Annotation of wikisrc/kyua/import.mdwn, revision 1.3
1.1 jmmv 1: [[!meta title="Kyua: The way into NetBSD"]]
2: [[!toc levels=2]]
4: **Project owner: [Julio Merino](mailto:jmmv@NetBSD.org).**
5: **Status: draft as of 2012-08-31; not even sent for review yet.**
7: The import of Kyua into NetBSD to replace the deprecated ATF tools is
8: planned to happen in NetBSD 7.0. The ATF libraries will remain in place,
9: and as such no changes will happen to the tests that already live in
12: The transition from ATF to Kyua includes steps and tools to offer backwards
13: compatibility with users that may rely on the ATF tools shipped since
14: NetBSD 5.0. These backwards compatibility tools cover the most common use
15: cases but might not be perfect initially. Despite these provisions, this
16: plan should be executed well in advance the creation of the 7.0 branch to
17: ensure there is enough time in NetBSD-current to flesh out any major
1.3 ! jmmv 20: **Before moving on, please read the
1.2 jmmv 21: [[Kyua: An introduction for NetBSD users|/kyua]] page. You should be
1.1 jmmv 22: familiar with the structure of Kyua and its major components to be able to
1.3 ! jmmv 23: review this plan.**
1.1 jmmv 24:
25: # Proposed changes
27: The final user visible changes that this project will bring to NetBSD are
28: as follows. Because these change the availability of tools that have
29: already been shipped since 5.0, the modifications are staged to happen
30: across two major releases:
32: ## For NetBSD 7.0
34: * Addition of the new `kyua` command-line tool. This provides the runtime
35: engine for the test programs (`atf-run`'s functionality) and the ability
36: to generate reports of the results in plain text form and HTML form
37: (`atf-report`'s functionality). This tool is able to execute all the
38: existing test programs without modifications.
40: * Replacement of the `Atffile`s files in `/usr/tests` with `Kyuafile`s.
42: * Ability to generate HTML reports right form the base system, with such
43: reports hooked into the various continuous build systems.
45: * Replacement of the `atf-run` and `atf-report` tools with shell scripts
46: that provide backwards compatibility implementations based on `kyua`.
47: These can deal with old-style `Atffile`s so that users with custom test
48: suites can continue to run them.
50: * Introduction of a new `atf2kyua` script to convert old `Atffile`s to
51: `Kyuafile`s. This is used internally by the compatibility `atf-run`
52: script, but can also be invoked by the end user by hand to deal with his
53: own test suites. Could be placed into `libexec` if we do not want to
54: make this public.
56: ## For NetBSD 8.0
58: * Removal of the backwards compatibility `atf-run` and `atf-report`
59: scripts, as well as the supporting `atf2kyua` tool.
1.3 ! jmmv 61: # Why?
! 63: As mentioned in [[another page|/kyua]], Kyua should be seen as ATF 2.x even
! 64: though it carries a different name. It is the evolution of the previous
! 65: ATF tools (*the tools only*), but written in a more modular and flexible
! 66: way, and with a more reliable codebase. Therefore, you should consider
! 67: this project as the update of ATF to a newer version.
! 69: The `atf-run` and `atf-report` tools have effectively been in maintenance
! 70: mode for over a year already. None of the desired features (see the list
! 71: of open PRs) have been implemented on top of them, mostly because doing so
! 72: is building upon a broken implementation. Additionally, several developers
! 73: have had to implement their own test results dashboards due to
! 74: defficiencies in `atf-report`, effectively reinventing the wheel.
! 76: This update will permit the real removal of the obsolete tools, thus
! 77: allowing us to build additional features on top of Kyua without having to
! 78: worry about being compatible with `atf-run` (and thus adjusting this tool
! 79: to behave in the same manner).
! 81: Some possible answers to "Why not?" can be found later in this page.
! 83: ## And why now?
1.1 jmmv 84:
85: Kyua has existed for almost 3 years already, so you may be wondering why
86: this import is being proposed now. The major reasons are:
88: * Since 0.5, Kyua has now feature-parity with ATF. Replacing the ATF tools
89: with Kyua should not introduce functionality regressions.
91: * NetBSD-6 has been branched relatively recently, so we can expect NetBSD-7
92: to be far away enough in the future to provide plenty of time to
93: stabilize Kyua in NetBSD-current.
95: * The existence of the deprecated `atf-run` and `atf-report` tools hinders
96: the development of new features that require changes to the inteface of
97: ATF test programs. As an example: a highly requested feature is the
98: ability to change the timeout settings of the test programs (for the
99: benefit of old, slow platforms); doing this in the ATF codebase is
102: * Some developers continue to add old-style tests to the tree. While this
103: is suboptimal and against the policy dictated by core@, Kyua brings in a
104: mechanism to allow running these tests unmodified (i.e. without adding
105: any ATF calls in their code). This, in turn, will let such developers
106: add their tests more easily, and thus increase the tests coverage.
108: # The plan
110: This section details what the transition plan looks like and an estimate
111: timeline. As all things that depend on free time (opensource software
112: hacking, in this case), take the estimates with a grain of salt.
114: ## Discuss this plan in tech-userlevel@
116: If you are here it's possibly because a review request for this plan has
117: already been published and thus the plan has already begun. If
118: not... well, hold off for a little bit ;-)
120: This plan will be sent to the tech-userlevel@ mailing list asking for
121: comments. **Two weeks shall be allowed for initial discussion.** Depending
122: on the outcome, the plan and/or the software will need to be adjusted
123: (which in turn may require significant amounts of time not yet accounted
124: for). I'll be optimistic for now.
126: ## Get core@ approval
128: By policy, the import of any new software component into src requires core@
129: approval. Even though this will have been discussed at length in
130: tech-userlevel@ (as per the previous step), the final decider on this issue
131: will be core@. The corollary of this is that, if no consensus can be
132: reached in tech-userlevel@ regarding this plan, core@ will be asked to come
133: up with a decision.
135: If core@ approves the plan, the next steps shall start immediately. If
136: core@ disagrees, core@ will be asked to provide advice on the corrections
137: that should be made before the plan can be approved.
139: ## Import Kyua into src
1.2 jmmv 141: As the [[introductory page to Kyua|/kyua]] describes, Kyua has been
142: available in pkgsrc for a while and can be readily installed and used to
143: run the tests from `/usr/tests`.
1.1 jmmv 144:
145: However, because ATF lives in src, and because NetBSD aims to provide the
146: best environment for testing "out of the box", Kyua should be imported into
147: src just like ATF was. The major reasons for this, as have been explained
148: in the past, are: first, to allow any new deployment of NetBSD to be
149: validated right after installation and continuously afterwards; and,
150: second, to permit the execution of tests during development without having
151: to install any additional software.
153: The specific steps to perform this import are:
155: 1. Import Lutok into `external/bsd/lutok/`. This is a shared library that
156: wraps the native Lua C interface in C++. Lutok was originally part of
157: Kyua, and was split into its own package per the request of some users
158: that found this component useful on its own.
160: 1. Import Kyua into `external/bsd/kyua-cli/`. This yields a new kyua
161: binary in `/usr/bin`, a lot of test programs in `/usr/tests/kyua-cli`
162: (around 100) and some auxiliary files in `/usr/share/kyua`.
164: 1. Protect all products of Lutok and Kyua with the existing `MKATF` knob.
165: If desired, these could instead be protected by a new and separate
166: `MKKYUA` knob, but because the transition is only temporary, I do not
167: think it's worth the effort.
169: 1. Update `bsd.test.mk` to generate `Kyuafile`s *in addition to*
172: There is no real need to do this import in a branch given that this import
173: only adds new functionality without touching existing stuff.
175: All the preparatory work for the import can be done offline (in about two
176: weeks at most, given that I have mot of this ready). Aside from the code
177: changes, this will involve the validation of NetBSD/amd64, NetBSD/i386 and
178: NetBSD/macppc builds (which are the ports I have access to). If you
179: consider that some other tricky architecture should be build-tested
180: (sparc64?), let me know and I'll include it in the list.
182: The submission step to CVS, once all the code changes are ready locally,
183: and any post-commit validation checks will take a few hours. Any build
184: breakage should be addressed in a timely manner (and can possibly be
185: worked-around by setting `MKATF=no`).
187: ## Update documentation
1.2 jmmv 189: The [[Kyua: An introduction for NetBSD users|/kyua]] wiki page will be
1.1 jmmv 190: updated to explain how Kyua is bundled in NetBSD and how to use the bundled
1.2 jmmv 193: The [[Creating atf-based tests for NetBSD src|/tutorials/atf]] wiki page
194: will be updated to account for the differences in test programs execution
195: with Kyua instead of ATF.
1.1 jmmv 196:
197: The afterboot(8) and tests(7) manpages will be adjusted to mention `kyua`
198: instead of the ATF tools.
200: ## Adjust continuous testing systems to use Kyua
202: With `kyua` being part of the release sets, it is possible to adjust the
203: continuous test systems to make use of this tool in the test environments
204: without having to take any additional step.
206: I'll work with gson@ and pgoyette@ to adapt their continuous testing
207: machines to use the new built-in `kyua` binary instead of `atf-run` and
208: `atf-report`. I'm planning to do the necessary work to change `anita`
209: myself, and I expect to help them deploy the changes to their own systems.
210: Because Kyua and ATF will cohexist in the base system at this point,
211: migrating the continuous testing systems to Kyua can happen at its own
214: It might happen that Kyua misses some little detail needed by these
215: systems. In that case, this may require a new release of Kyua and a
216: reimport into the tree. Incremental imports with new features are much
217: easier than the original import described in here. Also, it will be
218: possible to cherry-pick any external changes into the tree without a
219: reimport (as has often been done in ATF).
221: ## User validation period
223: At this stage, **at least one month shall be given to the community** to
224: test the new tools and the new test results dashboards. Collect feedback
225: and address requests as appropriate.
227: ## Replace atf-run and atf-report with kyua-atf-compat
229: Import the Kyua-based `atf-run` and `atf-report` compatibility tools and
230: stop building the deprecated versions of these. The compatibility versions
231: are shipped in a separate `kyua-atf-compat` package, and thus this will be
232: imported into `external/bsd/kyua-atf-compat/`.
234: Once this is done, change `bsd.test.mk` to not generate `Atffile`s any
235: more, as the compatibility tools do not need them.
237: ## Drop atf-run and atf-report's code
239: Delete the native `atf-run` and `atf-report` utilities from the source tree
240: (and from the upstream repository). This also gets rid of a lot of
241: supporting helper code, which makes the various ATF libraries leaner (and
242: therefore benefits all test programs!).
244: ## Get rid of some ATF wrappers as a proof of concept
246: There are some test programs (e.g. the ones in the ipf test suite) that are
247: not ATF-based. To plug them into the NetBSD test suite, we have
248: implemented ATF-based wrappers that invoke the original plain tests. This
249: is certainly suboptimal, as this level of indirection hinders readability
250: and makes development harder. (In particular, it prevents some key
251: developers from contributing tests in the first place.)
253: Because Kyua has the ability to run these "plain" (non-ATF) test programs
254: directly, and because a bunch of developers are really looking forward to
255: this feature, I will convert a few tests to not include ATF wrappers as a
256: proof of concept of this feature. The ipf tests are probably a good choice
257: for this.
259: ## Remove atf-run and atf-report compatibility tools
261: This is still really far away in the future (needs to happen after NetBSD
262: 7.0 is branched), but I'm listing it for completeness of the plan. The
263: idea is to get rid of any ATF compatibility scripts by NetBSD 8.0. This
264: gives end users a full release cycle (that of 7) to adapt to Kyua while at
265: the same time being able to use the old-style tools.
267: # Possible concerns
269: This section attempts to collect the list of possible concerns and/or
270: objections that may come up during the review, together with an attempt of
271: rebuttal from my side.
273: ## Increased compilation time
275: The addition of any new component to src increases the build time of the
276: whole system. In the case of Kyua, this increase might be noticeable
277: *because of the large amount of test programs provided (roughly 100)*, all
278: of which are in C++.
280: Note that ATF also had a relatively large codebase and a bunch of tests
281: (although not as many). The "trick" was that the majority of these tests
282: were written in shell, and as such they did not increase the build time by
283: much. However, they significantly increased the run time of the test
284: suite, and they were less detailed (mostly integration tests, few unit
285: tests) than the Kyua tests.
287: In order to mitigate this issue, the build of all pieces of Kyua will be
288: protected by `MKATF` so that people allergic to C++ can avoid the whole
289: thing. Even more, there are some other additional provisions described
292: ## Introduction of more C++ code in base
294: First of all, why did I use C++? To make the implementation simpler and
295: safer (the RAII programming pattern is really useful in avoiding memory and
296: resource leaks with minimal effort). And C++ is part of the base system
297: and a supported language, so there was no reason not to do so. But that's
298: not the point of this item: if you don't like C++, this is not going to
299: convince you otherwise.
301: It's true that, if we count the number of lines, Kyua brings in more C++
302: code than what will eventually be dropped by the removal of the ATF tools.
303: However, because ATF was also C++, the import of Kyua itself does not make
304: the situation significantly worse.
306: Additionally, the two compilers we can really use (GCC and LLVM) already
307: use, or will soon use, C++ in their code base. It is unlikely that we will
308: be able to remove all C++ support from base anytime soon due to this, while
309: at the same time keeping support for all the ports that NetBSD has.
311: Long term, if the use of C++ proves to be a problem, there are a few things
312: that can be done to slowly get rid of it that have been floating my mind
313: recently. The first is the rewrite of the performance-critical parts of
314: Kyua in plain C. This would involve splitting the runtime engine of a
315: single test case in its own binary (which, for other reasons may be a good
316: idea on its own). The second is the rewrite of most user-interface code in
317: Lua, which in itself would bring some extensibility advantages to the
318: program. Both options will be investigated when the time permits. In the
319: meantime, the replacement of ATF with Kyua does not make things worse; it
320: just changes one chunk of code with another.
322: ## No need for Lutok as a public library
324: Kyua depends on Lutok, which is a C++ interface to Lua. The code of this
325: library was originally part of Kyua, but I split it into its own project
326: because some users asked for it.
328: If there is no desire to ship Lutok as a shared library in `/usr/lib`, we
329: can build Lutok as a static private library and link Kyua against it.
330: There is no need to install this as a shared library.
CVSweb for NetBSD wikisrc <wikimaster@NetBSD.org> software: FreeBSD-CVSweb