[PATCH RESEND] riscv: asid: Fixup stale TLB entry cause application crash

From: guoren@kernel.org
To: anup@brainfault.org, paul.walmsley@sifive.com,
	palmer@dabbelt.com, conor.dooley@microchip.com, heiko@sntech.de,
	philipp.tomsich@vrull.eu
Cc: linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org,
	Guo Ren <guoren@linux.alibaba.com>, Guo Ren <guoren@kernel.org>,
	Anup Patel <apatel@ventanamicro.com>,
	Palmer Dabbelt <palmer@rivosinc.com>
Subject: [PATCH RESEND] riscv: asid: Fixup stale TLB entry cause application crash
Date: Tue,  8 Nov 2022 05:20:44 -0500	[thread overview]
Message-ID: <20221108102044.3317793-1-guoren@kernel.org> (raw)

From: Guo Ren <guoren@linux.alibaba.com>

After use_asid_allocator enabled, the userspace application will
crash for stale tlb entry. Because only using cpumask_clear_cpu without
local_flush_tlb_all couldn't guarantee CPU's tlb entries fresh. Then
set_mm_asid would cause user space application get a stale value by
the stale tlb entry, but set_mm_noasid is okay.

Here is the symptom of the bug:
unhandled signal 11 code 0x1 (coredump)
   0x0000003fd6d22524 <+4>:     auipc   s0,0x70
   0x0000003fd6d22528 <+8>:     ld      s0,-148(s0) # 0x3fd6d92490
=> 0x0000003fd6d2252c <+12>:    ld      a5,0(s0)
(gdb) i r s0
s0          0x8082ed1cc3198b21       0x8082ed1cc3198b21
(gdb) x/16 0x3fd6d92490
0x3fd6d92490:   0xd80ac8a8      0x0000003f
The core dump file shows that the value of register s0 is wrong, but the
value in memory is right. This is because 'ld s0, -148(s0)' use a stale
mapping entry in TLB and got a wrong value from a stale physical
address.

When task run on CPU0, the task loaded/speculative-loaded the value of
address(0x3fd6d92490), and the first version of tlb mapping entry was
PTWed into CPU0's tlb.
When the task switched from CPU0 to CPU1 without local_tlb_flush_all
(because of asid), the task happened to write a value on address
(0x3fd6d92490). It caused do_page_fault -> wp_page_copy ->
ptep_clear_flush -> ptep_get_and_clear & flush_tlb_page.
The flush_tlb_page used mm_cpumask(mm) to determine which CPUs need
tlb flush, but CPU0 had cleared the CPU0's mm_cpumask in previous switch_mm.
So we only flushed the CPU1 tlb, and setted second version mapping
of the pte. When the task switch from CPU1 to CPU0 again, CPU0 still used a
stale tlb mapping entry which contained a wrong target physical address.
When the task happened to read that value, the bug would be raised.

Fixes: 65d4b9c53017 ("RISC-V: Implement ASID allocator")
Signed-off-by: Guo Ren <guoren@linux.alibaba.com>
Signed-off-by: Guo Ren <guoren@kernel.org>
Cc: Anup Patel <apatel@ventanamicro.com>
Cc: Palmer Dabbelt <palmer@rivosinc.com>
---
 arch/riscv/mm/context.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/arch/riscv/mm/context.c b/arch/riscv/mm/context.c
index 7acbfbd14557..8ad6c2493e93 100644
--- a/arch/riscv/mm/context.c
+++ b/arch/riscv/mm/context.c
@@ -317,7 +317,9 @@ void switch_mm(struct mm_struct *prev, struct mm_struct *next,
 	 */
 	cpu = smp_processor_id();
 
-	cpumask_clear_cpu(cpu, mm_cpumask(prev));
+	if (!static_branch_unlikely(&use_asid_allocator))
+		cpumask_clear_cpu(cpu, mm_cpumask(prev));
+
 	cpumask_set_cpu(cpu, mm_cpumask(next));
 
 	set_mm(next, cpu);
-- 
2.36.1

