From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id B2789C433EF for ; Tue, 5 Oct 2021 20:42:27 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 9BFA9610C8 for ; Tue, 5 Oct 2021 20:42:27 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236732AbhJEUoP (ORCPT ); Tue, 5 Oct 2021 16:44:15 -0400 Received: from mga07.intel.com ([134.134.136.100]:11170 "EHLO mga07.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236086AbhJEUnz (ORCPT ); Tue, 5 Oct 2021 16:43:55 -0400 X-IronPort-AV: E=McAfee;i="6200,9189,10128"; a="289354613" X-IronPort-AV: E=Sophos;i="5.85,349,1624345200"; d="scan'208";a="289354613" Received: from fmsmga008.fm.intel.com ([10.253.24.58]) by orsmga105.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 05 Oct 2021 13:42:04 -0700 X-IronPort-AV: E=Sophos;i="5.85,349,1624345200"; d="scan'208";a="523979698" Received: from alyee-mobl.amr.corp.intel.com (HELO skuppusw-desk1.amr.corp.intel.com) ([10.254.5.222]) by fmsmga008-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 05 Oct 2021 13:42:03 -0700 From: Kuppuswamy Sathyanarayanan To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , x86@kernel.org, Paolo Bonzini , David Hildenbrand , Andrea Arcangeli , Josh Poimboeuf , "H . Peter Anvin" Cc: Dave Hansen , Tony Luck , Dan Williams , Andi Kleen , Kirill Shutemov , Sean Christopherson , Kuppuswamy Sathyanarayanan , linux-kernel@vger.kernel.org Subject: [PATCH v7 09/10] x86/tdx: Handle in-kernel MMIO Date: Tue, 5 Oct 2021 13:41:35 -0700 Message-Id: <20211005204136.1812078-10-sathyanarayanan.kuppuswamy@linux.intel.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20211005204136.1812078-1-sathyanarayanan.kuppuswamy@linux.intel.com> References: <20211005204136.1812078-1-sathyanarayanan.kuppuswamy@linux.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: "Kirill A. Shutemov" In traditional VMs, MMIO is usually implemented by giving a guest access to a mapping which will cause a VMEXIT on access and then the VMM emulating the access. That's not possible in TDX guest because VMEXIT will expose the register state to the host. TDX guests don't trust the host and can't have its state exposed to the host. In TDX the MMIO regions are instead configured to trigger a #VE exception in the guest. The guest #VE handler then emulates the MMIO instruction inside the guest and converts them into a controlled TDCALL to the host, rather than completely exposing the state to the host. Currently, TDX only supports MMIO for instructions that are known to come from io.h macros (build_mmio_read/write()). For drivers that don't use the io.h macros or uses structure overlay to do MMIO are currently not supported in TDX guest (for example the MMIO based XAPIC is disable at runtime for TDX). This way of handling is similar to AMD SEV. Also, reasons for supporting #VE based MMIO in TDX guest are, * MMIO is widely used and more drivers will be added in the future. * To avoid annotating every TDX specific MMIO readl/writel etc. * If annotation is not preferred, an alternative way is required for every MMIO access in the kernel (even though 99.9% will never be used on TDX) which would be a complete waste and incredible binary bloat for nothing. Signed-off-by: Kirill A. Shutemov Reviewed-by: Andi Kleen Reviewed-by: Tony Luck Signed-off-by: Kuppuswamy Sathyanarayanan --- Changes since v6: * None Changes since v5: * None Changes since v4: * Changed order of variable declaration in tdx_handle_mmio(). * Changed tdg_* prefix with tdx_*. Changes since v3: * None Changes since v2: * None arch/x86/kernel/tdx.c | 108 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 108 insertions(+) diff --git a/arch/x86/kernel/tdx.c b/arch/x86/kernel/tdx.c index cd0fb5d14ad7..851ad143da03 100644 --- a/arch/x86/kernel/tdx.c +++ b/arch/x86/kernel/tdx.c @@ -6,6 +6,9 @@ #include #include +#include +#include +#include /* force_sig_fault() */ /* TDX Module call Leaf IDs */ #define TDGETVEINFO 3 @@ -207,6 +210,103 @@ static void tdx_handle_io(struct pt_regs *regs, u32 exit_qual) } } +static unsigned long tdx_mmio(int size, bool write, unsigned long addr, + unsigned long *val) +{ + struct tdx_hypercall_output out = {0}; + u64 err; + + err = _tdx_hypercall(EXIT_REASON_EPT_VIOLATION, size, write, + addr, *val, &out); + *val = out.r11; + return err; +} + +static int tdx_handle_mmio(struct pt_regs *regs, struct ve_info *ve) +{ + char buffer[MAX_INSN_SIZE]; + unsigned long *reg, val; + struct insn insn = {}; + enum mmio_type mmio; + int size, ret; + u8 sign_byte; + + if (user_mode(regs)) { + ret = insn_fetch_from_user(regs, buffer); + if (!ret) + return -EFAULT; + if (!insn_decode_from_regs(&insn, regs, buffer, ret)) + return -EFAULT; + } else { + ret = copy_from_kernel_nofault(buffer, (void *)regs->ip, + MAX_INSN_SIZE); + if (ret) + return -EFAULT; + insn_init(&insn, buffer, MAX_INSN_SIZE, 1); + insn_get_length(&insn); + } + + mmio = insn_decode_mmio(&insn, &size); + if (mmio == MMIO_DECODE_FAILED) + return -EFAULT; + + if (mmio != MMIO_WRITE_IMM && mmio != MMIO_MOVS) { + reg = insn_get_modrm_reg_ptr(&insn, regs); + if (!reg) + return -EFAULT; + } + + switch (mmio) { + case MMIO_WRITE: + memcpy(&val, reg, size); + ret = tdx_mmio(size, true, ve->gpa, &val); + break; + case MMIO_WRITE_IMM: + val = insn.immediate.value; + ret = tdx_mmio(size, true, ve->gpa, &val); + break; + case MMIO_READ: + ret = tdx_mmio(size, false, ve->gpa, &val); + if (ret) + break; + /* Zero-extend for 32-bit operation */ + if (size == 4) + *reg = 0; + memcpy(reg, &val, size); + break; + case MMIO_READ_ZERO_EXTEND: + ret = tdx_mmio(size, false, ve->gpa, &val); + if (ret) + break; + + /* Zero extend based on operand size */ + memset(reg, 0, insn.opnd_bytes); + memcpy(reg, &val, size); + break; + case MMIO_READ_SIGN_EXTEND: + ret = tdx_mmio(size, false, ve->gpa, &val); + if (ret) + break; + + if (size == 1) + sign_byte = (val & 0x80) ? 0xff : 0x00; + else + sign_byte = (val & 0x8000) ? 0xff : 0x00; + + /* Sign extend based on operand size */ + memset(reg, sign_byte, insn.opnd_bytes); + memcpy(reg, &val, size); + break; + case MMIO_MOVS: + case MMIO_DECODE_FAILED: + return -EFAULT; + } + + if (ret) + return -EFAULT; + return insn.length; +} + unsigned long tdx_get_ve_info(struct ve_info *ve) { struct tdx_module_output out = {0}; @@ -256,6 +356,14 @@ int tdx_handle_virtualization_exception(struct pt_regs *regs, case EXIT_REASON_IO_INSTRUCTION: tdx_handle_io(regs, ve->exit_qual); break; + case EXIT_REASON_EPT_VIOLATION: + /* Currently only MMIO triggers EPT violation */ + ve->instr_len = tdx_handle_mmio(regs, ve); + if (ve->instr_len < 0) { + pr_warn_once("MMIO failed\n"); + return -EFAULT; + } + break; default: pr_warn("Unexpected #VE: %lld\n", ve->exit_reason); return -EFAULT; -- 2.25.1