Fix a bug introduced in my earlier fault information pollution
prevention patch. That patch assumed that no kernel fault information
should ever be put in thread.error_code and .trap_no. It turns out
that die() reads those fields, so fatal kernel faults, as well as
userspace faults, need those fields set. The case where they aren't
set is kernelspace faults which are fixed up.
Thanks to Jan Beulich for his review and spotting this bug.
Signed-off-by: Jeff Dike <[EMAIL PROTECTED]>
--
arch/i386/kernel/traps.c | 11 +++++++++++
arch/x86_64/kernel/traps.c | 14 +++++++++++++-
2 files changed, 24 insertions(+), 1 deletion(-)
Index: linux-2.6/arch/i386/kernel/traps.c
===================================================================
--- linux-2.6.orig/arch/i386/kernel/traps.c
+++ linux-2.6/arch/i386/kernel/traps.c
@@ -607,6 +607,15 @@ fastcall void __kprobes do_general_prote
if (!user_mode(regs))
goto gp_in_kernel;
+ /*
+ * We want error_code and trap_no set for userspace faults and
+ * kernelspace faults which result in die(), but not
+ * kernelspace faults which are fixed up. die() gives the
+ * process no chance to handle the signal and notice the
+ * kernel fault information, so that won't result in polluting
+ * the information about previously queued, but not yet
+ * delivered, fault.
+ */
current->thread.error_code = error_code;
current->thread.trap_no = 13;
force_sig(SIGSEGV, current);
@@ -619,6 +628,8 @@ gp_in_vm86:
gp_in_kernel:
if (!fixup_exception(regs)) {
+ current->thread.error_code = error_code;
+ current->thread.trap_no = 13;
if (notify_die(DIE_GPF, "general protection fault", regs,
error_code, 13, SIGSEGV) == NOTIFY_STOP)
return;
Index: linux-2.6/arch/x86_64/kernel/traps.c
===================================================================
--- linux-2.6.orig/arch/x86_64/kernel/traps.c
+++ linux-2.6/arch/x86_64/kernel/traps.c
@@ -582,6 +582,15 @@ static void __kprobes do_trap(int trapnr
struct task_struct *tsk = current;
if (user_mode(regs)) {
+ /*
+ * We want error_code and trap_no set for userspace faults
+ * and kernelspace faults which result in die(), but
+ * not kernelspace faults which are fixed up. die()
+ * gives the process no chance to handle the signal
+ * and notice the kernel fault information, so that
+ * won't result in polluting the information about
+ * previously queued, but not yet delivered, fault.
+ */
tsk->thread.error_code = error_code;
tsk->thread.trap_no = trapnr;
@@ -605,8 +614,11 @@ static void __kprobes do_trap(int trapnr
fixup = search_exception_tables(regs->rip);
if (fixup)
regs->rip = fixup->fixup;
- else
+ else {
+ tsk->thread.error_code = error_code;
+ tsk->thread.trap_no = trapnr;
die(str, regs, error_code);
+ }
return;
}
}
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
