6R|gt*1%-Sxs{e!N)AG` z3tI<H*XMN-ehQ+FtByt0&*KMYu_a8%UxJ-yJ00#-X<wr@{m@xjwI=g+ZJZFnKHuXy zx@wuy?kl-fHZ>>LUZwQ-To-?fkqJ-MpMx!C8dRFrs;};Gv*^4o{Y-uVw+rG}6d^tN zv3EqT=;IOD`lnwY?VC;e9`)LxTUX^S@tyUs7HWjk!MgOsR}o;9p6K{9^c%gHhX=C% zAqLVv5au92$oE6soe5Nac1Wx33!dlq@q=?5ZR}I3NM|HS9T~53Fov-<E>4;aq3+HR z?khQ$?b=qe=QKKq0(+>acIN3FfMsaOTvEl%bHF0&EIvdcS&Uw^=sLO=$D0O7gKD8p zDSmK~%k5`L`SW;4NKNY&@MZEI@pZFLh2|?1FaEu|*t4?vNkJ2YT_DK^Zu{!BFjq<K z)#~$m@6+FNNAs_tA)AMQlBlRy&g}>3;%K`^)i%ra!Dg=g3M9Zaq)!VCxbc`U&(H0Z znd@9tA319x!Hhi5-g}n35Q_XkI{rSr{S6>InGL7Zxh4TSp)h>Fopn9EPU9Q9gv)VT z#IMJAPsgFL!VG6Z&N#BRAN|N3@2k1aaK`pvjWgtOhBbO$gj18!;!HE6y(AMK<S1BB z5(8lhe}d#Zc(kBe&>#!fG$TnIbmGBiny!N48ze|15)e537*OPixz`d(xmcw&75B?^ zs>VR+KC`or6*fR3ib<~iMWt)9_2dT!8TbJS1;K(3Kp%uN?X2P1I&wR!LX066GE7qe zObaT1_GN4V4q}=teNeqcX2hoTrN>bQjYN6vZ0lt|q;`BH|I`^5`AkjSm#eNJD#}W! z8?&u;xqvd!8Ll_<x<7~&H5_W0#&)I$vIIKPO#=xwJ^5f=5gJGEHk$04Uj!mHSUu;4 z_5YKZYe;E%2AUN?VCFzj3rf3LAlTy$k3i9dm9c{G(*(B6-pyvUt;D0c-L8%dknUap z#c@$4gIEN(8LV(u;2_p|qHprUMh8()+eo#!D?Lbmp-^>4ilhF1KNO3wjtlfAD{I6h zi{l^ia9TZk1*Qxaw~wTLrM!Dj`k3W*W736Ri!IA?DP1eu#sH3of_&%xI8R(BHNU(J zQt4{BvL1o5zJyJv1sfH0Acx(5iNz<xLv_Y<y}1bylwnRBOJNaMe$2YI_c<q-ZObY} zPxEA~k`RG7U^Z$^_xE*%?E+dxX6#e-YbKGhq3mEpaD(pfb0EQndPJ=9b|8JgH!GnL zv<vr1l>znDYMH9NMkEq~)aO~B!dS)+=5kitvJ>=o)mcy#4+lP-+pEbBl-)FFmoh8@ z716Sx7dcjD4O(e%HKqGr{|JOGa(fPOTwnEMY6WdDu>5KdAB_LQ4}(;h7X;SfAcFlP zyJyU!^@IG@4=MfpFYX!Kn$*&7GK*m&!_F41-~Yz>!n3fdzhFdpAwP=~jU(q_ayOBo zzgh01LwUTpPPJ@9uKj4!s2c4r68`2->siXXzhVNO2mW2;_td9F9k$ZM?<U?ge)_>< zOugcGcsy94YG^Q6cX$6-Jg79Q+1Y6iaX9Kxs`rUcBdfTI-XNb5)tVM~>^*fx7kbye zfxMlDIC9M)FB`C3B+0Qu6x1V;=%*5(0$o9Eh^#C#2HT`+-&<hR_UHJ;a^6D0H-b7K zER2D1=0-A;wblW7P-*5a6-m_YGWJ#t?-^8;WJb(Sbm-O2o0GN*+fXq6NzUR7Y+sZM z%QsBzxIMo43}cE&;;4Od3-P+~^7l{uKF-D~EAEm<@bsk>Ls(64@@Syr>lHEGiLSW( zS9jJ5^Bp7W90FLvv47ep98*R3ja^A_jrV+>{KNG|Yc95kYh%$@{_>}m_Rd@0KdOol z_i_Tt#BiRT-tF6gvKwzI&i%gM85dM6E@IEtm3GLoz1hRekRL9!ZFH^Vnmo4CsxL>3 zac7XD{vGE!vHb_j<81K^7L!P9)0;@4YVr@=5&OG(nNv+cEGCcsIcUK^xOdk>Po4TG zFBqhwSZz>;1Gi13ue;Q`>mkJ%O5`^5kB(W4ax|oC7W+uze~LmE2x1O)Z}=@yWO~Ps zv%z34{WOd<P%R{=RYlCKY)UU)wTV6A+{GJhK9r<_w^n!ko0`R$7a?)CQa|WmZ!E!! zka47W{4qsJdKlL%omf^}^!u~<4{D*e7al<);{e;_3%`BW)!w;yY?*bJ47ytst{@Q5 zL3Ox&4*+m#$<9f|<75s<m|gY{>tmoM3*A|ugkcCMsRHE>FqwLD+v<)I@rA#2k{{%N zoB}ea<WE{7_}{Nc^>b+Zj`*lyy`d9;KoF!t!w93eh1hu+zJ}wo+gRomFK%!at64(N zTmPxEGQVB?Tc`epgx+?5Kr5xa;98`@cOL*_YEN7mUZtnVLW>f#XIa>wOupK#|G0gT z?m4JoPg}j-NDfa8;N<lNgMJJ$m0asK@UqJ|w`&{pG&7*+3q_56(7>7_&b8zT(L-^^ z+tAq$)J<>tYbvy&J2F78)TP~9hh+w&SfSAfI%0qjVLL|}Ul~sH46@|~vC|TBiH|hg z4P=ts0cp?!Q-`r<A8QRrw|VkC-~~i6?-xU3U3_rWVN>lo_m@YD6>rirx}GI}gjqII z0hNKSqIfKHW8FaxwNZG+5Q5ar#xuz}8HgdB(O;N#es4Q57LYfAS`>jmA}K-X$s6~& zpK>|ryE~L#K}N-v&<A8<FJH~?iPXc%l}%`H<Hr9IeCD%~vay?=en<?22C0^KAPGP! z3js(rTyEB9*okj59n`RpaldBs`gcp)oGT~;yl%08J`#KRfPg}Dg{ExybIB<HIY3lY zgS3Nfycy(40dwII>$fH*qXs*F<%@;f($}(KZ7{9&J&O<(qK7h2&z-W>3bT`bT&xw& zZjs6HE-i<!HB<r@1X5~H=xQBi@<50~1VMTW9ZFXa_=m?d{n8IfZL>_~-NcC5E}w4$ ztOZ!o59-`;MAtwlqXEomE{LDzml(RtaBdr^s?LfxfH+Efr7k+INgmqS;2Vhqf^0z$ z`tHh$$3&Ey`=g+(3#c~89?jI6Z%XrzSB&}N1pQO$n!1mFJ2RyLct%{E;@z{MGH@=~ z0$_h2qvg5q%M~)%4+1Y#lc1kmM!le>57SiwM%Fa*uV*{xz_$qB6J4pQKZ9jLO$0&C zG?AK-FpJ=HJMz~jWn47A6M}+fHqL_G@R#pLZ*1#BVKBRM?PHY2l$jo|epvBctP0G? zM?S;n*2&tCEX3a9nxu7~ZvIw^{I)Huhd{6pVLun3YUzkOFaWdwpI3_xNxzTZkQRMG zGn)W@Bg`sF?y;_-7-bwS61gbGjYK{{a^Qrq{)JUrbvQq34-z9lE?p5VTzx?_nWh#t z#27Lb=_;M&>&nzm9;<0iL1Mj<>~Bq>r4lOZF|P7U98{W6n6NY2)V_r?6yWC`wZ5&b zMSBk@O>cYS#`F`O-!z4R<JexDIA}4?4}K?<<Q^qYHS})4?4d{Ku*<uVDyzWbE@&ae zQuNMXPB%U5?zVDzC`Sky!4yDE=TW?b_4hhJkk}*(GM#y6VAtG{^?))B-hgiV_hF;o zXD9`2dS7yrNK;{QZV-gR$$?46Z4=rbzd!@)LWOOx91!60cv8YKvv)b6|8iRQcLR0m z`I)gAXIwn6I`>r>b{jVI!vc|hFdQ1-=WkY2rlVs1Wb(^A$7+!(G}<R8+2>2-zLMZS zgPH;Gi6l60v$F*vzy0JXSU@HR>y}of+VXahhROg4gFXZ<{5vs((xu;&d4MH5U(fC@ z1#mb4e-&heKb7u%@c=*vxgE3yxX6GcpG&DeF;QD^`14Pv$3Xo$UO@k#re+-;ytV4y zutL6$02dgbca@(g!d&O@OWwbQQ2(^N=~`Ug{2x-N1CI@B#Kz?H7d18WLXJA|<!3Tj zSAPBeT~c2PWyS%s@-_O5N};6OvyPTmLI1VzJgnI16fC_U^LMPk#kTu5MC-4s4ps{{ z+FwvHxtQgCus?l7%6d5IOT+4Wn3OmCyU{1+ZY2nHzoL30&q?EfI<?Z&f5UTqb}L8A zHl)NSU2DAzWKpp4|0P?M`>qYs47ciGWyat5o-0{8WlC8bH^O#@r=W};e5UieTInU5 zjxk#(|F2Bh-{E;D0eTh2skWY<F2+F0%w161F<gPQgdSwRl*#W0=hxQa^)yCBqpo6Q z*=7{S1FUl&r2n=ndmA=&UD7R#(Hhk;ZUU-7p5grwvJ+8LdBrnAyA(PWQaaPfK{F(_ zv?n_BS+Oi}?cUQhhju=VOP9tdJ7g*#2!Oqpe3vd*fzlf|Y-v65tZY1$1jUlyjuhgH zMv==ArXl~98Koh5+FyGq?@PU#^PBU!lo*a4nSQ3dX~eU-va?Stq4LKs_^x+kKSUkQ zn~=%qIU7KB05o-Cu1>k#N1qZTSd`0?$7OLkIlC2s_R_!<=}B=+W!K8bt+6AZ6w0Xk zBN4lS^TWZ9LyiGesCIC8P$mOq6;v_m^g9+h@PlL$7C_^!e?zZ~TOA6b&<_Dx#S63- zh#Bi?JqgZ@Tf}LV!8K67gucv$^3m5rM}tW(CIoLwUH~8c8ZDJ?4N0g=3liZ{HVYx5 zNTmSVG{M~<2;v>YOFb>*udycZik`PeepE%c^78Stu51<&1;&M$sz3$D_X;a_vwYc# zOpfo_^PWyKbG=y2a%BbV*Wk6<x;kXd3*zr(>*hWIhHE$|6c>!l#z9sjR^bj3Nx!Nx z{`5PR4>`wmY+!Wd(%xB`j+Ksp!l98p3~Z`e9a3}HmQaw3_}~UR1Xw46w^90qNA6hh z&7OL9m4<>tLJrENuWx3hDdAKs&C4~j2gSF$5Hv5+$@$8MPMq^^k_)!tr6z=4^?JvQ z!!g65>&g}HMQzG#)!fbuR|p>3U}eU*q6I7Njy=t<Y^z#So_Ff5gU?97S3D48@aG4J z5*C)M;dtdz!z6?Zk2@t-D@UJo>vEhEOH+Y7ls&UEpfPBgzd@?AG>Ux85iJa@xF9JF zrZ03Ss68F=_3z01zoX-WKb_wA*-PKMyvC0GxC`}I>Svpv=L12$+M{dtN5a2?ANwGz ztk11kj%Z!+fHGO5PSG*2p0CRc<D!4MH{je>xkOi`%po@C0sZ5zhpq^|ZJ8pTu<6Ij zMR}6sTGx%-^`|Wi3KK8lBET5H#{wNOw0l0T4Olihw5Boc{absl0ZxWlg+i@oqka-- zNhpCNLOQ$B%nUUpP)AQpHqE*JJE3(kb@i=fguZW!hvZTRcC+pV27vlBKktIugkB00 zH^0XBOUd^7iJ<yB#Q~(SX;i}UX=>%fKqTXW+ZG|9S@fdV+&UKK0(4H#3r)T<Kx4Rt z@9B<O{5-0#?qB8Gt&^Xj{(r03rwPcvF<SFX6)y<tER9Xf8Q#6+Z+QP4YHQ`?c~|I- zKzkBY>fbp|<Tv2fOoEFs=Hu-i{jdWXHt^Dp`@sRK|5k$XEhxgLoV1;yRX>BFM$DTp zW?v6J3zaig8w{E)mKcJ9#a;E=$WQEih3RS+7DX=Yc!4tReWt1K;2I^?1JdiaZ2PuE zvIfjc>Y2nIeOW)^JI08f{KDP1q4x}n+pQiQLH{5P1<TFPAtPpig(o76Pq4@H*~$?c zU707J<E@l}ddu0v@{{0W8o(P3aM`al(|4%gw87^NI2TZ5-FtlIG?D{|I0YX%L}9k6 z$`2~li=d}(AQ~FNT{g9#MXp=`16yKk#k%x02nayq@G^$1!vsOR5*0oV0#X_p$+Gu$ zpxna>Bp(<5@pa}^Nhmbq<~%Sp%qV>)^C{Q{dIiG5JZLCi1NKW*$s;wrZM>5AdxGg- za&|C{gnm#rw4S;gZcl^vge}lz?lQU>vB-^WkxFO>1IQRp=YO8zg^NO6O6Y^46>)66 zA=S4IQ-9#&S>Cm8diC;<$c#r~!^HP3xJo?sRb1KD&=DSdcEOP_GdXOY1(%z3It@oR zO=P{wF0Q?r3&>p5Ko&Dpf?o$Z-J=i`+Ti9&avl=7xzpg9IbVI^#Cc^;_p}i^Ym#Or zJ|6M*SMG94i=$Czi>l@ITFnVB|Buz1fA6)wBl6zejzk&kn--Vv(1_4_WKvP8_dbfm z>6>EK0kHjQeip=gxtP*xxJg7Vh>d<l>_;4^<hgd5?n+Z*yC^Y;h^&cz&3t}g?^`UR z%s1cMe+wn<AvCNF=%D^7$FQ3&-GI)k88@+o#KybW;_rOoG(}!J4t&%DD>tjFJhquw zB<la6NVig<Vp?~8TXiE=Hv=F#V#ckGIF^9VP=9A1hce$!s?yU9yfU=OE1%Vfb!^0e z=1I3Y$gMzXz91KdZBW?FElF_8bh*Ga6p8(ypTSh$*_(Qw{vzAIRDgsCq*T2EXqj0u z=scjgI~$sV0~KWxQLBA3yaV)NKP}t2+uit~K4E9i-!|qktm4x-Fg5u}T}JxA=t^4z z(@-#t*d<)V=U;>6uOXYAw@PVP5Zv7Pv`-@DJNF><*5(!HO;^NzLWd`GH(Tsj!<q0A z9ngWiPzHZ7ANpcw&V4d=ybPQY4&uK%E|WwPk`*8<mQc_V5dgv5wz7F1cX5>m+_l{m zR5-k8oBh0O<KJiMq%BfUfe-O2jb5fueSy^`LfiVuQyqU|XlYxR2P&6?+i4f2zc?cn z?7Kt-bbl9Ii5(gA?VNTb+G6Qou|aRD(rGUepkT}%AO0*}sZBYdA^4Ko4pb1dLr*Cw zA72aV-;S-3h~sxQ(Q2MzQU|>zCB`M=yzCc{?yZn3LsK0zB-3F#qJR8v?dDic3ERy% zGF?|P4-_;IISpRW_zq2wSwPtUv1CqDnv|TDA$e(x2sG**b)aO!iay~n;0dZ1mucf9 z1~6lw=>*kl<gH)L?q}{Gyf#l9*N;uJ#kM7)1Jk*v)J@7?aj~~r`#<c@zpsN^Qs%Su zH!CraN%jiko%vhS>e&k>#9c8HRpcBxd~74(H{}&bnZ93{{lxw@__ax4=-EJ~w;k1E zCamx5aul6W7FvV_ZVw&po>E=I3`%?4Dk7`)jH2!Lnda|qn@>yg-<7CXmG2ko{~Z5b z_anYa4E>wRD}C;fVhT7f?ou$lT#o2fpqUyhn$0wJm#=qd__AXr&|?uM;hLj*pu*zP zKEZ466v8q2WrA=?;q`8s<U1j)zY`h5p5sk)GvAQ^tC8!BYI51yp$IC)0*F!%LZk_Z z^v(e(0i>uwIwA^Jf*eo;=}qt=J%T6*s6>I#g^*AJC|!C&5)56C9zd$}J7Ilyo%?<3 z`*Hu|U76Xl_ny6HKkqYZ%{z@&C&MAE<K$=KnT6CCjC$ybmcK1hTRp^@lI0TOyR(>D z6*Tfj^LK*T3)~WWtv+sTk8SwM=Fn*Kiu+c13Csy0A&;`RMsEu0P@cHkP=4ITcy06H ziB)3=Xv6C-HKzE1m6Xi7(xQS1#!3hSXQP?Dm3=YdS7FVd|DQ>48OjSwj9t|;OrY(@ zb1Vm%lQf?6wE0}3yLeXg&aEap7-9Z?+L9RsxF)sIO(HuE`R<MH=DIZ}DvdShvN6R~ zJ|Ge4s_Zkd)kF5!;DqVz1|^TkXkHsclB9f!h@femul$%^gL677i?13V+`?H{Sl~{~ zd4vqSu}d(#Ra1P9<CHRu3qd0|6ij~B9ruCFIMIURqC7J2mTN^~@O7iynY``6L(3cY zGYXfh$jbMB{Vo@;t+~qf$iTqkp_jFz1j<IUWc9^K{|k-H*6`bfI(+zCHodE{47%2i zi?vt0<c5O}83nb2E;^e&z8D3;n%{Vt&X*HN{H@oZGZ{C^h*Mxzh1kL!tmj9S7*uQ# zW6G_H()fihTAuN0F2#<MZy>jA^26c`;cX?%Zp<@{42|I8g)cRQ3GzB?CQbI4nOs}^ zwAnjyj6hAhaK{=|yJK8ZBSHQ-zHjrB_tOCmD#Ph<$R9bL$kdML&G`iR^muulj1I%C z_09SIDUQLT-clPxkQhF7g80ffu#@di^O}H%xZITE8r*+0GU$#z!>J%eWQq1p@yyk7 z@$%yF@)-YIyeRL=#&$6BY=z~5o{0YikX^_sv>AU<Qf#h5Qgh!0RtPnSm$#*{#I;P? zIo$Zr5)uBb)63~jbspA#;A;5dkE<6U<5{OEG6h5DiRfm`m{9gtxB72x{B5<<_Py1Q z7j?&FKd^E_>UgIziZE+pv}9kyxiLRB-RyVtR{p6BqfnLBRMWs41kI7UFCi(Y-8N0y zXh9trJ@0Xn6HI#e{+ifabGGx1*vn-9m?noLRJNz*uK9Auar&a?WNba$+Y-)izRckb z(G6#E%P~c+dzy--y>^ou&|Dz%Lh2%6><|b91Qr4T>L7F}@NGaN0|WvTbiIS79yNKD z=74(yb;I>JVNQ-hN1~|QkTx$R(qG-*HWP(|ooO0=?jJ`$b-=63%y@TYXKNISo^*nH zG4`Z1xOGRz&DmEX<(@2XRHsB$Em2>c{T!Nup}bjken&6XD@IS5lK+~Gh~j0x5&4E9 zIQJ;%bB(AY+QCM7zC7jUreG}wQ~|Ye{sCO}GCyfoiNNF!s$@RVw?KoS^x)JO<~Q8m zK8>w&*r|5Fsj9rQprO>h-Vj*{Ds`)TfF`K(TU&z{q^~_vu5M7m)$C#YVz_kw*jyH6 zYoSG}J7d{7nY`1h=+(!%?~Y_BeLOH}p|ocytw*vugSK9%rM+ZsGk5fl@trowp##j4 zaXJ7HWm=W6cF@;-EUxGlZQ>ea@7-ZMz1Nj>?4~~|&f)@uQ6TvggE?WZP9k(~%fk)} z$8aDpSIQ)cXSmbT&|uiH`f4f8SMw(8L=BIomA*2n{~|f~5KFfsRA{IDk>=nZ^Q3$^ zGTY<#`;)<s!B}GeDgHsf=-DyLF-^TQ?GWco`f7A;gb?S4)g<Dqq&*5rf^&3Kp~uTH z_Q31gpO#2~KqDdp!0(tW7e7Vd*p+Vcz)Hup)%U+!?teVF{D8dDw}R78pgjM~AD;bS zc$V|MNewLkLCJa#!zye=R3lL#;-}`i(zLhSKf5?U+LgNWTH-6np<eOZW(5D^-U_*m zD!0M(p6rbW(SsHtcjg%#WW%hiGy2sY<mywVzLbI$;}{!?CO6TRPA_R5GOEPyS8YWC zcqfGvA&?Y4LhxZ^bAY!-XbBME6XGLp?#S#oksfIgaq^`)?q86|uxthQ9{gU?OP%_I zorTgsR+f5G=y7jdWyjS!;nMASf@L47?3OEOl$Lt7*tJ$YlE?S6=9=fwJx^^aK)hw_ zzk+l7>Mo}SZ*m|=pP@yhPAMo1pzN)oGHZ|wY7HWjo`YXe%Ls$~ovPj(utK9F4CtZ6 z)DM*ucMlmpEX9Ev@tW*cJNM)|BmdM|@6+G=>!o<2!9{MDa|&cTbod*c(+L6Q1b!g| zS)5G$SZ&#GKcaCG*@LHUpJOE|2ECuQg`sDY40~r*cWurcA9R$FImyhpIG|aMuCd$i z&wqcmf=kPLg&9H1Y<=w(un<{4+!xB{0HAgWGd{h`i%Vx<0M(kDshwI$HRZV4s_9-F z^3qY^2}tFVBNqfhjIzyNIV;EVQG8Ga+zl#lg%j~G31y*4E_x?FJ+vrj7A-%0!qto; z4U}|~C$`s>CYfCMawJmn8MX}N1rtw@4>K?T7Oo%wSijl|oUkUcJ!fwkDnvS^G_&gh zxR^@d)g)&#!i)xG+8Z}hSE|PBc2A`;gA%LAz*N~2ScMSQedjBZ@KGJ-U7gHrgN+EE zzORmEe03p_C>BGTiG~-YE=wWB91{)UPF5*ter-Q|c+T+@<Y8|oPUhEr&?|pBOvK2R z=J!P;a%x-t(I_&rSByGbvB?bejNiUc%fj^xn}u{Tk*$Z$*>;i$ZE)U`lFF*HEI3ce z^<5r*fm;DrB;5bezBa)D#=6@gbkf(kU`;{@s=1y(>xsWgjrr7+H%;txGMk3_jhks5 zCNvm-faksKe~5i(fkLTe0Tn0XUJPfK4QvDk29dZqYqjok#R8PKy*cSsC@J7~m|a5~ z_&G3U6uA9OT57oM1~#3#@Z?=?(|%BZfdN%SVFUkZ)8^IfTlppRPCt2gu$q2qd{2hT zKYudp={qef1=3d`Qesb75b~s&yLP#Nvp!#kWMLXN5=Uez)apQ7Htg&(ITuF?^%}wm z8-i{9xe?)aqoV9=;U6hfs-c#X5IY)E{hcP6*`8v2EpbvpSvCJ=R+c(n?CiiC@c|Y; zGpx2XwtA<JBb?3*&=og#PYDh5Pa*2a<WJ;jw-y#wjOuA`N>)3F@W%Sa+%5M!*;DN8 zK5wmh<67>j?-V-Io51Wxs#KzzR=0y+aB$AEzE2itn}~`&;UmuSRI3?(H-1|IwjqSD z!^^uBpssf;P`WD_VLQFEc>;>IMa!xa#Rj?Jqc*Nx!fx*p4#nq|i!rt|k;Im}=)zJ; zYYdA_&v-qWpIn+Q79s~#yq>0@i2CGhGsMq7(x{T=V%6p|KpZM$3flTMv-Y68H!lqi z-@##TOBbTWU=8LhMWE*srrSA3yST=(y}0|7o`}lv#LK)p`^d4J<3;7K@BYHc8s(~! zzDr8VS86Ix0R}O3_n-Fe@o&Tn2(H)dFM=2IS`t`CoVz`#7h!=Kd+Mxa=N2>)a}v9K zkq5pp)a^Pi&Pg}%#+FH`Q!G!!&id$oAF0Hb%b&?tS3lcS5fN^enfY^V`x7_w$L3^r z^y(c;?HmiEv5?!P7)*hrfb~t)B8~-0ZK;lxgl1oVN<Z^6o6DDAxug1*CeAafHg<|E zop9}XF2wHhIBizY;bHuCG>S=qdx`}*ckebYYp0?;+>u817kDoAa9#K2;qBO1E?c+| z`+Q@zmXve7gOTVNQnVhHL;gFb4+z>5!2f{3eL+F$!QR>h?3b-ZCL{{13M_}&&oA%H zk3}I<&tF}d?RyD_bHzyK^ybmtTN+OI-3v;bH0}_Y1Fof|{bFdwsj_ma@%-?89Fs%g zHl|*Z^oUpUjF)WaEDg&Nz1KQl3nLu#KSwJmtUOM__^pjd#PrS(K{ja?n^zOo1+vsb z>=Gw&O^I(zW}5dud8_&$8zymI^2*BkI9!FdHee=irBx;nYKg(xIYB~iCskudTV9;R zlgZQO!KxX<J-MQ#N*!up7S`}NW3eCRPxHZA`IeCZTd);kuW9-OChhhu67t-Z_lJ|! z-FA<`-0PPuKNu~!x|$`)e@eqEFC@Vb#rx$bXR*V3tqyEn)@Zh9?zlG6OHJn@S#Cm1 zcP0jhx@Xc9_Om0qq{HpKm)WMLjI5*&Wn{4L4-gbk9sbj^PfzZL%5XAQf{N9`<Cm%* z#F?6$t<Yy<!{E!UqC*?1RFu&m!7f%Q%ATWd$(FR3jfD{!?>Qc5%q^m7SP7=u7C>29 zzOm886iJQ~q7CO|t)9Kr^yboX*iBVE)xA)kIv_fS{H)kqbNTdB2ZCCRWuEh+<CqUR zd|ElJKGjA9V>z2l%w|AoUDExC@DUQ>17ybfC#if$IN^~7*7n8Y@#I3hkQVaQrRBsG zipkr!Y5113&yv$e)Wg=-y@rfgn9b{((?#mC*z>MMC>bPQf_$sp&}(U$PG7(L(qbDZ z8CvnTs@TB3z@9AA1p>4JZwLUq`XIWx<2+Xi_p7Z-s_`=#9XWxnUHnLGs}KHOmV|&L zhgZxubuMA~T&u)*4|~q9rD6PxWGh3Ob~Y5QR`?cmlpHfk0Ph^PHgK{U1&FkC>g;xY zS2?!qnxy`h*4J_`UA)8w9Lvc<t`;+?oG{U`RXIM_1CrgYMd~qN9;ej)MB2x(>Jc<> zDl%i7u}H}8@S&dLwb<+$IiwD=#$Up5VXpDj@0)4k@nSv3tbnjy+FliyOWCbLaUioT zz0CYxcOlD{MK3Id-dJrHO&?s#*5C~Dh`*4mXhb+8#R!-aYU|ieHR=-VdWTbWGU}tr zdKZKL<ase110cTsD+biu<}M3p&@}l{)9?-GCGM%u0gEnlQe2_c6(ZY_wUEQ^TX=)f zG3D&c`!j$vgF0@C@EX_{MCzAxA|t$<R>diif^>>D1<!x{+@TG;q4Sy-$ti^iveyqx z8=2@-J1M3Xww#^J9>B-7(rRB%T)Exs^rw2y!L-T?r+PCH5W(<*m^uxY_<$?W;|ErR zs9S<uZvgpeU|-)FGli~t<b)aUfUm9UcupTm%<Z*5T6NoO#2pbLp78IJ77~EMgK&bM zzsWri|C`+Z57++>8AbuneI#Sdq*B4^mI5b{xEbgzkD8${o>Y;-jbo7BD}dqvxnz9+ z5=XlP@aON5e;fN>JBfeG966-$Eb}|l>#MF^lSrY45mz!bo#-h3asvVx?h<OuI)+P! z;uXzjug;ac><ENFVl2T-j=XA@?TKLW7Vohjs=p~>g#;u$9tV0KvFqo9$A4q^NX9JO z{sBWp<7eGFb0J$JTbo*2!%gmi86j)kDr3VqpGl+r{#huR0c7uvy4KUl)&wNg@}DMV z6|O1&6w@+>(sOsuV0L#b@A&r<-jH$0w@l!u&(UQ6-%uXyrT*Sp9mxRQ{|?B}#QtmF zkMtZ3IZ}J1=RZsnmxN3pT|yJQ&Zy&bmy$&fObI<<Nrna2Hs{$O*)BzxTOgQ!Joo0L gr~3b_jz^?JrV(@=j@;LB4fH}_2IkjF^zKCd3(G2*J^%m!

literal 0 HcmV?d00001

diff --git a/Documentation/dev-tools/kunit/start.rst b/Documentation/dev-tools/kunit/start.rst index 55f8df1abd40..5dd2c88fa2bd 100644 --- a/Documentation/dev-tools/kunit/start.rst +++ b/Documentation/dev-tools/kunit/start.rst @@ -240,6 +240,7 @@ Congrats! You just wrote your first KUnit test. Next Steps ==========

+* Documentation/dev-tools/kunit/architecture.rst - KUnit architecture. * Documentation/dev-tools/kunit/usage.rst - KUnit features. * Documentation/dev-tools/kunit/tips.rst - best practices with examples.

We now have dedicated pages on running tests. Therefore refocus the usage page on writing tests and add content from tips page and information on other architectures.

Signed-off-by: Harinder Singh sharinder@google.com --- Documentation/dev-tools/kunit/index.rst | 2 +- Documentation/dev-tools/kunit/start.rst | 2 +- Documentation/dev-tools/kunit/usage.rst | 570 ++++++++++-------------- 3 files changed, 247 insertions(+), 327 deletions(-)

diff --git a/Documentation/dev-tools/kunit/index.rst b/Documentation/dev-tools/kunit/index.rst index c0d1fd749cd2..76c9704d6a1a 100644 --- a/Documentation/dev-tools/kunit/index.rst +++ b/Documentation/dev-tools/kunit/index.rst @@ -102,7 +102,7 @@ How do I use it? * Documentation/dev-tools/kunit/architecture.rst - KUnit architecture. * Documentation/dev-tools/kunit/run_wrapper.rst - run kunit_tool. * Documentation/dev-tools/kunit/run_manual.rst - run tests without kunit_tool. -* Documentation/dev-tools/kunit/usage.rst - KUnit features. +* Documentation/dev-tools/kunit/usage.rst - write tests. * Documentation/dev-tools/kunit/tips.rst - best practices with examples. * Documentation/dev-tools/kunit/api/index.rst - KUnit APIs diff --git a/Documentation/dev-tools/kunit/start.rst b/Documentation/dev-tools/kunit/start.rst index af13f443c976..a858ab009944 100644 --- a/Documentation/dev-tools/kunit/start.rst +++ b/Documentation/dev-tools/kunit/start.rst @@ -243,7 +243,7 @@ Next Steps * Documentation/dev-tools/kunit/architecture.rst - KUnit architecture. * Documentation/dev-tools/kunit/run_wrapper.rst - run kunit_tool. * Documentation/dev-tools/kunit/run_manual.rst - run tests without kunit_tool. -* Documentation/dev-tools/kunit/usage.rst - KUnit features. +* Documentation/dev-tools/kunit/usage.rst - write tests. * Documentation/dev-tools/kunit/tips.rst - best practices with examples. * Documentation/dev-tools/kunit/api/index.rst - KUnit APIs diff --git a/Documentation/dev-tools/kunit/usage.rst b/Documentation/dev-tools/kunit/usage.rst index 63f1bb89ebf5..b321877797f0 100644 --- a/Documentation/dev-tools/kunit/usage.rst +++ b/Documentation/dev-tools/kunit/usage.rst @@ -1,57 +1,13 @@ .. SPDX-License-Identifier: GPL-2.0

-=========== -Using KUnit -=========== - -The purpose of this document is to describe what KUnit is, how it works, how it -is intended to be used, and all the concepts and terminology that are needed to -understand it. This guide assumes a working knowledge of the Linux kernel and -some basic knowledge of testing. - -For a high level introduction to KUnit, including setting up KUnit for your -project, see Documentation/dev-tools/kunit/start.rst. - -Organization of this document -============================= - -This document is organized into two main sections: Testing and Common Patterns. -The first covers what unit tests are and how to use KUnit to write them. The -second covers common testing patterns, e.g. how to isolate code and make it -possible to unit test code that was otherwise un-unit-testable. - -Testing -======= - -What is KUnit? --------------- - -"K" is short for "kernel" so "KUnit" is the "(Linux) Kernel Unit Testing -Framework." KUnit is intended first and foremost for writing unit tests; it is -general enough that it can be used to write integration tests; however, this is -a secondary goal. KUnit has no ambition of being the only testing framework for -the kernel; for example, it does not intend to be an end-to-end testing -framework. - -What is Unit Testing? ---------------------- - -A `unit test https://martinfowler.com/bliki/UnitTest.html`_ is a test that -tests code at the smallest possible scope, a *unit* of code. In the C -programming language that's a function. - -Unit tests should be written for all the publicly exposed functions in a -compilation unit; so that is all the functions that are exported in either a -*class* (defined below) or all functions which are **not** static. - Writing Tests -------------- +=============

Test Cases -~~~~~~~~~~ +----------

The fundamental unit in KUnit is the test case. A test case is a function with -the signature ``void (*)(struct kunit *test)``. It calls a function to be tested +the signature ``void (*)(struct kunit *test)``. It calls the function under test and then sets *expectations* for what should happen. For example:

.. code-block:: c @@ -65,18 +21,19 @@ and then sets *expectations* for what should happen. For example: KUNIT_FAIL(test, "This test never passes."); }

-In the above example ``example_test_success`` always passes because it does -nothing; no expectations are set, so all expectations pass. On the other hand -``example_test_failure`` always fails because it calls ``KUNIT_FAIL``, which is -a special expectation that logs a message and causes the test case to fail. +In the above example, ``example_test_success`` always passes because it does +nothing; no expectations are set, and therefore all expectations pass. On the +other hand ``example_test_failure`` always fails because it calls ``KUNIT_FAIL``, +which is a special expectation that logs a message and causes the test case to +fail.

Expectations ~~~~~~~~~~~~ -An *expectation* is a way to specify that you expect a piece of code to do -something in a test. An expectation is called like a function. A test is made -by setting expectations about the behavior of a piece of code under test; when -one or more of the expectations fail, the test case fails and information about -the failure is logged. For example: +An *expectation* specifies that we expect a piece of code to do something in a +test. An expectation is called like a function. A test is made by setting +expectations about the behavior of a piece of code under test. When one or more +expectations fail, the test case fails and information about the failure is +logged. For example:

.. code-block:: c

@@ -86,29 +43,28 @@ the failure is logged. For example: KUNIT_EXPECT_EQ(test, 2, add(1, 1)); }

-In the above example ``add_test_basic`` makes a number of assertions about the -behavior of a function called ``add``; the first parameter is always of type -``struct kunit *``, which contains information about the current test context; -the second parameter, in this case, is what the value is expected to be; the +In the above example, ``add_test_basic`` makes a number of assertions about the +behavior of a function called ``add``. The first parameter is always of type +``struct kunit *``, which contains information about the current test context. +The second parameter, in this case, is what the value is expected to be. The last value is what the value actually is. If ``add`` passes all of these expectations, the test case, ``add_test_basic`` will pass; if any one of these expectations fails, the test case will fail.

-It is important to understand that a test case *fails* when any expectation is -violated; however, the test will continue running, potentially trying other -expectations until the test case ends or is otherwise terminated. This is as -opposed to *assertions* which are discussed later. +A test case *fails* when any expectation is violated; however, the test will +continue to run, and try other expectations until the test case ends or is +otherwise terminated. This is as opposed to *assertions* which are discussed +later.

-To learn about more expectations supported by KUnit, see -Documentation/dev-tools/kunit/api/test.rst. +To learn about more KUnit expectations, see Documentation/dev-tools/kunit/api/test.rst.

.. note:: - A single test case should be pretty short, pretty easy to understand, - focused on a single behavior. + A single test case should be short, easy to understand, and focused on a + single behavior.

-For example, if we wanted to properly test the add function above, we would -create additional tests cases which would each test a different property that an -add function should have like this: +For example, if we want to rigorously test the ``add`` function above, create +additional tests cases which would test each property that an ``add`` function +should have as shown below:

.. code-block:: c

@@ -134,56 +90,43 @@ add function should have like this: KUNIT_EXPECT_EQ(test, INT_MIN, add(INT_MAX, 1)); }

-Notice how it is immediately obvious what all the properties that we are testing -for are. - Assertions ~~~~~~~~~~

-KUnit also has the concept of an *assertion*. An assertion is just like an -expectation except the assertion immediately terminates the test case if it is -not satisfied. - -For example: +An assertion is like an expectation, except that the assertion immediately +terminates the test case if the condition is not satisfied. For example:

.. code-block:: c

- static void mock_test_do_expect_default_return(struct kunit *test) + static void test_sort(struct kunit *test) { - struct mock_test_context *ctx = test->priv; - struct mock *mock = ctx->mock; - int param0 = 5, param1 = -5; - const char *two_param_types[] = {"int", "int"}; - const void *two_params[] = {&param0, &param1}; - const void *ret; - - ret = mock->do_expect(mock, - "test_printk", test_printk, - two_param_types, two_params, - ARRAY_SIZE(two_params)); - KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ret); - KUNIT_EXPECT_EQ(test, -4, *((int *) ret)); + int *a, i, r = 1; + a = kunit_kmalloc_array(test, TEST_LEN, sizeof(*a), GFP_KERNEL); + KUNIT_ASSERT_NOT_ERR_OR_NULL(test, a); + for (i = 0; i < TEST_LEN; i++) { + r = (r * 725861) % 6599; + a[i] = r; + } + sort(a, TEST_LEN, sizeof(*a), cmpint, NULL); + for (i = 0; i < TEST_LEN-1; i++) + KUNIT_EXPECT_LE(test, a[i], a[i + 1]); }

-In this example, the method under test should return a pointer to a value, so -if the pointer returned by the method is null or an errno, we don't want to -bother continuing the test since the following expectation could crash the test -case. `ASSERT_NOT_ERR_OR_NULL(...)` allows us to bail out of the test case if -the appropriate conditions have not been satisfied to complete the test. +In this example, the method under test should return pointer to a value. If the +pointer returns null or an errno, we want to stop the test since the following +expectation could crash the test case. `ASSERT_NOT_ERR_OR_NULL(...)` allows us +to bail out of the test case if the appropriate conditions are not satisfied to +complete the test.

Test Suites ~~~~~~~~~~~

-Now obviously one unit test isn't very helpful; the power comes from having -many test cases covering all of a unit's behaviors. Consequently it is common -to have many *similar* tests; in order to reduce duplication in these closely -related tests most unit testing frameworks - including KUnit - provide the -concept of a *test suite*. A *test suite* is just a collection of test cases -for a unit of code with a set up function that gets invoked before every test -case and then a tear down function that gets invoked after every test case -completes. - -Example: +We need many test cases covering all the unit's behaviors. It is common to have +many similar tests. In order to reduce duplication in these closely related +tests, most unit testing frameworks (including KUnit) provide the concept of a +*test suite*. A test suite is a collection of test cases for a unit of code +with a setup function that gets invoked before every test case and then a tear +down function that gets invoked after every test case completes. For example:

.. code-block:: c

@@ -202,23 +145,48 @@ Example: }; kunit_test_suite(example_test_suite);

-In the above example the test suite, ``example_test_suite``, would run the test -cases ``example_test_foo``, ``example_test_bar``, and ``example_test_baz``; -each would have ``example_test_init`` called immediately before it and would -have ``example_test_exit`` called immediately after it. +In the above example, the test suite ``example_test_suite`` would run the test +cases ``example_test_foo``, ``example_test_bar``, and ``example_test_baz``. Each +would have ``example_test_init`` called immediately before it and +``example_test_exit`` called immediately after it. ``kunit_test_suite(example_test_suite)`` registers the test suite with the KUnit test framework.

.. note:: - A test case will only be run if it is associated with a test suite. + A test case will only run if it is associated with a test suite.

-``kunit_test_suite(...)`` is a macro which tells the linker to put the specified -test suite in a special linker section so that it can be run by KUnit either -after late_init, or when the test module is loaded (depending on whether the -test was built in or not). +``kunit_test_suite(...)`` is a macro which tells the linker to put the +specified test suite in a special linker section so that it can be run by KUnit +either after ``late_init``, or when the test module is loaded (if the test was +built as a module).

-For more information on these types of things see the -Documentation/dev-tools/kunit/api/test.rst. +For more information, see Documentation/dev-tools/kunit/api/test.rst. + +Writing Tests For Other Architectures +------------------------------------- + +Always prefer tests that run on UML to tests that only run under a particular +architecture. In addition, prefer tests that run under QEMU or another easy +(and monetarily free) to obtain software environment to a specific piece of +hardware. + +Nevertheless, there are still valid reasons to write an architecture or +hardware specific test. For example, we might want to test code that really +belongs in ``arch/some-arch/*``. Even so, try to write the test so that it does +not depend on physical hardware. Some of our test cases may not need hardware, +only few tests actually require the hardware to test it. When hardware is not +available, instead of disabling tests, we can skip them. + +Now that we have narrowed down exactly what bits are hardware specific, the +actual procedure for writing and running the tests is same as writing normal +KUnit tests. + +.. important:: + We may have to reset hardware state. If this is not possible, we may only + be able to run one test case per invocation. + +.. TODO(brendanhiggins@google.com): Add an actual example of an architecture- + dependent KUnit test.

Common Patterns =============== @@ -226,43 +194,39 @@ Common Patterns Isolating Behavior ------------------

-The most important aspect of unit testing that other forms of testing do not -provide is the ability to limit the amount of code under test to a single unit. -In practice, this is only possible by being able to control what code gets run -when the unit under test calls a function and this is usually accomplished -through some sort of indirection where a function is exposed as part of an API -such that the definition of that function can be changed without affecting the -rest of the code base. In the kernel this primarily comes from two constructs, -classes, structs that contain function pointers that are provided by the -implementer, and architecture-specific functions which have definitions selected -at compile time. +Unit testing limits the amount of code under test to a single unit. It controls +what code gets run when the unit under test calls a function. Where a function +is exposed as part of an API such that the definition of that function can be +changed without affecting the rest of the code base. In the kernel, this comes +from two constructs: classes, structs. that contain function pointers provided +by the implementer and architecture specific functions which have definitions +selected at compile time.

Classes ~~~~~~~

Classes are not a construct that is built into the C programming language; -however, it is an easily derived concept. Accordingly, pretty much every project -that does not use a standardized object oriented library (like GNOME's GObject) -has their own slightly different way of doing object oriented programming; the -Linux kernel is no exception. +however, it is an easily derived concept. Accordingly, in most cases, every +project that does not use a standardized object oriented library (like GNOME's +GObject) has their own slightly different way of doing object oriented +programming; the Linux kernel is no exception.

The central concept in kernel object oriented programming is the class. In the kernel, a *class* is a struct that contains function pointers. This creates a contract between *implementers* and *users* since it forces them to use the -same function signature without having to call the function directly. In order -for it to truly be a class, the function pointers must specify that a pointer -to the class, known as a *class handle*, be one of the parameters; this makes -it possible for the member functions (also known as *methods*) to have access -to member variables (more commonly known as *fields*) allowing the same -implementation to have multiple *instances*. - -Typically a class can be *overridden* by *child classes* by embedding the -*parent class* in the child class. Then when a method provided by the child -class is called, the child implementation knows that the pointer passed to it is -of a parent contained within the child; because of this, the child can compute -the pointer to itself because the pointer to the parent is always a fixed offset -from the pointer to the child; this offset is the offset of the parent contained -in the child struct. For example: +same function signature without having to call the function directly. To be a +class, the function pointers must specify that a pointer to the class, known as +a *class handle*, be one of the parameters. Thus the member functions (also +known as *methods*) have access to member variables (also known as *fields*) +allowing the same implementation to have multiple *instances*. + +A class can be *overridden* by *child classes* by embedding the *parent class* +in the child class. Then when the child class *method* is called, the child +implementation knows that the pointer passed to it is of a parent contained +within the child. Thus, the child can compute the pointer to itself because the +pointer to the parent is always a fixed offset from the pointer to the child. +This offset is the offset of the parent contained in the child struct. For +example:

.. code-block:: c

@@ -290,8 +254,8 @@ in the child struct. For example: self->width = width; }

-In this example (as in most kernel code) the operation of computing the pointer -to the child from the pointer to the parent is done by ``container_of``. +In this example, computing the pointer to the child from the pointer to the +parent is done by ``container_of``.

Faking Classes ~~~~~~~~~~~~~~ @@ -300,14 +264,11 @@ In order to unit test a piece of code that calls a method in a class, the behavior of the method must be controllable, otherwise the test ceases to be a unit test and becomes an integration test.

-A fake just provides an implementation of a piece of code that is different than -what runs in a production instance, but behaves identically from the standpoint -of the callers; this is usually done to replace a dependency that is hard to -deal with, or is slow. - -A good example for this might be implementing a fake EEPROM that just stores the -"contents" in an internal buffer. For example, let's assume we have a class that -represents an EEPROM: +A fake class implements a piece of code that is different than what runs in a +production instance, but behaves identical from the standpoint of the callers. +This is done to replace a dependency that is hard to deal with, or is slow. For +example, implementing a fake EEPROM that stores the "contents" in an +internal buffer. Assume we have a class that represents an EEPROM:

.. code-block:: c

@@ -316,7 +277,7 @@ represents an EEPROM: ssize_t (*write)(struct eeprom *this, size_t offset, const char *buffer, size_t count); };

-And we want to test some code that buffers writes to the EEPROM: +We want to test code that buffers writes to the EEPROM:

.. code-block:: c

@@ -329,7 +290,7 @@ And we want to test some code that buffers writes to the EEPROM: struct eeprom_buffer *new_eeprom_buffer(struct eeprom *eeprom); void destroy_eeprom_buffer(struct eeprom *eeprom);

-We can easily test this code by *faking out* the underlying EEPROM: +We can test this code by *faking out* the underlying EEPROM:

.. code-block:: c

@@ -456,14 +417,14 @@ We can now use it to test ``struct eeprom_buffer``: destroy_eeprom_buffer(ctx->eeprom_buffer); }

-Testing against multiple inputs +Testing Against Multiple Inputs -------------------------------

-Testing just a few inputs might not be enough to have confidence that the code -works correctly, e.g. for a hash function. +Testing just a few inputs is not enough to ensure that the code works correctly, +for example: testing a hash function.

-In such cases, it can be helpful to have a helper macro or function, e.g. this -fictitious example for ``sha1sum(1)`` +We can write a helper macro or function. The function is called for each input. +For example, to test ``sha1sum(1)``, we can write:

.. code-block:: c

@@ -475,16 +436,15 @@ fictitious example for ``sha1sum(1)`` TEST_SHA1("hello world", "2aae6c35c94fcfb415dbe95f408b9ce91ee846ed"); TEST_SHA1("hello world!", "430ce34d020724ed75a196dfc2ad67c77772d169");

+Note the use of the ``_MSG`` version of ``KUNIT_EXPECT_STREQ`` to print a more +detailed error and make the assertions clearer within the helper macros.

-Note the use of ``KUNIT_EXPECT_STREQ_MSG`` to give more context when it fails -and make it easier to track down. (Yes, in this example, ``want`` is likely -going to be unique enough on its own). +The ``_MSG`` variants are useful when the same expectation is called multiple +times (in a loop or helper function) and thus the line number is not enough to +identify what failed, as shown below.

-The ``_MSG`` variants are even more useful when the same expectation is called -multiple times (in a loop or helper function) and thus the line number isn't -enough to identify what failed, like below. - -In some cases, it can be helpful to write a *table-driven test* instead, e.g. +In complicated cases, we recommend using a *table-driven test* compared to the +helper macro variation, for example:

.. code-block:: c

@@ -513,17 +473,18 @@ In some cases, it can be helpful to write a *table-driven test* instead, e.g. }

-There's more boilerplate involved, but it can: +There is more boilerplate code involved, but it can: + +* be more readable when there are multiple inputs/outputs (due to field names).

-* be more readable when there are multiple inputs/outputs thanks to field names, + * For example, see ``fs/ext4/inode-test.c``.

- * E.g. see ``fs/ext4/inode-test.c`` for an example of both. -* reduce duplication if test cases can be shared across multiple tests. +* reduce duplication if test cases are shared across multiple tests.

- * E.g. if we wanted to also test ``sha256sum``, we could add a ``sha256`` + * For example: if we want to test ``sha256sum``, we could add a ``sha256`` field and reuse ``cases``.

-* be converted to a "parameterized test", see below. +* be converted to a "parameterized test".

Parameterized Testing ~~~~~~~~~~~~~~~~~~~~~ @@ -531,7 +492,7 @@ Parameterized Testing The table-driven testing pattern is common enough that KUnit has special support for it.

-Reusing the same ``cases`` array from above, we can write the test as a +By reusing the same ``cases`` array from above, we can write the test as a "parameterized test" with the following.

.. code-block:: c @@ -582,193 +543,152 @@ Reusing the same ``cases`` array from above, we can write the test as a

.. _kunit-on-non-uml:

-KUnit on non-UML architectures -============================== - -By default KUnit uses UML as a way to provide dependencies for code under test. -Under most circumstances KUnit's usage of UML should be treated as an -implementation detail of how KUnit works under the hood. Nevertheless, there -are instances where being able to run architecture-specific code or test -against real hardware is desirable. For these reasons KUnit supports running on -other architectures. - -Running existing KUnit tests on non-UML architectures ------------------------------------------------------ +Exiting Early on Failed Expectations +------------------------------------

-There are some special considerations when running existing KUnit tests on -non-UML architectures: +We can use ``KUNIT_EXPECT_EQ`` to mark the test as failed and continue +execution. In some cases, it is unsafe to continue. We can use the +``KUNIT_ASSERT`` variant to exit on failure.

-* Hardware may not be deterministic, so a test that always passes or fails - when run under UML may not always do so on real hardware. -* Hardware and VM environments may not be hermetic. KUnit tries its best to - provide a hermetic environment to run tests; however, it cannot manage state - that it doesn't know about outside of the kernel. Consequently, tests that - may be hermetic on UML may not be hermetic on other architectures. -* Some features and tooling may not be supported outside of UML. -* Hardware and VMs are slower than UML. +.. code-block:: c

-None of these are reasons not to run your KUnit tests on real hardware; they are -only things to be aware of when doing so. + void example_test_user_alloc_function(struct kunit *test) + { + void *object = alloc_some_object_for_me();

-Currently, the KUnit Wrapper (``tools/testing/kunit/kunit.py``) (aka -kunit_tool) only fully supports running tests inside of UML and QEMU; however, -this is only due to our own time limitations as humans working on KUnit. It is -entirely possible to support other emulators and even actual hardware, but for -now QEMU and UML is what is fully supported within the KUnit Wrapper. Again, to -be clear, this is just the Wrapper. The actualy KUnit tests and the KUnit -library they are written in is fully architecture agnostic and can be used in -virtually any setup, you just won't have the benefit of typing a single command -out of the box and having everything magically work perfectly. + /* Make sure we got a valid pointer back. */ + KUNIT_ASSERT_NOT_ERR_OR_NULL(test, object); + do_something_with_object(object); + }

-Again, all core KUnit framework features are fully supported on all -architectures, and using them is straightforward: Most popular architectures -are supported directly in the KUnit Wrapper via QEMU. Currently, supported -architectures on QEMU include: +Allocating Memory +-----------------

-* i386 -* x86_64 -* arm -* arm64 -* alpha -* powerpc -* riscv -* s390 -* sparc +We can use ``kzalloc``, you should prefer ``kunit_kzalloc`` and KUnit will +ensure that the memory is freed once the test completes.

-In order to run KUnit tests on one of these architectures via QEMU with the -KUnit wrapper, all you need to do is specify the flags ``--arch`` and -``--cross_compile`` when invoking the KUnit Wrapper. For example, we could run -the default KUnit tests on ARM in the following manner (assuming we have an ARM -toolchain installed): +This is useful because it lets us use the ``KUNIT_ASSERT_EQ`` macros to exit +early from a test without having to worry about remembering to call ``kfree``. +For example:

-.. code-block:: bash +.. code-block:: c

- tools/testing/kunit/kunit.py run --timeout=60 --jobs=12 --arch=arm --cross_compile=arm-linux-gnueabihf- + void example_test_allocation(struct kunit *test) + { + char *buffer = kunit_kzalloc(test, 16, GFP_KERNEL); + /* Ensure allocation succeeded. */ + KUNIT_ASSERT_NOT_ERR_OR_NULL(test, buffer);

-Alternatively, if you want to run your tests on real hardware or in some other -emulation environment, all you need to do is to take your kunitconfig, your -Kconfig options for the tests you would like to run, and merge them into -whatever config your are using for your platform. That's it! + KUNIT_ASSERT_STREQ(test, buffer, ""); + }

-For example, let's say you have the following kunitconfig:

-.. code-block:: none +Testing Static Functions +------------------------

- CONFIG_KUNIT=y - CONFIG_KUNIT_EXAMPLE_TEST=y +If we do not want to expose functions or variables for testing, one option is to +conditionally ``#include`` the test file at the end of your .c file. For +example:

-If you wanted to run this test on an x86 VM, you might add the following config -options to your ``.config``: +.. code-block:: c

-.. code-block:: none + /* In my_file.c */

- CONFIG_KUNIT=y - CONFIG_KUNIT_EXAMPLE_TEST=y - CONFIG_SERIAL_8250=y - CONFIG_SERIAL_8250_CONSOLE=y + static int do_interesting_thing();

-All these new options do is enable support for a common serial console needed -for logging. + #ifdef CONFIG_MY_KUNIT_TEST + #include "my_kunit_test.c" + #endif

-Next, you could build a kernel with these tests as follows: +Injecting Test-Only Code +------------------------

+Similar to as shown above, we can add test-specific logic. For example:

-.. code-block:: bash +.. code-block:: c

- make ARCH=x86 olddefconfig - make ARCH=x86 + /* In my_file.h */

-Once you have built a kernel, you could run it on QEMU as follows: + #ifdef CONFIG_MY_KUNIT_TEST + /* Defined in my_kunit_test.c */ + void test_only_hook(void); + #else + void test_only_hook(void) { } + #endif

-.. code-block:: bash +This test-only code can be made more useful by accessing the current ``kunit_test`` +as shown in next section: *Accessing The Current Test*.

- qemu-system-x86_64 -enable-kvm \ - -m 1024 \ - -kernel arch/x86_64/boot/bzImage \ - -append 'console=ttyS0' \ - --nographic +Accessing The Current Test +--------------------------

-Interspersed in the kernel logs you might see the following: +In some cases, we need to call test-only code from outside the test file. +For example, see example in section *Injecting Test-Only Code* or if +we are providing a fake implementation of an ops struct. Using +``kunit_test`` field in ``task_struct``, we can access it via +``current->kunit_test``.

-.. code-block:: none +Below example includes how to implement "mocking":

- TAP version 14 - # Subtest: example - 1..1 - # example_simple_test: initializing - ok 1 - example_simple_test - ok 1 - example +.. code-block:: c

-Congratulations, you just ran a KUnit test on the x86 architecture! + #include <linux/sched.h> /* for current */

-In a similar manner, kunit and kunit tests can also be built as modules, -so if you wanted to run tests in this way you might add the following config -options to your ``.config``: + struct test_data { + int foo_result; + int want_foo_called_with; + };

-.. code-block:: none + static int fake_foo(int arg) + { + struct kunit *test = current->kunit_test; + struct test_data *test_data = test->priv;

- CONFIG_KUNIT=m - CONFIG_KUNIT_EXAMPLE_TEST=m + KUNIT_EXPECT_EQ(test, test_data->want_foo_called_with, arg); + return test_data->foo_result; + }

-Once the kernel is built and installed, a simple + static void example_simple_test(struct kunit *test) + { + /* Assume priv is allocated in the suite's .init */ + struct test_data *test_data = test->priv;

-.. code-block:: bash + test_data->foo_result = 42; + test_data->want_foo_called_with = 1;

- modprobe example-test + /* In a real test, we'd probably pass a pointer to fake_foo somewhere + * like an ops struct, etc. instead of calling it directly. */ + KUNIT_EXPECT_EQ(test, fake_foo(1), 42); + }

-...will run the tests.

-.. note:: - Note that you should make sure your test depends on ``KUNIT=y`` in Kconfig - if the test does not support module build. Otherwise, it will trigger - compile errors if ``CONFIG_KUNIT`` is ``m``. +Note: here we are able to get away with using ``test->priv``, but if we want +something more flexible we could use a named ``kunit_resource``, see +Documentation/dev-tools/kunit/api/test.rst.

-Writing new tests for other architectures ------------------------------------------ +Failing The Current Test +------------------------

-The first thing you must do is ask yourself whether it is necessary to write a -KUnit test for a specific architecture, and then whether it is necessary to -write that test for a particular piece of hardware. In general, writing a test -that depends on having access to a particular piece of hardware or software (not -included in the Linux source repo) should be avoided at all costs. +If we want to fail the current test, we can use ``kunit_fail_current_test(fmt, args...)`` +which is defined in ``<kunit/test-bug.h>`` and does not require pulling in ``<kunit/test.h>``. +For example, we have an option to enable some extra debug checks on some data +structures as shown below:

-Even if you only ever plan on running your KUnit test on your hardware -configuration, other people may want to run your tests and may not have access -to your hardware. If you write your test to run on UML, then anyone can run your -tests without knowing anything about your particular setup, and you can still -run your tests on your hardware setup just by compiling for your architecture. +.. code-block:: c

-.. important:: - Always prefer tests that run on UML to tests that only run under a particular - architecture, and always prefer tests that run under QEMU or another easy - (and monetarily free) to obtain software environment to a specific piece of - hardware. - -Nevertheless, there are still valid reasons to write an architecture or hardware -specific test: for example, you might want to test some code that really belongs -in ``arch/some-arch/*``. Even so, try your best to write the test so that it -does not depend on physical hardware: if some of your test cases don't need the -hardware, only require the hardware for tests that actually need it. - -Now that you have narrowed down exactly what bits are hardware specific, the -actual procedure for writing and running the tests is pretty much the same as -writing normal KUnit tests. One special caveat is that you have to reset -hardware state in between test cases; if this is not possible, you may only be -able to run one test case per invocation. + #include <kunit/test-bug.h>

-.. TODO(brendanhiggins@google.com): Add an actual example of an architecture- - dependent KUnit test. + #ifdef CONFIG_EXTRA_DEBUG_CHECKS + static void validate_my_data(struct data *data) + { + if (is_valid(data)) + return;

-KUnit debugfs representation -============================ -When kunit test suites are initialized, they create an associated directory -in ``/sys/kernel/debug/kunit/<test-suite>``. The directory contains one file + kunit_fail_current_test("data %p is invalid", data);

-- results: "cat results" displays results of each test case and the results - of the entire suite for the last test run. + /* Normal, non-KUnit, error reporting code here. */ + } + #else + static void my_debug_function(void) { } + #endif

-The debugfs representation is primarily of use when kunit test suites are -run in a native environment, either as modules or builtin. Having a way -to display results like this is valuable as otherwise results can be -intermixed with other events in dmesg output. The maximum size of each -results file is KUNIT_LOG_SIZE bytes (defined in ``include/kunit/test.h``).

Reword to align with other chapters.

Signed-off-by: Harinder Singh sharinder@google.com --- Documentation/dev-tools/kunit/faq.rst | 73 +++++++++++++-------------- 1 file changed, 36 insertions(+), 37 deletions(-)

diff --git a/Documentation/dev-tools/kunit/faq.rst b/Documentation/dev-tools/kunit/faq.rst index 5c6555d020f3..172e239791a8 100644 --- a/Documentation/dev-tools/kunit/faq.rst +++ b/Documentation/dev-tools/kunit/faq.rst @@ -4,56 +4,55 @@ Frequently Asked Questions ==========================

-How is this different from Autotest, kselftest, etc? -==================================================== +How is this different from Autotest, kselftest, and so on? +========================================================== KUnit is a unit testing framework. Autotest, kselftest (and some others) are not.

A `unit test https://martinfowler.com/bliki/UnitTest.html`_ is supposed to -test a single unit of code in isolation, hence the name. A unit test should be -the finest granularity of testing and as such should allow all possible code -paths to be tested in the code under test; this is only possible if the code -under test is very small and does not have any external dependencies outside of +test a single unit of code in isolation and hence the name *unit test*. A unit +test should be the finest granularity of testing and should allow all possible +code paths to be tested in the code under test. This is only possible if the +code under test is small and does not have any external dependencies outside of the test's control like hardware.

There are no testing frameworks currently available for the kernel that do not -require installing the kernel on a test machine or in a VM and all require -tests to be written in userspace and run on the kernel under test; this is true -for Autotest, kselftest, and some others, disqualifying any of them from being -considered unit testing frameworks. +require installing the kernel on a test machine or in a virtual machine. All +testing frameworks require tests to be written in userspace and run on the +kernel under test. This is true for Autotest, kselftest, and some others, +disqualifying any of them from being considered unit testing frameworks.

Does KUnit support running on architectures other than UML? ===========================================================

-Yes, well, mostly. +Yes, mostly.

-For the most part, the KUnit core framework (what you use to write the tests) -can compile to any architecture; it compiles like just another part of the +For the most part, the KUnit core framework (what we use to write the tests) +can compile to any architecture. It compiles like just another part of the kernel and runs when the kernel boots, or when built as a module, when the -module is loaded. However, there is some infrastructure, -like the KUnit Wrapper (``tools/testing/kunit/kunit.py``) that does not support -other architectures. +module is loaded. However, there is infrastructure, like the KUnit Wrapper +(``tools/testing/kunit/kunit.py``) that does not support other architectures.

-In short, this means that, yes, you can run KUnit on other architectures, but -it might require more work than using KUnit on UML. +In short, yes, you can run KUnit on other architectures, but it might require +more work than using KUnit on UML.

For more information, see :ref:`kunit-on-non-uml`.

-What is the difference between a unit test and these other kinds of tests? -========================================================================== +What is the difference between a unit test and other kinds of tests? +==================================================================== Most existing tests for the Linux kernel would be categorized as an integration test, or an end-to-end test.

-- A unit test is supposed to test a single unit of code in isolation, hence the - name. A unit test should be the finest granularity of testing and as such - should allow all possible code paths to be tested in the code under test; this - is only possible if the code under test is very small and does not have any - external dependencies outside of the test's control like hardware. +- A unit test is supposed to test a single unit of code in isolation. A unit + test should be the finest granularity of testing and, as such, allows all + possible code paths to be tested in the code under test. This is only possible + if the code under test is small and does not have any external dependencies + outside of the test's control like hardware. - An integration test tests the interaction between a minimal set of components, usually just two or three. For example, someone might write an integration test to test the interaction between a driver and a piece of hardware, or to test the interaction between the userspace libraries the kernel provides and - the kernel itself; however, one of these tests would probably not test the + the kernel itself. However, one of these tests would probably not test the entire kernel along with hardware interactions and interactions with the userspace. - An end-to-end test usually tests the entire system from the perspective of the @@ -62,26 +61,26 @@ test, or an end-to-end test. hardware with a production userspace and then trying to exercise some behavior that depends on interactions between the hardware, the kernel, and userspace.

-KUnit isn't working, what should I do? -====================================== +KUnit is not working, what should I do? +=======================================

Unfortunately, there are a number of things which can break, but here are some things to try.

-1. Try running ``./tools/testing/kunit/kunit.py run`` with the ``--raw_output`` +1. Run ``./tools/testing/kunit/kunit.py run`` with the ``--raw_output`` parameter. This might show details or error messages hidden by the kunit_tool parser. 2. Instead of running ``kunit.py run``, try running ``kunit.py config``, ``kunit.py build``, and ``kunit.py exec`` independently. This can help track down where an issue is occurring. (If you think the parser is at fault, you - can run it manually against stdin or a file with ``kunit.py parse``.) -3. Running the UML kernel directly can often reveal issues or error messages - kunit_tool ignores. This should be as simple as running ``./vmlinux`` after - building the UML kernel (e.g., by using ``kunit.py build``). Note that UML - has some unusual requirements (such as the host having a tmpfs filesystem - mounted), and has had issues in the past when built statically and the host - has KASLR enabled. (On older host kernels, you may need to run ``setarch - `uname -m` -R ./vmlinux`` to disable KASLR.) + can run it manually against ``stdin`` or a file with ``kunit.py parse``.) +3. Running the UML kernel directly can often reveal issues or error messages, + ``kunit_tool`` ignores. This should be as simple as running ``./vmlinux`` + after building the UML kernel (for example, by using ``kunit.py build``). + Note that UML has some unusual requirements (such as the host having a tmpfs + filesystem mounted), and has had issues in the past when built statically and + the host has KASLR enabled. (On older host kernels, you may need to run + ``setarch `uname -m` -R ./vmlinux`` to disable KASLR.) 4. Make sure the kernel .config has ``CONFIG_KUNIT=y`` and at least one test (e.g. ``CONFIG_KUNIT_EXAMPLE_TEST=y``). kunit_tool will keep its .config around, so you can see what config was used after running ``kunit.py run``.