ACPICA 20130823 Release
Subsystems (2)
cover ACPICA 20130823 Release
The 20130823 ACPICA kernel resident system updates is linuxized based on the pm/linux-next branch.
The patch set has passed a build/boot test on the following machines:
Dell Inspiron Mini 1010 (i386)
HP Compaq 8200 Elite SFF PC (x86-64)
With the following kernel configuration items enabled:
CONFIG_ACPI_AC
CONFIG_ACPI_BATTERY
CONFIG_ACPI_NUMA
CONFIG_ACPI_PROCESSOR
CONFIG_ACPI_THERMAL
CONFIG_ACPI_VIDEO
CONFIG_CPU_FREQ
CONFIG_I2C_SCMI
CONFIG_SENSORS_ACPI_POWER
CONFIG_PANASONIC_LAPTOP
CONFIG_XO15_EBOOK
CONFIG_DMAR_TABLE
CONFIG_INTEL_IOMMU
CONFIG_IRQ_REMAP
New divergences are introduced in this release due to the deletion of the following patches:
ACPICA: iASL/disassembler: Add support for external method reference
file.
ACPICA: Tables: Add multiple table instances support.
They are AcpiDump specific updates and will appear along with AcpiDump release during next release.
NOTE: Label indentation is not fixed in this series, it can be fixed
by another ACPICA commit that changes the indent option to do so.
Please review.
Bjorn Helgaas (1):
ACPICA: acpi_read: On error, do not modify the return value target
location.
Bob Moore (7):
ACPICA: Improve error message for "too many parent prefixes"
condition.
ACPICA: Debug output: small formatting update, no functional change.
ACPICA: Debugger: Prevent possible command line buffer overflow.
ACPICA: Validate start object for acpi_walk_namespace.
ACPICA: Debugger: Add new command to display full namespace
pathnames.
ACPICA: SCI Handlers: Update handler interface, eliminate unnecessary
argument.
ACPICA: Update version to 20130823.
David E. Box (1):
ACPICA: Operation region support: Never free the handler "context"
pointer.
Lv Zheng (4):
ACPICA: Add support for host-installed SCI handlers.
ACPICA: Tables: Cleanup table checksum verification code.
ACPICA: Tables: Cleanup RSDP signature codes.
ACPICA: Cleanup exception codes.
Naresh Bhat (1):
ACPICA: Linux-specific header: enable "aarch64" 64-bit build.
drivers/acpi/acpica/acdebug.h | 5 +- drivers/acpi/acpica/acevents.h | 6 +- drivers/acpi/acpica/acglobal.h | 18 +++-- drivers/acpi/acpica/aclocal.h | 8 +++ drivers/acpi/acpica/acnamesp.h | 6 ++ drivers/acpi/acpica/evgpeutil.c | 2 +- drivers/acpi/acpica/evmisc.c | 14 ++-- drivers/acpi/acpica/evregion.c | 23 ++++--- drivers/acpi/acpica/evsci.c | 78 ++++++++++++++++++++-- drivers/acpi/acpica/evxface.c | 139 +++++++++++++++++++++++++++++++++++++++ drivers/acpi/acpica/hwxface.c | 26 ++++---- drivers/acpi/acpica/nsaccess.c | 5 +- drivers/acpi/acpica/nsdump.c | 134 +++++++++++++++++++++++++++++++++++++ drivers/acpi/acpica/nsxfeval.c | 8 +++ drivers/acpi/acpica/tbinstal.c | 12 +--- drivers/acpi/acpica/tbprint.c | 12 +++- drivers/acpi/acpica/tbxfroot.c | 3 +- drivers/acpi/acpica/utdebug.c | 2 +- drivers/acpi/acpica/utglobal.c | 3 +- include/acpi/acexcep.h | 8 ++- include/acpi/acpixf.h | 15 +++-- include/acpi/actypes.h | 8 +++ include/acpi/platform/aclinux.h | 2 +- 23 files changed, 465 insertions(+), 72 deletions(-)
Signature · 1 line
1.7.10
PATCH 1/14 ACPICA: Improve error message for "too many parent prefixes" condition.
From: Bob Moore <[email protected]>
Emit the full offending pathname in the error message.
--- drivers/acpi/acpica/nsaccess.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/acpi/acpica/nsaccess.c +3 -2
@@ -424,8 +424,9 @@ acpi_ns_lookup(union acpi_generic_state *scope_info,/* Current scope has no parent scope */ ACPI_ERROR ((AE_INFO , -"ACPI path has too many parent prefixes (^) " -"- reached beyond root node" )); +"%s: Path has too many parent prefixes (^) " +"- reached beyond root node" , + pathname));return_ACPI_STATUS (AE_NOT_FOUND ); } }
Signature · 1 line
1.7.10
PATCH 2/14 ACPICA: acpi_read: On error, do not modify the return value target location.
In reply to Lv ZhengFrom: Bjorn Helgaas <[email protected]>
If an error happens in the middle of a split 32/32 64-bit I/O operation, do not modify the target of the return value pointer. Makes the code consistent with the rest of ACPICA. Bjorn Helgaas.
--- drivers/acpi/acpica/hwxface.c | 26 ++++++++++++++------------ 1 file changed, 14 insertions(+), 12 deletions(-)
drivers/acpi/acpica/hwxface.c +14 -12
@@ -119,7 +119,8 @@ ACPI_EXPORT_SYMBOL(acpi_reset)******************************************************************************/ acpi_status acpi_read (u64 *return_value,struct acpi_generic_address *reg) { -u32 value; +u32 value_lo; +u32 value_hi;u32 width;u64 address;acpi_status status;
@@ -137,13 +138,8 @@ acpi_status acpi_read(u64 *return_value, struct acpi_generic_address *reg)return (status); } -/* Initialize entire 64-bit return value to zero */ - - *return_value =0 ; - value =0 ; -/* -* Two address spaces supported: Memory or IO. PCI_Config is +* Two address spaces supported: Memory or I/O. PCI_Config is * not supported here because the GAS structure is insufficient */ if (reg->space_id ==ACPI_ADR_SPACE_SYSTEM_MEMORY ) {
@@ -155,29 +151,35 @@ acpi_status acpi_read(u64 *return_value, struct acpi_generic_address *reg) } }else {/* ACPI_ADR_SPACE_SYSTEM_IO, validated earlier */ + value_lo =0 ; + value_hi =0 ; + width = reg->bit_width;if (width ==64 ) { width =32 ;/* Break into two 32-bit transfers */ } status =acpi_hw_read_port ((acpi_io_address ) - address, &value, width); + address, &value_lo, width);if (ACPI_FAILURE (status)) {return (status); } - *return_value = value;if (reg->bit_width ==64 ) {/* Read the top 32 bits */ status =acpi_hw_read_port ((acpi_io_address) - (address +4 ), &value,32 ); + (address +4 ), &value_hi, +32 );if (ACPI_FAILURE (status)) {return (status); } - *return_value |= ((u64 )value <<32 ); } + +/* Set the return value only if status is AE_OK */ + + *return_value = (value_lo | ((u64 )value_hi <<32 )); } ACPI_DEBUG_PRINT((ACPI_DB_IO,
@@ -186,7 +188,7 @@ acpi_status acpi_read(u64 *return_value, struct acpi_generic_address *reg)ACPI_FORMAT_UINT64 (address), acpi_ut_get_region_name(reg->space_id))); -return (status); +return (AE_OK ); }ACPI_EXPORT_SYMBOL (acpi_read)
Signature · 1 line
1.7.10
PATCH 3/14 ACPICA: Debug output: small formatting update, no functional change.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
Allow for longer filenames in the module name output during trace operations.
--- drivers/acpi/acpica/utdebug.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/acpi/acpica/utdebug.c +1 -1
@@ -190,7 +190,7 @@ acpi_debug_print(u32 requested_debug_level,* Display the module name, current line number, thread ID (if requested), * current procedure nesting level, and the current procedure name */ -acpi_os_printf ("%8s-%04ld " , module_name, line_number); +acpi_os_printf ("%9s-%04ld " , module_name, line_number);if (ACPI_LV_THREADS & acpi_dbg_level) {acpi_os_printf ("[%u] " , (u32 )thread_id);
Signature · 1 line
1.7.10
PATCH 4/14 ACPICA: Linux-specific header: enable "aarch64" 64-bit build.
In reply to Lv ZhengFrom: Naresh Bhat <[email protected]>
Add support for the __aarch64__ define for 64-bit builds.
--- include/acpi/platform/aclinux.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
include/acpi/platform/aclinux.h +1 -1
@@ -87,7 +87,7 @@#define ACPI_FLUSH_CPU_CACHE ()#define ACPI_CAST_PTHREAD_T (pthread) ((acpi_thread_id) (pthread)) -#if defined(__ia64__) || defined(__x86_64__) +#if defined(__ia64__) || defined(__x86_64__) || defined(__aarch64__)#define ACPI_MACHINE_WIDTH 64#define COMPILER_DEPENDENT_INT64 long#define COMPILER_DEPENDENT_UINT64 unsigned long
Signature · 1 line
1.7.10
PATCH 5/14 ACPICA: Debugger: Prevent possible command line buffer overflow.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
Increase the size of a couple of the debugger line buffers. ACPICA BZ 1037.
--- drivers/acpi/acpica/acglobal.h | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-)
drivers/acpi/acpica/acglobal.h +10 -7
@@ -445,13 +445,6 @@ ACPI_EXTERN u8 acpi_gbl_db_opt_tables; ACPI_EXTERN u8 acpi_gbl_db_opt_stats; ACPI_EXTERN u8 acpi_gbl_db_opt_ini_methods; ACPI_EXTERN u8 acpi_gbl_db_opt_no_region_support; - -ACPI_EXTERN char *acpi_gbl_db_args[ACPI_DEBUGGER_MAX_ARGS]; -ACPI_EXTERN acpi_object_type acpi_gbl_db_arg_types[ACPI_DEBUGGER_MAX_ARGS]; -ACPI_EXTERN char acpi_gbl_db_line_buf[ACPI_DB_LINE_BUFFER_SIZE]; -ACPI_EXTERN char acpi_gbl_db_parsed_buf[ACPI_DB_LINE_BUFFER_SIZE]; -ACPI_EXTERN char acpi_gbl_db_scope_buf[80]; -ACPI_EXTERN char acpi_gbl_db_debug_filename[80]; ACPI_EXTERN u8 acpi_gbl_db_output_to_file; ACPI_EXTERN char *acpi_gbl_db_buffer; ACPI_EXTERN char *acpi_gbl_db_filename;
@@ -459,6 +452,16 @@ ACPI_EXTERN u32 acpi_gbl_db_debug_level; ACPI_EXTERN u32 acpi_gbl_db_console_debug_level; ACPI_EXTERN struct acpi_namespace_node *acpi_gbl_db_scope_node; +ACPI_EXTERN char *acpi_gbl_db_args[ACPI_DEBUGGER_MAX_ARGS]; +ACPI_EXTERN acpi_object_type acpi_gbl_db_arg_types[ACPI_DEBUGGER_MAX_ARGS]; + +/* These buffers should all be the same size */ + +ACPI_EXTERN char acpi_gbl_db_line_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_parsed_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_scope_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_debug_filename[ACPI_DB_LINE_BUFFER_SIZE]; +/* * Statistic globals */
Signature · 1 line
1.7.10
PATCH 6/14 ACPICA: Validate start object for acpi_walk_namespace.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
Perform a sanity check on the start object to prevent problems later. ACPICA BZ 1025.
--- drivers/acpi/acpica/nsxfeval.c | 8 ++++++++ 1 file changed, 8 insertions(+)
drivers/acpi/acpica/nsxfeval.c +8 -0
@@ -605,11 +605,19 @@ acpi_walk_namespace(acpi_object_type type, goto unlock_and_exit; } +/* Now we can validate the starting node */ + +if (!acpi_ns_validate_handle (start_object)) { + status =AE_BAD_PARAMETER ; + goto unlock_and_exit2; + } + status =acpi_ns_walk_namespace (type, start_object, max_depth,ACPI_NS_WALK_UNLOCK , descending_callback, ascending_callback, context, return_value); + unlock_and_exit2: (void )acpi_ut_release_mutex (ACPI_MTX_NAMESPACE ); unlock_and_exit:
Signature · 1 line
1.7.10
PATCH 7/14 ACPICA: Add support for host-installed SCI handlers.
In reply to Lv ZhengThis change adds support to allow hosts to install System Control Interrupt handlers. Certain ACPI functionality requires the host to handle raw SCIs. For example, the "SCI Doorbell" that is defined for memory power state support requires the host device driver to handle SCIs to examine if the doorbell has been activated. Multiple SCI handlers can be installed to allow for future expansion. Debugger support is included.
Lv Zheng, Bob Moore. ACPICA BZ 1032.
--- drivers/acpi/acpica/acdebug.h | 3 +- drivers/acpi/acpica/acevents.h | 6 +- drivers/acpi/acpica/acglobal.h | 1 + drivers/acpi/acpica/aclocal.h | 8 +++ drivers/acpi/acpica/evgpeutil.c | 2 +- drivers/acpi/acpica/evmisc.c | 14 ++-- drivers/acpi/acpica/evsci.c | 80 ++++++++++++++++++++-- drivers/acpi/acpica/evxface.c | 139 +++++++++++++++++++++++++++++++++++++++ drivers/acpi/acpica/utglobal.c | 3 +- include/acpi/acpixf.h | 13 +++- include/acpi/actypes.h | 3 + 11 files changed, 249 insertions(+), 23 deletions(-)
drivers/acpi/acpica/acdebug.h +2 -1
@@ -113,11 +113,12 @@ void acpi_db_display_handlers(void); ACPI_HW_DEPENDENT_RETURN_VOID(void acpi_db_generate_gpe(char *gpe_arg,char *block_arg)) + ACPI_HW_DEPENDENT_RETURN_VOID(void acpi_db_generate_sci(void)) /* * dbconvert - miscellaneous conversion routines */ - acpi_status acpi_db_hex_char_to_value(int hex_char, u8 *return_value); +acpi_status acpi_db_hex_char_to_value(int hex_char, u8 *return_value); acpi_status acpi_db_convert_to_package(char *string, union acpi_object *object);
drivers/acpi/acpica/acevents.h +3 -3
@@ -242,11 +242,11 @@ acpi_ev_initialize_region(union acpi_operand_object *region_obj,*/ u32 ACPI_SYSTEM_XFACE acpi_ev_gpe_xrupt_handler(void *context); -u32 acpi_ev_install_sci_handler (void ); +u32 acpi_ev_sci_dispatch (void ); -acpi_status acpi_ev_remove_sci_handler (void ); +u32 acpi_ev_install_sci_handler (void ); -u32 acpi_ev_initialize_SCI (u32 program_SCI); +acpi_status acpi_ev_remove_all_sci_handlers (void ); ACPI_HW_DEPENDENT_RETURN_VOID(void acpi_ev_terminate(void))#endif /* __ACEVENTS_H__ */
drivers/acpi/acpica/acglobal.h +1 -0
@@ -269,6 +269,7 @@ ACPI_EXTERN acpi_table_handler acpi_gbl_table_handler; ACPI_EXTERN void *acpi_gbl_table_handler_context; ACPI_EXTERN struct acpi_walk_state *acpi_gbl_breakpoint_walk; ACPI_EXTERN acpi_interface_handler acpi_gbl_interface_handler; +ACPI_EXTERN struct acpi_sci_handler_info *acpi_gbl_sci_handler_list;/* Owner ID support */
drivers/acpi/acpica/aclocal.h +8 -0
@@ -398,6 +398,14 @@ struct acpi_simple_repair_info {* ****************************************************************************/ +/* Dispatch info for each host-installed SCI handler */ + +struct acpi_sci_handler_info { +struct acpi_sci_handler_info *next; +acpi_sci_handler address;/* Address of handler */ +void *context;/* Context to be passed to handler */ +}; +/* Dispatch info for each GPE -- either a method or handler, cannot be both */ struct acpi_gpe_handler_info {
drivers/acpi/acpica/evgpeutil.c +1 -1
@@ -196,7 +196,7 @@ acpi_ev_get_gpe_device(struct acpi_gpe_xrupt_info *gpe_xrupt_info,* * FUNCTION: acpi_ev_get_gpe_xrupt_block * -* PARAMETERS: interrupt_number - Interrupt for a GPE block +* PARAMETERS: interrupt_number - Interrupt for a GPE block * * RETURN: A GPE interrupt block *
drivers/acpi/acpica/evmisc.c +7 -7
@@ -264,13 +264,6 @@ void acpi_ev_terminate(void) status =acpi_ev_walk_gpe_list (acpi_hw_disable_gpe_block,NULL ); -/* Remove SCI handler */ - - status =acpi_ev_remove_sci_handler (); -if (ACPI_FAILURE (status)) { -ACPI_ERROR ((AE_INFO ,"Could not remove SCI handler" )); - } - status =acpi_ev_remove_global_lock_handler ();if (ACPI_FAILURE (status)) { ACPI_ERROR((AE_INFO,
@@ -280,6 +273,13 @@ void acpi_ev_terminate(void) acpi_gbl_events_initialized = FALSE; } +/* Remove SCI handlers */ + + status =acpi_ev_remove_all_sci_handlers (); +if (ACPI_FAILURE (status)) { +ACPI_ERROR ((AE_INFO ,"Could not remove SCI handler" )); + } +/* Deallocate all handler objects installed within GPE info structs */ status =acpi_ev_walk_gpe_list (acpi_ev_delete_gpe_handlers,NULL );
drivers/acpi/acpica/evsci.c +73 -7
131 lines not on this page · raw message
drivers/acpi/acpica/evxface.c +139 -0
151 lines not on this page · raw message
drivers/acpi/acpica/utglobal.c +2 -1
@@ -291,7 +291,7 @@ acpi_status acpi_ut_init_globals(void)#if (!ACPI_REDUCED_HARDWARE ) -/* GPE support */ +/* GPE/SCI support */ acpi_gbl_all_gpes_initialized = FALSE; acpi_gbl_gpe_xrupt_list_head =NULL ;
@@ -300,6 +300,7 @@ acpi_status acpi_ut_init_globals(void) acpi_current_gpe_count =0 ; acpi_gbl_global_event_handler =NULL ; + acpi_gbl_sci_handler_list =NULL ;#endif /* !ACPI_REDUCED_HARDWARE */
include/acpi/acpixf.h +10 -3
@@ -280,9 +280,16 @@ acpi_status acpi_install_initialization_handler(acpi_init_handler handler, u32 function); ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status - acpi_install_global_event_handler - (acpi_gbl_event_handler handler, void *context)) - + acpi_install_sci_handler(acpi_sci_handler + address, + void *context)) +ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status + acpi_remove_sci_handler(acpi_sci_handler + address)) +ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status + acpi_install_global_event_handler + (acpi_gbl_event_handler handler, + void *context)) ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status acpi_install_fixed_event_handler(u32 acpi_event,
include/acpi/actypes.h +3 -0
@@ -946,6 +946,9 @@ typedef void* Various handlers and callback procedures */ typedef +u32 (*acpi_sci_handler ) (u32 interrupt_number,void *context); + +typedef void (*acpi_gbl_event_handler ) (u32 event_type,acpi_handle device,u32 event_number,void *context);
Signature · 1 line
1.7.10
PATCH 9/14 ACPICA: Debugger: Add new command to display full namespace pathnames.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
Paths command displays the full pathname and object type for the entire namespace. Alternative to the Namespace command.
--- drivers/acpi/acpica/acdebug.h | 2 + drivers/acpi/acpica/acnamesp.h | 6 ++ drivers/acpi/acpica/nsdump.c | 134 ++++++++++++++++++++++++++++++++++++++++ 3 files changed, 142 insertions(+)
drivers/acpi/acpica/acdebug.h +2 -0
@@ -155,6 +155,8 @@ void acpi_db_set_scope(char *name);void acpi_db_dump_namespace (char *start_arg,char *depth_arg); +void acpi_db_dump_namespace_paths (void ); +void acpi_db_dump_namespace_by_owner (char *owner_arg,char *depth_arg);acpi_status acpi_db_find_name_in_namespace (char *name_arg);
drivers/acpi/acpica/acnamesp.h +6 -0
@@ -213,6 +213,12 @@ acpi_ns_dump_objects(acpi_object_type type,u8 display_type, u32 max_depth, acpi_owner_id owner_id, acpi_handle start_handle); + +void +acpi_ns_dump_object_paths (acpi_object_type type, +u8 display_type, +u32 max_depth, +acpi_owner_id owner_id,acpi_handle start_handle);#endif /* ACPI_FUTURE_USAGE */ /*
drivers/acpi/acpica/nsdump.c +134 -0
146 lines not on this page · raw message
Signature · 1 line
1.7.10
PATCH 10/14 ACPICA: Tables: Cleanup table checksum verification code.
In reply to Lv ZhengThis patch reduces code redundancy by moving the FACS/S3PT checksum verification skip logic into acpi_tb_verify_checksum() and other calls of this function also get benefit from this change. Lv Zheng.
--- drivers/acpi/acpica/tbinstal.c | 12 +++--------- drivers/acpi/acpica/tbprint.c | 10 ++++++++++ 2 files changed, 13 insertions(+), 9 deletions(-)
drivers/acpi/acpica/tbinstal.c +3 -9
@@ -80,16 +80,10 @@ acpi_status acpi_tb_verify_table(struct acpi_table_desc *table_desc) } } -/* FACS is the odd table, has no standard ACPI header and no checksum */ +/* Always calculate checksum, ignore bad checksum if requested */ -if (!ACPI_COMPARE_NAME (&table_desc->signature,ACPI_SIG_FACS )) { - -/* Always calculate checksum, ignore bad checksum if requested */ - - status = -acpi_tb_verify_checksum (table_desc->pointer, - table_desc->length); - } + status = +acpi_tb_verify_checksum (table_desc->pointer, table_desc->length);return_ACPI_STATUS (status); }
drivers/acpi/acpica/tbprint.c +10 -0
@@ -190,6 +190,16 @@ acpi_status acpi_tb_verify_checksum(struct acpi_table_header *table, u32 length) {u8 checksum; +/* +* FACS/S3PT: +* They are the odd tables, have no standard ACPI header and no checksum +*/ + +if (ACPI_COMPARE_NAME (table->signature,ACPI_SIG_S3PT ) || +ACPI_COMPARE_NAME (table->signature,ACPI_SIG_FACS )) { +return (AE_OK ); + } +/* Compute the checksum on the table */ checksum =acpi_tb_checksum (ACPI_CAST_PTR (u8, table), length);
Signature · 1 line
1.7.10
PATCH 11/14 ACPICA: Tables: Cleanup RSDP signature codes.
In reply to Lv ZhengThis patch introduces new macors to handle RSDP signature and cleans up the affected codes. Lv Zheng.
--- drivers/acpi/acpica/tbprint.c | 2 +- drivers/acpi/acpica/tbxfroot.c | 3 +-- include/acpi/actypes.h | 5 +++++ 3 files changed, 7 insertions(+), 3 deletions(-)
drivers/acpi/acpica/tbprint.c +1 -1
@@ -138,7 +138,7 @@ acpi_tb_print_table_header(acpi_physical_address address,ACPI_INFO ((AE_INFO ,"%4.4s %p %05X" , header->signature,ACPI_CAST_PTR (void, address), header->length)); - }else if (ACPI_COMPARE_NAME (header->signature,ACPI_SIG_RSDP )) { + }else if (ACPI_VALIDATE_RSDP_SIG (header->signature)) {/* RSDP has no common fields */
drivers/acpi/acpica/tbxfroot.c +1 -2
@@ -68,8 +68,7 @@ acpi_status acpi_tb_validate_rsdp(struct acpi_table_rsdp *rsdp)* Note: Sometimes there exists more than one RSDP in memory; the valid * RSDP has a valid checksum, all others have an invalid checksum. */ -if (ACPI_STRNCMP ((char *)rsdp->signature,ACPI_SIG_RSDP , -sizeof (ACPI_SIG_RSDP ) -1 ) !=0 ) { +if (!ACPI_VALIDATE_RSDP_SIG (rsdp->signature)) {/* Nope, BAD Signature */
include/acpi/actypes.h +5 -0
@@ -474,6 +474,11 @@ typedef u64 acpi_integer;#define ACPI_MOVE_NAME (dest,src) (ACPI_STRNCPY (ACPI_CAST_PTR (char, (dest)), ACPI_CAST_PTR (char, (src)), ACPI_NAME_SIZE))#endif +/* Support for the special RSDP signature (8 characters) */ + +#define ACPI_VALIDATE_RSDP_SIG (a) (!ACPI_STRNCMP (ACPI_CAST_PTR (char, (a)), ACPI_SIG_RSDP, 8)) +#define ACPI_MAKE_RSDP_SIG (dest) (ACPI_MEMCPY (ACPI_CAST_PTR (char, (dest)), ACPI_SIG_RSDP, 8)) +/******************************************************************************* * * Miscellaneous constants
Signature · 1 line
1.7.10
PATCH 8/14 ACPICA: Operation region support: Never free the handler "context" pointer.
In reply to Lv ZhengFrom: "David E. Box" <[email protected]>
This change removes some dangerous code that attempts to free the handler context pointer in some (rare) circumstances. The owner of the handler owns this pointer and the ACPICA code should never touch it. Although not seen to be an issue in any kernel, it did show up as a problem under AcpiExec. Also, set the internal storage field for the context pointer to zero when the region is deactivated, simply for sanity. David Box.
--- drivers/acpi/acpica/evregion.c | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-)
drivers/acpi/acpica/evregion.c +13 -10
@@ -217,16 +217,11 @@ acpi_ev_address_space_dispatch(union acpi_operand_object *region_obj,if (!(region_obj->region.flags &AOPOBJ_SETUP_COMPLETE )) { region_obj->region.flags |=AOPOBJ_SETUP_COMPLETE ; -if (region_obj2->extra.region_context) { - -/* The handler for this region was already installed */ - -ACPI_FREE (region_context); - }else { -/* -* Save the returned context for use in all accesses to -* this particular region -*/ +/* +* Save the returned context for use in all accesses to +* the handler for this particular region +*/ +if (!(region_obj2->extra.region_context)) { region_obj2->extra.region_context = region_context; }
@@ -402,6 +397,14 @@ acpi_ev_detach_region(union acpi_operand_object *region_obj, handler_obj->address_space. context, region_context); +/* +* region_context should have been released by the deactivate +* operation. We don't need access to it anymore here. +*/ +if (region_context) { + *region_context =NULL ; + } +/* Init routine may fail, Just ignore errors */ if (ACPI_FAILURE (status)) {
Signature · 1 line
1.7.10
PATCH 12/14 ACPICA: Cleanup exception codes.
In reply to Lv ZhengThis patch adds AE_ACCESS for EACCES or EPERM. Some error prompts are also cleaned up in this patch. Lv Zheng.
--- include/acpi/acexcep.h | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)
include/acpi/acexcep.h +5 -3
@@ -125,8 +125,9 @@ struct acpi_exception_info {#define AE_NO_HANDLER EXCEP_ENV (0x001A)#define AE_OWNER_ID_LIMIT EXCEP_ENV (0x001B)#define AE_NOT_CONFIGURED EXCEP_ENV (0x001C) +#define AE_ACCESS EXCEP_ENV (0x001D) -#define AE_CODE_ENV_MAX 0x001C +#define AE_CODE_ENV_MAX 0x001D/* * Programmer exceptions
@@ -227,7 +228,7 @@ static const struct acpi_exception_info acpi_gbl_exception_names_env[] = {EXCEP_TXT ("AE_NO_ACPI_TABLES" ,"ACPI tables could not be found" ),EXCEP_TXT ("AE_NO_NAMESPACE" ,"A namespace has not been loaded" ),EXCEP_TXT ("AE_NO_MEMORY" ,"Insufficient dynamic memory" ), -EXCEP_TXT ("AE_NOT_FOUND" ,"The name was not found in the namespace" ), +EXCEP_TXT ("AE_NOT_FOUND" ,"A requested entity is not found" ),EXCEP_TXT ("AE_NOT_EXIST" ,"A required entity does not exist" ),EXCEP_TXT ("AE_ALREADY_EXISTS" ,"An entity already exists" ), EXCEP_TXT("AE_TYPE", "The object type is incorrect"),
@@ -259,7 +260,8 @@ static const struct acpi_exception_info acpi_gbl_exception_names_env[] = {EXCEP_TXT ("AE_OWNER_ID_LIMIT" ,"There are no more Owner IDs available for ACPI tables or control methods" ),EXCEP_TXT ("AE_NOT_CONFIGURED" , -"The interface is not part of the current subsystem configuration" ) +"The interface is not part of the current subsystem configuration" ), +EXCEP_TXT ("AE_ACCESS" ,"Permission denied for the requested operation" ) }; static const struct acpi_exception_info acpi_gbl_exception_names_pgm[] = {
Signature · 1 line
1.7.10
PATCH 14/14 ACPICA: Update version to 20130823.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
Version 20130823.
--- include/acpi/acpixf.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
include/acpi/acpixf.h +1 -1
@@ -46,7 +46,7 @@/* Current ACPICA subsystem version in YYYYMMDD format */ -#define ACPI_CA_VERSION 0x20130725 +#define ACPI_CA_VERSION 0x20130823#include <acpi/acconfig.h> #include <acpi/actypes.h>
Signature · 1 line
1.7.10
PATCH 13/14 ACPICA: SCI Handlers: Update handler interface, eliminate unnecessary argument.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
The SCI interrupt number is not needed for the SCI handlers, and was just unnecessary overhead.
--- drivers/acpi/acpica/evsci.c | 4 +--- include/acpi/actypes.h | 2 +- 2 files changed, 2 insertions(+), 4 deletions(-)
drivers/acpi/acpica/evsci.c +1 -3
@@ -87,9 +87,7 @@ u32 acpi_ev_sci_dispatch(void)/* Invoke the installed handler (at interrupt level) */ - int_status |= sci_handler->address ((u32 )acpi_gbl_FADT. - sci_interrupt, - sci_handler->context); + int_status |= sci_handler->address (sci_handler->context); sci_handler = sci_handler->next; }
include/acpi/actypes.h +1 -1
@@ -951,7 +951,7 @@ typedef void* Various handlers and callback procedures */ typedef -u32 (*acpi_sci_handler ) (u32 interrupt_number,void *context); +u32 (*acpi_sci_handler ) (void *context);typedef void (*acpi_gbl_event_handler) (u32 event_type,
Signature · 1 line
1.7.10
Len Brown
reviewerRe: [PATCH 05/14] ACPICA: Debugger: Prevent possible command line buffer overflow.
In reply to Lv ZhengQuoted 17 lines · Lv Zheng
From: Bob Moore <[email protected]>
Increase the size of a couple of the debugger line buffers. ACPICA BZ 1037.
Signed-off-by Bob Moore <[email protected]>Signed-off-by Lv Zheng <[email protected]>--- drivers/acpi/acpica/acglobal.h | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-)drivers/acpi/acpica/acglobal.h +0
-0
drivers/acpi/acpica/acglobal.h ACPI_EXTERN u8 acpi_gbl_db_opt_tables; · line 454
Quoted 10 lines
drivers/acpi/acpica/acglobal.h +0
-7@@ -445,13 +445,6 @@ ACPI_EXTERN u8 acpi_gbl_db_opt_tables; ACPI_EXTERN u8 acpi_gbl_db_opt_stats; ACPI_EXTERN u8 acpi_gbl_db_opt_ini_methods; ACPI_EXTERN u8 acpi_gbl_db_opt_no_region_support; - -ACPI_EXTERN char *acpi_gbl_db_args[ACPI_DEBUGGER_MAX_ARGS]; -ACPI_EXTERN acpi_object_type acpi_gbl_db_arg_types[ACPI_DEBUGGER_MAX_ARGS]; -ACPI_EXTERN char acpi_gbl_db_line_buf[ACPI_DB_LINE_BUFFER_SIZE]; -ACPI_EXTERN char acpi_gbl_db_parsed_buf[ACPI_DB_LINE_BUFFER_SIZE]; -ACPI_EXTERN char acpi_gbl_db_scope_buf[80]; -ACPI_EXTERN char acpi_gbl_db_debug_filename[80];
If I read the code correctly, these are all inside #ifdefs that are not defined in kernel mode,
and so the kernel will not be impacted by this buffer growth from 80 to 512, and this patch is just to keep shared w/ acpiexec the same, yes?
thanks,
Len Brown, Intel Open Source Technology Center
Quoted 22 lines
drivers/acpi/acpica/acglobal.h +10
-1ACPI_EXTERN u8 acpi_gbl_db_output_to_file; ACPI_EXTERN char *acpi_gbl_db_buffer; ACPI_EXTERN char *acpi_gbl_db_filename;@@ -459,6 +452,16 @@ ACPI_EXTERN u32 acpi_gbl_db_debug_level; ACPI_EXTERN u32 acpi_gbl_db_console_debug_level; ACPI_EXTERN struct acpi_namespace_node *acpi_gbl_db_scope_node; +ACPI_EXTERN char *acpi_gbl_db_args[ACPI_DEBUGGER_MAX_ARGS]; +ACPI_EXTERN acpi_object_type acpi_gbl_db_arg_types[ACPI_DEBUGGER_MAX_ARGS]; + +/* These buffers should all be the same size */ + +ACPI_EXTERN char acpi_gbl_db_line_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_parsed_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_scope_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_debug_filename[ACPI_DB_LINE_BUFFER_SIZE]; +/* * Statistic globals */ --1.7.10
Len Brown
reviewerRe: [PATCH 06/14] ACPICA: Validate start object for acpi_walk_namespace.
In reply to Lv ZhengThis was filed an fixed very quickly.
Is it something we see in the field?
The patch is small, should we mark it for -stable?
thanks
Quoted 43 lines · Lv Zheng
From: Bob Moore <[email protected]>
Perform a sanity check on the start object to prevent problems later. ACPICA BZ 1025.
Signed-off-by Bob Moore <[email protected]>Signed-off-by Lv Zheng <[email protected]>--- drivers/acpi/acpica/nsxfeval.c | 8 ++++++++ 1 file changed, 8 insertions(+)drivers/acpi/acpica/nsxfeval.c +8
-1@@ -605,11 +605,19 @@ acpi_walk_namespace(acpi_object_type type, goto unlock_and_exit; } +/* Now we can validate the starting node */ + +if (!acpi_ns_validate_handle (start_object)) { + status =AE_BAD_PARAMETER ; + goto unlock_and_exit2; + } + status =acpi_ns_walk_namespace (type, start_object, max_depth,ACPI_NS_WALK_UNLOCK , descending_callback, ascending_callback, context, return_value); + unlock_and_exit2: (void )acpi_ut_release_mutex (ACPI_MTX_NAMESPACE ); unlock_and_exit: --1.7.10
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Signature · 1 line
Len Brown, Intel Open Source Technology Center
Len Brown
reviewerRe: [PATCH 07/14] ACPICA: Add support for host-installed SCI handlers.
In reply to Lv ZhengQuoted 9 lines · Lv Zheng
This change adds support to allow hosts to install System Control Interrupt handlers. Certain ACPI functionality requires the host to handle raw SCIs. For example, the "SCI Doorbell" that is defined for memory power state support requires the host device driver to handle SCIs to examine if the doorbell has been activated. Multiple SCI handlers can be installed to allow for future expansion. Debugger support is included.
Lv Zheng, Bob Moore. ACPICA BZ 1032.
Quoted 3 lines
Signed-off-by Lv Zheng <[email protected]>Signed-off-by Bob Moore <[email protected]>
The bug report references PCC -- is that _PCC events, or Processor Clocking Control?
We think we support both of those before this patch -- so how can PCC require this change?
thanks,
-Len
Len Brown
reviewerRe: [PATCH 11/14] ACPICA: Tables: Cleanup RSDP signature codes.
In reply to Lv ZhengQuoted 43 lines · Lv Zheng
This patch introduces new macors to handle RSDP signature and cleans up the affected codes. Lv Zheng.
Signed-off-by Lv Zheng <[email protected]>Signed-off-by Bob Moore <[email protected]>--- drivers/acpi/acpica/tbprint.c | 2 +- drivers/acpi/acpica/tbxfroot.c | 3 +-- include/acpi/actypes.h | 5 +++++ 3 files changed, 7 insertions(+), 3 deletions(-)drivers/acpi/acpica/tbprint.c +1
-1@@ -138,7 +138,7 @@ acpi_tb_print_table_header(acpi_physical_address address,ACPI_INFO ((AE_INFO ,"%4.4s %p %05X" , header->signature,ACPI_CAST_PTR (void, address), header->length)); - }else if (ACPI_COMPARE_NAME (header->signature,ACPI_SIG_RSDP )) { + }else if (ACPI_VALIDATE_RSDP_SIG (header->signature)) {/* RSDP has no common fields */ drivers/acpi/acpica/tbxfroot.c +1
-2@@ -68,8 +68,7 @@ acpi_status acpi_tb_validate_rsdp(struct acpi_table_rsdp *rsdp)* Note: Sometimes there exists more than one RSDP in memory; the valid * RSDP has a valid checksum, all others have an invalid checksum. */ -if (ACPI_STRNCMP ((char *)rsdp->signature,ACPI_SIG_RSDP , -sizeof (ACPI_SIG_RSDP ) -1 ) !=0 ) { +if (!ACPI_VALIDATE_RSDP_SIG (rsdp->signature)) {/* Nope, BAD Signature */ include/acpi/actypes.h +0
-0
include/acpi/actypes.h typedef u64 acpi_integer; · line 480
Quoted 7 lines
include/acpi/actypes.h +4
-0@@ -474,6 +474,11 @@ typedef u64 acpi_integer;#define ACPI_MOVE_NAME (dest,src) (ACPI_STRNCPY (ACPI_CAST_PTR (char, (dest)), ACPI_CAST_PTR (char, (src)), ACPI_NAME_SIZE))#endif +/* Support for the special RSDP signature (8 characters) */ + +#define ACPI_VALIDATE_RSDP_SIG (a) (!ACPI_STRNCMP (ACPI_CAST_PTR (char, (a)), ACPI_SIG_RSDP, 8)) +#define ACPI_MAKE_RSDP_SIG (dest) (ACPI_MEMCPY (ACPI_CAST_PTR (char, (dest)), ACPI_SIG_RSDP, 8))
is ACPI_MAKE_RSDP_SIGI(), defined, but not used?
If so, don't define it.
Len Brown, Intel Open Source Technology Center
this series is...
with satisfactory reply to the individual questions asked.
thanks,
Len Brown, Intel Open Source Technology Center
RE: [PATCH 07/14] ACPICA: Add support for host-installed SCI handlers.
In reply to Len BrownQuoted 21 lines
From: [email protected] [mailto:[email protected]] On Behalf Of Len Brown
Sent: Tuesday, September 17, 2013 8:48 AM
To: Zheng, Lv
Cc: Wysocki, Rafael J; Moore, Robert; Brown, Len; linux acpi
Subject: Re: [PATCH 07/14] ACPICA: Add support for host-installed SCI handlers.Quoted 9 lines · Lv Zheng
This change adds support to allow hosts to install System Control Interrupt handlers. Certain ACPI functionality requires the host to handle raw SCIs. For example, the "SCI Doorbell" that is defined for memory power state support requires the host device driver to handle SCIs to examine if the doorbell has been activated. Multiple SCI handlers can be installed to allow for future expansion. Debugger support is included.
Lv Zheng, Bob Moore. ACPICA BZ 1032.Quoted 3 lines
Signed-off-by Lv Zheng <[email protected]>Signed-off-by Bob Moore <[email protected]>
Quoted 4 lines
The bug report references PCC -- is that
_PCCevents, or Processor Clocking Control?
We think we support both of those before this patch -- so how can PCC require this change?
The PCC here refers to Platform Communications Channel (PCC) which is defined in the Chap 14 of the ACPI specification. It said that "the PCC completion check is not factored into the ACPICA in upstream kernel yet".
So the installable SCI handlers are expected.
Thanks
-Lv
Quoted 3 lines
thanks,
-Len
RE: [PATCH 11/14] ACPICA: Tables: Cleanup RSDP signature codes.
In reply to Len BrownQuoted 51 lines
From: [email protected] [mailto:[email protected]] On Behalf Of Len Brown
Sent: Tuesday, September 17, 2013 8:52 AMQuoted 43 lines · Lv Zheng
This patch introduces new macors to handle RSDP signature and cleans up the affected codes. Lv Zheng.
Signed-off-by Lv Zheng <[email protected]>Signed-off-by Bob Moore <[email protected]>--- drivers/acpi/acpica/tbprint.c | 2 +- drivers/acpi/acpica/tbxfroot.c | 3 +-- include/acpi/actypes.h | 5 +++++ 3 files changed, 7 insertions(+), 3 deletions(-)drivers/acpi/acpica/tbprint.c +1
-1@@ -138,7 +138,7 @@ acpi_tb_print_table_header(acpi_physical_address address,ACPI_INFO ((AE_INFO ,"%4.4s %p %05X" , header->signature,ACPI_CAST_PTR (void, address), header->length)); - }else if (ACPI_COMPARE_NAME (header->signature,ACPI_SIG_RSDP )) { + }else if (ACPI_VALIDATE_RSDP_SIG (header->signature)) {/* RSDP has no common fields */ drivers/acpi/acpica/tbxfroot.c +1
-2@@ -68,8 +68,7 @@ acpi_status acpi_tb_validate_rsdp(struct acpi_table_rsdp *rsdp)* Note: Sometimes there exists more than one RSDP in memory; the valid * RSDP has a valid checksum, all others have an invalid checksum. */ -if (ACPI_STRNCMP ((char *)rsdp->signature,ACPI_SIG_RSDP , -sizeof (ACPI_SIG_RSDP ) -1 ) !=0 ) { +if (!ACPI_VALIDATE_RSDP_SIG (rsdp->signature)) {/* Nope, BAD Signature */ include/acpi/actypes.h +0
-0include/acpi/actypes.h typedef u64 acpi_integer; · line 474
Quoted 1 line
include/acpi/actypes.h +0
-0@@ -474,6 +474,11 @@ typedef u64 acpi_integer;#define ACPI_MOVE_NAME (dest,src) (ACPI_STRNCPY (ACPI_CAST_PTR (char, (dest)), ACPI_CAST_PTR (char, (src)),
ACPI_NAME_SIZE))Quoted 2 lines
include/acpi/actypes.h +0
-0#endif
Quoted 7 lines
include/acpi/actypes.h typedef u64 acpi_integer; · line 480
Quoted 4 lines
include/acpi/actypes.h +4
-0+/* Support for the special RSDP signature (8 characters) */ + +#define ACPI_VALIDATE_RSDP_SIG (a) (!ACPI_STRNCMP (ACPI_CAST_PTR (char, (a)), ACPI_SIG_RSDP, 8)) +#define ACPI_MAKE_RSDP_SIG (dest) (ACPI_MEMCPY (ACPI_CAST_PTR (char, (dest)), ACPI_SIG_RSDP, 8))is
ACPI_MAKE_RSDP_SIGI(), defined, but not used?
If so, don't define it.
The macro is used by ACPICA user space utilities like AcpiDump. We've a plan to update the AcpiDump in the tools/acpi folder. So it's likely we'll see users of this macro in the kernel tree.
And, if we delete the piece of codes like this, then it will generate new source code differences between Linux and ACPICA, which will hurt the automation of the ACPICA release process. I'm sorry for that.
Thanks and best regards
-Lv
Quoted 2 lines
Len Brown, Intel Open Source Technology Center
RE: [PATCH 05/14] ACPICA: Debugger: Prevent possible command line buffer overflow.
In reply to Len BrownQuoted 28 lines
From: [email protected] [mailto:[email protected]] On Behalf Of Len Brown
Sent: Tuesday, September 17, 2013 8:43 AMQuoted 24 lines · Lv Zheng
From: Bob Moore <[email protected]>
Increase the size of a couple of the debugger line buffers. ACPICA BZ 1037.
Signed-off-by Bob Moore <[email protected]>Signed-off-by Lv Zheng <[email protected]>--- drivers/acpi/acpica/acglobal.h | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-)drivers/acpi/acpica/acglobal.h +0
-4@@ -445,13 +445,6 @@ ACPI_EXTERN u8 acpi_gbl_db_opt_tables; ACPI_EXTERN u8 acpi_gbl_db_opt_stats; ACPI_EXTERN u8 acpi_gbl_db_opt_ini_methods; ACPI_EXTERN u8 acpi_gbl_db_opt_no_region_support; - -ACPI_EXTERN char *acpi_gbl_db_args[ACPI_DEBUGGER_MAX_ARGS]; -ACPI_EXTERN acpi_object_type acpi_gbl_db_arg_types[ACPI_DEBUGGER_MAX_ARGS]; -ACPI_EXTERN char acpi_gbl_db_line_buf[ACPI_DB_LINE_BUFFER_SIZE];
Quoted 8 lines
drivers/acpi/acpica/acglobal.h ACPI_EXTERN u8 acpi_gbl_db_opt_tables; · line 454
Quoted 3 lines
drivers/acpi/acpica/acglobal.h +0
-3-ACPI_EXTERN char acpi_gbl_db_parsed_buf[ACPI_DB_LINE_BUFFER_SIZE]; -ACPI_EXTERN char acpi_gbl_db_scope_buf[80]; -ACPI_EXTERN char acpi_gbl_db_debug_filename[80];If I read the code correctly, these are all inside #ifdefs that are not defined in kernel mode,
and so the kernel will not be impacted by this buffer growth from 80 to 512, and this patch is just to keep shared w/ acpiexec the same, yes?
Yes, the codes belongs to the AML debugger which is defined by ACPI specification, implemented in the ACPICA but not shipped in the Linux kernel. We've a plan to port the debugger/disassmebler to the kernel.
So it's likely we'll see users of such globals in the kernel tree.
Currently, kernel behaviors are not affected by this patch.
And, if we delete the piece of codes like this, then it will generate new source code differences between Linux and ACPICA, which will hurt the automation of the ACPICA release process. I'm sorry for that.
Thanks and best regards
-Lv
Quoted 30 lines
thanks,
Len Brown, Intel Open Source Technology CenterQuoted 21 lines
drivers/acpi/acpica/acglobal.h +10
-1ACPI_EXTERN u8 acpi_gbl_db_output_to_file; ACPI_EXTERN char *acpi_gbl_db_buffer; ACPI_EXTERN char *acpi_gbl_db_filename;@@ -459,6 +452,16 @@ ACPI_EXTERN u32 acpi_gbl_db_debug_level; ACPI_EXTERN u32 acpi_gbl_db_console_debug_level; ACPI_EXTERN struct acpi_namespace_node *acpi_gbl_db_scope_node; +ACPI_EXTERN char *acpi_gbl_db_args[ACPI_DEBUGGER_MAX_ARGS]; +ACPI_EXTERN acpi_object_type acpi_gbl_db_arg_types[ACPI_DEBUGGER_MAX_ARGS]; + +/* These buffers should all be the same size */ + +ACPI_EXTERN char acpi_gbl_db_line_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_parsed_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_scope_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_debug_filename[ACPI_DB_LINE_BUFFER_SIZE]; +/* * Statistic globals */ --Quoted 1 line
1.7.10
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
RE: [PATCH 06/14] ACPICA: Validate start object for acpi_walk_namespace.
In reply to Len BrownHi, Bob
Could you help to answer this?
Thanks in advance.
Best regards
-Lv
Quoted 62 lines
From: [email protected] [mailto:[email protected]] On Behalf Of Len Brown
Sent: Tuesday, September 17, 2013 8:44 AMThis was filed an fixed very quickly.
Is it something we see in the field?
The patch is small, should we mark it for -stable?thanks
Quoted 39 lines · Lv Zheng
From: Bob Moore <[email protected]>
Perform a sanity check on the start object to prevent problems later. ACPICA BZ 1025.
Signed-off-by Bob Moore <[email protected]>Signed-off-by Lv Zheng <[email protected]>--- drivers/acpi/acpica/nsxfeval.c | 8 ++++++++ 1 file changed, 8 insertions(+)drivers/acpi/acpica/nsxfeval.c +8
-1@@ -605,11 +605,19 @@ acpi_walk_namespace(acpi_object_type type, goto unlock_and_exit; } +/* Now we can validate the starting node */ + +if (!acpi_ns_validate_handle (start_object)) { + status =AE_BAD_PARAMETER ; + goto unlock_and_exit2; + } + status =acpi_ns_walk_namespace (type, start_object, max_depth,ACPI_NS_WALK_UNLOCK , descending_callback, ascending_callback, context, return_value); + unlock_and_exit2: (void )acpi_ut_release_mutex (ACPI_MTX_NAMESPACE ); unlock_and_exit: --1.7.10
Quoted 4 lines
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html--
Len Brown, Intel Open Source Technology Center
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
RE: [PATCH 06/14] ACPICA: Validate start object for acpi_walk_namespace.
Just adds a sanity check. I don't think we see it, otherwise the caller would fault dead in its tracks.
Quoted 80 lines
-----Original Message-----
From: Zheng, Lv
Sent: Monday, September 16, 2013 6:26 PM
To: Len Brown
Cc: Wysocki, Rafael J; Moore, Robert; Brown, Len; linux acpi
Subject: RE: [PATCH 06/14] ACPICA: Validate start object foracpi_walk_namespace.Hi, Bob
Could you help to answer this?
Thanks in advance.Best regards
-LvQuoted 31 lines
From: [email protected] [mailto:[email protected]] On Behalf Of Len Brown
Sent: Tuesday, September 17, 2013 8:44 AMThis was filed an fixed very quickly.
Is it something we see in the field?
The patch is small, should we mark it for -stable?thanks
Quoted 20 lines · Lv Zheng
From: Bob Moore <[email protected]>
Perform a sanity check on the start object to prevent problems later. ACPICA BZ 1025.
Signed-off-by Bob Moore <[email protected]>Signed-off-by Lv Zheng <[email protected]>--- drivers/acpi/acpica/nsxfeval.c | 8 ++++++++ 1 file changed, 8 insertions(+)(quoted diff) +0
-0
b/drivers/acpi/acpica/nsxfeval.cindex b38b4b0..481a6b4 100644drivers/acpi/acpica/nsxfeval.c +0
-0@@ -605,11 +605,19 @@ acpi_walk_namespace(acpi_object_type type, goto unlock_and_exit; }Quoted 10 lines
drivers/acpi/acpica/nsxfeval.c acpi_walk_namespace(acpi_object_type type,
Quoted 10 lines
drivers/acpi/acpica/nsxfeval.c +7
-0+/* Now we can validate the starting node */ + +if (!acpi_ns_validate_handle (start_object)) { + status =AE_BAD_PARAMETER ; + goto unlock_and_exit2; + } + status = acpi_ns_walk_namespace(type, start_object, max_depth, ACPI_NS_WALK_UNLOCK, descending_callback,
ascending_callback,Quoted 22 lines
Quoted 9 lines
drivers/acpi/acpica/nsxfeval.c +1
-1context, return_value); + unlock_and_exit2: (void )acpi_ut_release_mutex (ACPI_MTX_NAMESPACE ); unlock_and_exit: --1.7.10
Quoted 4 lines
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html--
Len Brown, Intel Open Source Technology Center
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
RE: [PATCH 07/14] ACPICA: Add support for host-installed SCI handlers.
In reply to Len BrownQuoted 23 lines
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Len Brown
Sent: Monday, September 16, 2013 5:48 PM
To: Zheng, Lv
Cc: Wysocki, Rafael J; Moore, Robert; Brown, Len; linux acpi
Subject: Re: [PATCH 07/14] ACPICA: Add support for host-installed SCI handlers.Quoted 9 lines · Lv Zheng
This change adds support to allow hosts to install System Control Interrupt handlers. Certain ACPI functionality requires the host to handle raw SCIs. For example, the "SCI Doorbell" that is defined for memory power state support requires the host device driver to handle SCIs to examine if the doorbell has been activated. Multiple SCI handlers can be installed to allow for future expansion. Debugger support is included.
Lv Zheng, Bob Moore. ACPICA BZ 1032.Quoted 3 lines
Signed-off-by Lv Zheng <[email protected]>Signed-off-by Bob Moore <[email protected]>
Quoted 4 lines
The bug report references PCC -- is that
_PCCevents, or Processor Clocking Control?
We think we support both of those before this patch -- so how can PCC require this change?
Platform Communication Channel, via the MPST table (new for ACPI 5.0)
When the PCC subchannel responds to the host, it issues an SCI and the host must probe the subchannel for channel status.
Quoted 3 lines
thanks,
-Len
RE: [PATCH 05/14] ACPICA: Debugger: Prevent possible command line buffer overflow.
In reply to Zheng, LvQuoted 1 line
Currently, kernel behaviors are not affected by this patch.
This is exactly what you want to write in the commit message.
Imagine, for a minute, a distro scanning thousands of patches and reading the title of this one -- they would be alarmed by the phrase "buffer overflow" on a kernel patch -- so you need to tell them not to be alarmed.
thanks,
-Len
RE: [PATCH 06/14] ACPICA: Validate start object for acpi_walk_namespace.
In reply to Moore, RobertQuoted 1 line
Just adds a sanity check.
Ideal if commit messages answer the question:
"Is this patch important?"
"If yes, what bad things happen if I don't have it in my kernel?"
Remember, the commit messages have customers.
Nice to have a BZ referenced, but this BZ entry
had no such information, and
"prevent problems later" didn't answer that key question either.
thanks,
-Len
RE: [PATCH 06/14] ACPICA: Validate start object for acpi_walk_namespace.
I would also say something like:
"Adds additional input parameter validation"
Quoted 21 lines
-----Original Message-----
From: Brown, Len
Sent: Tuesday, September 17, 2013 12:46 PM
To: Moore, Robert; Zheng, Lv
Cc: Wysocki, Rafael J; linux acpi
Subject: RE: [PATCH 06/14] ACPICA: Validate start object foracpi_walk_namespace.Quoted 1 line
Just adds a sanity check.
Ideal if commit messages answer the question:
"Is this patch important?"
"If yes, what bad things happen if I don't have it in my kernel?"Remember, the commit messages have customers.
Nice to have a BZ referenced, but this BZ entry had no such information, and "prevent problems later" didn't answer that key question either.thanks,
-Len
RE: [PATCH 05/14] ACPICA: Debugger: Prevent possible command line buffer overflow.
In reply to Brown, LenOK, I'll add this line to the subject.
Thanks and best regards
-Lv
Quoted 14 lines
From: Brown, Len
Sent: Wednesday, September 18, 2013 3:31 AMQuoted 1 line
Currently, kernel behaviors are not affected by this patch.
This is exactly what you want to write in the commit message.
Imagine, for a minute, a distro scanning thousands of patches and reading the title of this one -- they would be alarmed by the phrase "buffer overflow" on a kernel patch -- so you need to tell them not to be alarmed.
thanks,
-Len
RE: [PATCH 06/14] ACPICA: Validate start object for acpi_walk_namespace.
In reply to Brown, LenOK, I'll clarify that currently kernel is not suffering from this issue in the patch description.
Thanks and best regards
-Lv
Quoted 21 lines
-----Original Message-----
From: Brown, Len
Sent: Wednesday, September 18, 2013 3:46 AM
To: Moore, Robert; Zheng, Lv
Cc: Wysocki, Rafael J; linux acpi
Subject: RE: [PATCH 06/14] ACPICA: Validate start object foracpi_walk_namespace.Quoted 1 line
Just adds a sanity check.
Ideal if commit messages answer the question:
"Is this patch important?"
"If yes, what bad things happen if I don't have it in my kernel?"Remember, the commit messages have customers.
Nice to have a BZ referenced, but this BZ entry
had no such information, and
"prevent problems later" didn't answer that key question either.thanks,
-Len
v2 cover ACPICA 20130823 Release
In reply to Lv ZhengThe 20130823 ACPICA kernel resident system updates is linuxized based on the pm/linux-next branch.
The patch set has passed a build/boot test on the following machines:
Dell Inspiron Mini 1010 (i386)
HP Compaq 8200 Elite SFF PC (x86-64)
With the following kernel configuration items enabled:
CONFIG_ACPI_AC
CONFIG_ACPI_BATTERY
CONFIG_ACPI_NUMA
CONFIG_ACPI_PROCESSOR
CONFIG_ACPI_THERMAL
CONFIG_ACPI_VIDEO
CONFIG_CPU_FREQ
CONFIG_I2C_SCMI
CONFIG_SENSORS_ACPI_POWER
CONFIG_PANASONIC_LAPTOP
CONFIG_XO15_EBOOK
CONFIG_DMAR_TABLE
CONFIG_INTEL_IOMMU
CONFIG_IRQ_REMAP
Bjorn Helgaas (1):
ACPICA: acpi_read: On error, do not modify the return value target
location.
Bob Moore (7):
ACPICA: Improve error message for "too many parent prefixes"
condition.
ACPICA: Debug output: small formatting update, no functional change.
ACPICA: Debugger: Prevent possible command line buffer overflow,
kernel behavior is not affected.
ACPICA: Validate start object for acpi_walk_namespace.
ACPICA: Debugger: Add new command to display full namespace
pathnames.
ACPICA: SCI Handlers: Update handler interface, eliminate unnecessary
argument.
ACPICA: Update version to 20130823.
David E. Box (1):
ACPICA: Operation region support: Never free the handler "context"
pointer.
Lv Zheng (4):
ACPICA: Add support for host-installed SCI handlers.
ACPICA: Tables: Cleanup table checksum verification code.
ACPICA: Tables: Cleanup RSDP signature codes.
ACPICA: Cleanup exception codes.
Naresh Bhat (1):
ACPICA: Linux-specific header: enable "aarch64" 64-bit build.
drivers/acpi/acpica/acdebug.h | 5 +- drivers/acpi/acpica/acevents.h | 6 +- drivers/acpi/acpica/acglobal.h | 18 +++-- drivers/acpi/acpica/aclocal.h | 8 +++ drivers/acpi/acpica/acnamesp.h | 6 ++ drivers/acpi/acpica/evgpeutil.c | 2 +- drivers/acpi/acpica/evmisc.c | 14 ++-- drivers/acpi/acpica/evregion.c | 23 ++++--- drivers/acpi/acpica/evsci.c | 78 ++++++++++++++++++++-- drivers/acpi/acpica/evxface.c | 139 +++++++++++++++++++++++++++++++++++++++ drivers/acpi/acpica/hwxface.c | 26 ++++---- drivers/acpi/acpica/nsaccess.c | 5 +- drivers/acpi/acpica/nsdump.c | 134 +++++++++++++++++++++++++++++++++++++ drivers/acpi/acpica/nsxfeval.c | 8 +++ drivers/acpi/acpica/tbinstal.c | 12 +--- drivers/acpi/acpica/tbprint.c | 12 +++- drivers/acpi/acpica/tbxfroot.c | 3 +- drivers/acpi/acpica/utdebug.c | 2 +- drivers/acpi/acpica/utglobal.c | 3 +- include/acpi/acexcep.h | 8 ++- include/acpi/acpixf.h | 15 +++-- include/acpi/actypes.h | 8 +++ include/acpi/platform/aclinux.h | 2 +- 23 files changed, 465 insertions(+), 72 deletions(-)
Signature · 1 line
1.7.10
PATCH v2 1/14 ACPICA: Improve error message for "too many parent prefixes" condition.
From: Bob Moore <[email protected]>
Emit the full offending pathname in the error message.
--- drivers/acpi/acpica/nsaccess.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/acpi/acpica/nsaccess.c +3 -2
@@ -424,8 +424,9 @@ acpi_ns_lookup(union acpi_generic_state *scope_info,/* Current scope has no parent scope */ ACPI_ERROR ((AE_INFO , -"ACPI path has too many parent prefixes (^) " -"- reached beyond root node" )); +"%s: Path has too many parent prefixes (^) " +"- reached beyond root node" , + pathname));return_ACPI_STATUS (AE_NOT_FOUND ); } }
Signature · 1 line
1.7.10
PATCH v2 2/14 ACPICA: acpi_read: On error, do not modify the return value target location.
In reply to Lv ZhengFrom: Bjorn Helgaas <[email protected]>
If an error happens in the middle of a split 32/32 64-bit I/O operation, do not modify the target of the return value pointer. Makes the code consistent with the rest of ACPICA. Bjorn Helgaas.
--- drivers/acpi/acpica/hwxface.c | 26 ++++++++++++++------------ 1 file changed, 14 insertions(+), 12 deletions(-)
drivers/acpi/acpica/hwxface.c +14 -12
@@ -119,7 +119,8 @@ ACPI_EXPORT_SYMBOL(acpi_reset)******************************************************************************/ acpi_status acpi_read (u64 *return_value,struct acpi_generic_address *reg) { -u32 value; +u32 value_lo; +u32 value_hi;u32 width;u64 address;acpi_status status;
@@ -137,13 +138,8 @@ acpi_status acpi_read(u64 *return_value, struct acpi_generic_address *reg)return (status); } -/* Initialize entire 64-bit return value to zero */ - - *return_value =0 ; - value =0 ; -/* -* Two address spaces supported: Memory or IO. PCI_Config is +* Two address spaces supported: Memory or I/O. PCI_Config is * not supported here because the GAS structure is insufficient */ if (reg->space_id ==ACPI_ADR_SPACE_SYSTEM_MEMORY ) {
@@ -155,29 +151,35 @@ acpi_status acpi_read(u64 *return_value, struct acpi_generic_address *reg) } }else {/* ACPI_ADR_SPACE_SYSTEM_IO, validated earlier */ + value_lo =0 ; + value_hi =0 ; + width = reg->bit_width;if (width ==64 ) { width =32 ;/* Break into two 32-bit transfers */ } status =acpi_hw_read_port ((acpi_io_address ) - address, &value, width); + address, &value_lo, width);if (ACPI_FAILURE (status)) {return (status); } - *return_value = value;if (reg->bit_width ==64 ) {/* Read the top 32 bits */ status =acpi_hw_read_port ((acpi_io_address) - (address +4 ), &value,32 ); + (address +4 ), &value_hi, +32 );if (ACPI_FAILURE (status)) {return (status); } - *return_value |= ((u64 )value <<32 ); } + +/* Set the return value only if status is AE_OK */ + + *return_value = (value_lo | ((u64 )value_hi <<32 )); } ACPI_DEBUG_PRINT((ACPI_DB_IO,
@@ -186,7 +188,7 @@ acpi_status acpi_read(u64 *return_value, struct acpi_generic_address *reg)ACPI_FORMAT_UINT64 (address), acpi_ut_get_region_name(reg->space_id))); -return (status); +return (AE_OK ); }ACPI_EXPORT_SYMBOL (acpi_read)
Signature · 1 line
1.7.10
PATCH v2 3/14 ACPICA: Debug output: small formatting update, no functional change.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
Allow for longer filenames in the module name output during trace operations.
--- drivers/acpi/acpica/utdebug.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/acpi/acpica/utdebug.c +1 -1
@@ -190,7 +190,7 @@ acpi_debug_print(u32 requested_debug_level,* Display the module name, current line number, thread ID (if requested), * current procedure nesting level, and the current procedure name */ -acpi_os_printf ("%8s-%04ld " , module_name, line_number); +acpi_os_printf ("%9s-%04ld " , module_name, line_number);if (ACPI_LV_THREADS & acpi_dbg_level) {acpi_os_printf ("[%u] " , (u32 )thread_id);
Signature · 1 line
1.7.10
PATCH v2 4/14 ACPICA: Linux-specific header: enable "aarch64" 64-bit build.
In reply to Lv ZhengFrom: Naresh Bhat <[email protected]>
Add support for the __aarch64__ define for 64-bit builds.
--- include/acpi/platform/aclinux.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
include/acpi/platform/aclinux.h +1 -1
@@ -87,7 +87,7 @@#define ACPI_FLUSH_CPU_CACHE ()#define ACPI_CAST_PTHREAD_T (pthread) ((acpi_thread_id) (pthread)) -#if defined(__ia64__) || defined(__x86_64__) +#if defined(__ia64__) || defined(__x86_64__) || defined(__aarch64__)#define ACPI_MACHINE_WIDTH 64#define COMPILER_DEPENDENT_INT64 long#define COMPILER_DEPENDENT_UINT64 unsigned long
Signature · 1 line
1.7.10
PATCH v2 5/14 ACPICA: Debugger: Prevent possible command line buffer overflow, kernel behavior is not affected.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
Increase the size of a couple of the debugger line buffers. ACPICA BZ 1037.
The debugger related code is not in the kernel so the behavior of the kernel is not affected.
--- drivers/acpi/acpica/acglobal.h | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-)
drivers/acpi/acpica/acglobal.h +10 -7
@@ -445,13 +445,6 @@ ACPI_EXTERN u8 acpi_gbl_db_opt_tables; ACPI_EXTERN u8 acpi_gbl_db_opt_stats; ACPI_EXTERN u8 acpi_gbl_db_opt_ini_methods; ACPI_EXTERN u8 acpi_gbl_db_opt_no_region_support; - -ACPI_EXTERN char *acpi_gbl_db_args[ACPI_DEBUGGER_MAX_ARGS]; -ACPI_EXTERN acpi_object_type acpi_gbl_db_arg_types[ACPI_DEBUGGER_MAX_ARGS]; -ACPI_EXTERN char acpi_gbl_db_line_buf[ACPI_DB_LINE_BUFFER_SIZE]; -ACPI_EXTERN char acpi_gbl_db_parsed_buf[ACPI_DB_LINE_BUFFER_SIZE]; -ACPI_EXTERN char acpi_gbl_db_scope_buf[80]; -ACPI_EXTERN char acpi_gbl_db_debug_filename[80]; ACPI_EXTERN u8 acpi_gbl_db_output_to_file; ACPI_EXTERN char *acpi_gbl_db_buffer; ACPI_EXTERN char *acpi_gbl_db_filename;
@@ -459,6 +452,16 @@ ACPI_EXTERN u32 acpi_gbl_db_debug_level; ACPI_EXTERN u32 acpi_gbl_db_console_debug_level; ACPI_EXTERN struct acpi_namespace_node *acpi_gbl_db_scope_node; +ACPI_EXTERN char *acpi_gbl_db_args[ACPI_DEBUGGER_MAX_ARGS]; +ACPI_EXTERN acpi_object_type acpi_gbl_db_arg_types[ACPI_DEBUGGER_MAX_ARGS]; + +/* These buffers should all be the same size */ + +ACPI_EXTERN char acpi_gbl_db_line_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_parsed_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_scope_buf[ACPI_DB_LINE_BUFFER_SIZE]; +ACPI_EXTERN char acpi_gbl_db_debug_filename[ACPI_DB_LINE_BUFFER_SIZE]; +/* * Statistic globals */
Signature · 1 line
1.7.10
PATCH v2 6/14 ACPICA: Validate start object for acpi_walk_namespace.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
Perform a sanity check on the start object to prevent problems later. ACPICA BZ 1025.
This patch only adds additional input parameter validation, no actual kernel suffering has been discovered.
--- drivers/acpi/acpica/nsxfeval.c | 8 ++++++++ 1 file changed, 8 insertions(+)
drivers/acpi/acpica/nsxfeval.c +8 -0
@@ -605,11 +605,19 @@ acpi_walk_namespace(acpi_object_type type, goto unlock_and_exit; } +/* Now we can validate the starting node */ + +if (!acpi_ns_validate_handle (start_object)) { + status =AE_BAD_PARAMETER ; + goto unlock_and_exit2; + } + status =acpi_ns_walk_namespace (type, start_object, max_depth,ACPI_NS_WALK_UNLOCK , descending_callback, ascending_callback, context, return_value); + unlock_and_exit2: (void )acpi_ut_release_mutex (ACPI_MTX_NAMESPACE ); unlock_and_exit:
Signature · 1 line
1.7.10
PATCH v2 7/14 ACPICA: Add support for host-installed SCI handlers.
In reply to Lv ZhengThis change adds support to allow hosts to install System Control Interrupt handlers. Certain ACPI functionality requires the host to handle raw SCIs. For example, the "SCI Doorbell" that is defined for memory power state support requires the host device driver to handle SCIs to examine if the doorbell has been activated. Multiple SCI handlers can be installed to allow for future expansion. Debugger support is included.
Lv Zheng, Bob Moore. ACPICA BZ 1032.
Bug summary:
It is reported when the PCC (Platform Communication Channel, via MPST table, defined in ACPI specification 5.0) subchannel responds to the host, it issues an SCI and the host must probe the subchannel for channel status.
--- drivers/acpi/acpica/acdebug.h | 3 +- drivers/acpi/acpica/acevents.h | 6 +- drivers/acpi/acpica/acglobal.h | 1 + drivers/acpi/acpica/aclocal.h | 8 +++ drivers/acpi/acpica/evgpeutil.c | 2 +- drivers/acpi/acpica/evmisc.c | 14 ++-- drivers/acpi/acpica/evsci.c | 80 ++++++++++++++++++++-- drivers/acpi/acpica/evxface.c | 139 +++++++++++++++++++++++++++++++++++++++ drivers/acpi/acpica/utglobal.c | 3 +- include/acpi/acpixf.h | 13 +++- include/acpi/actypes.h | 3 + 11 files changed, 249 insertions(+), 23 deletions(-)
drivers/acpi/acpica/acdebug.h +2 -1
@@ -113,11 +113,12 @@ void acpi_db_display_handlers(void); ACPI_HW_DEPENDENT_RETURN_VOID(void acpi_db_generate_gpe(char *gpe_arg,char *block_arg)) + ACPI_HW_DEPENDENT_RETURN_VOID(void acpi_db_generate_sci(void)) /* * dbconvert - miscellaneous conversion routines */ - acpi_status acpi_db_hex_char_to_value(int hex_char, u8 *return_value); +acpi_status acpi_db_hex_char_to_value(int hex_char, u8 *return_value); acpi_status acpi_db_convert_to_package(char *string, union acpi_object *object);
drivers/acpi/acpica/acevents.h +3 -3
@@ -242,11 +242,11 @@ acpi_ev_initialize_region(union acpi_operand_object *region_obj,*/ u32 ACPI_SYSTEM_XFACE acpi_ev_gpe_xrupt_handler(void *context); -u32 acpi_ev_install_sci_handler (void ); +u32 acpi_ev_sci_dispatch (void ); -acpi_status acpi_ev_remove_sci_handler (void ); +u32 acpi_ev_install_sci_handler (void ); -u32 acpi_ev_initialize_SCI (u32 program_SCI); +acpi_status acpi_ev_remove_all_sci_handlers (void ); ACPI_HW_DEPENDENT_RETURN_VOID(void acpi_ev_terminate(void))#endif /* __ACEVENTS_H__ */
drivers/acpi/acpica/acglobal.h +1 -0
@@ -269,6 +269,7 @@ ACPI_EXTERN acpi_table_handler acpi_gbl_table_handler; ACPI_EXTERN void *acpi_gbl_table_handler_context; ACPI_EXTERN struct acpi_walk_state *acpi_gbl_breakpoint_walk; ACPI_EXTERN acpi_interface_handler acpi_gbl_interface_handler; +ACPI_EXTERN struct acpi_sci_handler_info *acpi_gbl_sci_handler_list;/* Owner ID support */
drivers/acpi/acpica/aclocal.h +8 -0
@@ -398,6 +398,14 @@ struct acpi_simple_repair_info {* ****************************************************************************/ +/* Dispatch info for each host-installed SCI handler */ + +struct acpi_sci_handler_info { +struct acpi_sci_handler_info *next; +acpi_sci_handler address;/* Address of handler */ +void *context;/* Context to be passed to handler */ +}; +/* Dispatch info for each GPE -- either a method or handler, cannot be both */ struct acpi_gpe_handler_info {
drivers/acpi/acpica/evgpeutil.c +1 -1
@@ -196,7 +196,7 @@ acpi_ev_get_gpe_device(struct acpi_gpe_xrupt_info *gpe_xrupt_info,* * FUNCTION: acpi_ev_get_gpe_xrupt_block * -* PARAMETERS: interrupt_number - Interrupt for a GPE block +* PARAMETERS: interrupt_number - Interrupt for a GPE block * * RETURN: A GPE interrupt block *
drivers/acpi/acpica/evmisc.c +7 -7
@@ -264,13 +264,6 @@ void acpi_ev_terminate(void) status =acpi_ev_walk_gpe_list (acpi_hw_disable_gpe_block,NULL ); -/* Remove SCI handler */ - - status =acpi_ev_remove_sci_handler (); -if (ACPI_FAILURE (status)) { -ACPI_ERROR ((AE_INFO ,"Could not remove SCI handler" )); - } - status =acpi_ev_remove_global_lock_handler ();if (ACPI_FAILURE (status)) { ACPI_ERROR((AE_INFO,
@@ -280,6 +273,13 @@ void acpi_ev_terminate(void) acpi_gbl_events_initialized = FALSE; } +/* Remove SCI handlers */ + + status =acpi_ev_remove_all_sci_handlers (); +if (ACPI_FAILURE (status)) { +ACPI_ERROR ((AE_INFO ,"Could not remove SCI handler" )); + } +/* Deallocate all handler objects installed within GPE info structs */ status =acpi_ev_walk_gpe_list (acpi_ev_delete_gpe_handlers,NULL );
drivers/acpi/acpica/evsci.c +73 -7
131 lines not on this page · raw message
drivers/acpi/acpica/evxface.c +139 -0
151 lines not on this page · raw message
drivers/acpi/acpica/utglobal.c +2 -1
@@ -291,7 +291,7 @@ acpi_status acpi_ut_init_globals(void)#if (!ACPI_REDUCED_HARDWARE ) -/* GPE support */ +/* GPE/SCI support */ acpi_gbl_all_gpes_initialized = FALSE; acpi_gbl_gpe_xrupt_list_head =NULL ;
@@ -300,6 +300,7 @@ acpi_status acpi_ut_init_globals(void) acpi_current_gpe_count =0 ; acpi_gbl_global_event_handler =NULL ; + acpi_gbl_sci_handler_list =NULL ;#endif /* !ACPI_REDUCED_HARDWARE */
include/acpi/acpixf.h +10 -3
@@ -280,9 +280,16 @@ acpi_status acpi_install_initialization_handler(acpi_init_handler handler, u32 function); ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status - acpi_install_global_event_handler - (acpi_gbl_event_handler handler, void *context)) - + acpi_install_sci_handler(acpi_sci_handler + address, + void *context)) +ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status + acpi_remove_sci_handler(acpi_sci_handler + address)) +ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status + acpi_install_global_event_handler + (acpi_gbl_event_handler handler, + void *context)) ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status acpi_install_fixed_event_handler(u32 acpi_event,
include/acpi/actypes.h +3 -0
@@ -946,6 +946,9 @@ typedef void* Various handlers and callback procedures */ typedef +u32 (*acpi_sci_handler ) (u32 interrupt_number,void *context); + +typedef void (*acpi_gbl_event_handler ) (u32 event_type,acpi_handle device,u32 event_number,void *context);
Signature · 1 line
1.7.10
PATCH v2 8/14 ACPICA: Operation region support: Never free the handler "context" pointer.
In reply to Lv ZhengFrom: "David E. Box" <[email protected]>
This change removes some dangerous code that attempts to free the handler context pointer in some (rare) circumstances. The owner of the handler owns this pointer and the ACPICA code should never touch it. Although not seen to be an issue in any kernel, it did show up as a problem under AcpiExec. Also, set the internal storage field for the context pointer to zero when the region is deactivated, simply for sanity. David Box.
--- drivers/acpi/acpica/evregion.c | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-)
drivers/acpi/acpica/evregion.c +13 -10
@@ -217,16 +217,11 @@ acpi_ev_address_space_dispatch(union acpi_operand_object *region_obj,if (!(region_obj->region.flags &AOPOBJ_SETUP_COMPLETE )) { region_obj->region.flags |=AOPOBJ_SETUP_COMPLETE ; -if (region_obj2->extra.region_context) { - -/* The handler for this region was already installed */ - -ACPI_FREE (region_context); - }else { -/* -* Save the returned context for use in all accesses to -* this particular region -*/ +/* +* Save the returned context for use in all accesses to +* the handler for this particular region +*/ +if (!(region_obj2->extra.region_context)) { region_obj2->extra.region_context = region_context; }
@@ -402,6 +397,14 @@ acpi_ev_detach_region(union acpi_operand_object *region_obj, handler_obj->address_space. context, region_context); +/* +* region_context should have been released by the deactivate +* operation. We don't need access to it anymore here. +*/ +if (region_context) { + *region_context =NULL ; + } +/* Init routine may fail, Just ignore errors */ if (ACPI_FAILURE (status)) {
Signature · 1 line
1.7.10
PATCH v2 9/14 ACPICA: Debugger: Add new command to display full namespace pathnames.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
Paths command displays the full pathname and object type for the entire namespace. Alternative to the Namespace command.
--- drivers/acpi/acpica/acdebug.h | 2 + drivers/acpi/acpica/acnamesp.h | 6 ++ drivers/acpi/acpica/nsdump.c | 134 ++++++++++++++++++++++++++++++++++++++++ 3 files changed, 142 insertions(+)
drivers/acpi/acpica/acdebug.h +2 -0
@@ -155,6 +155,8 @@ void acpi_db_set_scope(char *name);void acpi_db_dump_namespace (char *start_arg,char *depth_arg); +void acpi_db_dump_namespace_paths (void ); +void acpi_db_dump_namespace_by_owner (char *owner_arg,char *depth_arg);acpi_status acpi_db_find_name_in_namespace (char *name_arg);
drivers/acpi/acpica/acnamesp.h +6 -0
@@ -213,6 +213,12 @@ acpi_ns_dump_objects(acpi_object_type type,u8 display_type, u32 max_depth, acpi_owner_id owner_id, acpi_handle start_handle); + +void +acpi_ns_dump_object_paths (acpi_object_type type, +u8 display_type, +u32 max_depth, +acpi_owner_id owner_id,acpi_handle start_handle);#endif /* ACPI_FUTURE_USAGE */ /*
drivers/acpi/acpica/nsdump.c +134 -0
146 lines not on this page · raw message
Signature · 1 line
1.7.10
PATCH v2 10/14 ACPICA: Tables: Cleanup table checksum verification code.
In reply to Lv ZhengThis patch reduces code redundancy by moving the FACS/S3PT checksum verification skip logic into acpi_tb_verify_checksum() and other calls of this function also get benefit from this change. Lv Zheng.
--- drivers/acpi/acpica/tbinstal.c | 12 +++--------- drivers/acpi/acpica/tbprint.c | 10 ++++++++++ 2 files changed, 13 insertions(+), 9 deletions(-)
drivers/acpi/acpica/tbinstal.c +3 -9
@@ -80,16 +80,10 @@ acpi_status acpi_tb_verify_table(struct acpi_table_desc *table_desc) } } -/* FACS is the odd table, has no standard ACPI header and no checksum */ +/* Always calculate checksum, ignore bad checksum if requested */ -if (!ACPI_COMPARE_NAME (&table_desc->signature,ACPI_SIG_FACS )) { - -/* Always calculate checksum, ignore bad checksum if requested */ - - status = -acpi_tb_verify_checksum (table_desc->pointer, - table_desc->length); - } + status = +acpi_tb_verify_checksum (table_desc->pointer, table_desc->length);return_ACPI_STATUS (status); }
drivers/acpi/acpica/tbprint.c +10 -0
@@ -190,6 +190,16 @@ acpi_status acpi_tb_verify_checksum(struct acpi_table_header *table, u32 length) {u8 checksum; +/* +* FACS/S3PT: +* They are the odd tables, have no standard ACPI header and no checksum +*/ + +if (ACPI_COMPARE_NAME (table->signature,ACPI_SIG_S3PT ) || +ACPI_COMPARE_NAME (table->signature,ACPI_SIG_FACS )) { +return (AE_OK ); + } +/* Compute the checksum on the table */ checksum =acpi_tb_checksum (ACPI_CAST_PTR (u8, table), length);
Signature · 1 line
1.7.10
PATCH v2 11/14 ACPICA: Tables: Cleanup RSDP signature codes.
In reply to Lv ZhengThis patch introduces new macors to handle RSDP signature and cleans up the affected codes. Lv Zheng.
Some updates are only used for ACPICA utilities which are not shipped in the kernel yet.
--- drivers/acpi/acpica/tbprint.c | 2 +- drivers/acpi/acpica/tbxfroot.c | 3 +-- include/acpi/actypes.h | 5 +++++ 3 files changed, 7 insertions(+), 3 deletions(-)
drivers/acpi/acpica/tbprint.c +1 -1
@@ -138,7 +138,7 @@ acpi_tb_print_table_header(acpi_physical_address address,ACPI_INFO ((AE_INFO ,"%4.4s %p %05X" , header->signature,ACPI_CAST_PTR (void, address), header->length)); - }else if (ACPI_COMPARE_NAME (header->signature,ACPI_SIG_RSDP )) { + }else if (ACPI_VALIDATE_RSDP_SIG (header->signature)) {/* RSDP has no common fields */
drivers/acpi/acpica/tbxfroot.c +1 -2
@@ -68,8 +68,7 @@ acpi_status acpi_tb_validate_rsdp(struct acpi_table_rsdp *rsdp)* Note: Sometimes there exists more than one RSDP in memory; the valid * RSDP has a valid checksum, all others have an invalid checksum. */ -if (ACPI_STRNCMP ((char *)rsdp->signature,ACPI_SIG_RSDP , -sizeof (ACPI_SIG_RSDP ) -1 ) !=0 ) { +if (!ACPI_VALIDATE_RSDP_SIG (rsdp->signature)) {/* Nope, BAD Signature */
include/acpi/actypes.h +5 -0
@@ -474,6 +474,11 @@ typedef u64 acpi_integer;#define ACPI_MOVE_NAME (dest,src) (ACPI_STRNCPY (ACPI_CAST_PTR (char, (dest)), ACPI_CAST_PTR (char, (src)), ACPI_NAME_SIZE))#endif +/* Support for the special RSDP signature (8 characters) */ + +#define ACPI_VALIDATE_RSDP_SIG (a) (!ACPI_STRNCMP (ACPI_CAST_PTR (char, (a)), ACPI_SIG_RSDP, 8)) +#define ACPI_MAKE_RSDP_SIG (dest) (ACPI_MEMCPY (ACPI_CAST_PTR (char, (dest)), ACPI_SIG_RSDP, 8)) +/******************************************************************************* * * Miscellaneous constants
Signature · 1 line
1.7.10
PATCH v2 12/14 ACPICA: Cleanup exception codes.
In reply to Lv ZhengThis patch adds AE_ACCESS for EACCES or EPERM. Some error prompts are also cleaned up in this patch. Lv Zheng.
--- include/acpi/acexcep.h | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)
include/acpi/acexcep.h +5 -3
@@ -125,8 +125,9 @@ struct acpi_exception_info {#define AE_NO_HANDLER EXCEP_ENV (0x001A)#define AE_OWNER_ID_LIMIT EXCEP_ENV (0x001B)#define AE_NOT_CONFIGURED EXCEP_ENV (0x001C) +#define AE_ACCESS EXCEP_ENV (0x001D) -#define AE_CODE_ENV_MAX 0x001C +#define AE_CODE_ENV_MAX 0x001D/* * Programmer exceptions
@@ -227,7 +228,7 @@ static const struct acpi_exception_info acpi_gbl_exception_names_env[] = {EXCEP_TXT ("AE_NO_ACPI_TABLES" ,"ACPI tables could not be found" ),EXCEP_TXT ("AE_NO_NAMESPACE" ,"A namespace has not been loaded" ),EXCEP_TXT ("AE_NO_MEMORY" ,"Insufficient dynamic memory" ), -EXCEP_TXT ("AE_NOT_FOUND" ,"The name was not found in the namespace" ), +EXCEP_TXT ("AE_NOT_FOUND" ,"A requested entity is not found" ),EXCEP_TXT ("AE_NOT_EXIST" ,"A required entity does not exist" ),EXCEP_TXT ("AE_ALREADY_EXISTS" ,"An entity already exists" ), EXCEP_TXT("AE_TYPE", "The object type is incorrect"),
@@ -259,7 +260,8 @@ static const struct acpi_exception_info acpi_gbl_exception_names_env[] = {EXCEP_TXT ("AE_OWNER_ID_LIMIT" ,"There are no more Owner IDs available for ACPI tables or control methods" ),EXCEP_TXT ("AE_NOT_CONFIGURED" , -"The interface is not part of the current subsystem configuration" ) +"The interface is not part of the current subsystem configuration" ), +EXCEP_TXT ("AE_ACCESS" ,"Permission denied for the requested operation" ) }; static const struct acpi_exception_info acpi_gbl_exception_names_pgm[] = {
Signature · 1 line
1.7.10
PATCH v2 13/14 ACPICA: SCI Handlers: Update handler interface, eliminate unnecessary argument.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
The SCI interrupt number is not needed for the SCI handlers, and was just unnecessary overhead.
--- drivers/acpi/acpica/evsci.c | 4 +--- include/acpi/actypes.h | 2 +- 2 files changed, 2 insertions(+), 4 deletions(-)
drivers/acpi/acpica/evsci.c +1 -3
@@ -87,9 +87,7 @@ u32 acpi_ev_sci_dispatch(void)/* Invoke the installed handler (at interrupt level) */ - int_status |= sci_handler->address ((u32 )acpi_gbl_FADT. - sci_interrupt, - sci_handler->context); + int_status |= sci_handler->address (sci_handler->context); sci_handler = sci_handler->next; }
include/acpi/actypes.h +1 -1
@@ -951,7 +951,7 @@ typedef void* Various handlers and callback procedures */ typedef -u32 (*acpi_sci_handler ) (u32 interrupt_number,void *context); +u32 (*acpi_sci_handler ) (void *context);typedef void (*acpi_gbl_event_handler) (u32 event_type,
Signature · 1 line
1.7.10
PATCH v2 14/14 ACPICA: Update version to 20130823.
In reply to Lv ZhengFrom: Bob Moore <[email protected]>
Version 20130823.
--- include/acpi/acpixf.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
include/acpi/acpixf.h +1 -1
@@ -46,7 +46,7 @@/* Current ACPICA subsystem version in YYYYMMDD format */ -#define ACPI_CA_VERSION 0x20130725 +#define ACPI_CA_VERSION 0x20130823#include <acpi/acconfig.h> #include <acpi/actypes.h>
Signature · 1 line
1.7.10
Thanks for posting the refreshed series.
It looks ready for linux-next to me.
cheers,
-Len
Rafael J. Wysocki
AppliedQuoted 43 lines · Lv Zheng
The 20130823 ACPICA kernel resident system updates is linuxized based on the pm/linux-next branch.
The patch set has passed a build/boot test on the following machines:
Dell Inspiron Mini 1010 (i386)
HP Compaq 8200 Elite SFF PC (x86-64)
With the following kernel configuration items enabled:
CONFIG_ACPI_AC
CONFIG_ACPI_BATTERY
CONFIG_ACPI_NUMA
CONFIG_ACPI_PROCESSOR
CONFIG_ACPI_THERMAL
CONFIG_ACPI_VIDEO
CONFIG_CPU_FREQ
CONFIG_I2C_SCMI
CONFIG_SENSORS_ACPI_POWER
CONFIG_PANASONIC_LAPTOP
CONFIG_XO15_EBOOK
CONFIG_DMAR_TABLE
CONFIG_INTEL_IOMMU
CONFIG_IRQ_REMAPBjorn Helgaas (1):
ACPICA:acpi_read: On error, do not modify the return value target
location.Bob Moore (7):
ACPICA: Improve error message for "too many parent prefixes"
condition.
ACPICA: Debug output: small formatting update, no functional change.
ACPICA: Debugger: Prevent possible command line buffer overflow,
kernel behavior is not affected.
ACPICA: Validate start object foracpi_walk_namespace.
ACPICA: Debugger: Add new command to display full namespace
pathnames.
ACPICA: SCI Handlers: Update handler interface, eliminate unnecessary
argument.
ACPICA: Update version to 20130823.David E. Box (1):
ACPICA: Operation region support: Never free the handler "context"
pointer.
Quoted 8 lines
Lv Zheng (4):
ACPICA: Add support for host-installed SCI handlers.
ACPICA: Tables: Cleanup table checksum verification code.
ACPICA: Tables: Cleanup RSDP signature codes.
ACPICA: Cleanup exception codes.Naresh Bhat (1):
ACPICA: Linux-specific header: enable "aarch64" 64-bit build.
Queued up for 3.13, thanks Lv!
Quoted 59 lines
-----Original Message-----
From: Rafael J. Wysocki [mailto:[email protected]]
Sent: Thursday, September 26, 2013 12:01 AM
To: Zheng, Lv
Cc: Wysocki, Rafael J; Moore, Robert; Brown, Len; [email protected]
Subject: Re: [PATCH v2 00/14] ACPICA 20130823 ReleaseQuoted 49 lines · Lv Zheng
The 20130823 ACPICA kernel resident system updates is linuxized based on the pm/linux-next branch.
The patch set has passed a build/boot test on the following machines:
Dell Inspiron Mini 1010 (i386)
HP Compaq 8200 Elite SFF PC (x86-64)
With the following kernel configuration items enabled:
CONFIG_ACPI_AC
CONFIG_ACPI_BATTERY
CONFIG_ACPI_NUMA
CONFIG_ACPI_PROCESSOR
CONFIG_ACPI_THERMAL
CONFIG_ACPI_VIDEO
CONFIG_CPU_FREQ
CONFIG_I2C_SCMI
CONFIG_SENSORS_ACPI_POWER
CONFIG_PANASONIC_LAPTOP
CONFIG_XO15_EBOOK
CONFIG_DMAR_TABLE
CONFIG_INTEL_IOMMU
CONFIG_IRQ_REMAPBjorn Helgaas (1):
ACPICA:acpi_read: On error, do not modify the return value target
location.Bob Moore (7):
ACPICA: Improve error message for "too many parent prefixes"
condition.
ACPICA: Debug output: small formatting update, no functional change.
ACPICA: Debugger: Prevent possible command line buffer overflow,
kernel behavior is not affected.
ACPICA: Validate start object foracpi_walk_namespace.
ACPICA: Debugger: Add new command to display full namespace
pathnames.
ACPICA: SCI Handlers: Update handler interface, eliminate unnecessary
argument.
ACPICA: Update version to 20130823.David E. Box (1):
ACPICA: Operation region support: Never free the handler "context"
pointer.Lv Zheng (4):
ACPICA: Add support for host-installed SCI handlers.
ACPICA: Tables: Cleanup table checksum verification code.
ACPICA: Tables: Cleanup RSDP signature codes.
ACPICA: Cleanup exception codes.
Quoted 4 lines
Quoted 2 lines
Naresh Bhat (1):
ACPICA: Linux-specific header: enable "aarch64" 64-bit build.Queued up for 3.13, thanks Lv!
Thanks for the help.
Cheers
-Lv