Skip to contentSkip to conversation

intel_txt: Intel(R) TXT reboot/halt shutdown support

RFC PATCH NAK Joseph Cihula 5 participants 14 messages Jul 1, 2009 – Aug 26, 2009

Subsystems (7)
x86 ACPI PCI Security (LSM) +3 more subsystems
lore
Share & tools
lore

Joseph Cihula

author
#0 lore raw

RFC PATCH v6 2/4 intel_txt: Intel(R) TXT reboot/halt shutdown support

Support for graceful handling of kernel reboots after an Intel(R) TXT launch.

Without this patch, attempting to reboot or halt the system will cause the TXT hardware to lock memory upon system restart because the secrets-in-memory flag that was set on launch was never cleared. This will in turn cause BIOS to execute a TXT Authenticated Code Module (ACM) that will scrub all of memory and then unlock it. Depending on the amount of memory in the system and its type, this may take some time.

This patch creates a 1:1 address mapping to the tboot module and then calls back into tboot so that it may properly and securely clean up system state and clear the secrets-in-memory flag. When it has completed these steps, the tboot module will reboot or halt the system.

 arch/x86/kernel/reboot.c |    8 ++++++++
 init/main.c              |    3 +++
 2 files changed, 11 insertions(+)
Signed-off-by Joseph Cihula <[email protected]>
Signed-off-by Shane Wang <[email protected]>

---

arch/x86/kernel/reboot.c +8 -0
diff -uprN -X linus-2.6.git-0629/Documentation/dontdiff linus-2.6.git-0629/arch/x86/kernel/reboot.c linus-2.6.git-0629-txt/arch/x86/kernel/reboot.c
@@ -24,6 +24,8 @@
 # include <asm/iommu.h>
 #endif
 
+#include <asm/tboot.h>
+
 /*
  * Power off function, if any
  */

@@ -460,6 +462,8 @@ static void native_machine_emergency_res
 	if (reboot_emergency)
 		emergency_vmx_disable_all();
 
+	tboot_shutdown(TB_SHUTDOWN_REBOOT);
+
 	/* Tell the BIOS if we want cold or warm reboot */
 	*((unsigned short *)__va(0x472)) = reboot_mode;
 

@@ -586,6 +590,8 @@ static void native_machine_halt(void)
 	/* stop other cpus and apics */
 	machine_shutdown();
 
+	tboot_shutdown(TB_SHUTDOWN_HALT);
+
 	/* stop this cpu */
 	stop_this_cpu(NULL);
 }

@@ -597,6 +603,8 @@ static void native_machine_power_off(voi
 			machine_shutdown();
 		pm_power_off();
 	}
+	/* a fallback in case there is no PM info available */
+	tboot_shutdown(TB_SHUTDOWN_HALT);
 }
 
 struct machine_ops machine_ops = {
init/main.c +3 -0
diff -uprN -X linus-2.6.git-0629/Documentation/dontdiff linus-2.6.git-0629/init/main.c linus-2.6.git-0629-txt/init/main.c
@@ -73,6 +73,7 @@
 #include <asm/io.h>
 #include <asm/bugs.h>
 #include <asm/setup.h>
+#include <asm/tboot.h>
 #include <asm/sections.h>
 #include <asm/cacheflush.h>
 

@@ -715,6 +716,8 @@ asmlinkage void __init start_kernel(void
 
 	ftrace_init();
 
+	tboot_create_trampoline();
+
 	/* Do the rest non-__init'ed, we're now alive */
 	rest_init();
 }
#1 lore raw
Quoted 12 lines · Joseph Cihula

diff -uprN -X linus-2.6.git-0629/Documentation/dontdiff linus-2.6.git-0629/init/main.c linus-2.6.git-0629-txt/init/main.c

init/main.c +1 -0
@@ -73,6 +73,7 @@
 #include <asm/io.h>
 #include <asm/bugs.h>
 #include <asm/setup.h>
+#include <asm/tboot.h>
 #include <asm/sections.h>
 #include <asm/cacheflush.h>

init/main.c asmlinkage void __init start_kernel(void · line 723

Quoted 8 lines
init/main.c +2 -0
@@ -715,6 +716,8 @@ asmlinkage void __init start_kernel(void
 
 	ftrace_init();
 
+	tboot_create_trampoline();
+
 	/* Do the rest non-__init'ed, we're now alive */
 	rest_init();
 }

hm, this breaks the kernel build on every non-x86 architecture - none of which has asm/tboot.h.

Ingo

#2 lore raw
Quoted 8 lines · Ingo Molnar

init/main.c asmlinkage void __init start_kernel(void · line 723

Quoted 5 lines
init/main.c +2 -0
+	tboot_create_trampoline();
+
 	/* Do the rest non-__init'ed, we're now alive */
 	rest_init();
 }

hm, this breaks the kernel build on every non-x86 architecture - none of which has asm/tboot.h.

also, tboot modifies kernel/cpu.c (and that too breaks the build) - why? I've excluded x86/txt from tip:master for now.

Ingo

#3 lore raw

Hi Ingo & hpa,

I have made some change on the previous tboot patch to fix the build errors. Please comment.

Thanks.
Shane

Quoted 11 lines · Ingo Molnar
Quoted 8 lines · Ingo Molnar

Review

Quoted 5 lines
(quoted diff) +2 -0
+	tboot_create_trampoline();
+
 	/* Do the rest non-__init'ed, we're now alive */
 	rest_init();
 }

hm, this breaks the kernel build on every non-x86 architecture - none of which has asm/tboot.h.

Quoted 4 lines

also, tboot modifies kernel/cpu.c (and that too breaks the build) - why? I've excluded x86/txt from tip:master for now.

Ingo

drivers/acpi/acpica/hwsleep.c +5 -0
diff -r 04f249f00303 drivers/acpi/acpica/hwsleep.c
@@ -45,7 +45,12 @@
 #include <acpi/acpi.h>
 #include "accommon.h"
 #include "actables.h"
+
+#ifdef CONFIG_X86
 #include <asm/tboot.h>
+#else
+#define tboot_sleep(sleep_state, pm1a_control, pm1b_control) do { } while (0)
+#endif
 
 #define _COMPONENT          ACPI_HARDWARE
 ACPI_MODULE_NAME("hwsleep")
drivers/pci/dmar.c +5 -0
diff -r 04f249f00303 drivers/pci/dmar.c
@@ -33,7 +33,12 @@
 #include <linux/timer.h>
 #include <linux/irq.h>
 #include <linux/interrupt.h>
+
+#ifdef CONFIG_X86
 #include <asm/tboot.h>
+#else
+#define tboot_get_dmar_table(x) (x)
+#endif
 
 #undef PREFIX
 #define PREFIX "DMAR:"
drivers/pci/intel-iommu.c +6 -1
diff -r 04f249f00303 drivers/pci/intel-iommu.c
@@ -38,9 +38,14 @@
 #include <linux/intel-iommu.h>
 #include <linux/sysdev.h>
 #include <asm/cacheflush.h>
-#include <asm/tboot.h>
 #include <asm/iommu.h>
 #include "pci.h"
+
+#ifdef CONFIG_X86
+#include <asm/tboot.h>
+#else
+#define tboot_force_iommu() 0
+#endif
 
 #define ROOT_SIZE		VTD_PAGE_SIZE
 #define CONTEXT_SIZE		VTD_PAGE_SIZE
init/main.c +7 -1
diff -r 04f249f00303 init/main.c
@@ -73,13 +73,19 @@
 #include <asm/io.h>
 #include <asm/bugs.h>
 #include <asm/setup.h>
-#include <asm/tboot.h>
 #include <asm/sections.h>
 #include <asm/cacheflush.h>
 
 #ifdef CONFIG_X86_LOCAL_APIC
 #include <asm/smp.h>
 #endif
+
+#ifdef CONFIG_X86
+#include <asm/tboot.h>
+#else
+#define tboot_create_trampoline() do { } while (0)
+#endif
+
 
 static int kernel_init(void *);
kernel/cpu.c +5 -0
diff -r 04f249f00303 kernel/cpu.c
@@ -14,7 +14,12 @@
 #include <linux/kthread.h>
 #include <linux/stop_machine.h>
 #include <linux/mutex.h>
+
+#ifdef CONFIG_X86
 #include <asm/tboot.h>
+#else
+#define tboot_wait_for_aps(x) 0
+#endif
 
 #ifdef CONFIG_SMP
 /* Serializes the updates to cpu_online_mask, cpu_present_mask */
reviewer
#4 lore raw
Quoted 8 lines · Shane

Hi Ingo & hpa,

I have made some change on the previous tboot patch to fix the build errors. Please comment.

Thanks.
Shane

This code addresses the build errors, but not in the right manner.

We should not have to have tboot-specific #ifdefs in tons and tons of arch-generic files; we need to figure out how to move them into code flow already in arch-specific code, or we need to very carefully figure out what additional arch-specific operations are needed at a higher level.

-hpa

Signature · 2 lines
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.
#5 lore raw
In reply to Wang, Shane
Quoted 4 lines · Shane

Hi Ingo & hpa,

I have made some change on the previous tboot patch to fix the build errors. Please comment.

The standard way to do this would be to define a linux/tboot.h that does the ifdef and in the else path define dummy inlines for these functions.

-Andi

#6 lore raw

PATCH 1/1 txt: fix the build errors on non-X86 platforms

This patch moves tboot.h from asm to linux to fix the build errors on non-X86 platforms.

Signed-off-by Shane Wang <[email protected]>
arch/x86/include/asm/tboot.h deleted +0 -197
diff -r 04f249f00303 arch/x86/include/asm/tboot.h
@@ -1,197 +0,0 @@
-/*
- * tboot.h: shared data structure with tboot and kernel and functions
- *          used by kernel for runtime support of Intel(R) Trusted
- *          Execution Technology
- *
- * Copyright (c) 2006-2009, Intel Corporation
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc.,
- * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
- *
- */
-
-#ifndef _ASM_TBOOT_H
-#define _ASM_TBOOT_H
-
-#include <acpi/acpi.h>
-
-/* these must have the values from 0-5 in this order */
-enum {
-	TB_SHUTDOWN_REBOOT = 0,
-	TB_SHUTDOWN_S5,
-	TB_SHUTDOWN_S4,
-	TB_SHUTDOWN_S3,
-	TB_SHUTDOWN_HALT,
-	TB_SHUTDOWN_WFS
-};
-
-#ifdef CONFIG_INTEL_TXT
-
-/* used to communicate between tboot and the launched kernel */
-
-#define TB_KEY_SIZE             64   /* 512 bits */
-
-#define MAX_TB_MAC_REGIONS      32
-
-struct tboot_mac_region {
-	u64  start;         /* must be 64 byte -aligned */
-	u32  size;          /* must be 64 byte -granular */
-} __packed;
-
-/* GAS - Generic Address Structure (ACPI 2.0+) */
-struct tboot_acpi_generic_address {
-	u8  space_id;
-	u8  bit_width;
-	u8  bit_offset;
-	u8  access_width;
-	u64 address;
-} __packed;
-
-/*
- * combines Sx info from FADT and FACS tables per ACPI 2.0+ spec
- * (http://www.acpi.info/)
- */
-struct tboot_acpi_sleep_info {
-	struct tboot_acpi_generic_address pm1a_cnt_blk;
-	struct tboot_acpi_generic_address pm1b_cnt_blk;
-	struct tboot_acpi_generic_address pm1a_evt_blk;
-	struct tboot_acpi_generic_address pm1b_evt_blk;
-	u16 pm1a_cnt_val;
-	u16 pm1b_cnt_val;
-	u64 wakeup_vector;
-	u32 vector_width;
-	u64 kernel_s3_resume_vector;
-} __packed;
-
-/*
- * shared memory page used for communication between tboot and kernel
- */
-struct tboot {
-	/*
-	 * version 3+ fields:
-	 */
-
-	/* TBOOT_UUID */
-	u8 uuid[16];
-
-	/* version number: 5 is current */
-	u32 version;
-
-	/* physical addr of tb_log_t log */
-	u32 log_addr;
-
-	/*
-	 * physical addr of entry point for tboot shutdown and
-	 * type of shutdown (TB_SHUTDOWN_*) being requested
-	 */
-	u32 shutdown_entry;
-	u32 shutdown_type;
-
-	/* kernel-specified ACPI info for Sx shutdown */
-	struct tboot_acpi_sleep_info acpi_sinfo;
-
-	/* tboot location in memory (physical) */
-	u32 tboot_base;
-	u32 tboot_size;
-
-	/* memory regions (phys addrs) for tboot to MAC on S3 */
-	u8 num_mac_regions;
-	struct tboot_mac_region mac_regions[MAX_TB_MAC_REGIONS];
-
-
-	/*
-	 * version 4+ fields:
-	 */
-
-	/* symmetric key for use by kernel; will be encrypted on S3 */
-	u8 s3_key[TB_KEY_SIZE];
-
-
-	/*
-	 * version 5+ fields:
-	 */
-
-	/* used to 4byte-align num_in_wfs */
-	u8 reserved_align[3];
-
-	/* number of processors in wait-for-SIPI */
-	u32 num_in_wfs;
-} __packed;
-
-/*
- * UUID for tboot data struct to facilitate matching
- * defined as {663C8DFF-E8B3-4b82-AABF-19EA4D057A08} by tboot, which is
- * represented as {} in the char array used here
- */
-#define TBOOT_UUID	{0xff, 0x8d, 0x3c, 0x66, 0xb3, 0xe8, 0x82, 0x4b, 0xbf,\
-			 0xaa, 0x19, 0xea, 0x4d, 0x5, 0x7a, 0x8}
-
-extern struct tboot *tboot;
-
-static inline int tboot_enabled(void)
-{
-	return tboot != NULL;
-}
-
-extern void tboot_probe(void);
-extern void tboot_create_trampoline(void);
-extern void tboot_shutdown(u32 shutdown_type);
-extern void tboot_sleep(u8 sleep_state, u32 pm1a_control, u32 pm1b_control);
-extern int tboot_wait_for_aps(int num_aps);
-extern struct acpi_table_header *tboot_get_dmar_table(
-				      struct acpi_table_header *dmar_tbl);
-extern int tboot_force_iommu(void);
-
-#else     /* CONFIG_INTEL_TXT */
-
-static inline int tboot_enabled(void)
-{
-	return 0;
-}
-
-static inline void tboot_probe(void)
-{
-}
-
-static inline void tboot_create_trampoline(void)
-{
-}
-
-static inline void tboot_shutdown(u32 shutdown_type)
-{
-}
-
-static inline void tboot_sleep(u8 sleep_state, u32 pm1a_control,
-			       u32 pm1b_control)
-{
-}
-
-static inline int tboot_wait_for_aps(int num_aps)
-{
-	return 0;
-}
-
-static inline struct acpi_table_header *tboot_get_dmar_table(
-					struct acpi_table_header *dmar_tbl)
-{
-	return dmar_tbl;
-}
-
-static inline int tboot_force_iommu(void)
-{
-	return 0;
-}
-
-#endif /* !CONFIG_INTEL_TXT */
-
-#endif /* _ASM_TBOOT_H */
arch/x86/kernel/reboot.c +1 -2
diff -r 04f249f00303 arch/x86/kernel/reboot.c
@@ -4,6 +4,7 @@
 #include <linux/pm.h>
 #include <linux/efi.h>
 #include <linux/dmi.h>
+#include <linux/tboot.h>
 #include <acpi/reboot.h>
 #include <asm/io.h>
 #include <asm/apic.h>

@@ -23,8 +24,6 @@
 #else
 # include <asm/iommu.h>
 #endif
-
-#include <asm/tboot.h>
 
 /*
  * Power off function, if any
arch/x86/kernel/setup.c +1 -2
diff -r 04f249f00303 arch/x86/kernel/setup.c
@@ -66,6 +66,7 @@
 
 #include <linux/percpu.h>
 #include <linux/crash_dump.h>
+#include <linux/tboot.h>
 
 #include <video/edid.h>
 

@@ -144,8 +145,6 @@ struct boot_params __initdata boot_param
 #else
 struct boot_params boot_params;
 #endif
-
-#include <asm/tboot.h>
 
 /*
  * Machine setup..
arch/x86/kernel/smpboot.c +1 -1
diff -r 04f249f00303 arch/x86/kernel/smpboot.c
@@ -47,6 +47,7 @@
 #include <linux/bootmem.h>
 #include <linux/err.h>
 #include <linux/nmi.h>
+#include <linux/tboot.h>
 
 #include <asm/acpi.h>
 #include <asm/desc.h>

@@ -62,7 +63,6 @@
 #include <asm/vmi.h>
 #include <asm/apic.h>
 #include <asm/setup.h>
-#include <asm/tboot.h>
 #include <asm/uv/uv.h>
 #include <asm/debugreg.h>
 #include <linux/mc146818rtc.h>
arch/x86/kernel/tboot.c +1 -1
diff -r 04f249f00303 arch/x86/kernel/tboot.c
@@ -27,6 +27,7 @@
 #include <linux/dmar.h>
 #include <linux/pfn.h>
 #include <linux/mm.h>
+#include <linux/tboot.h>
 
 #include <asm/trampoline.h>
 #include <asm/processor.h>

@@ -36,7 +37,6 @@
 #include <asm/fixmap.h>
 #include <asm/proto.h>
 #include <asm/setup.h>
-#include <asm/tboot.h>
 #include <asm/e820.h>
 #include <asm/io.h>
drivers/acpi/acpica/hwsleep.c +1 -1
diff -r 04f249f00303 drivers/acpi/acpica/hwsleep.c
@@ -45,7 +45,7 @@
 #include <acpi/acpi.h>
 #include "accommon.h"
 #include "actables.h"
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #define _COMPONENT          ACPI_HARDWARE
 ACPI_MODULE_NAME("hwsleep")
drivers/pci/dmar.c +1 -1
diff -r 04f249f00303 drivers/pci/dmar.c
@@ -33,7 +33,7 @@
 #include <linux/timer.h>
 #include <linux/irq.h>
 #include <linux/interrupt.h>
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #undef PREFIX
 #define PREFIX "DMAR:"
drivers/pci/intel-iommu.c +1 -1
diff -r 04f249f00303 drivers/pci/intel-iommu.c
@@ -37,8 +37,8 @@
 #include <linux/iommu.h>
 #include <linux/intel-iommu.h>
 #include <linux/sysdev.h>
+#include <linux/tboot.h>
 #include <asm/cacheflush.h>
-#include <asm/tboot.h>
 #include <asm/iommu.h>
 #include "pci.h"
include/linux/tboot.h new file +169 -0

169 lines not on this page · raw message

init/main.c +1 -1
diff -r 04f249f00303 init/main.c
@@ -68,12 +68,12 @@
 #include <linux/async.h>
 #include <linux/kmemcheck.h>
 #include <linux/kmemtrace.h>
+#include <linux/tboot.h>
 #include <trace/boot.h>
 
 #include <asm/io.h>
 #include <asm/bugs.h>
 #include <asm/setup.h>
-#include <asm/tboot.h>
 #include <asm/sections.h>
 #include <asm/cacheflush.h>
kernel/cpu.c +1 -1
diff -r 04f249f00303 kernel/cpu.c
@@ -14,7 +14,7 @@
 #include <linux/kthread.h>
 #include <linux/stop_machine.h>
 #include <linux/mutex.h>
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #ifdef CONFIG_SMP
 /* Serializes the updates to cpu_online_mask, cpu_present_mask */
#7 lore raw

Re: [PATCH] txt: fix the build errors on non-X86 platforms

Quoted 7 lines · Shane Wang

This patch moves tboot.h from asm to linux to fix the build errors on non-X86 platforms.

Signed-off-by Shane Wang <[email protected]>

diff -r 04f249f00303 arch/x86/include/asm/tboot.h

[ small nit: please always generate diffstat information for your
patches as well, so that one can see which files are changed and by how much. ]

This patch looks better, but i have to question why tboot modifies generic code at all.

i've attached those generic-code changes below. The init/main.c one could sure be done in x86 arch init code, or via an initcall, right?

Regarding kernel/cpu.c. Tthis code in tboot_wait_for_aps() looks suspicious:

int tboot_wait_for_aps(int num_aps)

{
        unsigned long timeout;

        if (!tboot_enabled())
                return 0;

        timeout = jiffies + AP_WAIT_TIMEOUT*HZ;
        while (atomic_read((atomic_t *)&tboot->num_in_wfs) != num_aps &&
               time_before(jiffies, timeout))
                cpu_relax();

        return time_before(jiffies, timeout) ? 0 : 1;
}

the return code looks a bit racy - what if an AP came back just in the final moment. It should return whether num_in_wfs == num_aps.

But more importantly, why does this have to be done in generic code in kernel/smp.c? Why doesnt the x86 arch level bit of _cpu_down() check whether the CPU goes down. (or, if there's no proper signalling for that one in the tboot protocol - the _cpu_down() code in x86 could call tboot_wait_for_aps() if num_online_cpus() == 1 - no need to change generic code here.

Also, as i pointed it out in prior review, this depends on line:

+config INTEL_TXT
+ bool "Enable Intel(R) Trusted Execution Technology (Intel(R) TXT)"
+ depends on EXPERIMENTAL && X86 && DMAR && ACPI

should be turned into the standard:

depends on HAVE_INTEL_TXT

line, where arch/x86/Kconfig selects HAVE_INTEL_TXT. As we do it for other options as well.

Ingo

---------------->

 init/main.c  |    3 +++
 kernel/cpu.c |    7 ++++++-
 2 files changed, 9 insertions(+), 1 deletion(-)
init/main.c +3 -0
index 2d9d6bd..f8e9124 100644
@@ -73,6 +73,7 @@
 #include <asm/io.h>
 #include <asm/bugs.h>
 #include <asm/setup.h>
+#include <asm/tboot.h>
 #include <asm/sections.h>
 #include <asm/cacheflush.h>
 

@@ -715,6 +716,8 @@ asmlinkage void __init start_kernel(void)
 
 	ftrace_init();
 
+	tboot_create_trampoline();
+
 	/* Do the rest non-__init'ed, we're now alive */
 	rest_init();
 }
kernel/cpu.c +6 -1
index 8ce1004..ff071e0 100644
@@ -14,6 +14,7 @@
 #include <linux/kthread.h>
 #include <linux/stop_machine.h>
 #include <linux/mutex.h>
+#include <asm/tboot.h>
 
 #ifdef CONFIG_SMP
 /* Serializes the updates to cpu_online_mask, cpu_present_mask */

@@ -376,7 +377,7 @@ static cpumask_var_t frozen_cpus;
 
 int disable_nonboot_cpus(void)
 {
-	int cpu, first_cpu, error;
+	int cpu, first_cpu, error, num_cpus = 0;
 
 	error = stop_machine_create();
 	if (error)

@@ -391,6 +392,7 @@ int disable_nonboot_cpus(void)
 	for_each_online_cpu(cpu) {
 		if (cpu == first_cpu)
 			continue;
+		num_cpus++;
 		error = _cpu_down(cpu, 1);
 		if (!error) {
 			cpumask_set_cpu(cpu, frozen_cpus);

@@ -401,6 +403,9 @@ int disable_nonboot_cpus(void)
 			break;
 		}
 	}
+	/* ensure all CPUs have gone into wait-for-SIPI */
+	error |= tboot_wait_for_aps(num_cpus);
+
 	if (!error) {
 		BUG_ON(num_online_cpus() > 1);
 		/* Make sure the CPUs won't be enabled by someone else */
#8 lore raw

Re: [PATCH] intel_txt: fix the build errors of intel_txt patch on non-X86 platforms

Hi

Forget the previous patch. I misundertood Andi's comments. It should be this one. Please comment.

Thanks.
Shane

---
 arch/x86/Kconfig              |    4 +++
 drivers/acpi/acpica/hwsleep.c |    2 -
 drivers/pci/dmar.c            |    2 -
 drivers/pci/intel-iommu.c     |    2 -
 include/linux/tboot.h         |   35 ++++++++++++++++++++++++++++++++
 init/main.c                   |    2 -
 kernel/cpu.c                  |    2 -
 security/Kconfig              |    2 -
 8 files changed, 45 insertions(+), 6 deletions(-)
Signed-off-by Shane Wang <[email protected]>
arch/x86/Kconfig +4 -0
diff -r e5406357eaf2 arch/x86/Kconfig
@@ -179,6 +179,10 @@ config ARCH_SUPPORTS_OPTIMIZED_INLINING
 
 config ARCH_SUPPORTS_DEBUG_PAGEALLOC
 	def_bool y
+
+config ARCH_HAS_INTEL_TXT
+	def_bool y
+	depends on EXPERIMENTAL && DMAR && ACPI
 
 # Use the generic interrupt handling code in kernel/irq/:
 config GENERIC_HARDIRQS
drivers/acpi/acpica/hwsleep.c +1 -1
diff -r e5406357eaf2 drivers/acpi/acpica/hwsleep.c
@@ -45,7 +45,7 @@
 #include <acpi/acpi.h>
 #include "accommon.h"
 #include "actables.h"
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #define _COMPONENT          ACPI_HARDWARE
 ACPI_MODULE_NAME("hwsleep")
drivers/pci/dmar.c +1 -1
diff -r e5406357eaf2 drivers/pci/dmar.c
@@ -33,7 +33,7 @@
 #include <linux/timer.h>
 #include <linux/irq.h>
 #include <linux/interrupt.h>
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #undef PREFIX
 #define PREFIX "DMAR:"
drivers/pci/intel-iommu.c +1 -1
diff -r e5406357eaf2 drivers/pci/intel-iommu.c
@@ -37,8 +37,8 @@
 #include <linux/iommu.h>
 #include <linux/intel-iommu.h>
 #include <linux/sysdev.h>
+#include <linux/tboot.h>
 #include <asm/cacheflush.h>
-#include <asm/tboot.h>
 #include <asm/iommu.h>
 #include "pci.h"
include/linux/tboot.h new file +35 -0
diff -r e5406357eaf2 include/linux/tboot.h
@@ -0,0 +1,35 @@
+/*
+ * Copyright (c) 2006-2009, Intel Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ */
+
+#ifndef _LINUX_TBOOT_H
+#define _LINUX_TBOOT_H
+
+#ifdef CONFIG_ARCH_HAS_INTEL_TXT
+#include <asm/tboot.h>
+#else
+
+#define tboot_sleep(sleep_state, pm1a_control, pm1b_control)	\
+					do { } while (0)
+#define tboot_get_dmar_table(dmar_tbl)	(dmar_tbl)
+#define tboot_force_iommu()		0
+#define tboot_create_trampoline()	do { } while (0)
+#define tboot_wait_for_aps(num_aps)	0
+
+#endif /* !CONFIG_ARCH_HAS_INTEL_TXT */
+
+#endif /* _LINUX_TBOOT_H */
init/main.c +1 -1
diff -r e5406357eaf2 init/main.c
@@ -68,12 +68,12 @@
 #include <linux/async.h>
 #include <linux/kmemcheck.h>
 #include <linux/kmemtrace.h>
+#include <linux/tboot.h>
 #include <trace/boot.h>
 
 #include <asm/io.h>
 #include <asm/bugs.h>
 #include <asm/setup.h>
-#include <asm/tboot.h>
 #include <asm/sections.h>
 #include <asm/cacheflush.h>
kernel/cpu.c +1 -1
diff -r e5406357eaf2 kernel/cpu.c
@@ -14,7 +14,7 @@
 #include <linux/kthread.h>
 #include <linux/stop_machine.h>
 #include <linux/mutex.h>
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #ifdef CONFIG_SMP
 /* Serializes the updates to cpu_online_mask, cpu_present_mask */
security/Kconfig +1 -1
diff -r e5406357eaf2 security/Kconfig
@@ -131,7 +131,7 @@ config LSM_MMAP_MIN_ADDR
 
 config INTEL_TXT
 	bool "Enable Intel(R) Trusted Execution Technology (Intel(R) TXT)"
-	depends on EXPERIMENTAL && X86 && DMAR && ACPI
+	depends on ARCH_HAS_INTEL_TXT
 	help
 	  This option enables support for booting the kernel with the
 	  Trusted Boot (tboot) module. This will utilize
Quoted 4 lines

This patch looks better, but i have to question why tboot modifies generic code at all.

i've attached those generic-code changes below. The init/main.c one could sure be done in x86 arch init code, or via an initcall, right?

As long as the page table is set up and the memory is initialized, since the tboot code is only to set up 1:1 mapping page table for later use. Do you mean setup_arch() in setup.c? I will try.

Quoted 18 lines

Regarding kernel/cpu.c. Tthis code in tboot_wait_for_aps() looks suspicious:

int tboot_wait_for_aps(int num_aps)

{
        unsigned long timeout;

        if (!tboot_enabled())
                return 0;

        timeout = jiffies + AP_WAIT_TIMEOUT*HZ;
        while (atomic_read((atomic_t *)&tboot->num_in_wfs) != num_aps &&
               time_before(jiffies, timeout))
                cpu_relax();

        return time_before(jiffies, timeout) ? 0 : 1;
}
Quoted 1 line

the return code looks a bit racy - what if an AP came back just in the final moment. It should return whether num_in_wfs == num_aps.

Yes;-) "return num_in_wfs == num_aps ? 1 : 0" should be better, right?

Quoted 3 lines

But more importantly, why does this have to be done in generic code in kernel/smp.c? Why doesnt the x86 arch level bit of _cpu_down() check whether the CPU goes down. (or, if there's no proper signalling for that one in the tboot protocol - the _cpu_down() code in x86 could call tboot_wait_for_aps() if num_online_cpus() == 1 - no need to change generic code here.

Which c file do you mentioned about _cpu_down()? I only can find _cpu_down() in kernel/cpu.c.

#9 lore raw

Re: [PATCH] intel_txt: fix the build errors of intel_txt patch on non-X86 platforms

Quoted 101 lines · Shane Wang

Hi

Forget the previous patch. I misundertood Andi's comments. It should be this one. Please comment.

Thanks.
Shane

---
 arch/x86/Kconfig              |    4 +++
 drivers/acpi/acpica/hwsleep.c |    2 -
 drivers/pci/dmar.c            |    2 -
 drivers/pci/intel-iommu.c     |    2 -
 include/linux/tboot.h         |   35 ++++++++++++++++++++++++++++++++
 init/main.c                   |    2 -
 kernel/cpu.c                  |    2 -
 security/Kconfig              |    2 -
 8 files changed, 45 insertions(+), 6 deletions(-)
Signed-off-by Shane Wang <[email protected]>

diff -r e5406357eaf2 arch/x86/Kconfig

arch/x86/Kconfig +4 -0
@@ -179,6 +179,10 @@ config ARCH_SUPPORTS_OPTIMIZED_INLINING
 
 config ARCH_SUPPORTS_DEBUG_PAGEALLOC
 	def_bool y
+
+config ARCH_HAS_INTEL_TXT
+	def_bool y
+	depends on EXPERIMENTAL && DMAR && ACPI
 
 # Use the generic interrupt handling code in kernel/irq/:
 config GENERIC_HARDIRQS

diff -r e5406357eaf2 drivers/acpi/acpica/hwsleep.c

drivers/acpi/acpica/hwsleep.c +1 -1
@@ -45,7 +45,7 @@
 #include <acpi/acpi.h>
 #include "accommon.h"
 #include "actables.h"
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #define _COMPONENT          ACPI_HARDWARE
 ACPI_MODULE_NAME("hwsleep")

diff -r e5406357eaf2 drivers/pci/dmar.c

drivers/pci/dmar.c +1 -1
@@ -33,7 +33,7 @@
 #include <linux/timer.h>
 #include <linux/irq.h>
 #include <linux/interrupt.h>
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #undef PREFIX
 #define PREFIX "DMAR:"

diff -r e5406357eaf2 drivers/pci/intel-iommu.c

drivers/pci/intel-iommu.c +1 -1
@@ -37,8 +37,8 @@
 #include <linux/iommu.h>
 #include <linux/intel-iommu.h>
 #include <linux/sysdev.h>
+#include <linux/tboot.h>
 #include <asm/cacheflush.h>
-#include <asm/tboot.h>
 #include <asm/iommu.h>
 #include "pci.h"

diff -r e5406357eaf2 include/linux/tboot.h

include/linux/tboot.h new file +23 -0
@@ -0,0 +1,35 @@
+/*
+ * Copyright (c) 2006-2009, Intel Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ */
+
+#ifndef _LINUX_TBOOT_H
+#define _LINUX_TBOOT_H
+
+#ifdef CONFIG_ARCH_HAS_INTEL_TXT
+#include <asm/tboot.h>

include/linux/tboot.h line 33

Quoted 10 lines
include/linux/tboot.h new file +10 -0
+#else
+
+#define tboot_sleep(sleep_state, pm1a_control, pm1b_control)	\
+					do { } while (0)
+#define tboot_get_dmar_table(dmar_tbl)	(dmar_tbl)
+#define tboot_force_iommu()		0
+#define tboot_create_trampoline()	do { } while (0)
+#define tboot_wait_for_aps(num_aps)	0
+
+#endif /* !CONFIG_ARCH_HAS_INTEL_TXT */

If then these should be dummy inlines - but in any case, do you expect tboot to be used by non-x86 too?

Quoted 34 lines
include/linux/tboot.h new file +1 -0
+#endif /* _LINUX_TBOOT_H */

diff -r e5406357eaf2 init/main.c

init/main.c +1 -1
@@ -68,12 +68,12 @@
 #include <linux/async.h>
 #include <linux/kmemcheck.h>
 #include <linux/kmemtrace.h>
+#include <linux/tboot.h>
 #include <trace/boot.h>
 
 #include <asm/io.h>
 #include <asm/bugs.h>
 #include <asm/setup.h>
-#include <asm/tboot.h>
 #include <asm/sections.h>
 #include <asm/cacheflush.h>

diff -r e5406357eaf2 kernel/cpu.c

kernel/cpu.c +1 -1
@@ -14,7 +14,7 @@
 #include <linux/kthread.h>
 #include <linux/stop_machine.h>
 #include <linux/mutex.h>
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #ifdef CONFIG_SMP
 /* Serializes the updates to cpu_online_mask, cpu_present_mask */

diff -r e5406357eaf2 security/Kconfig

security/Kconfig config LSM_MMAP_MIN_ADDR · line 134

Quoted 5 lines
security/Kconfig +1 -1
@@ -131,7 +131,7 @@ config LSM_MMAP_MIN_ADDR
 
 config INTEL_TXT
 	bool "Enable Intel(R) Trusted Execution Technology (Intel(R) TXT)"
-	depends on EXPERIMENTAL && X86 && DMAR && ACPI
+	depends on ARCH_HAS_INTEL_TXT

Nit: what i suggested was HAVE_INTEL_TXT - that is the standard kbuild solution these days. (ARCH_HAS_* is legacy-ish)

Quoted 10 lines
Quoted 6 lines

This patch looks better, but i have to question why tboot modifies generic code at all.

i've attached those generic-code changes below. The init/main.c one could sure be done in x86 arch init code, or via an initcall, right?

As long as the page table is set up and the memory is initialized, since the tboot code is only to set up 1:1 mapping page table for later use. Do you mean setup_arch() in setup.c? I will try.

yes, something like that. (i have not checked the init dependencies, so other places might be more suitable - but setup_arch() definitely sounds good. You obviously want to unlock your hardware ASAP.)

Quoted 15 lines
Quoted 15 lines

Regarding kernel/cpu.c. Tthis code in tboot_wait_for_aps() looks suspicious:

int tboot_wait_for_aps(int num_aps)

{
        unsigned long timeout;

        if (!tboot_enabled())
                return 0;

        timeout = jiffies + AP_WAIT_TIMEOUT*HZ;
        while (atomic_read((atomic_t *)&tboot->num_in_wfs) != num_aps &&
               time_before(jiffies, timeout))
                cpu_relax();
Quoted 7 lines
Quoted 5 lines
        return time_before(jiffies, timeout) ? 0 : 1;
}

the return code looks a bit racy - what if an AP came back just in the final moment. It should return whether num_in_wfs == num_aps.

Yes;-) "return num_in_wfs == num_aps ? 1 : 0" should be better, right?

better yes - cleaner, not ;-)

jiffies loops are dubious anyway - what if this code ever gets called from an irqs-off section? If you want to break out then please use mdelay based loops and print a warning message if the code has to break out after the timeout and the APs have not hand-shaken with us properly.

Quoted 9 lines
Quoted 7 lines

But more importantly, why does this have to be done in generic code in kernel/smp.c? Why doesnt the x86 arch level bit of _cpu_down() check whether the CPU goes down. (or, if there's no proper signalling for that one in the tboot protocol - the _cpu_down() code in x86 could call tboot_wait_for_aps() if num_online_cpus() == 1 - no need to change generic code here.

Which c file do you mentioned about _cpu_down()? I only can find _cpu_down() in kernel/cpu.c.

_cpu_down() is a thin wrapper and there's architecture code that does the real thing: __cpu_disable() / smp_ops.cpu_disable() on x86.

Plus there's a CPU notifier chain action as well you can register to via the kernel/cpu.c APIs, and provide a CPU_DYING handler.

(I'd argue for the latter in this case, it's more generic and can stay local to the tboot code.)

Ingo

reviewer
#10 lore raw

Re: [PATCH] intel_txt: fix the build errors of intel_txt patch on non-X86 platforms

Quoted 109 lines · Ingo Molnar
Quoted 106 lines · Shane Wang

Hi

Forget the previous patch. I misundertood Andi's comments. It should be this one. Please comment.

Thanks.
Shane

---
 arch/x86/Kconfig              |    4 +++
 drivers/acpi/acpica/hwsleep.c |    2 -
 drivers/pci/dmar.c            |    2 -
 drivers/pci/intel-iommu.c     |    2 -
 include/linux/tboot.h         |   35 ++++++++++++++++++++++++++++++++
 init/main.c                   |    2 -
 kernel/cpu.c                  |    2 -
 security/Kconfig              |    2 -
 8 files changed, 45 insertions(+), 6 deletions(-)
Signed-off-by Shane Wang <[email protected]>

diff -r e5406357eaf2 arch/x86/Kconfig

arch/x86/Kconfig +4 -0
@@ -179,6 +179,10 @@ config ARCH_SUPPORTS_OPTIMIZED_INLINING
 
 config ARCH_SUPPORTS_DEBUG_PAGEALLOC
 	def_bool y
+
+config ARCH_HAS_INTEL_TXT
+	def_bool y
+	depends on EXPERIMENTAL && DMAR && ACPI
 
 # Use the generic interrupt handling code in kernel/irq/:
 config GENERIC_HARDIRQS

diff -r e5406357eaf2 drivers/acpi/acpica/hwsleep.c

drivers/acpi/acpica/hwsleep.c +1 -1
@@ -45,7 +45,7 @@
 #include <acpi/acpi.h>
 #include "accommon.h"
 #include "actables.h"
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #define _COMPONENT          ACPI_HARDWARE
 ACPI_MODULE_NAME("hwsleep")

diff -r e5406357eaf2 drivers/pci/dmar.c

drivers/pci/dmar.c +1 -1
@@ -33,7 +33,7 @@
 #include <linux/timer.h>
 #include <linux/irq.h>
 #include <linux/interrupt.h>
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #undef PREFIX
 #define PREFIX "DMAR:"

diff -r e5406357eaf2 drivers/pci/intel-iommu.c

drivers/pci/intel-iommu.c +1 -1
@@ -37,8 +37,8 @@
 #include <linux/iommu.h>
 #include <linux/intel-iommu.h>
 #include <linux/sysdev.h>
+#include <linux/tboot.h>
 #include <asm/cacheflush.h>
-#include <asm/tboot.h>
 #include <asm/iommu.h>
 #include "pci.h"

diff -r e5406357eaf2 include/linux/tboot.h

include/linux/tboot.h new file +28 -0
@@ -0,0 +1,35 @@
+/*
+ * Copyright (c) 2006-2009, Intel Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ */
+
+#ifndef _LINUX_TBOOT_H
+#define _LINUX_TBOOT_H
+
+#ifdef CONFIG_ARCH_HAS_INTEL_TXT
+#include <asm/tboot.h>
+#else
+
+#define tboot_sleep(sleep_state, pm1a_control, pm1b_control)	\
+					do { } while (0)
+#define tboot_get_dmar_table(dmar_tbl)	(dmar_tbl)
Quoted 9 lines

include/linux/tboot.h line 33

Quoted 5 lines
include/linux/tboot.h new file +5 -0
+#define tboot_force_iommu()		0
+#define tboot_create_trampoline()	do { } while (0)
+#define tboot_wait_for_aps(num_aps)	0
+
+#endif /* !CONFIG_ARCH_HAS_INTEL_TXT */

If then these should be dummy inlines - but in any case, do you expect tboot to be used by non-x86 too?

It would have to be IA64 ... I can't ask Tony since he's out, but I would assume it is x86-specific at this point.

-hpa

#11 lore raw

RE: [PATCH] intel_txt: fix the build errors of intel_txt patch on non-X86 platforms

Quoted 9 lines
Quoted 4 lines

If then these should be dummy inlines - but in any case, do you expect tboot to be used by non-x86 too?

It would have to be IA64 ... I can't ask Tony since he's out, but I would assume it is x86-specific at this point.

-hpa

Yes, to my knowledge, it is x86 at this point. TXT-SX is the server extension also for ia32 only.

Shane

#12 lore raw

Re: [PATCH] intel_txt: fix the build errors of intel_txt patch on non-X86 platforms

Quoted 5 lines · Wang, Shane
Quoted 5 lines
Quoted 4 lines

If then these should be dummy inlines - but in any case, do you expect tboot to be used by non-x86 too?

Quoted 7 lines
Quoted 4 lines

It would have to be IA64 ... I can't ask Tony since he's out, but I would assume it is x86-specific at this point.

-hpa

Yes, to my knowledge, it is x86 at this point. TXT-SX is the server extension also for ia32 only.

Then i'd suggest to keep it simple and clean for now and move those very few lines to arch/x86 files only. It will be easy to generalize it later on, should the need arise.

Ingo

#13 lore raw

Re: [PATCH] intel_txt: fix the build errors of intel_txt patch on non-X86 platforms

This patch moves tboot.h from asm to linux to fix the build errors of intel_txt patch on non-X86 platforms, and removes the tboot code from generic code init/main.c and kernel/cpu.c.

--

 arch/x86/Kconfig              |    4
 arch/x86/include/asm/tboot.h  |  197 --------------------------------
 arch/x86/kernel/reboot.c      |    3
 arch/x86/kernel/setup.c       |    3
 arch/x86/kernel/smpboot.c     |    2
 arch/x86/kernel/tboot.c       |   58 +++++++--
 drivers/acpi/acpica/hwsleep.c |    2
 drivers/pci/dmar.c            |    2
 drivers/pci/intel-iommu.c     |    2
 include/linux/tboot.h         |  164 ++++++++++++++++++++++++++
 init/main.c                   |    3
 kernel/cpu.c                  |    6
 security/Kconfig              |    2
 13 files changed, 223 insertions(+), 225 deletions(-)
Signed-off-by Shane Wang <[email protected]>
arch/x86/Kconfig +4 -0
diff -r e5406357eaf2 arch/x86/Kconfig
@@ -179,6 +179,10 @@ config ARCH_SUPPORTS_OPTIMIZED_INLINING
 
 config ARCH_SUPPORTS_DEBUG_PAGEALLOC
 	def_bool y
+
+config HAVE_INTEL_TXT
+	def_bool y
+	depends on EXPERIMENTAL && DMAR && ACPI
 
 # Use the generic interrupt handling code in kernel/irq/:
 config GENERIC_HARDIRQS
arch/x86/include/asm/tboot.h deleted +0 -197

197 lines not on this page · raw message

arch/x86/kernel/reboot.c +1 -2
diff -r e5406357eaf2 arch/x86/kernel/reboot.c
@@ -4,6 +4,7 @@
 #include <linux/pm.h>
 #include <linux/efi.h>
 #include <linux/dmi.h>
+#include <linux/tboot.h>
 #include <acpi/reboot.h>
 #include <asm/io.h>
 #include <asm/apic.h>

@@ -23,8 +24,6 @@
 #else
 # include <asm/iommu.h>
 #endif
-
-#include <asm/tboot.h>
 
 /*
  * Power off function, if any
arch/x86/kernel/setup.c +1 -2
diff -r e5406357eaf2 arch/x86/kernel/setup.c
@@ -66,6 +66,7 @@
 
 #include <linux/percpu.h>
 #include <linux/crash_dump.h>
+#include <linux/tboot.h>
 
 #include <video/edid.h>
 

@@ -144,8 +145,6 @@ struct boot_params __initdata boot_param
 #else
 struct boot_params boot_params;
 #endif
-
-#include <asm/tboot.h>
 
 /*
  * Machine setup..
arch/x86/kernel/smpboot.c +1 -1
diff -r e5406357eaf2 arch/x86/kernel/smpboot.c
@@ -47,6 +47,7 @@
 #include <linux/bootmem.h>
 #include <linux/err.h>
 #include <linux/nmi.h>
+#include <linux/tboot.h>
 
 #include <asm/acpi.h>
 #include <asm/desc.h>

@@ -62,7 +63,6 @@
 #include <asm/vmi.h>
 #include <asm/apic.h>
 #include <asm/setup.h>
-#include <asm/tboot.h>
 #include <asm/uv/uv.h>
 #include <asm/debugreg.h>
 #include <linux/mc146818rtc.h>
arch/x86/kernel/tboot.c +47 -11

97 lines not on this page · raw message

drivers/acpi/acpica/hwsleep.c +1 -1
diff -r e5406357eaf2 drivers/acpi/acpica/hwsleep.c
@@ -45,7 +45,7 @@
 #include <acpi/acpi.h>
 #include "accommon.h"
 #include "actables.h"
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #define _COMPONENT          ACPI_HARDWARE
 ACPI_MODULE_NAME("hwsleep")
drivers/pci/dmar.c +1 -1
diff -r e5406357eaf2 drivers/pci/dmar.c
@@ -33,7 +33,7 @@
 #include <linux/timer.h>
 #include <linux/irq.h>
 #include <linux/interrupt.h>
-#include <asm/tboot.h>
+#include <linux/tboot.h>
 
 #undef PREFIX
 #define PREFIX "DMAR:"
drivers/pci/intel-iommu.c +1 -1
diff -r e5406357eaf2 drivers/pci/intel-iommu.c
@@ -37,8 +37,8 @@
 #include <linux/iommu.h>
 #include <linux/intel-iommu.h>
 #include <linux/sysdev.h>
+#include <linux/tboot.h>
 #include <asm/cacheflush.h>
-#include <asm/tboot.h>
 #include <asm/iommu.h>
 #include "pci.h"
include/linux/tboot.h new file +164 -0

164 lines not on this page · raw message

init/main.c +0 -3
diff -r e5406357eaf2 init/main.c
@@ -73,7 +73,6 @@
 #include <asm/io.h>
 #include <asm/bugs.h>
 #include <asm/setup.h>
-#include <asm/tboot.h>
 #include <asm/sections.h>
 #include <asm/cacheflush.h>
 

@@ -716,8 +715,6 @@ asmlinkage void __init start_kernel(void
 
 	ftrace_init();
 
-	tboot_create_trampoline();
-
 	/* Do the rest non-__init'ed, we're now alive */
 	rest_init();
 }
kernel/cpu.c +1 -5
diff -r e5406357eaf2 kernel/cpu.c
@@ -14,7 +14,6 @@
 #include <linux/kthread.h>
 #include <linux/stop_machine.h>
 #include <linux/mutex.h>
-#include <asm/tboot.h>
 
 #ifdef CONFIG_SMP
 /* Serializes the updates to cpu_online_mask, cpu_present_mask */

@@ -377,7 +376,7 @@ static cpumask_var_t frozen_cpus;
 
 int disable_nonboot_cpus(void)
 {
-	int cpu, first_cpu, error, num_cpus = 0;
+	int cpu, first_cpu, error;
 
 	error = stop_machine_create();
 	if (error)

@@ -392,7 +391,6 @@ int disable_nonboot_cpus(void)
 	for_each_online_cpu(cpu) {
 		if (cpu == first_cpu)
 			continue;
-		num_cpus++;
 		error = _cpu_down(cpu, 1);
 		if (!error) {
 			cpumask_set_cpu(cpu, frozen_cpus);

@@ -403,8 +401,6 @@ int disable_nonboot_cpus(void)
 			break;
 		}
 	}
-	/* ensure all CPUs have gone into wait-for-SIPI */
-	error |= tboot_wait_for_aps(num_cpus);
 
 	if (!error) {
 		BUG_ON(num_online_cpus() > 1);
security/Kconfig +1 -1
diff -r e5406357eaf2 security/Kconfig
@@ -131,7 +131,7 @@ config LSM_MMAP_MIN_ADDR
 
 config INTEL_TXT
 	bool "Enable Intel(R) Trusted Execution Technology (Intel(R) TXT)"
-	depends on EXPERIMENTAL && X86 && DMAR && ACPI
+	depends on HAVE_INTEL_TXT
 	help
 	  This option enables support for booting the kernel with the
 	  Trusted Boot (tboot) module. This will utilize
Quoted 14 lines · Ingo Molnar
Quoted 8 lines · Wang, Shane
Quoted 7 lines
Quoted 3 lines

If then these should be dummy inlines - but in any case, do you expect tboot to be used by non-x86 too?

It would have to be IA64 ... I can't ask Tony since he's out, but I would assume it is x86-specific at this point.

-hpa

Yes, to my knowledge, it is x86 at this point. TXT-SX is the server extension also for ia32 only.

Then i'd suggest to keep it simple and clean for now and move those very few lines to arch/x86 files only. It will be easy to generalize it later on, should the need arise.

Ingo