Skip to contentSkip to conversation

KVM: x86: Fix stalled descriptor table accesses to MMIO

PATCH New weichenchen 2 participants 2 messages Nov 19

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 KVM: x86: Fix stalled descriptor table accesses to MMIO

Use the emulated memory accessors for the explicit operands of LGDT, LIDT, SGDT and SIDT. When the operand has no backing memslot, the standard accessors return X86EMUL_IO_NEEDED without queuing an MMIO fragment. LIDT can repeat the generic operand read at the same RIP, and SIDT can re-enter emulation without sending a write to userspace.

Check the segment bounds of both fields, then read the whole descriptor in one emulated access. On a single page, this lets KVM use the faulting GPA for both fields without translating the base field after userspace has handled part of the read. For a split operand, translate both pages before exiting to userspace. Use the emulator's read cache to resume after MMIO and skip the redundant generic source read.

Use the destination operand for common writeback of SGDT and SIDT. This keeps the entire descriptor available until userspace completes a ten-byte MMIO write.

Fixes c3cd7ffaf57a ("KVM: x86 emulator: x86_emulate_insn() return -1 only in case of emulation failure")
Fixes 129a72a0d3c8 ("KVM: x86: Introduce segmented_write_std")

---
Changes in v2:
- Read the whole LGDT/LIDT operand in one read_emulated() call after validating both fields. This avoids a second GVA translation after userspace has handled the first part of a same-page read, as pointed
out by Sashiko.
- v1: https://lore.kernel.org/kvm/[email protected]/

On Intel VMX/EPT, the 64-bit LIDT and SIDT probes reached the following HLT with the expected data for RAM and MMIO operands. For same-page MMIO, both transferred eight bytes followed by two. A TDP-style mock with a saved GPA counted one GPA lookup and zero software translations through completion for both loads at all three operand sizes.

Six KVM selftests passed, including a split RAM/MMIO LIDT and a segment limit fault before any userspace MMIO exit. Eight kvm-unit-tests suites at a2891d9528ef reported 445 passes, five skips and no failures. Both vendor modules built with KVM_WERROR. Runtime testing used Intel.

Codex assisted with the emulator changes and regression testing.

 arch/x86/kvm/emulate.c | 49 +++++++++++++++++++-----------------------
 1 file changed, 22 insertions(+), 27 deletions(-)
arch/x86/kvm/emulate.c +22 -27
index c1b2128218..3a9a67f4b0 100644
@@ -918,23 +918,6 @@ static void *decode_register(struct x86_emulate_ctxt *ctxt, u8 modrm_reg,
 	return p;
 }
 
-static int read_descriptor(struct x86_emulate_ctxt *ctxt,
-			   struct segmented_address addr,
-			   u16 *size, unsigned long *address, int op_bytes)
-{
-	int rc;
-
-	if (op_bytes == 2)
-		op_bytes = 3;
-	*address = 0;
-	rc = segmented_read_std(ctxt, addr, size, 2);
-	if (rc != X86EMUL_CONTINUE)
-		return rc;
-	addr.ea += 2;
-	rc = segmented_read_std(ctxt, addr, address, op_bytes);
-	return rc;
-}
-
 EM_ASM_2(add);
 EM_ASM_2(or);
 EM_ASM_2(adc);
@@ -3447,10 +3430,9 @@ static int emulate_store_desc_ptr(struct x86_emulate_ctxt *ctxt,
 		ctxt->op_bytes = 4;
 		desc_ptr.address &= 0x00ffffff;
 	}
-	/* Disable writeback. */
-	ctxt->dst.type = OP_NONE;
-	return segmented_write_std(ctxt, ctxt->dst.addr.mem,
-				   &desc_ptr, 2 + ctxt->op_bytes);
+	ctxt->dst.bytes = 2 + ctxt->op_bytes;
+	memcpy(ctxt->dst.valptr, &desc_ptr, ctxt->dst.bytes);
+	return X86EMUL_CONTINUE;
 }
 
 static int em_sgdt(struct x86_emulate_ctxt *ctxt)
@@ -3465,14 +3447,27 @@ static int em_sidt(struct x86_emulate_ctxt *ctxt)
 
 static int em_lgdt_lidt(struct x86_emulate_ctxt *ctxt, bool lgdt)
 {
-	struct desc_ptr desc_ptr;
+	struct segmented_address addr = ctxt->src.addr.mem;
+	struct desc_ptr desc_ptr = {};
+	unsigned long limit_addr, base_addr;
+	unsigned int base_bytes, desc_bytes;
 	int rc;
 
 	if (ctxt->mode == X86EMUL_MODE_PROT64)
 		ctxt->op_bytes = 8;
-	rc = read_descriptor(ctxt, ctxt->src.addr.mem,
-			     &desc_ptr.size, &desc_ptr.address,
-			     ctxt->op_bytes);
+	base_bytes = ctxt->op_bytes == 2 ? 3 : ctxt->op_bytes;
+	desc_bytes = sizeof(desc_ptr.size) + base_bytes;
+
+	/* Check both fields before an MMIO read can have side effects. */
+	rc = linearize(ctxt, addr, sizeof(desc_ptr.size), false, &limit_addr);
+	if (rc != X86EMUL_CONTINUE)
+		return rc;
+	addr.ea += sizeof(desc_ptr.size);
+	rc = linearize(ctxt, addr, base_bytes, false, &base_addr);
+	if (rc != X86EMUL_CONTINUE)
+		return rc;
+
+	rc = read_emulated(ctxt, limit_addr, &desc_ptr, desc_bytes);
 	if (rc != X86EMUL_CONTINUE)
 		return rc;
 	if (ctxt->mode == X86EMUL_MODE_PROT64 &&
@@ -4074,8 +4069,8 @@ static const struct opcode group6[] = {
 static const struct group_dual group7 = { {
 	II(Mov | DstMem,			em_sgdt, sgdt),
 	II(Mov | DstMem,			em_sidt, sidt),
-	II(SrcMem | Priv,			em_lgdt, lgdt),
-	II(SrcMem | Priv,			em_lidt, lidt),
+	II(SrcMem | Priv | NoAccess,		em_lgdt, lgdt),
+	II(SrcMem | Priv | NoAccess,		em_lidt, lidt),
 	II(SrcNone | DstMem | Mov,		em_smsw, smsw), N,
 	II(SrcMem16 | Mov | Priv,		em_lmsw, lmsw),
 	II(SrcMem | ByteOp | Priv | NoAccess,	em_invlpg, invlpg),

base-commit: cbe54fc3462b3949ab27b9ae81fb9ddb509040c7

Signature · 1 line
2.53.0
#1 lore raw
Thank you for your contribution! Sashiko AI review found 1 potential issue to consider: - [Medium] KVM: x86: unmasked addr.ea increment in em_lgdt_lidt() breaks 32-bit wrap --

Thank you for your contribution! Sashiko AI review found 1 potential issue to consider:
- [Medium] KVM: x86: unmasked addr.ea increment in em_lgdt_lidt() breaks 32-bit wrap
--

commit 14df3ccaae0dc59c054ba57d9b0335ad6207effb
Author: weichenchen <[email protected]>

KVM: x86: Fix stalled descriptor table accesses to MMIO

This commit switches LGDT, LIDT, SGDT, and SIDT operand accesses from standard memory accessors to emulated memory accessors and common writeback so MMIO operands do not stall emulation.

Quoted 4 lines
arch/x86/kvm/emulate.c +0 -0
index c1b21282187f3..3a9a67f4b0d7c 100644

[ ... ]

Quoted 19 lines
arch/x86/kvm/emulate.c +7 -4
@@ -3465,14 +3447,27 @@ static int em_sidt(struct x86_emulate_ctxt *ctxt)
 
 static int em_lgdt_lidt(struct x86_emulate_ctxt *ctxt, bool lgdt)
 {
-	struct desc_ptr desc_ptr;
+	struct segmented_address addr = ctxt->src.addr.mem;
+	struct desc_ptr desc_ptr = {};
+	unsigned long limit_addr, base_addr;
+	unsigned int base_bytes, desc_bytes;
 	int rc;
 
 	if (ctxt->mode == X86EMUL_MODE_PROT64)
 		ctxt->op_bytes = 8;
-	rc = read_descriptor(ctxt, ctxt->src.addr.mem,
-			     &desc_ptr.size, &desc_ptr.address,
-			     ctxt->op_bytes);
+	base_bytes = ctxt->op_bytes == 2 ? 3 : ctxt->op_bytes;
+	desc_bytes = sizeof(desc_ptr.size) + base_bytes;
+

arch/x86/kvm/emulate.c static int em_sidt(struct x86_emulate_ctxt *ctxt) · line 3470

Quoted 10 lines
arch/x86/kvm/emulate.c +10 -0
+	/* Check both fields before an MMIO read can have side effects. */
+	rc = linearize(ctxt, addr, sizeof(desc_ptr.size), false, &limit_addr);
+	if (rc != X86EMUL_CONTINUE)
+		return rc;
+	addr.ea += sizeof(desc_ptr.size);
+	rc = linearize(ctxt, addr, base_bytes, false, &base_addr);
+	if (rc != X86EMUL_CONTINUE)
+		return rc;
+
+	rc = read_emulated(ctxt, limit_addr, &desc_ptr, desc_bytes);

[Severity: Medium]
Does incrementing addr.ea without 32-bit masking break 32-bit address wraparound for LGDT/LIDT at the top of a 4GB segment?

Because addr.ea is a 64-bit unsigned long, an operand at 0xfffffffe or 0xffffffff in 16-bit or 32-bit mode overflows addr.ea to 0x100000000 or 0x100000001, causing the second linearize() call to fail the addr.ea > lim check and inject a spurious #GP(0) or #SS(0) when lim == 0xffffffff.

A single linearize(ctxt, addr, desc_bytes, false, &limit_addr) call (or segmented_read(ctxt, ctxt->src.addr.mem, &desc_ptr, desc_bytes)) already checks that the full descriptor fits within the segment limit when lim != 0xffffffff while preserving 32-bit wraparound when lim == 0xffffffff, and avoids the unused base_addr variable.

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1