More intuitive FINALIZE-INHERITANCE behaviour ... the change for 0.9.11 fixing forward references and TYPEP made finalize-inheritance call itself recursively on subclasses. This broke clg, which is now fixed, but still... ... might as well fix the odd behaviour, and add a test for it.
Fix bug reported by Levente Meszaros sbcl-devel 2004-04-19: cache filling with negative offset. ... when precomputing cache contents, don't include classes with invalid wrappers. ... irony of ironies: I'm pretty sure that there's a bug lurking somewhere else in any case, because the problem was being exhibited in the computation of a cache for SLOT-BOUNDP-USING-CLASS, which doesn't (or shouldn't) use a cacheing dfun, but instead has its own specialized dfun which basically calls the boundp function from the slot definition. Hmm... ... comments and minor tidying in cache.lisp
* package lock violations from lexical operations always cause runtime PROGRAM-ERRORs * better EXTRA_CFLAGS handling in SB-GROVEL
* If we want to delete a file we must have write-permissions... * ...elide empty environment-variable from gcc arguments in def-to-lisp.lisp
* Check for correct "expand" in canonicalize-whitespace, and skip canonicalization if it doesn't seem right. (Windows "expand" is something quite different.) * RUN-PROGRAM now always returns a process structure, which reports the exit-code of the process when :WAIT was true. :WAIT nil process-structures still keep their :RUNNING status indefinitely on Windows, though. * FIND-EXECUTABLE-IN-SEARCH-PATH actually searches the path, and adds "exe" as :TYPE if :TYPE is missing on Windows. * ASDF:RUN-SHELL-COMMAND searches for Bourne-shell on Windows, as there is no default location. * SB-GROVEL directly runs gcc and the groveler instead of indirecting via shell, and the groveler directly writes to the lisp-file instead of via stdout and shell redirection. * Hack SB-POSIX till it builds and passes all applicable tests on Windows. Mostly this involved plenty of #-win32, but a few tests needed to be adjusted for the delication Microsoft constitution. * Implement COPY-STREAM in ASDF-INSTALL so that it doesn't have to depend on SB-EXECUTABLE. * Take the .exe suffix into account when installing over an old SBCL on MSYS. * Adjust UNPARSE-NATIVE-WIN32-NAMESTRING slightly: Windows OS functions like stat don't like to have directory names ending with a slash. This is good enough to run unpatched Slime with *COMMUNICATION-STYLE* NIL, and build & pass tests with all contribs except SB-SIMPLE-STREAMS.
Misc. * Support SETF function names in SB-INTROSPECT:FUNCTION-ARGLIST. (Patch by Todd Mokros on sbcl-devel) * Minor performance improvement in the x86-64 assembler. (Patch by Lutz Euler on sbcl-devel) * Add a recursive TRACE :ENCAPSULATE NIL (supposed to be part of a recent Solaris/x86 commit, but somehow didn't make it through) * Add a failing FILE-POSITION test on utf-8 streams. (Reported by Lutz Euler)
Make the test suite pass on Solaris/x86. * Fix a number of bashisms in test/*.sh * :ENCAPSULATE NIL tracing: ... In the breakpoint handling internals use the signal context FP/PC directly to construct the frame, instead of walking through the backtrace until a matching frame is found. ... Kill the single-stepper remains in x86(-64)-arch.c. Turning on processor single-stepping with signal context frobbing was causing extra trace traps, and the code for handling those was presumably already lost a long time ago. * Floating point exception handling: ... Define os_context_fp_control, and use it instead of the stub implementation in x86-vm.lisp. * Mark the usual backtrace tests as expected to fail on this OS too