From 1bc623167f63249fcff5c4b271886e98479b0c5a Mon Sep 17 00:00:00 2001 From: Joe Groff Date: Thu, 3 Nov 2011 20:19:01 -0700 Subject: [PATCH] cpu.x86.bootstrap: don't safepoint in a safepoint Safepointing in the signal-handler primitive causes a crash if the safepoint is guarded and the signal being handled was raised by a safepoint. Put our own RET in the subprimitive to bail out before the standard epilog gets written out. This tentatively appears to fix #349. --- basis/cpu/x86/bootstrap.factor | 1 + 1 file changed, 1 insertion(+) diff --git a/basis/cpu/x86/bootstrap.factor b/basis/cpu/x86/bootstrap.factor index 570c7c87c5..fc80a75e87 100644 --- a/basis/cpu/x86/bootstrap.factor +++ b/basis/cpu/x86/bootstrap.factor @@ -112,6 +112,7 @@ big-endian off temp0 vm-reg vm-signal-handler-addr-offset [+] MOV temp0 CALL frame-size jit-signal-handler-epilog + 0 RET ] \ signal-handler define-sub-primitive : leaf-frame-size ( -- n ) 4 bootstrap-cells ;