From: Masami Hiramatsu Date: Thu, 5 Feb 2009 22:12:39 +0000 (-0500) Subject: prevent kprobes from catching spurious page faults X-Git-Tag: v2.6.28.5~48 X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=5f1d5882000a8d13f244165ae8747029958ab3df;p=thirdparty%2Fkernel%2Fstable.git prevent kprobes from catching spurious page faults commit 9be260a646bf76fa418ee519afa10196b3164681 upstream. Prevent kprobes from catching spurious faults which will cause infinite recursive page-fault and memory corruption by stack overflow. Signed-off-by: Masami Hiramatsu Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index 46f4cb168ac2f..68dc0609b896b 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -601,8 +601,6 @@ void __kprobes do_page_fault(struct pt_regs *regs, unsigned long error_code) si_code = SEGV_MAPERR; - if (notify_page_fault(regs)) - return; if (unlikely(kmmio_fault(regs, address))) return; @@ -632,6 +630,9 @@ void __kprobes do_page_fault(struct pt_regs *regs, unsigned long error_code) if (spurious_fault(address, error_code)) return; + /* kprobes don't want to hook the spurious faults. */ + if (notify_page_fault(regs)) + return; /* * Don't take the mm semaphore here. If we fixup a prefetch * fault we could otherwise deadlock. @@ -639,6 +640,9 @@ void __kprobes do_page_fault(struct pt_regs *regs, unsigned long error_code) goto bad_area_nosemaphore; } + /* kprobes don't want to hook the spurious faults. */ + if (notify_page_fault(regs)) + return; /* * It's safe to allow irq's after cr2 has been saved and the