Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Message-Id: <c9b86c174b96a0d4aac9b9036d7fcfb459214e52.1579885400.git.christophe.leroy@c-s.fr>
From:   Christophe Leroy <christophe.leroy@c-s.fr>
Subject: [PATCH v2] powerpc/32: Warn and return ENOSYS on syscalls from kernel
To:     Benjamin Herrenschmidt <benh@kernel.crashing.org>,
        Paul Mackerras <paulus@samba.org>,
        Michael Ellerman <mpe@ellerman.id.au>
Cc:     linux-kernel@vger.kernel.org, linuxppc-dev@lists.ozlabs.org
Date:   Fri, 24 Jan 2020 17:03:49 +0000 (UTC)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Since commit b86fb88855ea ("powerpc/32: implement fast entry for
syscalls on non BOOKE") and commit 1a4b739bbb4f ("powerpc/32:
implement fast entry for syscalls on BOOKE"), syscalls from
kernel are unexpected and can have catastrophic consequences
as it will destroy the kernel stack.

Test MSR_PR on syscall entry. In case syscall is from kernel,
emit a warning and return ENOSYS error.

Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
---
v2: Rebased on powerpc/next-test, ie on top of VMAP_STACK series
---
 arch/powerpc/kernel/entry_32.S   | 26 ++++++++++++++++++++++++++
 arch/powerpc/kernel/head_32.h    | 18 +++++++++++-------
 arch/powerpc/kernel/head_booke.h |  5 ++++-
 3 files changed, 41 insertions(+), 8 deletions(-)

diff --git a/arch/powerpc/kernel/entry_32.S b/arch/powerpc/kernel/entry_32.S
index 3795654d15d1..73b80143ffac 100644
--- a/arch/powerpc/kernel/entry_32.S
+++ b/arch/powerpc/kernel/entry_32.S
@@ -574,6 +574,32 @@ syscall_exit_work:
 	bl	do_syscall_trace_leave
 	b	ret_from_except_full
 
+	/*
+	 * System call was called from kernel. We get here with SRR1 in r9.
+	 * Mark the exception as recoverable once we have retrieved SRR0,
+	 * trap a warning and return ENOSYS with CR[SO] set.
+	 */
+	.globl	ret_from_kernel_syscall
+ret_from_kernel_syscall:
+	mfspr	r11, SPRN_SRR0
+#if !defined(CONFIG_4xx) && !defined(CONFIG_BOOKE)
+	LOAD_REG_IMMEDIATE(r12, MSR_KERNEL & ~(MSR_IR|MSR_DR))
+	MTMSRD(r12)
+#endif
+
+0:	trap
+	EMIT_BUG_ENTRY 0b,__FILE__,__LINE__, BUGFLAG_WARNING
+
+	li	r3, ENOSYS
+	crset	so
+#if defined(CONFIG_PPC_8xx) && defined(CONFIG_PERF_EVENTS)
+	mtspr	SPRN_NRI, r0
+#endif
+	mtspr	SPRN_SRR1, r9
+	mtspr	SPRN_SRR0, r11
+	SYNC
+	RFI
+
 /*
  * The fork/clone functions need to copy the full register set into
  * the child process. Therefore we need to save all the nonvolatile
diff --git a/arch/powerpc/kernel/head_32.h b/arch/powerpc/kernel/head_32.h
index a6a5fbbf8504..5f3cfc9ef1b6 100644
--- a/arch/powerpc/kernel/head_32.h
+++ b/arch/powerpc/kernel/head_32.h
@@ -112,13 +112,17 @@
 .macro SYSCALL_ENTRY trapno
 	mfspr	r12,SPRN_SPRG_THREAD
 #ifdef CONFIG_VMAP_STACK
-	mfspr	r9, SPRN_SRR0
-	mfspr	r11, SPRN_SRR1
-	stw	r9, SRR0(r12)
-	stw	r11, SRR1(r12)
+	mfspr	r11, SPRN_SRR0
+	mfspr	r9, SPRN_SRR1
+	stw	r11, SRR0(r12)
+	stw	r9, SRR1(r12)
+#else
+	mfspr	r9, SPRN_SRR1
 #endif
 	mfcr	r10
+	andi.	r11, r9, MSR_PR
 	lwz	r11,TASK_STACK-THREAD(r12)
+	beq-	99f
 	rlwinm	r10,r10,0,4,2	/* Clear SO bit in CR */
 	addi	r11, r11, THREAD_SIZE - INT_FRAME_SIZE
 #ifdef CONFIG_VMAP_STACK
@@ -128,15 +132,14 @@
 #endif
 	tovirt_vmstack r12, r12
 	tophys_novmstack r11, r11
-	mflr	r9
 	stw	r10,_CCR(r11)		/* save registers */
-	stw	r9, _LINK(r11)
+	mflr	r10
+	stw	r10, _LINK(r11)
 #ifdef CONFIG_VMAP_STACK
 	lwz	r10, SRR0(r12)
 	lwz	r9, SRR1(r12)
 #else
 	mfspr	r10,SPRN_SRR0
-	mfspr	r9,SPRN_SRR1
 #endif
 	stw	r1,GPR1(r11)
 	stw	r1,0(r11)
@@ -209,6 +212,7 @@
 	mtspr	SPRN_SRR0,r11
 	SYNC
 	RFI				/* jump to handler, enable MMU */
+99:	b	ret_from_kernel_syscall
 .endm
 
 .macro save_dar_dsisr_on_stack reg1, reg2, sp
diff --git a/arch/powerpc/kernel/head_booke.h b/arch/powerpc/kernel/head_booke.h
index 37fc84ed90e3..bd2e5ed8dd50 100644
--- a/arch/powerpc/kernel/head_booke.h
+++ b/arch/powerpc/kernel/head_booke.h
@@ -104,16 +104,18 @@ FTR_SECTION_ELSE
 #ifdef CONFIG_KVM_BOOKE_HV
 ALT_FTR_SECTION_END_IFSET(CPU_FTR_EMB_HV)
 #endif
+	mfspr	r9, SPRN_SRR1
 	BOOKE_CLEAR_BTB(r11)
+	andi.	r11, r9, MSR_PR
 	lwz	r11, TASK_STACK - THREAD(r10)
 	rlwinm	r12,r12,0,4,2	/* Clear SO bit in CR */
+	beq-	99f
 	ALLOC_STACK_FRAME(r11, THREAD_SIZE - INT_FRAME_SIZE)
 	stw	r12, _CCR(r11)		/* save various registers */
 	mflr	r12
 	stw	r12,_LINK(r11)
 	mfspr	r12,SPRN_SRR0
 	stw	r1, GPR1(r11)
-	mfspr	r9,SPRN_SRR1
 	stw	r1, 0(r11)
 	mr	r1, r11
 	stw	r12,_NIP(r11)
@@ -176,6 +178,7 @@ ALT_FTR_SECTION_END_IFSET(CPU_FTR_EMB_HV)
 	mtspr	SPRN_SRR0,r11
 	SYNC
 	RFI				/* jump to handler, enable MMU */
+99:	b	ret_from_kernel_syscall
 .endm
 
 /* To handle the additional exception priority levels on 40x and Book-E
-- 
2.25.0