* [PATCH 0/6] ACPI v5 adjustments
@ 2013-02-21 12:49 Jan Beulich
2013-02-21 12:58 ` [PATCH 1/6] ACPI 5.0: Basic support for FADT version 5 Jan Beulich
` (6 more replies)
0 siblings, 7 replies; 9+ messages in thread
From: Jan Beulich @ 2013-02-21 12:49 UTC (permalink / raw)
To: xen-devel
The first three of these are cloned from Linux counterparts, and the
last one adjusts behavior we would have needed for v4 already.
1: ACPI 5.0: Basic support for FADT version 5
2: ACPI 5.0: Implement hardware-reduced option
3: ACPICA: Update for larger ACPI 5 FADT size
4: ACPI: support v5 (reduced HW) sleep interface
5: x86: honor ACPI indicating absence of CMOS RTC
6: honor ACPI v4 FADT flags
Signed-off-by: Jan Beulich <jbeulich@suse.com>
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH 1/6] ACPI 5.0: Basic support for FADT version 5
2013-02-21 12:49 [PATCH 0/6] ACPI v5 adjustments Jan Beulich
@ 2013-02-21 12:58 ` Jan Beulich
2013-02-21 13:03 ` Jan Beulich
2013-02-21 12:59 ` [PATCH 2/6] ACPI 5.0: Implement hardware-reduced option Jan Beulich
` (5 subsequent siblings)
6 siblings, 1 reply; 9+ messages in thread
From: Jan Beulich @ 2013-02-21 12:58 UTC (permalink / raw)
To: xen-devel
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/xen/drivers/acpi/tables/tbfadt.c
+++ b/xen/drivers/acpi/tables/tbfadt.c
@@ -222,12 +222,13 @@ void __init acpi_tb_create_local_fadt(st
/*
* Check if the FADT is larger than the largest table that we expect
- * (the ACPI 2.0/3.0 version). If so, truncate the table, and issue
+ * (the ACPI 5.0 version). If so, truncate the table, and issue
* a warning.
*/
if (length > sizeof(struct acpi_table_fadt)) {
ACPI_WARNING((AE_INFO,
- "FADT (revision %u) is longer than ACPI 2.0 version, truncating length 0x%X to 0x%zX",
+ "FADT (revision %u) is longer than ACPI 5.0 version,"
+ " truncating length %u to %zu",
table->revision, (unsigned)length,
sizeof(struct acpi_table_fadt)));
}
--- a/xen/include/acpi/actbl.h
+++ b/xen/include/acpi/actbl.h
@@ -255,6 +255,8 @@ struct acpi_table_fadt {
struct acpi_generic_address xpm_timer_block; /* 64-bit Extended Power Mgt Timer Ctrl Reg Blk address */
struct acpi_generic_address xgpe0_block; /* 64-bit Extended General Purpose Event 0 Reg Blk address */
struct acpi_generic_address xgpe1_block; /* 64-bit Extended General Purpose Event 1 Reg Blk address */
+ struct acpi_generic_address sleep_control; /* 64-bit Sleep Control register */
+ struct acpi_generic_address sleep_status; /* 64-bit Sleep Status register */
};
/* Masks for FADT Boot Architecture Flags (boot_flags) */
@@ -264,6 +266,7 @@ struct acpi_table_fadt {
#define ACPI_FADT_NO_VGA (1<<2) /* 02: [V4] It is not safe to probe for VGA hardware */
#define ACPI_FADT_NO_MSI (1<<3) /* 03: [V4] Message Signaled Interrupts (MSI) must not be enabled */
#define ACPI_FADT_NO_ASPM (1<<4) /* 04: [V4] PCIe ASPM control must not be enabled */
+#define ACPI_FADT_NO_CMOS_RTC (1<<5) /* 05: [V5] No CMOS real-time clock present */
#define FADT2_REVISION_ID 3
@@ -289,6 +292,8 @@ struct acpi_table_fadt {
#define ACPI_FADT_REMOTE_POWER_ON (1<<17) /* 17: [V4] System is compatible with remote power on (ACPI 3.0) */
#define ACPI_FADT_APIC_CLUSTER (1<<18) /* 18: [V4] All local APICs must use cluster model (ACPI 3.0) */
#define ACPI_FADT_APIC_PHYSICAL (1<<19) /* 19: [V4] All local x_aPICs must use physical dest mode (ACPI 3.0) */
+#define ACPI_FADT_HW_REDUCED (1<<20) /* 20: [V5] ACPI hardware is not implemented (ACPI 5.0) */
+#define ACPI_FADT_LOW_POWER_S0 (1<<21) /* 21: [V5] S0 power savings are equal or better than S3 (ACPI 5.0) */
/* Values for preferred_profile (Preferred Power Management Profiles) */
@@ -299,14 +304,16 @@ enum acpi_prefered_pm_profiles {
PM_WORKSTATION = 3,
PM_ENTERPRISE_SERVER = 4,
PM_SOHO_SERVER = 5,
- PM_APPLIANCE_PC = 6
+ PM_APPLIANCE_PC = 6,
+ PM_PERFORMANCE_SERVER = 7,
+ PM_TABLET = 8
};
/* Reset to default packing */
#pragma pack()
-#define ACPI_FADT_OFFSET(f) (u8) ACPI_OFFSET (struct acpi_table_fadt, f)
+#define ACPI_FADT_OFFSET(f) (u16) ACPI_OFFSET (struct acpi_table_fadt, f)
/*
* Get the remaining ACPI tables
@@ -324,12 +331,15 @@ enum acpi_prefered_pm_profiles {
* FADT is the bottom line as to what the version really is.
*
* For reference, the values below are as follows:
- * FADT V1 size: 0x74
- * FADT V2 size: 0x84
- * FADT V3+ size: 0xF4
+ * FADT V1 size: 0x074
+ * FADT V2 size: 0x084
+ * FADT V3 size: 0x0F4
+ * FADT V4 size: 0x0F4
+ * FADT V5 size: 0x10C
*/
#define ACPI_FADT_V1_SIZE (u32) (ACPI_FADT_OFFSET (flags) + 4)
#define ACPI_FADT_V2_SIZE (u32) (ACPI_FADT_OFFSET (reserved4[0]) + 3)
-#define ACPI_FADT_V3_SIZE (u32) (sizeof (struct acpi_table_fadt))
+#define ACPI_FADT_V3_SIZE (u32) (ACPI_FADT_OFFSET (sleep_control))
+#define ACPI_FADT_V5_SIZE (u32) (sizeof (struct acpi_table_fadt))
#endif /* __ACTBL_H__ */
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH 2/6] ACPI 5.0: Implement hardware-reduced option
2013-02-21 12:49 [PATCH 0/6] ACPI v5 adjustments Jan Beulich
2013-02-21 12:58 ` [PATCH 1/6] ACPI 5.0: Basic support for FADT version 5 Jan Beulich
@ 2013-02-21 12:59 ` Jan Beulich
2013-02-21 12:59 ` [PATCH 3/6] ACPICA: Update for larger ACPI 5 FADT size Jan Beulich
` (4 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Jan Beulich @ 2013-02-21 12:59 UTC (permalink / raw)
To: xen-devel
[-- Attachment #1: Type: text/plain, Size: 3266 bytes --]
If HW-reduced flag is set in the FADT, do not attempt to access
or initialize any ACPI hardware, including SCI and global lock.
No FACS will be present.
Signed-off-by: Bob Moore <robert.moore@intel.com>
Also adjust acpi_fadt_parse_sleep_info().
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/xen/arch/x86/acpi/boot.c
+++ b/xen/arch/x86/acpi/boot.c
@@ -328,6 +328,9 @@ acpi_fadt_parse_sleep_info(struct acpi_t
struct acpi_table_facs *facs = NULL;
uint64_t facs_pa;
+ if (fadt->flags & ACPI_FADT_HW_REDUCED)
+ goto bad;
+
acpi_fadt_copy_address(pm1a_cnt, pm1a_control, pm1_control);
acpi_fadt_copy_address(pm1b_cnt, pm1b_control, pm1_control);
acpi_fadt_copy_address(pm1a_evt, pm1a_event, pm1_event);
@@ -351,6 +354,8 @@ acpi_fadt_parse_sleep_info(struct acpi_t
fadt->facs, facs_pa);
facs_pa = (uint64_t)fadt->facs;
}
+ if (!facs_pa)
+ goto bad;
facs = (struct acpi_table_facs *)
__acpi_map_table(facs_pa, sizeof(struct acpi_table_facs));
--- a/xen/include/acpi/acglobal.h
+++ b/xen/include/acpi/acglobal.h
@@ -78,6 +78,13 @@ ACPI_EXTERN struct acpi_table_fadt acpi_
ACPI_EXTERN struct acpi_generic_address acpi_gbl_xpm1a_enable;
ACPI_EXTERN struct acpi_generic_address acpi_gbl_xpm1b_enable;
+/*
+ * ACPI 5.0 introduces the concept of a "reduced hardware platform", meaning
+ * that the ACPI hardware is no longer required. A flag in the FADT indicates
+ * a reduced HW machine, and that flag is duplicated here for convenience.
+ */
+ACPI_EXTERN u8 acpi_gbl_reduced_hardware;
+
/*****************************************************************************
*
* Miscellaneous globals
--- a/xen/drivers/acpi/tables/tbfadt.c
+++ b/xen/drivers/acpi/tables/tbfadt.c
@@ -197,8 +197,13 @@ void __init acpi_tb_parse_fadt(acpi_nati
acpi_tb_install_table((acpi_physical_address) acpi_gbl_FADT.Xdsdt,
flags, ACPI_SIG_DSDT, ACPI_TABLE_INDEX_DSDT);
- acpi_tb_install_table((acpi_physical_address) acpi_gbl_FADT.Xfacs,
- flags, ACPI_SIG_FACS, ACPI_TABLE_INDEX_FACS);
+ /* If Hardware Reduced flag is set, there is no FACS */
+
+ if (!acpi_gbl_reduced_hardware) {
+ acpi_tb_install_table((acpi_physical_address) acpi_gbl_FADT.
+ Xfacs, flags, ACPI_SIG_FACS,
+ ACPI_TABLE_INDEX_FACS);
+ }
}
/*******************************************************************************
@@ -242,6 +247,13 @@ void __init acpi_tb_create_local_fadt(st
ACPI_MEMCPY(&acpi_gbl_FADT, table,
ACPI_MIN(length, sizeof(struct acpi_table_fadt)));
+ /* Take a copy of the Hardware Reduced flag */
+
+ acpi_gbl_reduced_hardware = FALSE;
+ if (acpi_gbl_FADT.flags & ACPI_FADT_HW_REDUCED) {
+ acpi_gbl_reduced_hardware = TRUE;
+ }
+
/*
* 1) Convert the local copy of the FADT to the common internal format
* 2) Validate some of the important values within the FADT
@@ -401,6 +413,12 @@ static void __init acpi_tb_validate_fadt
u8 length;
acpi_native_uint i;
+ /* If Hardware Reduced flag is set, we are all done */
+
+ if (acpi_gbl_reduced_hardware) {
+ return;
+ }
+
/* Examine all of the 64-bit extended address fields (X fields) */
for (i = 0; i < ACPI_FADT_INFO_ENTRIES; i++) {
[-- Attachment #2: ACPI-v5-reduced-HW.patch --]
[-- Type: text/plain, Size: 3307 bytes --]
ACPI 5.0: Implement hardware-reduced option
If HW-reduced flag is set in the FADT, do not attempt to access
or initialize any ACPI hardware, including SCI and global lock.
No FACS will be present.
Signed-off-by: Bob Moore <robert.moore@intel.com>
Also adjust acpi_fadt_parse_sleep_info().
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/xen/arch/x86/acpi/boot.c
+++ b/xen/arch/x86/acpi/boot.c
@@ -328,6 +328,9 @@ acpi_fadt_parse_sleep_info(struct acpi_t
struct acpi_table_facs *facs = NULL;
uint64_t facs_pa;
+ if (fadt->flags & ACPI_FADT_HW_REDUCED)
+ goto bad;
+
acpi_fadt_copy_address(pm1a_cnt, pm1a_control, pm1_control);
acpi_fadt_copy_address(pm1b_cnt, pm1b_control, pm1_control);
acpi_fadt_copy_address(pm1a_evt, pm1a_event, pm1_event);
@@ -351,6 +354,8 @@ acpi_fadt_parse_sleep_info(struct acpi_t
fadt->facs, facs_pa);
facs_pa = (uint64_t)fadt->facs;
}
+ if (!facs_pa)
+ goto bad;
facs = (struct acpi_table_facs *)
__acpi_map_table(facs_pa, sizeof(struct acpi_table_facs));
--- a/xen/include/acpi/acglobal.h
+++ b/xen/include/acpi/acglobal.h
@@ -78,6 +78,13 @@ ACPI_EXTERN struct acpi_table_fadt acpi_
ACPI_EXTERN struct acpi_generic_address acpi_gbl_xpm1a_enable;
ACPI_EXTERN struct acpi_generic_address acpi_gbl_xpm1b_enable;
+/*
+ * ACPI 5.0 introduces the concept of a "reduced hardware platform", meaning
+ * that the ACPI hardware is no longer required. A flag in the FADT indicates
+ * a reduced HW machine, and that flag is duplicated here for convenience.
+ */
+ACPI_EXTERN u8 acpi_gbl_reduced_hardware;
+
/*****************************************************************************
*
* Miscellaneous globals
--- a/xen/drivers/acpi/tables/tbfadt.c
+++ b/xen/drivers/acpi/tables/tbfadt.c
@@ -197,8 +197,13 @@ void __init acpi_tb_parse_fadt(acpi_nati
acpi_tb_install_table((acpi_physical_address) acpi_gbl_FADT.Xdsdt,
flags, ACPI_SIG_DSDT, ACPI_TABLE_INDEX_DSDT);
- acpi_tb_install_table((acpi_physical_address) acpi_gbl_FADT.Xfacs,
- flags, ACPI_SIG_FACS, ACPI_TABLE_INDEX_FACS);
+ /* If Hardware Reduced flag is set, there is no FACS */
+
+ if (!acpi_gbl_reduced_hardware) {
+ acpi_tb_install_table((acpi_physical_address) acpi_gbl_FADT.
+ Xfacs, flags, ACPI_SIG_FACS,
+ ACPI_TABLE_INDEX_FACS);
+ }
}
/*******************************************************************************
@@ -242,6 +247,13 @@ void __init acpi_tb_create_local_fadt(st
ACPI_MEMCPY(&acpi_gbl_FADT, table,
ACPI_MIN(length, sizeof(struct acpi_table_fadt)));
+ /* Take a copy of the Hardware Reduced flag */
+
+ acpi_gbl_reduced_hardware = FALSE;
+ if (acpi_gbl_FADT.flags & ACPI_FADT_HW_REDUCED) {
+ acpi_gbl_reduced_hardware = TRUE;
+ }
+
/*
* 1) Convert the local copy of the FADT to the common internal format
* 2) Validate some of the important values within the FADT
@@ -401,6 +413,12 @@ static void __init acpi_tb_validate_fadt
u8 length;
acpi_native_uint i;
+ /* If Hardware Reduced flag is set, we are all done */
+
+ if (acpi_gbl_reduced_hardware) {
+ return;
+ }
+
/* Examine all of the 64-bit extended address fields (X fields) */
for (i = 0; i < ACPI_FADT_INFO_ENTRIES; i++) {
[-- Attachment #3: Type: text/plain, Size: 126 bytes --]
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH 3/6] ACPICA: Update for larger ACPI 5 FADT size
2013-02-21 12:49 [PATCH 0/6] ACPI v5 adjustments Jan Beulich
2013-02-21 12:58 ` [PATCH 1/6] ACPI 5.0: Basic support for FADT version 5 Jan Beulich
2013-02-21 12:59 ` [PATCH 2/6] ACPI 5.0: Implement hardware-reduced option Jan Beulich
@ 2013-02-21 12:59 ` Jan Beulich
2013-02-21 13:00 ` [PATCH 4/6] ACPI: support v5 (reduced HW) sleep interface Jan Beulich
` (3 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Jan Beulich @ 2013-02-21 12:59 UTC (permalink / raw)
To: xen-devel
[-- Attachment #1: Type: text/plain, Size: 1405 bytes --]
FADT is now larger than 256 bytes, so all FADT offsets must be
changed from 8 bits to 16 bits.
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/xen/drivers/acpi/tables/tbfadt.c
+++ b/xen/drivers/acpi/tables/tbfadt.c
@@ -62,13 +62,14 @@ static void acpi_tb_validate_fadt(void);
typedef struct acpi_fadt_info {
char *name;
- u8 target;
- u8 source;
- u8 length;
+ u16 target;
+ u16 source;
+ u16 length;
u8 type;
} acpi_fadt_info;
+#define ACPI_FADT_OPTIONAL 0
#define ACPI_FADT_REQUIRED 1
#define ACPI_FADT_SEPARATE_LENGTH 2
@@ -79,7 +80,7 @@ static struct acpi_fadt_info __initdata
{"Pm1bEventBlock", ACPI_FADT_OFFSET(xpm1b_event_block),
ACPI_FADT_OFFSET(pm1b_event_block),
- ACPI_FADT_OFFSET(pm1_event_length), 0},
+ ACPI_FADT_OFFSET(pm1_event_length), ACPI_FADT_OPTIONAL},
{"Pm1aControlBlock", ACPI_FADT_OFFSET(xpm1a_control_block),
ACPI_FADT_OFFSET(pm1a_control_block),
@@ -87,7 +88,7 @@ static struct acpi_fadt_info __initdata
{"Pm1bControlBlock", ACPI_FADT_OFFSET(xpm1b_control_block),
ACPI_FADT_OFFSET(pm1b_control_block),
- ACPI_FADT_OFFSET(pm1_control_length), 0},
+ ACPI_FADT_OFFSET(pm1_control_length), ACPI_FADT_OPTIONAL},
{"Pm2ControlBlock", ACPI_FADT_OFFSET(xpm2_control_block),
ACPI_FADT_OFFSET(pm2_control_block),
[-- Attachment #2: ACPI-v5-FADT-offsets.patch --]
[-- Type: text/plain, Size: 1445 bytes --]
ACPICA: Update for larger ACPI 5 FADT size
FADT is now larger than 256 bytes, so all FADT offsets must be
changed from 8 bits to 16 bits.
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/xen/drivers/acpi/tables/tbfadt.c
+++ b/xen/drivers/acpi/tables/tbfadt.c
@@ -62,13 +62,14 @@ static void acpi_tb_validate_fadt(void);
typedef struct acpi_fadt_info {
char *name;
- u8 target;
- u8 source;
- u8 length;
+ u16 target;
+ u16 source;
+ u16 length;
u8 type;
} acpi_fadt_info;
+#define ACPI_FADT_OPTIONAL 0
#define ACPI_FADT_REQUIRED 1
#define ACPI_FADT_SEPARATE_LENGTH 2
@@ -79,7 +80,7 @@ static struct acpi_fadt_info __initdata
{"Pm1bEventBlock", ACPI_FADT_OFFSET(xpm1b_event_block),
ACPI_FADT_OFFSET(pm1b_event_block),
- ACPI_FADT_OFFSET(pm1_event_length), 0},
+ ACPI_FADT_OFFSET(pm1_event_length), ACPI_FADT_OPTIONAL},
{"Pm1aControlBlock", ACPI_FADT_OFFSET(xpm1a_control_block),
ACPI_FADT_OFFSET(pm1a_control_block),
@@ -87,7 +88,7 @@ static struct acpi_fadt_info __initdata
{"Pm1bControlBlock", ACPI_FADT_OFFSET(xpm1b_control_block),
ACPI_FADT_OFFSET(pm1b_control_block),
- ACPI_FADT_OFFSET(pm1_control_length), 0},
+ ACPI_FADT_OFFSET(pm1_control_length), ACPI_FADT_OPTIONAL},
{"Pm2ControlBlock", ACPI_FADT_OFFSET(xpm2_control_block),
ACPI_FADT_OFFSET(pm2_control_block),
[-- Attachment #3: Type: text/plain, Size: 126 bytes --]
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH 4/6] ACPI: support v5 (reduced HW) sleep interface
2013-02-21 12:49 [PATCH 0/6] ACPI v5 adjustments Jan Beulich
` (2 preceding siblings ...)
2013-02-21 12:59 ` [PATCH 3/6] ACPICA: Update for larger ACPI 5 FADT size Jan Beulich
@ 2013-02-21 13:00 ` Jan Beulich
2013-02-21 13:01 ` [PATCH 5/6] x86: honor ACPI indicating absence of CMOS RTC Jan Beulich
` (2 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Jan Beulich @ 2013-02-21 13:00 UTC (permalink / raw)
To: xen-devel
[-- Attachment #1: Type: text/plain, Size: 11747 bytes --]
Note that this also fixes a broken input check in acpi_enter_sleep()
(previously validating the sleep->pm1[ab]_cnt_val relationship based
on acpi_sinfo.pm1b_cnt_val, which however gets set only subsequently).
Also adjust a few minor issues with the pre-v5 handling in
acpi_fadt_parse_sleep_info().
Signed-off-by: Jan Beulich <jbeulich@suse.com>
---
A question is whether, rather than passing the raw values read from
_Sx to XENPF_enter_acpi_sleep, we should again have Dom0 pass the
values ready to be written into the registers - that would make things
consistent with the legacy case, but add further dependency of the
hypervisor on Dom0 behavior where it's not really needed. I think that
it was a mistake for the original variant to have Dom0 pass massaged
values to the hypervisor.
--- a/xen/arch/x86/acpi/boot.c
+++ b/xen/arch/x86/acpi/boot.c
@@ -310,14 +310,15 @@ static int __init acpi_invalidate_bgrt(s
#ifdef CONFIG_ACPI_SLEEP
#define acpi_fadt_copy_address(dst, src, len) do { \
- if (fadt->header.revision >= FADT2_REVISION_ID) \
+ if (fadt->header.revision >= FADT2_REVISION_ID && \
+ fadt->header.length >= ACPI_FADT_V2_SIZE) \
acpi_sinfo.dst##_blk = fadt->x##src##_block; \
if (!acpi_sinfo.dst##_blk.address) { \
acpi_sinfo.dst##_blk.address = fadt->src##_block; \
acpi_sinfo.dst##_blk.space_id = ACPI_ADR_SPACE_SYSTEM_IO; \
acpi_sinfo.dst##_blk.bit_width = fadt->len##_length << 3; \
acpi_sinfo.dst##_blk.bit_offset = 0; \
- acpi_sinfo.dst##_blk.access_width = 0; \
+ acpi_sinfo.dst##_blk.access_width = fadt->len##_length; \
} \
} while (0)
@@ -328,6 +329,41 @@ acpi_fadt_parse_sleep_info(struct acpi_t
struct acpi_table_facs *facs = NULL;
uint64_t facs_pa;
+ if (fadt->header.revision >= 5 &&
+ fadt->header.length >= ACPI_FADT_V5_SIZE) {
+ acpi_sinfo.sleep_control = fadt->sleep_control;
+ acpi_sinfo.sleep_status = fadt->sleep_status;
+
+ printk(KERN_INFO PREFIX
+ "v5 SLEEP INFO: control[%d:%"PRIx64"],"
+ " status[%d:%"PRIx64"]\n",
+ acpi_sinfo.sleep_control.space_id,
+ acpi_sinfo.sleep_control.address,
+ acpi_sinfo.sleep_status.space_id,
+ acpi_sinfo.sleep_status.address);
+
+ if ((fadt->sleep_control.address &&
+ (fadt->sleep_control.bit_offset ||
+ fadt->sleep_control.bit_width !=
+ fadt->sleep_control.access_width * 8)) ||
+ (fadt->sleep_status.address &&
+ (fadt->sleep_status.bit_offset ||
+ fadt->sleep_status.bit_width !=
+ fadt->sleep_status.access_width * 8))) {
+ printk(KERN_WARNING PREFIX
+ "Invalid sleep control/status register data:"
+ " %#x:%#x:%#x %#x:%#x:%#x\n",
+ fadt->sleep_control.bit_offset,
+ fadt->sleep_control.bit_width,
+ fadt->sleep_control.access_width,
+ fadt->sleep_status.bit_offset,
+ fadt->sleep_status.bit_width,
+ fadt->sleep_status.access_width);
+ fadt->sleep_control.address = 0;
+ fadt->sleep_status.address = 0;
+ }
+ }
+
if (fadt->flags & ACPI_FADT_HW_REDUCED)
goto bad;
@@ -337,7 +373,7 @@ acpi_fadt_parse_sleep_info(struct acpi_t
acpi_fadt_copy_address(pm1b_evt, pm1b_event, pm1_event);
printk(KERN_INFO PREFIX
- "ACPI SLEEP INFO: pm1x_cnt[%"PRIx64",%"PRIx64"], "
+ "SLEEP INFO: pm1x_cnt[%"PRIx64",%"PRIx64"], "
"pm1x_evt[%"PRIx64",%"PRIx64"]\n",
acpi_sinfo.pm1a_cnt_blk.address,
acpi_sinfo.pm1b_cnt_blk.address,
@@ -384,11 +420,14 @@ acpi_fadt_parse_sleep_info(struct acpi_t
acpi_sinfo.vector_width = 32;
printk(KERN_INFO PREFIX
- " wakeup_vec[%"PRIx64"], vec_size[%x]\n",
+ " wakeup_vec[%"PRIx64"], vec_size[%x]\n",
acpi_sinfo.wakeup_vector, acpi_sinfo.vector_width);
return;
bad:
- memset(&acpi_sinfo, 0, sizeof(acpi_sinfo));
+ memset(&acpi_sinfo, 0,
+ offsetof(struct acpi_sleep_info, sleep_control));
+ memset(&acpi_sinfo.sleep_status + 1, 0,
+ (long)(&acpi_sinfo + 1) - (long)(&acpi_sinfo.sleep_status + 1));
}
#endif
--- a/xen/arch/x86/acpi/power.c
+++ b/xen/arch/x86/acpi/power.c
@@ -239,23 +239,47 @@ static long enter_state_helper(void *dat
*/
int acpi_enter_sleep(struct xenpf_enter_acpi_sleep *sleep)
{
- if ( !acpi_sinfo.pm1a_cnt_blk.address )
+ if ( sleep->flags & XENPF_ACPI_SLEEP_EXTENDED )
+ {
+ if ( !acpi_sinfo.sleep_control.address ||
+ !acpi_sinfo.sleep_status.address )
+ return -EPERM;
+
+ if ( sleep->flags & ~XENPF_ACPI_SLEEP_EXTENDED )
+ return -EINVAL;
+
+ if ( sleep->val_a > ACPI_SLEEP_TYPE_MAX ||
+ (sleep->val_b != ACPI_SLEEP_TYPE_INVALID &&
+ sleep->val_b > ACPI_SLEEP_TYPE_MAX) )
+ return -ERANGE;
+
+ acpi_sinfo.sleep_type_a = sleep->val_a;
+ acpi_sinfo.sleep_type_b = sleep->val_b;
+
+ acpi_sinfo.sleep_extended = 1;
+ }
+
+ else if ( !acpi_sinfo.pm1a_cnt_blk.address )
return -EPERM;
/* Sanity check */
- if ( acpi_sinfo.pm1b_cnt_val &&
- ((sleep->pm1a_cnt_val ^ sleep->pm1b_cnt_val) &
- ACPI_BITMASK_SLEEP_ENABLE) )
+ else if ( sleep->val_b &&
+ ((sleep->val_a ^ sleep->val_b) & ACPI_BITMASK_SLEEP_ENABLE) )
{
gdprintk(XENLOG_ERR, "Mismatched pm1a/pm1b setting.");
return -EINVAL;
}
- if ( sleep->flags )
+ else if ( sleep->flags )
return -EINVAL;
- acpi_sinfo.pm1a_cnt_val = sleep->pm1a_cnt_val;
- acpi_sinfo.pm1b_cnt_val = sleep->pm1b_cnt_val;
+ else
+ {
+ acpi_sinfo.pm1a_cnt_val = sleep->val_a;
+ acpi_sinfo.pm1b_cnt_val = sleep->val_b;
+ acpi_sinfo.sleep_extended = 0;
+ }
+
acpi_sinfo.sleep_state = sleep->sleep_state;
return continue_hypercall_on_cpu(0, enter_state_helper, &acpi_sinfo);
@@ -266,6 +290,13 @@ static int acpi_get_wake_status(void)
uint32_t val;
acpi_status status;
+ if ( acpi_sinfo.sleep_extended )
+ {
+ status = acpi_hw_register_read(ACPI_REGISTER_SLEEP_STATUS, &val);
+
+ return ACPI_FAILURE(status) ? 0 : val & ACPI_X_WAKE_STATUS;
+ }
+
/* Wake status is the 15th bit of PM1 status register. (ACPI spec 3.0) */
status = acpi_hw_register_read(ACPI_REGISTER_PM1_STATUS, &val);
if ( ACPI_FAILURE(status) )
@@ -335,19 +366,33 @@ acpi_status acpi_enter_sleep_state(u8 sl
ACPI_FLUSH_CPU_CACHE();
- status = acpi_hw_register_write(ACPI_REGISTER_PM1A_CONTROL,
- acpi_sinfo.pm1a_cnt_val);
- if ( ACPI_FAILURE(status) )
- return_ACPI_STATUS(AE_ERROR);
-
- if ( acpi_sinfo.pm1b_cnt_blk.address )
+ if ( acpi_sinfo.sleep_extended )
{
- status = acpi_hw_register_write(ACPI_REGISTER_PM1B_CONTROL,
- acpi_sinfo.pm1b_cnt_val);
- if ( ACPI_FAILURE(status) )
- return_ACPI_STATUS(AE_ERROR);
+ /*
+ * Set the SLP_TYP and SLP_EN bits.
+ *
+ * Note: We only use the first value returned by the \_Sx method
+ * (acpi_sinfo.sleep_type_a) - As per ACPI specification.
+ */
+ u8 sleep_type_value =
+ ((acpi_sinfo.sleep_type_a << ACPI_X_SLEEP_TYPE_POSITION) &
+ ACPI_X_SLEEP_TYPE_MASK) | ACPI_X_SLEEP_ENABLE;
+
+ status = acpi_hw_register_write(ACPI_REGISTER_SLEEP_CONTROL,
+ sleep_type_value);
+ }
+ else
+ {
+ status = acpi_hw_register_write(ACPI_REGISTER_PM1A_CONTROL,
+ acpi_sinfo.pm1a_cnt_val);
+ if ( !ACPI_FAILURE(status) && acpi_sinfo.pm1b_cnt_blk.address )
+ status = acpi_hw_register_write(ACPI_REGISTER_PM1B_CONTROL,
+ acpi_sinfo.pm1b_cnt_val);
}
+ if ( ACPI_FAILURE(status) )
+ return_ACPI_STATUS(AE_ERROR);
+
/* Wait until we enter sleep state, and spin until we wake */
while ( !acpi_get_wake_status() )
continue;
--- a/xen/drivers/acpi/hwregs.c
+++ b/xen/drivers/acpi/hwregs.c
@@ -365,6 +365,14 @@ acpi_hw_register_read(u32 register_id, u
acpi_os_read_port(acpi_gbl_FADT.smi_command, &value1, 8);
break;
+ case ACPI_REGISTER_SLEEP_STATUS:
+
+ status =
+ acpi_hw_low_level_read(acpi_gbl_FADT.sleep_status.bit_width,
+ &value1,
+ &acpi_gbl_FADT.sleep_status);
+ break;
+
default:
ACPI_DEBUG_PRINT((AE_INFO, "Unknown Register ID: %X", register_id));
status = AE_BAD_PARAMETER;
@@ -525,6 +533,14 @@ acpi_status acpi_hw_register_write(u32 r
acpi_os_write_port(acpi_gbl_FADT.smi_command, value, 8);
break;
+ case ACPI_REGISTER_SLEEP_CONTROL:
+
+ status =
+ acpi_hw_low_level_write(acpi_gbl_FADT.sleep_control.bit_width,
+ value,
+ &acpi_gbl_FADT.sleep_control);
+ break;
+
default:
status = AE_BAD_PARAMETER;
break;
--- a/xen/include/acpi/aclocal.h
+++ b/xen/include/acpi/aclocal.h
@@ -128,6 +128,8 @@ struct acpi_bit_register_info {
#define ACPI_REGISTER_PM_TIMER 0x07
#define ACPI_REGISTER_PROCESSOR_BLOCK 0x08
#define ACPI_REGISTER_SMI_COMMAND_BLOCK 0x09
+#define ACPI_REGISTER_SLEEP_CONTROL 0x0a
+#define ACPI_REGISTER_SLEEP_STATUS 0x0b
/* Masks used to access the bit_registers */
--- a/xen/include/acpi/actbl.h
+++ b/xen/include/acpi/actbl.h
@@ -309,6 +309,13 @@ enum acpi_prefered_pm_profiles {
PM_TABLET = 8
};
+/* Values for sleep_status and sleep_control registers (V5 FADT) */
+
+#define ACPI_X_WAKE_STATUS 0x80
+#define ACPI_X_SLEEP_TYPE_MASK 0x1C
+#define ACPI_X_SLEEP_TYPE_POSITION 0x02
+#define ACPI_X_SLEEP_ENABLE 0x20
+
/* Reset to default packing */
#pragma pack()
--- a/xen/include/asm-x86/acpi.h
+++ b/xen/include/asm-x86/acpi.h
@@ -126,11 +126,20 @@ struct acpi_sleep_info {
struct acpi_generic_address pm1b_cnt_blk;
struct acpi_generic_address pm1a_evt_blk;
struct acpi_generic_address pm1b_evt_blk;
- uint16_t pm1a_cnt_val;
- uint16_t pm1b_cnt_val;
+ struct acpi_generic_address sleep_control;
+ struct acpi_generic_address sleep_status;
+ union {
+ uint16_t pm1a_cnt_val;
+ uint8_t sleep_type_a;
+ };
+ union {
+ uint16_t pm1b_cnt_val;
+ uint8_t sleep_type_b;
+ };
uint32_t sleep_state;
uint64_t wakeup_vector;
uint32_t vector_width;
+ bool_t sleep_extended;
};
#endif /* CONFIG_ACPI_SLEEP */
--- a/xen/include/public/platform.h
+++ b/xen/include/public/platform.h
@@ -290,10 +290,16 @@ DEFINE_XEN_GUEST_HANDLE(xenpf_firmware_i
#define XENPF_enter_acpi_sleep 51
struct xenpf_enter_acpi_sleep {
/* IN variables */
+#if __XEN_INTERFACE_VERSION__ < 0x00040300
uint16_t pm1a_cnt_val; /* PM1a control value. */
uint16_t pm1b_cnt_val; /* PM1b control value. */
+#else
+ uint16_t val_a; /* PM1a control / sleep type A. */
+ uint16_t val_b; /* PM1b control / sleep type B. */
+#endif
uint32_t sleep_state; /* Which state to enter (Sn). */
- uint32_t flags; /* Must be zero. */
+#define XENPF_ACPI_SLEEP_EXTENDED 0x00000001
+ uint32_t flags; /* XENPF_ACPI_SLEEP_*. */
};
typedef struct xenpf_enter_acpi_sleep xenpf_enter_acpi_sleep_t;
DEFINE_XEN_GUEST_HANDLE(xenpf_enter_acpi_sleep_t);
[-- Attachment #2: ACPI-v5-sleep.patch --]
[-- Type: text/plain, Size: 11792 bytes --]
ACPI: support v5 (reduced HW) sleep interface
Note that this also fixes a broken input check in acpi_enter_sleep()
(previously validating the sleep->pm1[ab]_cnt_val relationship based
on acpi_sinfo.pm1b_cnt_val, which however gets set only subsequently).
Also adjust a few minor issues with the pre-v5 handling in
acpi_fadt_parse_sleep_info().
Signed-off-by: Jan Beulich <jbeulich@suse.com>
---
A question is whether, rather than passing the raw values read from
_Sx to XENPF_enter_acpi_sleep, we should again have Dom0 pass the
values ready to be written into the registers - that would make things
consistent with the legacy case, but add further dependency of the
hypervisor on Dom0 behavior where it's not really needed. I think that
it was a mistake for the original variant to have Dom0 pass massaged
values to the hypervisor.
--- a/xen/arch/x86/acpi/boot.c
+++ b/xen/arch/x86/acpi/boot.c
@@ -310,14 +310,15 @@ static int __init acpi_invalidate_bgrt(s
#ifdef CONFIG_ACPI_SLEEP
#define acpi_fadt_copy_address(dst, src, len) do { \
- if (fadt->header.revision >= FADT2_REVISION_ID) \
+ if (fadt->header.revision >= FADT2_REVISION_ID && \
+ fadt->header.length >= ACPI_FADT_V2_SIZE) \
acpi_sinfo.dst##_blk = fadt->x##src##_block; \
if (!acpi_sinfo.dst##_blk.address) { \
acpi_sinfo.dst##_blk.address = fadt->src##_block; \
acpi_sinfo.dst##_blk.space_id = ACPI_ADR_SPACE_SYSTEM_IO; \
acpi_sinfo.dst##_blk.bit_width = fadt->len##_length << 3; \
acpi_sinfo.dst##_blk.bit_offset = 0; \
- acpi_sinfo.dst##_blk.access_width = 0; \
+ acpi_sinfo.dst##_blk.access_width = fadt->len##_length; \
} \
} while (0)
@@ -328,6 +329,41 @@ acpi_fadt_parse_sleep_info(struct acpi_t
struct acpi_table_facs *facs = NULL;
uint64_t facs_pa;
+ if (fadt->header.revision >= 5 &&
+ fadt->header.length >= ACPI_FADT_V5_SIZE) {
+ acpi_sinfo.sleep_control = fadt->sleep_control;
+ acpi_sinfo.sleep_status = fadt->sleep_status;
+
+ printk(KERN_INFO PREFIX
+ "v5 SLEEP INFO: control[%d:%"PRIx64"],"
+ " status[%d:%"PRIx64"]\n",
+ acpi_sinfo.sleep_control.space_id,
+ acpi_sinfo.sleep_control.address,
+ acpi_sinfo.sleep_status.space_id,
+ acpi_sinfo.sleep_status.address);
+
+ if ((fadt->sleep_control.address &&
+ (fadt->sleep_control.bit_offset ||
+ fadt->sleep_control.bit_width !=
+ fadt->sleep_control.access_width * 8)) ||
+ (fadt->sleep_status.address &&
+ (fadt->sleep_status.bit_offset ||
+ fadt->sleep_status.bit_width !=
+ fadt->sleep_status.access_width * 8))) {
+ printk(KERN_WARNING PREFIX
+ "Invalid sleep control/status register data:"
+ " %#x:%#x:%#x %#x:%#x:%#x\n",
+ fadt->sleep_control.bit_offset,
+ fadt->sleep_control.bit_width,
+ fadt->sleep_control.access_width,
+ fadt->sleep_status.bit_offset,
+ fadt->sleep_status.bit_width,
+ fadt->sleep_status.access_width);
+ fadt->sleep_control.address = 0;
+ fadt->sleep_status.address = 0;
+ }
+ }
+
if (fadt->flags & ACPI_FADT_HW_REDUCED)
goto bad;
@@ -337,7 +373,7 @@ acpi_fadt_parse_sleep_info(struct acpi_t
acpi_fadt_copy_address(pm1b_evt, pm1b_event, pm1_event);
printk(KERN_INFO PREFIX
- "ACPI SLEEP INFO: pm1x_cnt[%"PRIx64",%"PRIx64"], "
+ "SLEEP INFO: pm1x_cnt[%"PRIx64",%"PRIx64"], "
"pm1x_evt[%"PRIx64",%"PRIx64"]\n",
acpi_sinfo.pm1a_cnt_blk.address,
acpi_sinfo.pm1b_cnt_blk.address,
@@ -384,11 +420,14 @@ acpi_fadt_parse_sleep_info(struct acpi_t
acpi_sinfo.vector_width = 32;
printk(KERN_INFO PREFIX
- " wakeup_vec[%"PRIx64"], vec_size[%x]\n",
+ " wakeup_vec[%"PRIx64"], vec_size[%x]\n",
acpi_sinfo.wakeup_vector, acpi_sinfo.vector_width);
return;
bad:
- memset(&acpi_sinfo, 0, sizeof(acpi_sinfo));
+ memset(&acpi_sinfo, 0,
+ offsetof(struct acpi_sleep_info, sleep_control));
+ memset(&acpi_sinfo.sleep_status + 1, 0,
+ (long)(&acpi_sinfo + 1) - (long)(&acpi_sinfo.sleep_status + 1));
}
#endif
--- a/xen/arch/x86/acpi/power.c
+++ b/xen/arch/x86/acpi/power.c
@@ -239,23 +239,47 @@ static long enter_state_helper(void *dat
*/
int acpi_enter_sleep(struct xenpf_enter_acpi_sleep *sleep)
{
- if ( !acpi_sinfo.pm1a_cnt_blk.address )
+ if ( sleep->flags & XENPF_ACPI_SLEEP_EXTENDED )
+ {
+ if ( !acpi_sinfo.sleep_control.address ||
+ !acpi_sinfo.sleep_status.address )
+ return -EPERM;
+
+ if ( sleep->flags & ~XENPF_ACPI_SLEEP_EXTENDED )
+ return -EINVAL;
+
+ if ( sleep->val_a > ACPI_SLEEP_TYPE_MAX ||
+ (sleep->val_b != ACPI_SLEEP_TYPE_INVALID &&
+ sleep->val_b > ACPI_SLEEP_TYPE_MAX) )
+ return -ERANGE;
+
+ acpi_sinfo.sleep_type_a = sleep->val_a;
+ acpi_sinfo.sleep_type_b = sleep->val_b;
+
+ acpi_sinfo.sleep_extended = 1;
+ }
+
+ else if ( !acpi_sinfo.pm1a_cnt_blk.address )
return -EPERM;
/* Sanity check */
- if ( acpi_sinfo.pm1b_cnt_val &&
- ((sleep->pm1a_cnt_val ^ sleep->pm1b_cnt_val) &
- ACPI_BITMASK_SLEEP_ENABLE) )
+ else if ( sleep->val_b &&
+ ((sleep->val_a ^ sleep->val_b) & ACPI_BITMASK_SLEEP_ENABLE) )
{
gdprintk(XENLOG_ERR, "Mismatched pm1a/pm1b setting.");
return -EINVAL;
}
- if ( sleep->flags )
+ else if ( sleep->flags )
return -EINVAL;
- acpi_sinfo.pm1a_cnt_val = sleep->pm1a_cnt_val;
- acpi_sinfo.pm1b_cnt_val = sleep->pm1b_cnt_val;
+ else
+ {
+ acpi_sinfo.pm1a_cnt_val = sleep->val_a;
+ acpi_sinfo.pm1b_cnt_val = sleep->val_b;
+ acpi_sinfo.sleep_extended = 0;
+ }
+
acpi_sinfo.sleep_state = sleep->sleep_state;
return continue_hypercall_on_cpu(0, enter_state_helper, &acpi_sinfo);
@@ -266,6 +290,13 @@ static int acpi_get_wake_status(void)
uint32_t val;
acpi_status status;
+ if ( acpi_sinfo.sleep_extended )
+ {
+ status = acpi_hw_register_read(ACPI_REGISTER_SLEEP_STATUS, &val);
+
+ return ACPI_FAILURE(status) ? 0 : val & ACPI_X_WAKE_STATUS;
+ }
+
/* Wake status is the 15th bit of PM1 status register. (ACPI spec 3.0) */
status = acpi_hw_register_read(ACPI_REGISTER_PM1_STATUS, &val);
if ( ACPI_FAILURE(status) )
@@ -335,19 +366,33 @@ acpi_status acpi_enter_sleep_state(u8 sl
ACPI_FLUSH_CPU_CACHE();
- status = acpi_hw_register_write(ACPI_REGISTER_PM1A_CONTROL,
- acpi_sinfo.pm1a_cnt_val);
- if ( ACPI_FAILURE(status) )
- return_ACPI_STATUS(AE_ERROR);
-
- if ( acpi_sinfo.pm1b_cnt_blk.address )
+ if ( acpi_sinfo.sleep_extended )
{
- status = acpi_hw_register_write(ACPI_REGISTER_PM1B_CONTROL,
- acpi_sinfo.pm1b_cnt_val);
- if ( ACPI_FAILURE(status) )
- return_ACPI_STATUS(AE_ERROR);
+ /*
+ * Set the SLP_TYP and SLP_EN bits.
+ *
+ * Note: We only use the first value returned by the \_Sx method
+ * (acpi_sinfo.sleep_type_a) - As per ACPI specification.
+ */
+ u8 sleep_type_value =
+ ((acpi_sinfo.sleep_type_a << ACPI_X_SLEEP_TYPE_POSITION) &
+ ACPI_X_SLEEP_TYPE_MASK) | ACPI_X_SLEEP_ENABLE;
+
+ status = acpi_hw_register_write(ACPI_REGISTER_SLEEP_CONTROL,
+ sleep_type_value);
+ }
+ else
+ {
+ status = acpi_hw_register_write(ACPI_REGISTER_PM1A_CONTROL,
+ acpi_sinfo.pm1a_cnt_val);
+ if ( !ACPI_FAILURE(status) && acpi_sinfo.pm1b_cnt_blk.address )
+ status = acpi_hw_register_write(ACPI_REGISTER_PM1B_CONTROL,
+ acpi_sinfo.pm1b_cnt_val);
}
+ if ( ACPI_FAILURE(status) )
+ return_ACPI_STATUS(AE_ERROR);
+
/* Wait until we enter sleep state, and spin until we wake */
while ( !acpi_get_wake_status() )
continue;
--- a/xen/drivers/acpi/hwregs.c
+++ b/xen/drivers/acpi/hwregs.c
@@ -365,6 +365,14 @@ acpi_hw_register_read(u32 register_id, u
acpi_os_read_port(acpi_gbl_FADT.smi_command, &value1, 8);
break;
+ case ACPI_REGISTER_SLEEP_STATUS:
+
+ status =
+ acpi_hw_low_level_read(acpi_gbl_FADT.sleep_status.bit_width,
+ &value1,
+ &acpi_gbl_FADT.sleep_status);
+ break;
+
default:
ACPI_DEBUG_PRINT((AE_INFO, "Unknown Register ID: %X", register_id));
status = AE_BAD_PARAMETER;
@@ -525,6 +533,14 @@ acpi_status acpi_hw_register_write(u32 r
acpi_os_write_port(acpi_gbl_FADT.smi_command, value, 8);
break;
+ case ACPI_REGISTER_SLEEP_CONTROL:
+
+ status =
+ acpi_hw_low_level_write(acpi_gbl_FADT.sleep_control.bit_width,
+ value,
+ &acpi_gbl_FADT.sleep_control);
+ break;
+
default:
status = AE_BAD_PARAMETER;
break;
--- a/xen/include/acpi/aclocal.h
+++ b/xen/include/acpi/aclocal.h
@@ -128,6 +128,8 @@ struct acpi_bit_register_info {
#define ACPI_REGISTER_PM_TIMER 0x07
#define ACPI_REGISTER_PROCESSOR_BLOCK 0x08
#define ACPI_REGISTER_SMI_COMMAND_BLOCK 0x09
+#define ACPI_REGISTER_SLEEP_CONTROL 0x0a
+#define ACPI_REGISTER_SLEEP_STATUS 0x0b
/* Masks used to access the bit_registers */
--- a/xen/include/acpi/actbl.h
+++ b/xen/include/acpi/actbl.h
@@ -309,6 +309,13 @@ enum acpi_prefered_pm_profiles {
PM_TABLET = 8
};
+/* Values for sleep_status and sleep_control registers (V5 FADT) */
+
+#define ACPI_X_WAKE_STATUS 0x80
+#define ACPI_X_SLEEP_TYPE_MASK 0x1C
+#define ACPI_X_SLEEP_TYPE_POSITION 0x02
+#define ACPI_X_SLEEP_ENABLE 0x20
+
/* Reset to default packing */
#pragma pack()
--- a/xen/include/asm-x86/acpi.h
+++ b/xen/include/asm-x86/acpi.h
@@ -126,11 +126,20 @@ struct acpi_sleep_info {
struct acpi_generic_address pm1b_cnt_blk;
struct acpi_generic_address pm1a_evt_blk;
struct acpi_generic_address pm1b_evt_blk;
- uint16_t pm1a_cnt_val;
- uint16_t pm1b_cnt_val;
+ struct acpi_generic_address sleep_control;
+ struct acpi_generic_address sleep_status;
+ union {
+ uint16_t pm1a_cnt_val;
+ uint8_t sleep_type_a;
+ };
+ union {
+ uint16_t pm1b_cnt_val;
+ uint8_t sleep_type_b;
+ };
uint32_t sleep_state;
uint64_t wakeup_vector;
uint32_t vector_width;
+ bool_t sleep_extended;
};
#endif /* CONFIG_ACPI_SLEEP */
--- a/xen/include/public/platform.h
+++ b/xen/include/public/platform.h
@@ -290,10 +290,16 @@ DEFINE_XEN_GUEST_HANDLE(xenpf_firmware_i
#define XENPF_enter_acpi_sleep 51
struct xenpf_enter_acpi_sleep {
/* IN variables */
+#if __XEN_INTERFACE_VERSION__ < 0x00040300
uint16_t pm1a_cnt_val; /* PM1a control value. */
uint16_t pm1b_cnt_val; /* PM1b control value. */
+#else
+ uint16_t val_a; /* PM1a control / sleep type A. */
+ uint16_t val_b; /* PM1b control / sleep type B. */
+#endif
uint32_t sleep_state; /* Which state to enter (Sn). */
- uint32_t flags; /* Must be zero. */
+#define XENPF_ACPI_SLEEP_EXTENDED 0x00000001
+ uint32_t flags; /* XENPF_ACPI_SLEEP_*. */
};
typedef struct xenpf_enter_acpi_sleep xenpf_enter_acpi_sleep_t;
DEFINE_XEN_GUEST_HANDLE(xenpf_enter_acpi_sleep_t);
[-- Attachment #3: Type: text/plain, Size: 126 bytes --]
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH 5/6] x86: honor ACPI indicating absence of CMOS RTC
2013-02-21 12:49 [PATCH 0/6] ACPI v5 adjustments Jan Beulich
` (3 preceding siblings ...)
2013-02-21 13:00 ` [PATCH 4/6] ACPI: support v5 (reduced HW) sleep interface Jan Beulich
@ 2013-02-21 13:01 ` Jan Beulich
2013-02-21 13:01 ` [PATCH 6/6] honor ACPI v4 FADT flags Jan Beulich
2013-02-21 13:12 ` [PATCH 0/6] ACPI v5 adjustments Keir Fraser
6 siblings, 0 replies; 9+ messages in thread
From: Jan Beulich @ 2013-02-21 13:01 UTC (permalink / raw)
To: xen-devel
[-- Attachment #1: Type: text/plain, Size: 509 bytes --]
On such systems we can boot through EFI only.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/xen/arch/x86/time.c
+++ b/xen/arch/x86/time.c
@@ -754,6 +754,9 @@ static unsigned long get_cmos_time(void)
return res;
}
+ if ( unlikely(acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_CMOS_RTC) )
+ panic("System without CMOS RTC must be booted from EFI\n");
+
spin_lock_irqsave(&rtc_lock, flags);
/* read RTC exactly on falling edge of update flag */
[-- Attachment #2: ACPI-v5-no-CMOS-RTC.patch --]
[-- Type: text/plain, Size: 553 bytes --]
x86: honor ACPI indicating absence of CMOS RTC
On such systems we can boot through EFI only.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/xen/arch/x86/time.c
+++ b/xen/arch/x86/time.c
@@ -754,6 +754,9 @@ static unsigned long get_cmos_time(void)
return res;
}
+ if ( unlikely(acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_CMOS_RTC) )
+ panic("System without CMOS RTC must be booted from EFI\n");
+
spin_lock_irqsave(&rtc_lock, flags);
/* read RTC exactly on falling edge of update flag */
[-- Attachment #3: Type: text/plain, Size: 126 bytes --]
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH 6/6] honor ACPI v4 FADT flags
2013-02-21 12:49 [PATCH 0/6] ACPI v5 adjustments Jan Beulich
` (4 preceding siblings ...)
2013-02-21 13:01 ` [PATCH 5/6] x86: honor ACPI indicating absence of CMOS RTC Jan Beulich
@ 2013-02-21 13:01 ` Jan Beulich
2013-02-21 13:12 ` [PATCH 0/6] ACPI v5 adjustments Keir Fraser
6 siblings, 0 replies; 9+ messages in thread
From: Jan Beulich @ 2013-02-21 13:01 UTC (permalink / raw)
To: xen-devel
[-- Attachment #1: Type: text/plain, Size: 5368 bytes --]
- force use of physical APIC mode if indicated so (as we don't support
xAPIC cluster mode, the respective flag is taken to force physical
mode too)
- don't use MSI if indicated so (implies no IOMMU)
Both can be overridden on the command line, for the MSI case this at
once adds a new command line option allowing to turn off PCI MSI (IOMMU
and HPET are unaffected by this).
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/docs/misc/xen-command-line.markdown
+++ b/docs/misc/xen-command-line.markdown
@@ -639,6 +639,13 @@ limit is ignored by Xen.
Specify if the MMConfig space should be enabled.
+### msi
+> `= <boolean>`
+
+> Default: `true`
+
+Force Xen to (not) use PCI-MSI, even if ACPI FADT says otherwise.
+
### mwait-idle
> `= <boolean>`
--- a/xen/arch/x86/genapic/bigsmp.c
+++ b/xen/arch/x86/genapic/bigsmp.c
@@ -40,7 +40,14 @@ static struct dmi_system_id __initdata b
static __init int probe_bigsmp(void)
{
- if (!def_to_bigsmp)
+ /*
+ * We don't implement cluster mode, so force use of
+ * physical mode in both cases.
+ */
+ if (acpi_gbl_FADT.flags &
+ (ACPI_FADT_APIC_CLUSTER | ACPI_FADT_APIC_PHYSICAL))
+ def_to_bigsmp = 1;
+ else if (!def_to_bigsmp)
dmi_check_system(bigsmp_dmi_table);
return def_to_bigsmp;
}
--- a/xen/arch/x86/genapic/x2apic.c
+++ b/xen/arch/x86/genapic/x2apic.c
@@ -30,9 +30,6 @@
#include <xen/smp.h>
#include <asm/mach-default/mach_mpparse.h>
-static bool_t __initdata x2apic_phys; /* By default we use logical cluster mode. */
-boolean_param("x2apic_phys", x2apic_phys);
-
static DEFINE_PER_CPU_READ_MOSTLY(u32, cpu_2_logical_apicid);
static DEFINE_PER_CPU_READ_MOSTLY(cpumask_t *, cluster_cpus);
static cpumask_t *cluster_cpus_spare;
@@ -223,8 +220,14 @@ static struct notifier_block x2apic_cpu_
.notifier_call = update_clusterinfo
};
+static s8 __initdata x2apic_phys = -1; /* By default we use logical cluster mode. */
+boolean_param("x2apic_phys", x2apic_phys);
+
const struct genapic *__init apic_x2apic_probe(void)
{
+ if ( x2apic_phys < 0 )
+ x2apic_phys = !!(acpi_gbl_FADT.flags & ACPI_FADT_APIC_PHYSICAL);
+
if ( x2apic_phys )
return &apic_x2apic_phys;
--- a/xen/arch/x86/hpet.c
+++ b/xen/arch/x86/hpet.c
@@ -381,6 +381,9 @@ static void __init hpet_fsb_cap_lookup(v
u32 id;
unsigned int i, num_chs;
+ if ( unlikely(acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_MSI) )
+ return;
+
id = hpet_read32(HPET_ID);
num_chs = ((id & HPET_ID_NUMBER) >> HPET_ID_NUMBER_SHIFT);
--- a/xen/arch/x86/msi.c
+++ b/xen/arch/x86/msi.c
@@ -32,6 +32,9 @@
#include <xen/iommu.h>
#include <xsm/xsm.h>
+static s8 __read_mostly use_msi = -1;
+boolean_param("msi", use_msi);
+
/* bitmap indicate which fixed map is free */
static DEFINE_SPINLOCK(msix_fixmap_lock);
static DECLARE_BITMAP(msix_fixmap_pages, FIX_MSIX_MAX_PAGES);
@@ -994,6 +997,9 @@ int pci_prepare_msix(u16 seg, u8 bus, u8
unsigned int pos = pci_find_cap_offset(seg, bus, slot, func,
PCI_CAP_ID_MSIX);
+ if ( !use_msi )
+ return 0;
+
if ( !pos )
return -ENODEV;
@@ -1029,6 +1035,9 @@ int pci_enable_msi(struct msi_info *msi,
{
ASSERT(spin_is_locked(&pcidevs_lock));
+ if ( !use_msi )
+ return -EPERM;
+
return msi->table_base ? __pci_enable_msix(msi, desc) :
__pci_enable_msi(msi, desc);
}
@@ -1074,7 +1083,10 @@ int pci_restore_msi_state(struct pci_dev
ASSERT(spin_is_locked(&pcidevs_lock));
- if (!pdev)
+ if ( !use_msi )
+ return -EOPNOTSUPP;
+
+ if ( !pdev )
return -EINVAL;
ret = xsm_resource_setup_pci(XSM_PRIV, (pdev->seg << 16) | (pdev->bus << 8) | pdev->devfn);
@@ -1133,7 +1145,7 @@ unsigned int pci_msix_get_table_len(stru
func = PCI_FUNC(pdev->devfn);
pos = pci_find_cap_offset(seg, bus, slot, func, PCI_CAP_ID_MSIX);
- if ( !pos )
+ if ( !pos || !use_msi )
return 0;
control = pci_conf_read16(seg, bus, slot, func, msix_control_reg(pos));
@@ -1170,6 +1182,11 @@ static struct notifier_block msi_cpu_nfb
void __init early_msi_init(void)
{
+ if ( use_msi < 0 )
+ use_msi = !(acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_MSI);
+ if ( !use_msi )
+ return;
+
register_cpu_notifier(&msi_cpu_nfb);
if ( msi_cpu_callback(&msi_cpu_nfb, CPU_UP_PREPARE, NULL) &
NOTIFY_STOP_MASK )
--- a/xen/drivers/passthrough/amd/iommu_acpi.c
+++ b/xen/drivers/passthrough/amd/iommu_acpi.c
@@ -1094,5 +1094,8 @@ int __init amd_iommu_get_ivrs_dev_entrie
int __init amd_iommu_update_ivrs_mapping_acpi(void)
{
+ if ( unlikely(acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_MSI) )
+ return -EPERM;
+
return acpi_table_parse(ACPI_SIG_IVRS, parse_ivrs_table);
}
--- a/xen/drivers/passthrough/vtd/iommu.c
+++ b/xen/drivers/passthrough/vtd/iommu.c
@@ -2115,6 +2115,12 @@ int __init intel_vtd_setup(void)
goto error;
}
+ if ( unlikely(acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_MSI) )
+ {
+ ret = -EPERM;
+ goto error;
+ }
+
platform_quirks_init();
/* We enable the following features only if they are supported by all VT-d
[-- Attachment #2: ACPI-v4-flags.patch --]
[-- Type: text/plain, Size: 5392 bytes --]
honor ACPI v4 FADT flags
- force use of physical APIC mode if indicated so (as we don't support
xAPIC cluster mode, the respective flag is taken to force physical
mode too)
- don't use MSI if indicated so (implies no IOMMU)
Both can be overridden on the command line, for the MSI case this at
once adds a new command line option allowing to turn off PCI MSI (IOMMU
and HPET are unaffected by this).
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/docs/misc/xen-command-line.markdown
+++ b/docs/misc/xen-command-line.markdown
@@ -639,6 +639,13 @@ limit is ignored by Xen.
Specify if the MMConfig space should be enabled.
+### msi
+> `= <boolean>`
+
+> Default: `true`
+
+Force Xen to (not) use PCI-MSI, even if ACPI FADT says otherwise.
+
### mwait-idle
> `= <boolean>`
--- a/xen/arch/x86/genapic/bigsmp.c
+++ b/xen/arch/x86/genapic/bigsmp.c
@@ -40,7 +40,14 @@ static struct dmi_system_id __initdata b
static __init int probe_bigsmp(void)
{
- if (!def_to_bigsmp)
+ /*
+ * We don't implement cluster mode, so force use of
+ * physical mode in both cases.
+ */
+ if (acpi_gbl_FADT.flags &
+ (ACPI_FADT_APIC_CLUSTER | ACPI_FADT_APIC_PHYSICAL))
+ def_to_bigsmp = 1;
+ else if (!def_to_bigsmp)
dmi_check_system(bigsmp_dmi_table);
return def_to_bigsmp;
}
--- a/xen/arch/x86/genapic/x2apic.c
+++ b/xen/arch/x86/genapic/x2apic.c
@@ -30,9 +30,6 @@
#include <xen/smp.h>
#include <asm/mach-default/mach_mpparse.h>
-static bool_t __initdata x2apic_phys; /* By default we use logical cluster mode. */
-boolean_param("x2apic_phys", x2apic_phys);
-
static DEFINE_PER_CPU_READ_MOSTLY(u32, cpu_2_logical_apicid);
static DEFINE_PER_CPU_READ_MOSTLY(cpumask_t *, cluster_cpus);
static cpumask_t *cluster_cpus_spare;
@@ -223,8 +220,14 @@ static struct notifier_block x2apic_cpu_
.notifier_call = update_clusterinfo
};
+static s8 __initdata x2apic_phys = -1; /* By default we use logical cluster mode. */
+boolean_param("x2apic_phys", x2apic_phys);
+
const struct genapic *__init apic_x2apic_probe(void)
{
+ if ( x2apic_phys < 0 )
+ x2apic_phys = !!(acpi_gbl_FADT.flags & ACPI_FADT_APIC_PHYSICAL);
+
if ( x2apic_phys )
return &apic_x2apic_phys;
--- a/xen/arch/x86/hpet.c
+++ b/xen/arch/x86/hpet.c
@@ -381,6 +381,9 @@ static void __init hpet_fsb_cap_lookup(v
u32 id;
unsigned int i, num_chs;
+ if ( unlikely(acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_MSI) )
+ return;
+
id = hpet_read32(HPET_ID);
num_chs = ((id & HPET_ID_NUMBER) >> HPET_ID_NUMBER_SHIFT);
--- a/xen/arch/x86/msi.c
+++ b/xen/arch/x86/msi.c
@@ -32,6 +32,9 @@
#include <xen/iommu.h>
#include <xsm/xsm.h>
+static s8 __read_mostly use_msi = -1;
+boolean_param("msi", use_msi);
+
/* bitmap indicate which fixed map is free */
static DEFINE_SPINLOCK(msix_fixmap_lock);
static DECLARE_BITMAP(msix_fixmap_pages, FIX_MSIX_MAX_PAGES);
@@ -994,6 +997,9 @@ int pci_prepare_msix(u16 seg, u8 bus, u8
unsigned int pos = pci_find_cap_offset(seg, bus, slot, func,
PCI_CAP_ID_MSIX);
+ if ( !use_msi )
+ return 0;
+
if ( !pos )
return -ENODEV;
@@ -1029,6 +1035,9 @@ int pci_enable_msi(struct msi_info *msi,
{
ASSERT(spin_is_locked(&pcidevs_lock));
+ if ( !use_msi )
+ return -EPERM;
+
return msi->table_base ? __pci_enable_msix(msi, desc) :
__pci_enable_msi(msi, desc);
}
@@ -1074,7 +1083,10 @@ int pci_restore_msi_state(struct pci_dev
ASSERT(spin_is_locked(&pcidevs_lock));
- if (!pdev)
+ if ( !use_msi )
+ return -EOPNOTSUPP;
+
+ if ( !pdev )
return -EINVAL;
ret = xsm_resource_setup_pci(XSM_PRIV, (pdev->seg << 16) | (pdev->bus << 8) | pdev->devfn);
@@ -1133,7 +1145,7 @@ unsigned int pci_msix_get_table_len(stru
func = PCI_FUNC(pdev->devfn);
pos = pci_find_cap_offset(seg, bus, slot, func, PCI_CAP_ID_MSIX);
- if ( !pos )
+ if ( !pos || !use_msi )
return 0;
control = pci_conf_read16(seg, bus, slot, func, msix_control_reg(pos));
@@ -1170,6 +1182,11 @@ static struct notifier_block msi_cpu_nfb
void __init early_msi_init(void)
{
+ if ( use_msi < 0 )
+ use_msi = !(acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_MSI);
+ if ( !use_msi )
+ return;
+
register_cpu_notifier(&msi_cpu_nfb);
if ( msi_cpu_callback(&msi_cpu_nfb, CPU_UP_PREPARE, NULL) &
NOTIFY_STOP_MASK )
--- a/xen/drivers/passthrough/amd/iommu_acpi.c
+++ b/xen/drivers/passthrough/amd/iommu_acpi.c
@@ -1094,5 +1094,8 @@ int __init amd_iommu_get_ivrs_dev_entrie
int __init amd_iommu_update_ivrs_mapping_acpi(void)
{
+ if ( unlikely(acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_MSI) )
+ return -EPERM;
+
return acpi_table_parse(ACPI_SIG_IVRS, parse_ivrs_table);
}
--- a/xen/drivers/passthrough/vtd/iommu.c
+++ b/xen/drivers/passthrough/vtd/iommu.c
@@ -2115,6 +2115,12 @@ int __init intel_vtd_setup(void)
goto error;
}
+ if ( unlikely(acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_MSI) )
+ {
+ ret = -EPERM;
+ goto error;
+ }
+
platform_quirks_init();
/* We enable the following features only if they are supported by all VT-d
[-- Attachment #3: Type: text/plain, Size: 126 bytes --]
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH 1/6] ACPI 5.0: Basic support for FADT version 5
2013-02-21 12:58 ` [PATCH 1/6] ACPI 5.0: Basic support for FADT version 5 Jan Beulich
@ 2013-02-21 13:03 ` Jan Beulich
0 siblings, 0 replies; 9+ messages in thread
From: Jan Beulich @ 2013-02-21 13:03 UTC (permalink / raw)
To: xen-devel
[-- Attachment #1: Type: text/plain, Size: 229 bytes --]
>>> On 21.02.13 at 13:58, "Jan Beulich" <JBeulich@suse.com> wrote:
> Signed-off-by: Bob Moore <robert.moore@intel.com>
> Signed-off-by: Jan Beulich <jbeulich@suse.com>
Patch now also attached, if needed by anyone.
Jan
[-- Attachment #2: ACPI-v5-FADT.patch --]
[-- Type: text/plain, Size: 4097 bytes --]
ACPI 5.0: Basic support for FADT version 5
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
--- a/xen/drivers/acpi/tables/tbfadt.c
+++ b/xen/drivers/acpi/tables/tbfadt.c
@@ -222,12 +222,13 @@ void __init acpi_tb_create_local_fadt(st
/*
* Check if the FADT is larger than the largest table that we expect
- * (the ACPI 2.0/3.0 version). If so, truncate the table, and issue
+ * (the ACPI 5.0 version). If so, truncate the table, and issue
* a warning.
*/
if (length > sizeof(struct acpi_table_fadt)) {
ACPI_WARNING((AE_INFO,
- "FADT (revision %u) is longer than ACPI 2.0 version, truncating length 0x%X to 0x%zX",
+ "FADT (revision %u) is longer than ACPI 5.0 version,"
+ " truncating length %u to %zu",
table->revision, (unsigned)length,
sizeof(struct acpi_table_fadt)));
}
--- a/xen/include/acpi/actbl.h
+++ b/xen/include/acpi/actbl.h
@@ -255,6 +255,8 @@ struct acpi_table_fadt {
struct acpi_generic_address xpm_timer_block; /* 64-bit Extended Power Mgt Timer Ctrl Reg Blk address */
struct acpi_generic_address xgpe0_block; /* 64-bit Extended General Purpose Event 0 Reg Blk address */
struct acpi_generic_address xgpe1_block; /* 64-bit Extended General Purpose Event 1 Reg Blk address */
+ struct acpi_generic_address sleep_control; /* 64-bit Sleep Control register */
+ struct acpi_generic_address sleep_status; /* 64-bit Sleep Status register */
};
/* Masks for FADT Boot Architecture Flags (boot_flags) */
@@ -264,6 +266,7 @@ struct acpi_table_fadt {
#define ACPI_FADT_NO_VGA (1<<2) /* 02: [V4] It is not safe to probe for VGA hardware */
#define ACPI_FADT_NO_MSI (1<<3) /* 03: [V4] Message Signaled Interrupts (MSI) must not be enabled */
#define ACPI_FADT_NO_ASPM (1<<4) /* 04: [V4] PCIe ASPM control must not be enabled */
+#define ACPI_FADT_NO_CMOS_RTC (1<<5) /* 05: [V5] No CMOS real-time clock present */
#define FADT2_REVISION_ID 3
@@ -289,6 +292,8 @@ struct acpi_table_fadt {
#define ACPI_FADT_REMOTE_POWER_ON (1<<17) /* 17: [V4] System is compatible with remote power on (ACPI 3.0) */
#define ACPI_FADT_APIC_CLUSTER (1<<18) /* 18: [V4] All local APICs must use cluster model (ACPI 3.0) */
#define ACPI_FADT_APIC_PHYSICAL (1<<19) /* 19: [V4] All local x_aPICs must use physical dest mode (ACPI 3.0) */
+#define ACPI_FADT_HW_REDUCED (1<<20) /* 20: [V5] ACPI hardware is not implemented (ACPI 5.0) */
+#define ACPI_FADT_LOW_POWER_S0 (1<<21) /* 21: [V5] S0 power savings are equal or better than S3 (ACPI 5.0) */
/* Values for preferred_profile (Preferred Power Management Profiles) */
@@ -299,14 +304,16 @@ enum acpi_prefered_pm_profiles {
PM_WORKSTATION = 3,
PM_ENTERPRISE_SERVER = 4,
PM_SOHO_SERVER = 5,
- PM_APPLIANCE_PC = 6
+ PM_APPLIANCE_PC = 6,
+ PM_PERFORMANCE_SERVER = 7,
+ PM_TABLET = 8
};
/* Reset to default packing */
#pragma pack()
-#define ACPI_FADT_OFFSET(f) (u8) ACPI_OFFSET (struct acpi_table_fadt, f)
+#define ACPI_FADT_OFFSET(f) (u16) ACPI_OFFSET (struct acpi_table_fadt, f)
/*
* Get the remaining ACPI tables
@@ -324,12 +331,15 @@ enum acpi_prefered_pm_profiles {
* FADT is the bottom line as to what the version really is.
*
* For reference, the values below are as follows:
- * FADT V1 size: 0x74
- * FADT V2 size: 0x84
- * FADT V3+ size: 0xF4
+ * FADT V1 size: 0x074
+ * FADT V2 size: 0x084
+ * FADT V3 size: 0x0F4
+ * FADT V4 size: 0x0F4
+ * FADT V5 size: 0x10C
*/
#define ACPI_FADT_V1_SIZE (u32) (ACPI_FADT_OFFSET (flags) + 4)
#define ACPI_FADT_V2_SIZE (u32) (ACPI_FADT_OFFSET (reserved4[0]) + 3)
-#define ACPI_FADT_V3_SIZE (u32) (sizeof (struct acpi_table_fadt))
+#define ACPI_FADT_V3_SIZE (u32) (ACPI_FADT_OFFSET (sleep_control))
+#define ACPI_FADT_V5_SIZE (u32) (sizeof (struct acpi_table_fadt))
#endif /* __ACTBL_H__ */
[-- Attachment #3: Type: text/plain, Size: 126 bytes --]
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH 0/6] ACPI v5 adjustments
2013-02-21 12:49 [PATCH 0/6] ACPI v5 adjustments Jan Beulich
` (5 preceding siblings ...)
2013-02-21 13:01 ` [PATCH 6/6] honor ACPI v4 FADT flags Jan Beulich
@ 2013-02-21 13:12 ` Keir Fraser
6 siblings, 0 replies; 9+ messages in thread
From: Keir Fraser @ 2013-02-21 13:12 UTC (permalink / raw)
To: Jan Beulich, xen-devel
On 21/02/2013 12:49, "Jan Beulich" <JBeulich@suse.com> wrote:
> The first three of these are cloned from Linux counterparts, and the
> last one adjusts behavior we would have needed for v4 already.
>
> 1: ACPI 5.0: Basic support for FADT version 5
> 2: ACPI 5.0: Implement hardware-reduced option
> 3: ACPICA: Update for larger ACPI 5 FADT size
> 4: ACPI: support v5 (reduced HW) sleep interface
> 5: x86: honor ACPI indicating absence of CMOS RTC
> 6: honor ACPI v4 FADT flags
>
> Signed-off-by: Jan Beulich <jbeulich@suse.com>
Looks alright to me.
Acked-by: Keir Fraser <keir@xen.org>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2013-02-21 13:12 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-02-21 12:49 [PATCH 0/6] ACPI v5 adjustments Jan Beulich
2013-02-21 12:58 ` [PATCH 1/6] ACPI 5.0: Basic support for FADT version 5 Jan Beulich
2013-02-21 13:03 ` Jan Beulich
2013-02-21 12:59 ` [PATCH 2/6] ACPI 5.0: Implement hardware-reduced option Jan Beulich
2013-02-21 12:59 ` [PATCH 3/6] ACPICA: Update for larger ACPI 5 FADT size Jan Beulich
2013-02-21 13:00 ` [PATCH 4/6] ACPI: support v5 (reduced HW) sleep interface Jan Beulich
2013-02-21 13:01 ` [PATCH 5/6] x86: honor ACPI indicating absence of CMOS RTC Jan Beulich
2013-02-21 13:01 ` [PATCH 6/6] honor ACPI v4 FADT flags Jan Beulich
2013-02-21 13:12 ` [PATCH 0/6] ACPI v5 adjustments Keir Fraser
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.