Annotation of wikisrc/kyua/import.mdwn, revision 1.13

1.1       jmmv        1: [[!meta title="Kyua: The way into NetBSD"]]
1.11      wiki        2: 
                      3: **Contents**
1.1       jmmv        5: [[!toc levels=2]]
                      7: **Project owner: [Julio Merino](**  
1.12      jmmv        8: **Status: Import of Kyua done as of 2013-02-26.  All new code still
1.13    ! jmmv        9: disabled by default.  Call for testing sent on 2013-02-28.**
1.1       jmmv       10: 
                     11: The import of Kyua into NetBSD to replace the deprecated ATF tools is
                     12: planned to happen in NetBSD 7.0.  The ATF libraries will remain in place,
                     13: and as such no changes will happen to the tests that already live in
                     14: `src/tests`.
                     16: The transition from ATF to Kyua includes steps and tools to offer backwards
                     17: compatibility with users that may rely on the ATF tools shipped since
                     18: NetBSD 5.0.  These backwards compatibility tools cover the most common use
                     19: cases but might not be perfect initially.  Despite these provisions, this
                     20: plan should be executed well in advance the creation of the 7.0 branch to
                     21: ensure there is enough time in NetBSD-current to flesh out any major
                     22: problems.
1.3       jmmv       24: **Before moving on, please read the
1.2       jmmv       25: [[Kyua: An introduction for NetBSD users|/kyua]] page.  You should be
1.1       jmmv       26: familiar with the structure of Kyua and its major components to be able to
1.3       jmmv       27: review this plan.**
1.1       jmmv       28: 
                     29: # Proposed changes
                     31: The final user visible changes that this project will bring to NetBSD are
                     32: as follows.  Because these change the availability of tools that have
                     33: already been shipped since 5.0, the modifications are staged to happen
                     34: across two major releases:
                     36: ## For NetBSD 7.0
                     38: * Addition of the new `kyua` command-line tool.  This provides the runtime
                     39:   engine for the test programs (`atf-run`'s functionality) and the ability
                     40:   to generate reports of the results in plain text form and HTML form
                     41:   (`atf-report`'s functionality).  This tool is able to execute all the
                     42:   existing test programs without modifications.
                     44: * Replacement of the `Atffile`s files in `/usr/tests` with `Kyuafile`s.
                     46: * Ability to generate HTML reports right form the base system, with such
                     47:   reports hooked into the various continuous build systems.
                     49: * Replacement of the `atf-run` and `atf-report` tools with shell scripts
                     50:   that provide backwards compatibility implementations based on `kyua`.
                     51:   These can deal with old-style `Atffile`s so that users with custom test
                     52:   suites can continue to run them.
                     54: * Introduction of a new `atf2kyua` script to convert old `Atffile`s to
                     55:   `Kyuafile`s.  This is used internally by the compatibility `atf-run`
                     56:   script, but can also be invoked by the end user by hand to deal with his
                     57:   own test suites.  Could be placed into `libexec` if we do not want to
                     58:   make this public.
                     60: ## For NetBSD 8.0
                     62: * Removal of the backwards compatibility `atf-run` and `atf-report`
                     63:   scripts, as well as the supporting `atf2kyua` tool.
1.3       jmmv       65: # Why?
                     67: As mentioned in [[another page|/kyua]], Kyua should be seen as ATF 2.x even
                     68: though it carries a different name.  It is the evolution of the previous
                     69: ATF tools (*the tools only*), but written in a more modular and flexible
                     70: way, and with a more reliable codebase.  Therefore, you should consider
                     71: this project as the update of ATF to a newer version.
                     73: The `atf-run` and `atf-report` tools have effectively been in maintenance
                     74: mode for over a year already.  None of the desired features (see the list
                     75: of open PRs) have been implemented on top of them, mostly because doing so
                     76: is building upon a broken implementation.  Additionally, several developers
                     77: have had to implement their own test results dashboards due to
                     78: defficiencies in `atf-report`, effectively reinventing the wheel.
                     80: This update will permit the real removal of the obsolete tools, thus
                     81: allowing us to build additional features on top of Kyua without having to
                     82: worry about being compatible with `atf-run` (and thus adjusting this tool
                     83: to behave in the same manner).
                     85: Some possible answers to "Why not?" can be found later in this page.
                     87: ## And why now?
1.1       jmmv       88: 
                     89: Kyua has existed for almost 3 years already, so you may be wondering why
                     90: this import is being proposed now.  The major reasons are:
                     92: * Since 0.5, Kyua has now feature-parity with ATF.  Replacing the ATF tools
                     93:   with Kyua should not introduce functionality regressions.
                     95: * NetBSD-6 has been branched relatively recently, so we can expect NetBSD-7
                     96:   to be far away enough in the future to provide plenty of time to
                     97:   stabilize Kyua in NetBSD-current.
                     99: * The existence of the deprecated `atf-run` and `atf-report` tools hinders
                    100:   the development of new features that require changes to the inteface of
                    101:   ATF test programs.  As an example: a highly requested feature is the
                    102:   ability to change the timeout settings of the test programs (for the
                    103:   benefit of old, slow platforms); doing this in the ATF codebase is
                    104:   tricky.
                    106: * Some developers continue to add old-style tests to the tree.  While this
                    107:   is suboptimal and against the policy dictated by core@, Kyua brings in a
                    108:   mechanism to allow running these tests unmodified (i.e. without adding
                    109:   any ATF calls in their code).  This, in turn, will let such developers
                    110:   add their tests more easily, and thus increase the tests coverage.
                    112: # The plan
                    114: This section details what the transition plan looks like and an estimate
                    115: timeline.  As all things that depend on free time (opensource software
                    116: hacking, in this case), take the estimates with a grain of salt.
                    118: ## Discuss this plan in tech-userlevel@
                    120: If you are here it's possibly because a review request for this plan has
1.4       jmmv      121: already been published and thus the plan has already begun.
1.1       jmmv      122: 
                    123: This plan will be sent to the tech-userlevel@ mailing list asking for
                    124: comments.  **Two weeks shall be allowed for initial discussion.** Depending
                    125: on the outcome, the plan and/or the software will need to be adjusted
                    126: (which in turn may require significant amounts of time not yet accounted
                    127: for).  I'll be optimistic for now.
                    129: ## Get core@ approval
                    131: By policy, the import of any new software component into src requires core@
                    132: approval.  Even though this will have been discussed at length in
                    133: tech-userlevel@ (as per the previous step), the final decider on this issue
                    134: will be core@.  The corollary of this is that, if no consensus can be
                    135: reached in tech-userlevel@ regarding this plan, core@ will be asked to come
                    136: up with a decision.
                    138: If core@ approves the plan, the next steps shall start immediately.  If
                    139: core@ disagrees, core@ will be asked to provide advice on the corrections
                    140: that should be made before the plan can be approved.
1.4       jmmv      142: It is hard to tell how long this step will last, but possibly account for 2
                    143: to 4 weeks.
1.5       jmmv      145: ## Address feedback as a new release
                    147: Publish a new Kyua release that collects all the feedback from the reviews
                    148: above.
                    150: The list of issues to be addressed can be found by querying the
                    151: [bug tracker]( for the
                    152: [Milestone=Release0.6](
                    153: keyword.  In particular, the following are the issues that have arisen from
                    154: the review:
1.9       jmmv      156: * [Issue 36]( Fix the
1.10      jmmv      157:   `help` command to not fail if the configuration file is bogus.  **DONE**
1.9       jmmv      158: 
                    159: * [Issue 37]( Simplify
                    160:   the syntax definition of configuration and `Kyuafile` files by removing
1.10      jmmv      161:   the format name.  **DONE**
1.9       jmmv      162: 
1.5       jmmv      163: * [Issue 40]( Provide
1.6       jmmv      164:   manpages instead of an info document.  **DONE**
1.5       jmmv      165: 
1.8       jmmv      166: * [Issue 47]( Implement
                    167:   independent testers, which reduces the amount of C++ code and avoids the
                    168:   need of modifying ``.  **DONE**
1.9       jmmv      170: * [Issue 57]( Generalized
                    171:   support for all metadata properties to plain test programs.  This is to
                    172:   make plain test programs more versatile by bringing them closer to feature
1.10      jmmv      173:   parity with ATF test programs.  **DONE**
1.9       jmmv      174: 
1.1       jmmv      175: ## Import Kyua into src
1.2       jmmv      177: As the [[introductory page to Kyua|/kyua]] describes, Kyua has been
                    178: available in pkgsrc for a while and can be readily installed and used to
                    179: run the tests from `/usr/tests`.
1.1       jmmv      180: 
                    181: However, because ATF lives in src, and because NetBSD aims to provide the
                    182: best environment for testing "out of the box", Kyua should be imported into
                    183: src just like ATF was.  The major reasons for this, as have been explained
                    184: in the past, are: first, to allow any new deployment of NetBSD to be
                    185: validated right after installation and continuously afterwards; and,
                    186: second, to permit the execution of tests during development without having
                    187: to install any additional software.
                    189: The specific steps to perform this import are:
                    191: 1. Import Lutok into `external/bsd/lutok/`.  This is a shared library that
                    192:    wraps the native Lua C interface in C++.  Lutok was originally part of
                    193:    Kyua, and was split into its own package per the request of some users
                    194:    that found this component useful on its own.
1.8       jmmv      196: 1. Import the Kyua testers into `external/bsd/kyua-testers/`.  This yields two
                    197:    new binaries in `/usr/libexec` (`kyua-atf-tester` and
                    198:    `kyua-plain-tester`) and a bunch of tests in `/usr/tests/kyua-testers`.
                    200: 1. Import the Kyua frontend into `external/bsd/kyua-cli/`.  This yields a
                    201:    new kyua binary in `/usr/bin`, a lot of test programs in
                    202:    `/usr/tests/kyua-cli` (around 100) and some auxiliary files in
                    203:    `/usr/share/kyua`.
1.1       jmmv      204: 
1.6       jmmv      205: 1. Protect all products of Lutok and Kyua with a new `MKKYUA` knob.  **Set
                    206:    `MKKYUA=no` by default.**  Once the ATF tools are removed, the existence
                    207:    of both the `MKATF` and `MKKYUA` knows will probably be confusing.  When
                    208:    that happens, we can revisit this decision by possibly replacing both
                    209:    with an `MKTESTS`.
1.1       jmmv      210: 
                    211: 1. Update `` to generate `Kyuafile`s *in addition to*
                    212:    `Atffile`s.
                    214: There is no real need to do this import in a branch given that this import
1.6       jmmv      215: only adds new functionality without touching existing stuff, and the new
                    216: code is disabled by default.
1.1       jmmv      217: 
                    218: All the preparatory work for the import can be done offline (in about two
                    219: weeks at most, given that I have mot of this ready).  Aside from the code
                    220: changes, this will involve the validation of NetBSD/amd64, NetBSD/i386 and
                    221: NetBSD/macppc builds (which are the ports I have access to).  If you
                    222: consider that some other tricky architecture should be build-tested
                    223: (sparc64?), let me know and I'll include it in the list.
1.8       jmmv      225: The submission to CVS will be prepared locally and performed on a package
                    226: basis (i.e. `lutok`, `kyua-testers` and `kyua-cli`, in this order).  These
                    227: are to be imported separately to simplify the review of the changes and to
                    228: allow me to better test every individual change locally.  There may be an
                    229: arbitrary amount of time between the submission of each package: this
                    230: should not be a problem because these modules are still disabled due to
                    231: `MKKYUA` being set to `no` by default.
1.1       jmmv      232: 
                    233: ## Adjust continuous testing systems to use Kyua
                    235: With `kyua` being part of the release sets, it is possible to adjust the
                    236: continuous test systems to make use of this tool in the test environments
                    237: without having to take any additional step.
                    239: I'll work with gson@ and pgoyette@ to adapt their continuous testing
                    240: machines to use the new built-in `kyua` binary instead of `atf-run` and
                    241: `atf-report`.  I'm planning to do the necessary work to change `anita`
                    242: myself, and I expect to help them deploy the changes to their own systems.
                    243: Because Kyua and ATF will cohexist in the base system at this point,
                    244: migrating the continuous testing systems to Kyua can happen at its own
                    245: peace.
                    247: It might happen that Kyua misses some little detail needed by these
                    248: systems.  In that case, this may require a new release of Kyua and a
                    249: reimport into the tree.  Incremental imports with new features are much
                    250: easier than the original import described in here.  Also, it will be
                    251: possible to cherry-pick any external changes into the tree without a
                    252: reimport (as has often been done in ATF).
1.4       jmmv      254: This step can take a few weeks of time, mostly due to the back and forth
                    255: between different people in different timezones.
1.13    ! jmmv      257: ## Flip MKKYUA to yes
1.6       jmmv      258: 
                    259: The previous steps imported Kyua but didn't enable its build by default so
                    260: that proper testing can be performed by the only people that care.  Once
                    261: basic testing (particularly build testing on a variety of platforms) is
                    262: performed, flip `MKKYUA=yes`.
                    264: ## Update documentation
                    266: The [[Kyua: An introduction for NetBSD users|/kyua]] wiki page will be
                    267: updated to explain how Kyua is bundled in NetBSD and how to use the bundled
                    268: version.
                    270: The [[Creating atf-based tests for NetBSD src|/tutorials/atf]] wiki page
                    271: will be updated to account for the differences in test programs execution
                    272: with Kyua instead of ATF.
                    274: The afterboot(8) and tests(7) manpages will be adjusted to mention `kyua`
                    275: instead of the ATF tools.
1.1       jmmv      277: ## User validation period
                    279: At this stage, **at least one month shall be given to the community** to
                    280: test the new tools and the new test results dashboards.  Collect feedback
1.5       jmmv      281: and address requests as appropriate (possibly by releasing and importing a
                    282: new version of Kyua).
                    284: One important thing to validate is that the results reported by `atf-run`
                    285: and `kyua test` match with each other.  I have already been validating this
                    286: with every public release of Kyua, but it has been a manual process.  To
                    287: make this more reliable, I will set up a continuous testing machine of my
                    288: own in which I will execute `atf-run` and `kyua test` in sequence (possibly
                    289: within anita) and will add an automatic comparison of the exit status of
                    290: each other.
1.1       jmmv      291: 
1.4       jmmv      292: *Because the old `atf-run` and `atf-report` tools have not yet been dropped
                    293: at this point, we can spend as much time as necessary on this phase to get
                    294: things right.*
1.1       jmmv      296: ## Replace atf-run and atf-report with kyua-atf-compat
                    298: Import the Kyua-based `atf-run` and `atf-report` compatibility tools and
                    299: stop building the deprecated versions of these.  The compatibility versions
                    300: are shipped in a separate `kyua-atf-compat` package, and thus this will be
                    301: imported into `external/bsd/kyua-atf-compat/`.
                    303: Once this is done, change `` to not generate `Atffile`s any
                    304: more, as the compatibility tools do not need them.
                    306: ## Drop atf-run and atf-report's code
                    308: Delete the native `atf-run` and `atf-report` utilities from the source tree
                    309: (and from the upstream repository).  This also gets rid of a lot of
                    310: supporting helper code, which makes the various ATF libraries leaner (and
                    311: therefore benefits all test programs!).
                    313: ## Get rid of some ATF wrappers as a proof of concept
                    315: There are some test programs (e.g. the ones in the ipf test suite) that are
                    316: not ATF-based.  To plug them into the NetBSD test suite, we have
                    317: implemented ATF-based wrappers that invoke the original plain tests.  This
                    318: is certainly suboptimal, as this level of indirection hinders readability
                    319: and makes development harder.  (In particular, it prevents some key
                    320: developers from contributing tests in the first place.)
                    322: Because Kyua has the ability to run these "plain" (non-ATF) test programs
                    323: directly, and because a bunch of developers are really looking forward to
                    324: this feature, I will convert a few tests to not include ATF wrappers as a
                    325: proof of concept of this feature.  The ipf tests are probably a good choice
                    326: for this.
                    328: ## Remove atf-run and atf-report compatibility tools
                    330: This is still really far away in the future (needs to happen after NetBSD
                    331: 7.0 is branched), but I'm listing it for completeness of the plan.  The
                    332: idea is to get rid of any ATF compatibility scripts by NetBSD 8.0.  This
                    333: gives end users a full release cycle (that of 7) to adapt to Kyua while at
                    334: the same time being able to use the old-style tools.
                    336: # Possible concerns
                    338: This section attempts to collect the list of possible concerns and/or
                    339: objections that may come up during the review, together with an attempt of
                    340: rebuttal from my side.
                    342: ## Increased compilation time
                    344: The addition of any new component to src increases the build time of the
                    345: whole system.  In the case of Kyua, this increase might be noticeable
                    346: *because of the large amount of test programs provided (roughly 100)*, all
                    347: of which are in C++.
                    349: Note that ATF also had a relatively large codebase and a bunch of tests
                    350: (although not as many).  The "trick" was that the majority of these tests
                    351: were written in shell, and as such they did not increase the build time by
                    352: much.  However, they significantly increased the run time of the test
                    353: suite, and they were less detailed (mostly integration tests, few unit
                    354: tests) than the Kyua tests.
                    356: In order to mitigate this issue, the build of all pieces of Kyua will be
1.5       jmmv      357: protected by `MKKYUA` so that people allergic to C++ can avoid the whole
1.1       jmmv      358: thing.  Even more, there are some other additional provisions described
                    359: below.
                    361: ## Introduction of more C++ code in base
                    363: First of all, why did I use C++?  To make the implementation simpler and
                    364: safer (the RAII programming pattern is really useful in avoiding memory and
                    365: resource leaks with minimal effort).  And C++ is part of the base system
                    366: and a supported language, so there was no reason not to do so.  But that's
1.4       jmmv      367: not the point of this item: if you dislike like C++, this is not going to
                    368: make you think I did right.
1.1       jmmv      369: 
                    370: It's true that, if we count the number of lines, Kyua brings in more C++
                    371: code than what will eventually be dropped by the removal of the ATF tools.
                    372: However, because ATF was also C++, the import of Kyua itself does not make
                    373: the situation significantly worse.
                    375: Additionally, the two compilers we can really use (GCC and LLVM) already
                    376: use, or will soon use, C++ in their code base.  It is unlikely that we will
                    377: be able to remove all C++ support from base anytime soon due to this, while
                    378: at the same time keeping support for all the ports that NetBSD has.
1.8       jmmv      380: Thanks to the
                    381: [testers project](, and
                    382: starting with Kyua 0.6, a lot of the tricky OS-specific code in Kyua has
                    383: been rewritten in plain C.  This paves the way to rewriting parts of the
                    384: now-simpler frontend in C or Lua, if the use of C++ proves to be a serious
                    385: problem in the future.
1.7       jmmv      386: 
                    387: In the short term, the replacement of ATF with Kyua does not make things
                    388: worse: this project just changes one chunk of code with another.
1.1       jmmv      389: 
                    390: ## No need for Lutok as a public library
                    392: Kyua depends on Lutok, which is a C++ interface to Lua.  The code of this
                    393: library was originally part of Kyua, but I split it into its own project
                    394: because some users asked for it.
                    396: If there is no desire to ship Lutok as a shared library in `/usr/lib`, we
                    397: can build Lutok as a static private library and link Kyua against it.
                    398: There is no need to install this as a shared library.

CVSweb for NetBSD wikisrc <> software: FreeBSD-CVSweb