- Feb 04, 2005
-
-
rtoy authored
MB. Martin Rydstrom says if the heap size is not a multiple of 8 MB on Solaris 10, he gets GC lossage with gencgc. Don't know why, but this works around the problem.
-
- Feb 03, 2005
-
-
rtoy authored
work on Solaris 10. (Eventually, he gets a GC lossage). But setting -dynamic-space-size 256 works fine. The difference is that the default is 256 MB - 32 KB. So change the default to just 256 MB. We don't understand why the old default causes funny things to happen on Solaris 10.
-
- Jan 13, 2005
-
-
fgilham authored
lisp.c into the correct <foo>-os.h files.
-
- Dec 24, 2004
-
-
rtoy authored
on Linux/x86, the Lisp signal handler for SIGSEGV overwrites the beginning of the control stack. For SIGSEGV we use an extra signal stack and when the signal isn't handled by the GC or by the stack overflow handler we call eventually call_into_lisp. call_into_lisp tests whether the current stack pointer points to the normal control stack. But the test fails because we use the extra signal stack, so call_into_lisp just sets the stack pointer to the beginning of the normal control stack. This is not much fun for debugging because the backtrace is truncated and it's also no longer possible to throw to the top-level-catcher because its frame was overwritten. I tried to fix that with the patch below. The idea is to switch back to the normal stack before calling the Lisp handler. The patch may not be very elegant but it is at least short :) .
-
- Oct 20, 2004
-
-
rtoy authored
10.3 was putting some malloc'ed areas in the old (bigger) static space, which causes cmucl not to run. Moving the space to 0x10000000 makes it no longer overlap. I don't understand the reason for this. OS X 10.2 doesn't have this problem. Use boot-2004-10-1-ppc.lisp to cross-compile this change.
-
- Oct 19, 2004
-
-
cwang authored
-
- Oct 16, 2004
-
-
rtoy authored
binding-stack to 128 MB each. Increase the heap to a maximum of 512 MB, but leave the default heap at 128 MB. The control-stack and binding-stack were moved down just below the heap. No bootstrap or cross-compiled required.
-
- Oct 15, 2004
-
-
cwang authored
-
- Sep 24, 2004
-
-
rtoy authored
with gencgc because the interrupt handler eventually calls Lisp code that might cons.
-
- Sep 21, 2004
-
-
emarsden authored
support. This is necessary for FILE-AUTHOR.
-
- Sep 14, 2004
-
-
rtoy authored
because it calls handle_breakpoint which eventually calls into Lisp. Gencgc allocation needs SIGILL enabled.
-
- Sep 11, 2004
-
-
rtoy authored
double-float-epsilon actually be epsilon on x86. No effect on other ports. The default precision is now 53-bit (double-float) instead of 64-bit (80-bit floats). However, to preserve C expectations, all calls to C have the precision set to 64-bit. This slows down calls to C, but we try to make syscalls and such fast by not changing precision for the call. By default ext:float-accuracy is 3. Use boot7.lisp to bootstrap.
-
- Aug 17, 2004
-
-
rtoy authored
executable and if the lisp.core is readable. o Add -debug-lisp-search so some information about how the search is progressing is displayed. o To support -debug-lisp-search, change the #ifdefs into code.
-
- Aug 06, 2004
- Aug 04, 2004
- Aug 02, 2004
-
-
cwang authored
Reported by Carl Shapiro
-
- Jul 30, 2004
-
-
rtoy authored
-
rtoy authored
-
rtoy authored
o Move the SC_OFFSET_{HI,LO} macros from sparc-lispregs to lispregs.h, so they can be used elsewhere. ppc-lispregs.h: o Define the register numbers so they can be used in the SC_OFFSET macros. ppc-assem.S: o Document what the byte 18<<2, 0, 0, 24 is supposed to do. o Use the SC_OFFSET macros in undefined_tramp instead of hard-wiring the register number.
-
- Jul 29, 2004
-
-
rtoy authored
the handler, which seems to be a bad idea since interrupt_handle_now has comments about alloc_sap while interrupts are disabled. Not 100% sure this is all right, but everything still appears to work.
-
rtoy authored
-
rtoy authored
#C(0d0 0d0) failing a typecase. Then doing a backtrace gave another error about some negative number not being a positive fixnum. The former error is caused by a bad definition of undefined_tramp. The latter error is caused by the debugger not being able to compute the offset correctly. (Fixes ported from sbcl.)
-
rtoy authored
doesn't like.
-
- Jul 27, 2004
- Jul 25, 2004
-
-
pmai authored
cleanup of the includes.
-
- Jul 24, 2004
-
-
rtoy authored
assertion, let's disable gc assertions on sparc now.
-
- Jul 20, 2004
- Jul 15, 2004
-
-
cwang authored
-
- Jul 14, 2004
- Jul 13, 2004
-
-
pmai authored
versions of all the required things, this commit separates the interrupt handler proper (now interrupt_handle_now_handler) from the actual work-horse (interrupt_handle_now), which is now only ever called from the two interrupt handlers, interrupt_handle_now_handler and maybe_now_maybe_later. This follows similar changes in SBCL, and is in fact needed in order to properly use sigreturn in signal handlers, as required to work around a problem with the G5.
-
- Jul 12, 2004
-
-
pmai authored
by the commit of the NetBSD support, and needed by the upcoming merge of the OS X/Darwin port.
-
- Jul 08, 2004
-
-
rtoy authored
* interrupt.c: oldcont isn't used. * interrupt.c, monitor.c, validate.c: Initialize some variables that the compiler thinks might be used without initialization.
-
rtoy authored
-
rtoy authored
* alloc.c, coreparse.c, monitor.c, validate.c, vars.c: #include some files to get declarations of functions. * coreparse.c, lisp.c, monitor.c: Fix printf warnings about args not matching the format string.
-
rtoy authored
conversions to pointers from ints. Correct the return type of sigsegv_handle_now.
-