Commit Graph

153 Commits (80b09cb075fae057d8c201dc2f10106b32c2f083)

Author SHA1 Message Date
slava 756cea6164 Grab native stack pointer from signal handler 2006-12-18 04:31:38 +00:00
slava 423a2f10fc AMD64 fixes 2006-12-18 02:25:43 +00:00
slava 2f3db7b389 #label nodes are now reported in compiled stack traces 2006-12-18 00:10:32 +00:00
slava e75774c59f Fix runtime compile warnings; working on native stack traces again 2006-12-17 22:40:21 +00:00
slava 07df0e40f9 Assorted fixes: sort out bignum division by zero, fix some unit tests 2006-12-14 07:35:22 +00:00
slava 8cfbd164f5 Remove startup messages from VM 2006-12-10 20:02:11 +00:00
slava 8f2bc485cc Remove erronous instruction from cpu-x86.S 2006-12-08 08:41:16 +00:00
slava 29c158f243 Add missing type check 2006-12-04 22:42:57 +00:00
slava 2faf88e593 Fix a typo and add a unit test 2006-12-04 05:30:19 +00:00
slava dc426c13f8 Resolve conflict 2006-12-04 05:28:03 +00:00
slava bdf2cc8abe Fix a compile warning 2006-12-04 05:06:45 +00:00
slava 936f7d8495 Fix a warning 2006-12-03 04:54:54 +00:00
slava c968767973 Add crazy paranoia to code GC and flush stderr on Windows 2006-12-03 04:54:19 +00:00
slava 320ee19b61 Fix bug in code GC free list handling 2006-12-04 05:20:54 +00:00
slava f0d353558e Check for NaN/infinite values in s48_double_to_bignum 2006-12-04 05:19:33 +00:00
slava 3716de2737 Simplify 2006-11-24 10:08:01 +00:00
slava f9f708739e Factor can now recover from certain out of memory conditions 2006-11-24 10:00:51 +00:00
slava 2d1c7f5e0f Minor robusifications 2006-11-24 09:50:25 +00:00
slava 29c3d757bd Also use mach exception handlers to trap SIGFPE 2006-11-23 21:35:36 +00:00
slava 5038554251 GC is not safe from a signal handler since the ds/rs registers might be invalid. So disable the GC in signal handlers 2006-11-22 06:24:55 +00:00
slava 12f61890ca -7 <byte-array> crash 2006-11-22 05:19:19 +00:00
slava f29e31c251 Comment out problematic/unfinished code 2006-11-21 23:42:59 +00:00
erg 077ebd7c6c fix compile errors on win32 2006-11-21 23:47:44 +00:00
slava aa23766df2 Fix issue when GC is called while allocating native stack trace 2006-11-21 21:19:17 +00:00
slava ab4df9d8e6 Fix recent regression 2006-11-21 10:17:46 +00:00
slava 0fb57e1e9e Fix crash bug introduced in 0.86 2006-11-20 08:43:03 +00:00
slava 4759883733 Fix a code GC race condition 2006-11-20 02:14:11 +00:00
slava 7d0e89a11f Add an assertion to the code GC code 2006-11-18 09:04:11 +00:00
slava 4852c894f2 Uncomment native stack pointer recording 2006-11-17 09:37:49 +00:00
slava 57f4d59b8c PowerPC float>fixnum intrinsic 2006-11-17 02:04:37 +00:00
slava e371ffcae6 Working on amd64 stack traces 2006-11-16 07:12:03 +00:00
slava de1ebb9f1a Trying to get stack traces going on OS X, but won't compile without -fomit-frame-pointer 2006-11-16 06:31:49 +00:00
slava c4ddcf54df x86 compiled stack traces working 2006-11-16 05:12:43 +00:00
slava c0038a28f9 Fix documentation typos, start working on compiled stack traces 2006-11-16 03:57:58 +00:00
slava c1b7fb46de Don't make this change yet 2006-11-17 07:10:14 +00:00
slava 7904ddcac2 Fix Objective-C exception handling 2006-11-17 07:01:05 +00:00
slava fde3ba3bb3 Fix a hang if you do 'clear drop datastack' 2006-11-17 06:51:29 +00:00
slava f67ee0f315 Kick off 0.87 with better callstack underflow handling 2006-11-15 02:40:54 +00:00
slava 1f8fa2c531 Fixes 2006-11-14 21:59:20 +00:00
slava 761dab4188 Fix sig11 when dividing large bignums 2006-11-09 23:10:02 +00:00
slava 7ee2128ff0 numbers-game has MAIN:, image loading cleanup 2006-11-09 03:13:33 +00:00
slava 75ee6ac549 PPC inline allocators almost working 2006-11-08 23:21:58 +00:00
slava d54e3baac8 First cut at bignum inline allocators 2006-11-07 05:22:34 +00:00
slava 7642d69352 Fix typo in fixnum* primitive 2006-11-07 23:04:16 +00:00
slava 4ee7c02798 Typo: box_float instead of box_double in floating point primitives 2006-11-04 22:04:22 +00:00
slava 48b288ceb9 Fix code heap saving if the last block in the heap is allocated 2006-11-04 21:51:18 +00:00
erg 6d3996ccfb fix win32 runtime compile errors 2006-11-04 16:52:57 +00:00
slava 861a6d32cf alien-invoke no longer generates calls to consing unbox_* functions -- instead we just unbox the parameter list first, in factor code 2006-11-03 21:39:37 +00:00
slava ab68570cc1 Fix callstack overflow handling 2006-11-03 21:14:53 +00:00
slava 4a8d65ccce Refactored the C type system in the alien interface and fixed some bugs 2006-11-03 08:54:30 +00:00