August 2020 - asdf-devel - mailman3.common-lisp.net

[PATCH] Add a TEST-OP-TEST-FAILURE condition for test libraries to sub-class
by Vladimir Sedach 26 Mar '21

26 Mar '21

Hello! I recently wrote a script to run FiveAM tests for one of my libraries on many different implementations on your own machine: https://gitlab.common-lisp.net/uri-template/uri-template2/blob/master/run-t… It would be really nice if I did not have to copy-paste that script to my other libraries, and instead could contribute a generalized version to Roswell (on which the script is based) and have it work for any project using any test library. What I would like to be able to do, for any system: (handler-case (asdf:test-system "system") (asdf:test-op-test-failure (condition) (princ condition uiop:*stderr*) (uiop:quit 1))) The attached patch adds an ASDF:TEST-OP-TEST-FAILURE condition that test libraries can inherit from. I also added the necessary functionality to FiveAM: https://github.com/sionescu/fiveam/pull/58 It should be easy to add similar functionality to other testing libraries. This will make test automation trivial, with few, if any, changes required to systems that use ASDF and the testing libraries. One thing I would like to discuss is which condition ASDF:TEST-OP-TEST-FAILURE should inherit from. It definitely should not be ERROR - there is no error in test-op, or in running the tests; test failures are a regular and expected occurrence of running tests. Also problematic is widespread abuse of HANDLER-CASE to catch any ERROR; I am afraid if signaled from popular test libraries it would break someone's test running code somewhere. WARNING seems like a nice parent condition, but maybe having ASDF:TEST-OP-TEST-FAILURE inherit from CONDITION is a better idea. Thoughts? Vladimir

4 15

A CFFI -ASDF integration bug
by Ilya Perminov 28 Aug '20

28 Aug '20

Hi, I found a ASDF-related CFFI bug a couple of days ago. Can anyone think of a good way of fixing it? A method in grovel/asdf.lisp adds output files of process-op to output files of compile-op: ;;; Declare the .o and .so files as compilation outputs, ;;; so they get picked up by bundle operations. #.(when (version<= "3.1.6" (asdf-version)) '(defmethod output-files ((op compile-op) (c wrapper-file)) (destructuring-bind (generated-lisp lib-file c-file o-file) (output-files 'process-op c) (declare (ignore generated-lisp c-file)) (multiple-value-bind (files translatedp) (call-next-method) (values (append files (list lib-file o-file)) translatedp))))) As a result inputs and outputs of the ops look like this: process-op: input: wrapper-file output: bindings-file.lisp file.c FILE.O FILE.SO compile-op: input: bindings-file.lisp output: bindings-file.fasl FILE.O FILE.SO The problem is that process-op generates file.o and file.so before it generates bindings-file.lisp. Thus compile-op gets re-executed all the time, because its output files file.o and file.so are always older than its input file bindings-file.lisp. And an execution of the compile-op does not change anything, because it does not really generate .o and .so. One way to fix it would be to "touch" .o and .so to get the right order of file modification times, but there may be a better way. CFFI bug report: https://bugs.launchpad.net/cffi/+bug/1889491 Thanks, Ilya

4 11

Re: A CFFI -ASDF integration bug
by Robert Goldman 04 Aug '20

04 Aug '20

On 4 Aug 2020, at 17:00, Ilya Perminov wrote: > I think there is a problem with the :around method approach - it is > not safe for an ASDF extension to define methods on ASDF's > operations/components, because other ASDF extensions may want to do > the same thing. Imagine there is CFFI* and I want to use both CFFI and > CFFI*. That is true -- that's why I always try to have either my own operation class or my own component class, or both. So to do this, you would want to have your own subclass of `system` that you wish to have work this way (e.g., `cffi-extended-system`) -- you generally do not want to add methods to the standard classes in ASDF, both because of potential collisions, and because you don't know what else they might interfere with. > > On Tue, Aug 4, 2020 at 2:35 PM Robert Goldman <rpgoldman(a)gmail.com> > wrote: >> >> On 4 Aug 2020, at 16:18, Ilya Perminov wrote: >> >> The around method will look something like this >> >> (defmethod input-files :around ((o asdf/bundle::gather-operation) (c >> system)) >> (unless (eq (asdf/bundle::bundle-type o) :no-output-file) >> (append (call-next-method) >> (remove-if-not >> (asdf/bundle::pathname-type-equal-function >> (asdf/bundle::bundle-pathname-type >> (asdf/bundle::gather-type o))) >> (mappend (lambda (c) (output-files >> 'process-op c)) >> (asdf/component:sub-components c >> :type 'wrapper-file)))))) >> >> To my taste it knows too much about ASDF guts, but maybe it is not >> too bad. >> >> I'm looking at the code now, and >> >> bundle-type is exported from asdf/bundle (but not asdf) (it could use >> a docstring if you can propose one) >> >> So is bundle-pathname-type >> >> pathname-type-equal-function is not exported, but it's a trivial >> combination of pathname-type and equalp >> >> gather-type is the only "really internal" function. (A block comment >> would be nice here, too.) >> >> gather-operation is not exported, but I think that's probably wrong >> on our part: I don't believe there should be hidden operations. If I >> get around to it, I will probably export it from the asdf (really >> asdf/interface) package, unless someone has a good argument for it >> continuing to be internal. >> >> So I don't think that is a lot of ASDF internals, especially not for >> customizing the ASDF protocol.

1 0

Re: A CFFI -ASDF integration bug
by Ilya Perminov 04 Aug '20

04 Aug '20

The around methed will look something like this (defmethod input-files :around ((o asdf/bundle::gather-operation) (c system)) (unless (eq (asdf/bundle::bundle-type o) :no-output-file) (append (call-next-method) (remove-if-not (asdf/bundle::pathname-type-equal-function (asdf/bundle::bundle-pathname-type (asdf/bundle::gather-type o))) (mappend (lambda (c) (output-files 'process-op c)) (asdf/component:sub-components c :type 'wrapper-file)))))) To my taste it knows too much about ASDF guts, but maybe it is not too bad.

2 3

Re: A CFFI -ASDF integration bug
by Ilya Perminov 04 Aug '20

04 Aug '20

On Tue, Aug 4, 2020 at 12:51 PM Robert Goldman <rpgoldman(a)sift.info> wrote: > I don't claim to understand this process, but wouldn't it be possible for you to make your own input-files :around method for gather-operation that would collect the outputs from the process-op's and add them to what you want? I think it is a reasonable option, at least I like it more than what CFFI does now. input-files for gather-operation is defined on SYSTEM, so the around method will have to fetch all WRAPPER-FILEs from a system. gather-type and probably some other functions, that the around method will need, are not exported from ASDF though. > That said, I can think of a simpler, and easier method, and that would be to override the operation-done-p method for the compile-op so that it knows that the .o and .so files are generated by the compile-op, and only pays attention to the relationship between bindings-file.lisp and bindings-file.fasl. I do not think it will work, because up-to-date check happens outside of operation-done-p. compute-action-stamp: (and all-present ;; if all filesystem effects are up-to-date up-to-date-p (operation-done-p o c) ;; and there's no invalidating reason. > One thing you didn't say was what it means that the compile-op is done over and over -- is it only compiling the bindings-file, or is it doing something to the object files? I don't believe it should change the .o and .so files, because those are already compiled by process-op, right? Yes, only the lisp file gets recompiled. compile-op does not do anything to the .o and .so files, they are just its fake outputs.

1 0

Apologies for slow progress...
by Robert Goldman 02 Aug '20

02 Aug '20

I'm afraid I have been very busy at work, but also the linux box I have been using as my Jenkins platform had a disk failure, taking it out of action. No data was lost, but my employer has been supplying this, and I have to wait in the queue, after business requirements, for the IT folks to get me a new server configured for this purpose. This has unavoidably hindered my work on ASDF. I'm open to suggestions about alternative resources for running these tests, but there are special challenges because I need to be able to put commercial software (Allegro CL and LispWorks) on any test server, which I think rules out using Travis. If anyone has hosting suggestions, please let me know: otherwise I'm afraid we'll be limping along till September, which is unfortunate, since we have a bunch of fixes and bug reports, and I would very much like to be able to get these wrapped up into a bug fix release. I hope you all are staying safe and healthy!

6 6