[PATCH] riscv: Support RANDOMIZE_KSTACK_OFFSET

[Palmer Dabbelt]

On Wed, 08 Nov 2023 15:52:34 PST (-0800), keescook at chromium.org wrote:
> On Wed, Nov 01, 2023 at 02:44:23PM +0800, Song Shuai wrote:
>> Inspired from arm64's implement -- commit 70918779aec9
>> ("arm64: entry: Enable random_kstack_offset support")
>>
>> Add support of kernel stack offset randomization while handling syscall,
>> the offset is defaultly limited by KSTACK_OFFSET_MAX() (i.e. 10 bits).
>>
>> In order to avoid trigger stack canaries (due to __builtin_alloca) and
>> slowing down the entry path, use __no_stack_protector attribute to
>> disable stack protector for do_trap_ecall_u() at the function level.
>>
>> Signed-off-by: Song Shuai <songshuaishuai at tinylab.org>
>
> I can't speak to the correctness of the entropy level, but the usage of
> the helpers looks correct to me.

As far as I can tell the comment matches how the system behaves.  I'm 
not sure if that much entropy is useful.  I was poking around for a bit 
to try and figure that out, but after reading that comment at the top of 
include/linux/randomize_kstack.h I decided to stop ;)

So aside from those whitespace errors Damien pointed out,

Reviewed-by: Palmer Dabbelt <palmer at rivosinc.com>

It's too late for the merge window for me, but

Acked-by: Palmer Dabbelt <palmer at rivosinc.com>

in case someone else wants to take it still.  Otherwise I'll try and 
remember to pick it up right after the merge window, but with Plumbers 
things might be a bit clunky.

> Reviewed-by: Kees Cook <keescook at chromium.org>
>
> -Kees
>
>> ---
>> Testing with randomize_kstack_offset=y cmdline, lkdtm/stack-entropy.sh
>> showed appropriate stack offset instead of zero.
>> ---
>>  arch/riscv/Kconfig        |  1 +
>>  arch/riscv/kernel/traps.c | 18 +++++++++++++++++-
>>  2 files changed, 18 insertions(+), 1 deletion(-)
>>
>> diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
>> index d607ab0f7c6d..0e843de33f0c 100644
>> --- a/arch/riscv/Kconfig
>> +++ b/arch/riscv/Kconfig
>> @@ -100,6 +100,7 @@ config RISCV
>>  	select HAVE_ARCH_KGDB_QXFER_PKT
>>  	select HAVE_ARCH_MMAP_RND_BITS if MMU
>>  	select HAVE_ARCH_MMAP_RND_COMPAT_BITS if COMPAT
>> +	select HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET
>>  	select HAVE_ARCH_SECCOMP_FILTER
>>  	select HAVE_ARCH_THREAD_STRUCT_WHITELIST
>>  	select HAVE_ARCH_TRACEHOOK
>> diff --git a/arch/riscv/kernel/traps.c b/arch/riscv/kernel/traps.c
>> index 19807c4d3805..3f869b2d47c3 100644
>> --- a/arch/riscv/kernel/traps.c
>> +++ b/arch/riscv/kernel/traps.c
>> @@ -6,6 +6,7 @@
>>  #include <linux/cpu.h>
>>  #include <linux/kernel.h>
>>  #include <linux/init.h>
>> +#include <linux/randomize_kstack.h>
>>  #include <linux/sched.h>
>>  #include <linux/sched/debug.h>
>>  #include <linux/sched/signal.h>
>> @@ -296,9 +297,11 @@ asmlinkage __visible __trap_section void do_trap_break(struct pt_regs *regs)
>>  	}
>>  }
>>
>> -asmlinkage __visible __trap_section void do_trap_ecall_u(struct pt_regs *regs)
>> +asmlinkage __visible __trap_section  __no_stack_protector
>> +void do_trap_ecall_u(struct pt_regs *regs)
>>  {
>>  	if (user_mode(regs)) {
>> +
>>  		long syscall = regs->a7;
>>
>>  		regs->epc += 4;
>> @@ -308,10 +311,23 @@ asmlinkage __visible __trap_section void do_trap_ecall_u(struct pt_regs *regs)
>>
>>  		syscall = syscall_enter_from_user_mode(regs, syscall);
>>
>> +		add_random_kstack_offset();
>> +
>>  		if (syscall >= 0 && syscall < NR_syscalls)
>>  			syscall_handler(regs, syscall);
>>  		else if (syscall != -1)
>>  			regs->a0 = -ENOSYS;
>> +		/*
>> +		 * Ultimately, this value will get limited by KSTACK_OFFSET_MAX(),
>> +		 * so the maximum stack offset is 1k bytes (10 bits).
>> +		 *
>> +		 * The actual entropy will be further reduced by the compiler when
>> +		 * applying stack alignment constraints: 16-byte (i.e. 4-bit) aligned
>> +		 * for RV32I or RV64I.
>> +		 *
>> +		 * The resulting 6 bits of entropy is seen in SP[9:4].
>> +		 */
>> +		choose_random_kstack_offset(get_random_u16());
>>
>>  		syscall_exit_to_user_mode(regs);
>>  	} else {
>> --
>> 2.20.1
>>