[tip:x86/mpx] x86/mpx: Do not use SIB.base if its value is 101b and ModRM.mod = 0

[tip-bot for Ricardo Neri]

Commit-ID:  4578f06fc93fb73c9c644ed838f4cdabbfdc4df1
Gitweb:     https://git.kernel.org/tip/4578f06fc93fb73c9c644ed838f4cdabbfdc4df1
Author:     Ricardo Neri <ricardo.neri-calderon@xxxxxxxxxxxxxxx>
AuthorDate: Fri, 27 Oct 2017 13:25:35 -0700
Committer:  Thomas Gleixner <tglx@xxxxxxxxxxxxx>
CommitDate: Wed, 1 Nov 2017 21:50:10 +0100

x86/mpx: Do not use SIB.base if its value is 101b and ModRM.mod = 0

Section 2.2.1.2 of the Intel 64 and IA-32 Architectures Software
Developer's Manual volume 2A states that if a SIB byte is used and
SIB.base is 101b and ModRM.mod is zero, then the base part of the base
part of the effective address computation is null. To signal this
situation, a -EDOM error is returned to indicate callers to ignore the
base value present in the register operand.

In this scenario, a 32-bit displacement follows the SIB byte. Displacement
is obtained when the instruction decoder parses the operands.

Signed-off-by: Ricardo Neri <ricardo.neri-calderon@xxxxxxxxxxxxxxx>
Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Reviewed-by: Borislav Petkov <bp@xxxxxxx>
Cc: Adan Hawthorn <adanhawthorn@xxxxxxxxx>
Cc: "Michael S. Tsirkin" <mst@xxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
Cc: ricardo.neri@xxxxxxxxx
Cc: Paul Gortmaker <paul.gortmaker@xxxxxxxxxxxxx>
Cc: Huang Rui <ray.huang@xxxxxxx>
Cc: Qiaowei Ren <qiaowei.ren@xxxxxxxxx>
Cc: Shuah Khan <shuah@xxxxxxxxxx>
Cc: Jonathan Corbet <corbet@xxxxxxx>
Cc: Jiri Slaby <jslaby@xxxxxxx>
Cc: Nathan Howard <liverlint@xxxxxxxxx>
Cc: "Ravi V. Shankar" <ravi.v.shankar@xxxxxxxxx>
Cc: Chris Metcalf <cmetcalf@xxxxxxxxxxxx>
Cc: Brian Gerst <brgerst@xxxxxxxxx>
Cc: Andy Lutomirski <luto@xxxxxxxxxx>
Cc: Colin Ian King <colin.king@xxxxxxxxxxxxx>
Cc: Chen Yucong <slaoub@xxxxxxxxx>
Cc: Adam Buchbinder <adam.buchbinder@xxxxxxxxx>
Cc: Vlastimil Babka <vbabka@xxxxxxx>
Cc: Lorenzo Stoakes <lstoakes@xxxxxxxxx>
Cc: Masami Hiramatsu <mhiramat@xxxxxxxxxx>
Cc: Joe Perches <joe@xxxxxxxxxxx>
Cc: Paolo Bonzini <pbonzini@xxxxxxxxxx>
Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Link: https://lkml.kernel.org/r/1509135945-13762-9-git-send-email-ricardo.neri-calderon@xxxxxxxxxxxxxxx

---
 arch/x86/mm/mpx.c | 28 ++++++++++++++++++++--------
 1 file changed, 20 insertions(+), 8 deletions(-)

diff --git a/arch/x86/mm/mpx.c b/arch/x86/mm/mpx.c
index 2ad1d4a..581a960 100644
--- a/arch/x86/mm/mpx.c
+++ b/arch/x86/mm/mpx.c
@@ -123,6 +123,14 @@ static int get_reg_offset(struct insn *insn, struct pt_regs *regs,
 
 	case REG_TYPE_BASE:
 		regno = X86_SIB_BASE(insn->sib.value);
+		/*
+		 * If ModRM.mod is 0 and SIB.base == 5, the base of the
+		 * register-indirect addressing is 0. In this case, a
+		 * 32-bit displacement follows the SIB byte.
+		 */
+		if (!X86_MODRM_MOD(insn->modrm.value) && regno == 5)
+			return -EDOM;
+
 		if (X86_REX_B(insn->rex_prefix.value))
 			regno += 8;
 		break;
@@ -164,16 +172,22 @@ static void __user *mpx_get_addr_ref(struct insn *insn, struct pt_regs *regs)
 		eff_addr = regs_get_register(regs, addr_offset);
 	} else {
 		if (insn->sib.nbytes) {
+			/*
+			 * Negative values in the base and index offset means
+			 * an error when decoding the SIB byte. Except -EDOM,
+			 * which means that the registers should not be used
+			 * in the address computation.
+			 */
 			base_offset = get_reg_offset(insn, regs, REG_TYPE_BASE);
-			if (base_offset < 0)
+			if (base_offset == -EDOM)
+				base = 0;
+			else if (base_offset < 0)
 				goto out;
+			else
+				base = regs_get_register(regs, base_offset);
 
 			indx_offset = get_reg_offset(insn, regs, REG_TYPE_INDEX);
-			/*
-			 * A negative offset generally means a error, except
-			 * -EDOM, which means that the contents of the register
-			 * should not be used as index.
-			 */
+
 			if (indx_offset == -EDOM)
 				indx = 0;
 			else if (indx_offset < 0)
@@ -181,8 +195,6 @@ static void __user *mpx_get_addr_ref(struct insn *insn, struct pt_regs *regs)
 			else
 				indx = regs_get_register(regs, indx_offset);
 
-			base = regs_get_register(regs, base_offset);
-
 			eff_addr = base + indx * (1 << X86_SIB_SCALE(sib));
 		} else {
 			addr_offset = get_reg_offset(insn, regs, REG_TYPE_RM);