November 2019 - armedbear-devel - mailman3.common-lisp.net

coerce a number to a short
by Brad Knotwell 12 Feb '20

12 Feb '20

I've been using ABCL to get access to a Java library and it's worked generally well. I have run into one problem I haven't been able to workaround. I need to make a call to a method that takes a short as an argument. I'm passing in a value--10--that fits in a short but ABCL can't find the method (presumedly due to a type issue). I've tried using jcoerce but it's not doing what I expect. What's the right way to do this? Thx. --Brad

3 10

On preparations to turn the Bear past 11
by Mark Evenson 02 Dec '19

02 Dec '19

Dear long-suffering fans of the Bear, Rejoice! for preparations for the release of abcl-1.6.0 to support openjdk{6,7,8,11} and beyond near completion. We have managed to get a [Continuous Integration (CI)][travis-builds] build mostly working via Travis CI integration with the Github repository. Anyone who forks us on Github with suitable monkeying around can get automated build and test results for their commits to their fork. The [current version of the tests][.travis.yml] attempt to run in environments that are the cross product of macOS/Linux with openjdk8/openjdk11. This isn't quite completely working yet as I haven't apparently figured out how to use jenv in Travis to specify the correct openjdk to invoke, but I'm close. For now, we at least getting test coverage for openjdk11 under Linux, and I hope to iron out the remaining kinks real soon now. It should be theoretically possible to [add Windows builds][windows-builds] as well, but I am planning on to tackling that in the near future. The CI tests have given me the confidence to next attempt to release abcl-1.6.0 via existing release engineering process in spite of the following problems: 1. While they seemingly pass in the CI, the CFFI linkages to CL+SSL libraries often fail in a spectacularly segmentation fault in practical use. With judicious refinement, I think we can figure out what is going but for now we should just accept this. 2. [A major regression from abcl-1.5.0][pathname-problems] concerning the use of CL:PATHNAME to refer to objects within a jar/zip archive has been discovered, but curiously it is seemingly not fatal to anything other than the execution of ABCL test suite. In analyzing the problems, I realized that we can considerably clean up our abstraction for a CL:PATHNAME which denotes an entry in an archive by allowing nested archives using the CL:PATHNAME-DEVICE component. To do this in a reasonably clean manner means a fair amount of modification on the Java side of our implementation by reflecting the various types of a CL:PATHNAME in that Java hierarchy as Erik Hülsmann suggested many years ago. While I have some promising preliminary patches towards this, due to the amount of outstanding effort, I feel we should get abcl-1.6.0 out first. Please holler loudly real soon if anyone has problems with releasing as things currently stand. I intend follow up with abcl-1.6.1 within a month to hopefully address the two major blockers in addition to any errata. [travis-builds]: https://travis-ci.org/armedbear/abcl/builds [.travis.yml]: https://github.com/armedbear/abcl/blob/master/.travis.yml [windows-builds]: https://twitter.com/ArmedBear/status/1197538417678192645 [cl+ssl-problems]: https://abcl.org/trac/ticket/464 [pathname-problems]: https://github.com/armedbear/abcl/commit/e962be5e0dd86335cc66415a0a417f9e4b… yers, Mark <evenson(a)panix.com> -- "A screaming comes across the sky. It has happened before but there is nothing to compare to it now."

3 2

ABCL 1.6.0 unleashed
by Mark Evenson 23 Nov '19

23 Nov '19

Against the falling canvas of 2019, the Bear has freely thrust [ABCL 1.6.0][abcl.org/releases] onto the [world stage][github]. This release fully supports building and running on Java 6, 7, 8, and 11 via the [openjdk runtimes][adoptjdk]. By [building from our source][source-build], it is [rumored that the User may utilize additional hosting platforms such as Java 13][sfp-mail]. Support for Microsoft Windows platforms has not been thoroughly tested at this point but we intend to add it our automatic cloud based build in the process of releasing a [stablizing ABCL 1.6.1][1.6.1] build due by the end of this year. Thanks to all past, present, and future contributors to the Bear; here's to [hackin' our way out of the prison of the Java language][jss] for over a decade. [github]: https://github.com/armedbear/abcl [abcl.org/releases] https://abcl.org/releases/1.6.0 [1.6.1]: https://abcl.org/trac/milestone/1.6.1 [source-build]: https://gitlab.common-lisp.net/abcl/abcl/tree/master/contrib/abcl-build [adoptjdk]: https://adoptopenjdk.net/ [sfp-mail]: https://mailman.common-lisp.net/pipermail/armedbear-devel/2019-November/004… -- "A screaming comes across the sky. It has happened before but there is nothing to compare to it now."

2 2

Patch for a couple of stack inconsistency bugs
by somewhat-functional-programmer 19 Nov '19

19 Nov '19

Hello list, Lately I have been starting to dive deeper into Common Lisp and have started to use ABCL more than SBCL or CCL lately. It is a very impressive project. I want to post a patch for review/comments and hopefully have it be worthwhile to eventually include it in ABCL. The patch attempts to fix a couple of stack inconsistency bugs in the compiler. I came across the stack inconsistency issue in one of my projects and started to try to find the root cause of the problem based on a nice minimal reproduction of the bug found in https://github.com/armedbear/abcl/issues/69. However, my particular bug was slightly different. It had to do with using a return-from in the cleanup form of an unwind-protect. The following two forms also result in a stack inconsistency problem (and are a more minimal reproduction of the bug my code introduced): (defun two-arg-fn (one two) (format t "Two args: ~S and ~S~%" one two)) (let ((fn (compile nil '(lambda () (two-arg-fn (block test-block (unwind-protect 30 (return-from test-block 8))) -1))))) (funcall fn)) My patch handles both the github issue and the stack inconsistency in the form above. It also fixes jvm::print-code to print string representations of values from the constant pool which I found useful in debugging the output. Anyhow, let me attempt to quickly summarize the stack inconsistency problem in general: - Certain common lisp control flow forms (tagbody/go/unwind-protect/block/return-from/throw/catch) require the use of JVM exceptions to implement in bytecode - When the JVM throws an exception, the operand stack is cleared and the exception is pushed onto the operand stack (see jvms8, 6.5/athrow, p378) - Therefore, any form which pushes values onto the operand stack for further use is confounded when these control flow forms are child forms - To properly handle these alternate control flows we need to save the result of the control flow form to a local variable in the stack frame (distinct from the operand stack, and not destroyed by an exception (well at least not until the exception passes /out/ of the method) and then reload for use by the parent form to push on the operand stack Take the case of the ash function (from the github issue). Bytecode for ash is emitted from jvm::p2-ash. It compiles its arguments to the operand stack, and is therefore vulnerable to the problem discussed above. Other low level functions (like + for example in p2-plus) use the following forms to overcome this issue: jvm::with-operand-accumulation and jvm::compile-operand. These forms save the results of "unsafe" forms (opstack unsafe) to "registers" (local variables in the stack frame). This technique allows for these complicated control flow forms to be a child form of + with no issues, but not the ash function (which does not do this). See my patch for how I added these already present with-operand-accumulation and compile-operand forms to ash so it is no longer vulnerable to stack inconsistency bugs. Generally, function calls in ABCL are not vulnerable to these stack inconsistency bugs. Function arguments are processed in jvm::process-args, and in this function, the opstack safety of child forms is checked, and values are saved to "registers" when a form is known to be unsafe. My case (the return-from in the cleanup form of an unwind-protect) simply wasn't properly being marked as opstack unsafe. I modified jvm::p1-unwind-protect to mark all direct children of the unwind-protect as opstack unsafe, which eliminated my problem. I believe there may have been confusion here in the code (speculation of course on my part) of 'protected' form referring to the form actually protected by the unwind-protect (which is totally different than being unsafe or needing opstack /protection/) (or I'm misinterpreting this function and potentially causing additional bugs!). The only other item in my patch is fixing how bytecode is printed for debugging. Basically, most items in a class constant pool are referenced with a 2 byte index, but one (ldc) uses a one byte index). This has been accounted for in the new function jvm::constant-pool-index. Let me know what you think, and again please review, there aren't many lines changed but I am new to the internals of the project. I think I ran all tests (ant abcl.test) but while the ant task completed successfully, my output complained of a missing dependency and I'm not sure what actually ran. -Mark The attached patch was produced against b3cfee6617e0c2c380d8675f3383d81e7758f358 from https://github.com/easye/abcl (latest master branch). [jvms8] https://docs.oracle.com/javase/specs/jvms/se8/jvms8.pdf

2 3

format ~$ may need consider -0.0
by dingd 16 Nov '19

16 Nov '19

this is what sbcl does * (format nil "~$" -0.0) "-0.00"

2 1

On the Bear turning to Java 11
by Mark Evenson 16 Nov '19

16 Nov '19

As the leaves fall in the Northern hemisphere, thoughts of the Bear turn to a rumored permanent slumber, and a complete absence of the possibility of new updates. The Bear [accepts][trac] [all][gitab] patches][github], so if you don't like the current implementation, please get out and push. Or pull? Or fast forward? The Bear is seemingly always confused with Git… As previously rumored, [patches for getting the Armed Bear Common Lisp implementation running on Java 11][github-with-java11] have landed in the [trunk][trunk-with-java11]. Unless anyone objects strongly, we will release [ABCL 1.6.0][abcl-1.6.0] as soon as possible with the intention to support the openjdk{6,7,8,11} runtimes. After stabilizing the 1.6 branch with the inevitable necessary and unknown updates, we can turn our attention to the [2.0 branch][abcl-2.0.0] which will introduce support for additional runtimes but drop support for emitting openjdk6 or openjdk7 compatible byte-code. As always, everyone's desires are welcome to be expressed. yers in CONS, Mark <evenson.not.org(a)gmail.com> [trunk-with-java11]: https://abcl.org/trac/changeset/15133 [github-with-java11]: https://github.com/armedbear/abcl/commit/a628e961a1dc33e1f225191a19494701d7… [github]: https://github.com/armedbear/abcl/ [trac]: https://abcl.org/trac/ [gitlab]: https://common-lisp.net/abcl/abcl/ [svn]: svn://abcl.org/svn/ [mailing-list]: <mailto:armedbear-devel@common-lisp.net> [abcl-1.6.0]: https://abcl.org/trac/milestone/1.6.0 [abcl-2.0.0]: https://abcl.org/trac/milestone/2.0.0 -- "A screaming comes across the sky. It has happened before but there is nothing to compare to it now."

2 1

Moving the Bear past Java 11
by Mark Evenson 14 Nov '19

14 Nov '19

As the leaves fall in the Northern hemisphere, thoughts of the Bear turn to a rumored permanent slumber, and a complete absence of the possibility of new updates. The Bear [accepts][trac] [all][gitab] patches][github], so if you don't like the current implementation, please get out and push. Or pull? Or fast forward? The Bear is seemingly always confused with Git… As previously rumored, [patches for getting the Armed Bear Common Lisp implementation running on Java 11][github-with-java11] have landed in the [trunk][trunk-with-java11]. Unless anyone objects strongly, we will release [ABCL 1.6.0][abcl-1.6.0] as soon as possible with the intention to support the openjdk{6,7,8,11} runtimes. After stabilizing the 1.6 branch with the inevitable necessary and unknown updates, we can turn our attention to the [2.0 branch][abcl-2.0.0] which will introduce support for additional runtimes but drop support for emitting openjdk6 or openjdk7 compatible byte-code. As always, everyone's desires are welcome to be expressed. yers in CONS, Mark <evenson.not.org(a)gmail.com> [trunk-with-java11]: https://abcl.org/trac/changeset/15133 [github-with-java11]: https://github.com/armedbear/abcl/commit/a628e961a1dc33e1f225191a19494701d7… [github]: https://github.com/armedbear/abcl/ [trac]: https://abcl.org/trac/ [gitlab]: https://gitlab.common-lisp.net/abcl/abcl/ [svn]: svn://abcl.org/svn/ [mailing-list]: <mailto:armedbear-devel@common-lisp.net> [abcl-1.6.0]: https://abcl.org/trac/milestone/1.6.0 [abcl-2.0.0]: https://abcl.org/trac/milestone/2.0.0 -- "No, this is not a disentanglement, but a progressive /knotting-into/."

1 0