From 243300b289ddf554db205173fbc041a12a3968eb Mon Sep 17 00:00:00 2001 From: Ingo Weinhold Date: Mon, 15 Sep 2008 13:18:11 +0000 Subject: [PATCH] For the time being allow page faults to happen when interrupts are disabled, as long as a fault handler is installed for the thread. It allows user_memcpy() to be invoked with interrupts disabled -- in this case it will simply fail, when the address is valid, but not mapped yet. This is a more desirable behavior for debug facilities, though in all other cases it is likely a bug. We should probably introduce a thread flag to discriminate these situations. git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@27530 a95241bf-73f2-0310-859d-f6bbb57e9c96 --- src/system/kernel/arch/x86/arch_int.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/src/system/kernel/arch/x86/arch_int.c b/src/system/kernel/arch/x86/arch_int.c index 16a663ac46..b362ce63b5 100644 --- a/src/system/kernel/arch/x86/arch_int.c +++ b/src/system/kernel/arch/x86/arch_int.c @@ -849,8 +849,19 @@ page_fault_exception(struct iframe* frame) "address %p from eip %p\n", (void *)cr2, (void *)frame->eip); return; } else if ((frame->flags & 0x200) == 0) { - // if the interrupts were disabled, and we are not running the kernel startup - // the page fault was not allowed to happen and we must panic + // interrupts disabled + + // If a page fault handler is installed, we're allowed to be here. + // TODO: Now we are generally allowing user_memcpy() with interrupts + // disabled, which in most cases is a bug. We should add some thread + // flag allowing to explicitly indicate that this handling is desired. + if (thread && thread->fault_handler != 0) { + frame->eip = thread->fault_handler; + return; + } + + // If we are not running the kernel startup the page fault was not + // allowed to happen and we must panic. panic("page fault, but interrupts were disabled. Touching address " "%p from eip %p\n", (void *)cr2, (void *)frame->eip); return;