Re: [PATCH v5 06/27] x86/cet: Control protection exception handler

From: Borislav Petkov
Date: Wed Nov 14 2018 - 13:45:08 EST


That subject needs a verb:

Subject: [PATCH v5 06/27] x86/cet: Add control protection exception handler

On Thu, Oct 11, 2018 at 08:15:02AM -0700, Yu-cheng Yu wrote:
> A control protection exception is triggered when a control flow transfer
> attempt violated shadow stack or indirect branch tracking constraints.
> For example, the return address for a RET instruction differs from the
> safe copy on the shadow stack; or a JMP instruction arrives at a non-
> ENDBR instruction.
>
> The control protection exception handler works in a similar way as the
> general protection fault handler.
>
> Signed-off-by: Yu-cheng Yu <yu-cheng.yu@xxxxxxxxx>
> ---
> arch/x86/entry/entry_64.S | 2 +-
> arch/x86/include/asm/traps.h | 3 ++
> arch/x86/kernel/idt.c | 4 ++
> arch/x86/kernel/signal_compat.c | 2 +-
> arch/x86/kernel/traps.c | 64 ++++++++++++++++++++++++++++++
> include/uapi/asm-generic/siginfo.h | 3 +-
> 6 files changed, 75 insertions(+), 3 deletions(-)

A *lot* of style problems here. Please use checkpatch and then common
sense to check your patches before sending. All those below are valid,
AFAICT:

WARNING: function definition argument 'struct pt_regs *' should also have an identifier name
#76: FILE: arch/x86/include/asm/traps.h:81:
+dotraplinkage void do_control_protection(struct pt_regs *, long);

WARNING: function definition argument 'long' should also have an identifier name
#76: FILE: arch/x86/include/asm/traps.h:81:
+dotraplinkage void do_control_protection(struct pt_regs *, long);

WARNING: static const char * array should probably be static const char * const
#124: FILE: arch/x86/kernel/traps.c:581:
+static const char *control_protection_err[] =

ERROR: that open brace { should be on the previous line
#125: FILE: arch/x86/kernel/traps.c:582:
+static const char *control_protection_err[] =
+{

WARNING: quoted string split across lines
#158: FILE: arch/x86/kernel/traps.c:615:
+ WARN_ONCE(1, "CET is disabled but got control "
+ "protection fault\n");

WARNING: Prefer printk_ratelimited or pr_<level>_ratelimited to printk_ratelimit
#165: FILE: arch/x86/kernel/traps.c:622:
+ printk_ratelimit()) {

WARNING: Avoid logging continuation uses where feasible
#176: FILE: arch/x86/kernel/traps.c:633:
+ pr_cont("\n");

ERROR: "(foo*)" should be "(foo *)"
#183: FILE: arch/x86/kernel/traps.c:640:
+ info.si_addr = (void __user*)uprobe_get_trap_addr(regs);


And now that patch doesn't even build anymore because of the siginfo
changes which came in during the merge window. I guess I'll wait for
your v6 patchset.

---
arch/x86/kernel/traps.c: In function âdo_control_protectionâ:
arch/x86/kernel/traps.c:627:16: error: passing argument 1 of âclear_siginfoâ from incompatible pointer type [-Werror=incompatible-pointer-types]
clear_siginfo(&info);
^~~~~
In file included from ./include/linux/sched/signal.h:6,
from ./include/linux/ptrace.h:7,
from ./include/linux/ftrace.h:14,
from ./include/linux/kprobes.h:42,
from arch/x86/kernel/traps.c:19:
./include/linux/signal.h:20:52: note: expected âkernel_siginfo_t *â {aka âstruct kernel_siginfo *â} but argument is of type âsiginfo_t *â {aka âstruct siginfo *â}
static inline void clear_siginfo(kernel_siginfo_t *info)
~~~~~~~~~~~~~~~~~~^~~~
arch/x86/kernel/traps.c:632:26: error: passing argument 2 of âforce_sig_infoâ from incompatible pointer type [-Werror=incompatible-pointer-types]
force_sig_info(SIGSEGV, &info, tsk);
^~~~~
In file included from ./include/linux/ptrace.h:7,
from ./include/linux/ftrace.h:14,
from ./include/linux/kprobes.h:42,
from arch/x86/kernel/traps.c:19:
./include/linux/sched/signal.h:327:32: note: expected âstruct kernel_siginfo *â but argument is of type âsiginfo_t *â {aka âstruct siginfo *â}
extern int force_sig_info(int, struct kernel_siginfo *, struct task_struct *);
^~~~~~~~~~~~~~~~~~~~~~~
cc1: some warnings being treated as errors
make[2]: *** [scripts/Makefile.build:291: arch/x86/kernel/traps.o] Error 1
make[1]: *** [scripts/Makefile.build:516: arch/x86/kernel] Error 2
make[1]: *** Waiting for unfinished jobs....
make: *** [Makefile:1060: arch/x86] Error 2
make: *** Waiting for unfinished jobs....

--
Regards/Gruss,
Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.