Skip to contentSkip to conversation

Add ACPI CPER firmware first error injection on ARM emulation

PATCH Superseded Mauro Carvalho Chehab 4 participants 29 messages Aug 25, 2024 – Oct 3, 2024

Subsystems (2)

v9 of 10 Superseded Latest is v10 →

lore
Share & tools
lore
#0 lore raw

v9 cover Add ACPI CPER firmware first error injection on ARM emulation

This series add support for injecting generic CPER records. Such records are generated outside QEMU via a provided script.

On this version, patches were reorganized to follow this pattern:

1. Addition of hest_add_le, mapping to the beginning of the HEST error source structure size. This is the part of the HEST table that contains the error source IDs to be notified;
2. GHESv2 code rework. It is basically a merge of all changes from v8, except for GPIO notify, QEMU notifier and renames;
3. cleanups and renames for hw/acpi/ghes.c;
4. GPIO GED device creation logic;
5. Logic to inject errors via QMP;
6. Error injection script

On this version, the logic now better navigates the source IDs, double-checking them at error injecting time. It is now easier to add other HEST types if ever needed.

Also, the source ID is now guest-OS specific: ARM will use SEA and GPIO, but x86 will likely use other notifiers and source IDs. Same will apply to other processor types.

---

v9:
- Patches reorganized to make easier for reviewers;
- source ID is now guest-OS specific;
- Some patches got a revision history since v8;
- Several minor cleanups.

v8:
- Fix one of the BIOS links that were incorrect;
- Changed mem error internal injection to use a common code;
- No more hardcoded values for CPER: instead of using just the payload at the QAPI, it now has the full raw CPER there;
- Error injection script now supports changing fields at the Generic Error Data section of the CPER;
- Several minor cleanups.

v7:
- Change the way offsets are calculated and used on HEST table. Now, it is compatible with migrations as all offsets are relative to the HEST table;
- GHES interface is now more generic: the entire CPER is sent via QMP, instead of just the payload;
- Some code cleanups to make the code more robust;
- The python script now uses QEMUMonitorProtocol class.

v6:
- PNP0C33 device creation moved to aml-build.c;
- acpi_ghes record functions now use ACPI notify parameter, instead of source ID;
- the number of source IDs is now automatically calculated;
- some code cleanups and function/var renames;
- some fixes and cleanups at the error injection script;
- ghes cper stub now produces an error if cper JSON is not compiled;
- Offset calculation logic for GHES was refactored;
- Updated documentation to reflect the GHES allocated size;
- Added a x-mpidr object for QOM usage;
- Added a patch making usage of x-mpidr field at ARM injection script;

v5:
- CPER guid is now passing as string;
- raw-data is now passed with base64 encode;
- Removed several GPIO left-overs from arm/virt.c changes;
- Lots of cleanups and improvements at the error injection script. It now better handles QMP dialog and doesn't print debug messages. Also, code was split on two modules, to make easier to add more error injection commands.

v4:
- CPER generation moved to happen outside QEMU;
- One patch adding support for mpidr query was removed.

v3:
- patch 1 cleanups with some comment changes and adding another place where the poweroff GPIO define should be used. No changes on other patches (except due to conflict resolution).

v2:
- added a new patch using a define for GPIO power pin;
- patch 2 changed to also use a define for generic error GPIO pin;
- a couple cleanups at patch 2 removing uneeded else clauses.

Example of generating a CPER record:

$ scripts/ghes_inject.py -d arm -p 0xdeadbeef
GUID: e19e3d16-bc11-11e4-9caa-c2051d5d46b0
Generic Error Status Block (20 bytes):
      00000000  01 00 00 00 00 00 00 00 00 00 00 00 90 00 00 00   ................
      00000010  00 00 00 00                                       ....

Generic Error Data Entry (72 bytes):

      00000000  16 3d 9e e1 11 bc e4 11 9c aa c2 05 1d 5d 46 b0   .=...........]F.
      00000010  00 00 00 00 00 03 00 00 48 00 00 00 00 00 00 00   ........H.......
      00000020  00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
      00000030  00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
      00000040  00 00 00 00 00 00 00 00                           ........

Payload (72 bytes):

      00000000  05 00 00 00 01 00 00 00 48 00 00 00 00 00 00 00   ........H.......
      00000010  00 00 00 80 00 00 00 00 10 05 0f 00 00 00 00 00   ................
      00000020  00 00 00 00 00 00 00 00 00 20 14 00 02 01 00 03   ......... ......
      00000030  0f 00 91 00 00 00 00 00 ef be ad de 00 00 00 00   ................
      00000040  ef be ad de 00 00 00 00                           ........

Error injected.

[    9.358364] {1}[Hardware Error]: Hardware error from APEI Generic Hardware Error Source: 1
[    9.359027] {1}[Hardware Error]: event severity: recoverable
[    9.359586] {1}[Hardware Error]:  Error 0, type: recoverable
[    9.360124] {1}[Hardware Error]:   section_type: ARM processor error
[    9.360561] {1}[Hardware Error]:   MIDR: 0x00000000000f0510
[    9.361160] {1}[Hardware Error]:   Multiprocessor Affinity Register (MPIDR): 0x0000000080000000
[    9.361643] {1}[Hardware Error]:   running state: 0x0
[    9.362142] {1}[Hardware Error]:   Power State Coordination Interface state: 0
[    9.362682] {1}[Hardware Error]:   Error info structure 0:
[    9.363030] {1}[Hardware Error]:   num errors: 2
[    9.363656] {1}[Hardware Error]:    error_type: 0x02: cache error
[    9.364163] {1}[Hardware Error]:    error_info: 0x000000000091000f
[    9.364834] {1}[Hardware Error]:     transaction type: Data Access
[    9.365599] {1}[Hardware Error]:     cache error, operation type: Data write
[    9.366441] {1}[Hardware Error]:     cache level: 2
[    9.367005] {1}[Hardware Error]:     processor context not corrupted
[    9.367753] {1}[Hardware Error]:    physical fault address: 0x00000000deadbeef
[    9.374267] Memory failure: 0xdeadb: recovery action for free buddy page: Recovered

Such script currently supports arm processor error CPER, but can easily be extended to other GHES notification types.

Mauro Carvalho Chehab (12):
acpi/ghes: add a firmware file with HEST address acpi/ghes: rework the logic to handle HEST source ID acpi/ghes: rename etc/hardware_error file macros acpi/ghes: better name GHES memory error function acpi/ghes: add a notifier to notify when error data is ready acpi/generic_event_device: add an APEI error device arm/virt: Wire up a GED error device for ACPI / GHES qapi/acpi-hest: add an interface to do generic CPER error injection
docs: acpi_hest_ghes: fix documentation for CPER size scripts/ghes_inject: add a script to generate GHES error inject target/arm: add an experimental mpidr arm cpu property object scripts/arm_processor_error.py: retrieve mpidr if not filled

 MAINTAINERS                            |  10 +
 docs/specs/acpi_hest_ghes.rst          |   6 +-
 hw/acpi/Kconfig                        |   5 +
 hw/acpi/aml-build.c                    |  10 +
 hw/acpi/generic_event_device.c         |   8 +
 hw/acpi/ghes-stub.c                    |   2 +-
 hw/acpi/ghes.c                         | 309 +++++++----
 hw/acpi/ghes_cper.c                    |  32 ++
 hw/acpi/ghes_cper_stub.c               |  19 +
 hw/acpi/meson.build                    |   2 +
 hw/arm/Kconfig                         |   5 +
 hw/arm/virt-acpi-build.c               |  12 +-
 hw/arm/virt.c                          |  12 +-
 include/hw/acpi/acpi_dev_interface.h   |   1 +
 include/hw/acpi/aml-build.h            |   2 +
 include/hw/acpi/generic_event_device.h |   1 +
 include/hw/acpi/ghes.h                 |  28 +-
 include/hw/arm/virt.h                  |  10 +
 qapi/acpi-hest.json                    |  36 ++
 qapi/meson.build                       |   1 +
 qapi/qapi-schema.json                  |   1 +
 scripts/arm_processor_error.py         | 388 ++++++++++++++
 scripts/ghes_inject.py                 |  51 ++
 scripts/qmp_helper.py                  | 702 +++++++++++++++++++++++++
 target/arm/cpu.c                       |   1 +
 target/arm/cpu.h                       |   1 +
 target/arm/helper.c                    |  10 +-
 target/arm/kvm.c                       |   3 +-
 28 files changed, 1539 insertions(+), 129 deletions(-)
 create mode 100644 hw/acpi/ghes_cper.c
 create mode 100644 hw/acpi/ghes_cper_stub.c
 create mode 100644 qapi/acpi-hest.json
 create mode 100644 scripts/arm_processor_error.py
 create mode 100755 scripts/ghes_inject.py
 create mode 100644 scripts/qmp_helper.py
Signature · 1 line
2.46.0
#1 lore raw

PATCH v9 1/12 acpi/ghes: add a firmware file with HEST address

Store HEST table address at GPA, placing its content at hest_addr_le variable.

---

Change from v8:
- hest_addr_lr is now pointing to the error source size and data.

---
 hw/acpi/ghes.c         | 15 +++++++++++++++
 include/hw/acpi/ghes.h |  1 +
 2 files changed, 16 insertions(+)
hw/acpi/ghes.c +15 -0
index e9511d9b8f71..529c14e3289f 100644
@@ -30,6 +30,7 @@
 
 #define ACPI_GHES_ERRORS_FW_CFG_FILE        "etc/hardware_errors"
 #define ACPI_GHES_DATA_ADDR_FW_CFG_FILE     "etc/hardware_errors_addr"
+#define ACPI_HEST_ADDR_FW_CFG_FILE          "etc/acpi_table_hest_addr"
 
 /* The max size in bytes for one error block */
 #define ACPI_GHES_MAX_RAW_DATA_LENGTH   (1 * KiB)
@@ -367,11 +368,22 @@ void acpi_build_hest(GArray *table_data, BIOSLinker *linker,
 
     acpi_table_begin(&table, table_data);
 
+    int hest_offset = table_data->len;
+
     /* Error Source Count */
     build_append_int_noprefix(table_data, ACPI_GHES_ERROR_SOURCE_COUNT, 4);
     build_ghes_v2(table_data, ACPI_HEST_SRC_ID_SEA, linker);
 
     acpi_table_end(linker, &table);
+
+    /*
+     * tell firmware to write into GPA the address of HEST via fw_cfg,
+     * once initialized.
+     */
+    bios_linker_loader_write_pointer(linker,
+                                     ACPI_HEST_ADDR_FW_CFG_FILE, 0,
+                                     sizeof(uint64_t),
+                                     ACPI_BUILD_TABLE_FILE, hest_offset);
 }
 
 void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
@@ -385,6 +397,9 @@ void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
     fw_cfg_add_file_callback(s, ACPI_GHES_DATA_ADDR_FW_CFG_FILE, NULL, NULL,
         NULL, &(ags->ghes_addr_le), sizeof(ags->ghes_addr_le), false);
 
+    fw_cfg_add_file_callback(s, ACPI_HEST_ADDR_FW_CFG_FILE, NULL, NULL,
+        NULL, &(ags->hest_addr_le), sizeof(ags->hest_addr_le), false);
+
     ags->present = true;
 }
 
include/hw/acpi/ghes.h +1 -0
index 674f6958e905..28b956acb19a 100644
@@ -63,6 +63,7 @@ enum {
 };
 
 typedef struct AcpiGhesState {
+    uint64_t hest_addr_le;
     uint64_t ghes_addr_le;
     bool present; /* True if GHES is present at all on this board */
 } AcpiGhesState;
Signature · 1 line
2.46.0
#2 lore raw

Re: [PATCH v9 01/12] acpi/ghes: add a firmware file with HEST address

Quoted 4 lines · On Sun, 25 Aug 2024 05:45:56 +0200 Mauro Carvalho Chehab…

Store HEST table address at GPA, placing its content at hest_addr_le variable.

This looks good to me.

in addition to this, it needs a patch on top to make sure that we migrate hest_addr_le.
See a08a64627b6b 'ACPI: Record the Generic Error Status Block address' and fixes on top of that for an example.

Quoted 68 lines

---

Change from v8:
- hest_addr_lr is now pointing to the error source size and data.

---
 hw/acpi/ghes.c         | 15 +++++++++++++++
 include/hw/acpi/ghes.h |  1 +
 2 files changed, 16 insertions(+)
hw/acpi/ghes.c +15 -0
index e9511d9b8f71..529c14e3289f 100644
@@ -30,6 +30,7 @@
 
 #define ACPI_GHES_ERRORS_FW_CFG_FILE        "etc/hardware_errors"
 #define ACPI_GHES_DATA_ADDR_FW_CFG_FILE     "etc/hardware_errors_addr"
+#define ACPI_HEST_ADDR_FW_CFG_FILE          "etc/acpi_table_hest_addr"
 
 /* The max size in bytes for one error block */
 #define ACPI_GHES_MAX_RAW_DATA_LENGTH   (1 * KiB)
@@ -367,11 +368,22 @@ void acpi_build_hest(GArray *table_data, BIOSLinker *linker,
 
     acpi_table_begin(&table, table_data);
 
+    int hest_offset = table_data->len;
+
     /* Error Source Count */
     build_append_int_noprefix(table_data, ACPI_GHES_ERROR_SOURCE_COUNT, 4);
     build_ghes_v2(table_data, ACPI_HEST_SRC_ID_SEA, linker);
 
     acpi_table_end(linker, &table);
+
+    /*
+     * tell firmware to write into GPA the address of HEST via fw_cfg,
+     * once initialized.
+     */
+    bios_linker_loader_write_pointer(linker,
+                                     ACPI_HEST_ADDR_FW_CFG_FILE, 0,
+                                     sizeof(uint64_t),
+                                     ACPI_BUILD_TABLE_FILE, hest_offset);
 }
 
 void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
@@ -385,6 +397,9 @@ void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
     fw_cfg_add_file_callback(s, ACPI_GHES_DATA_ADDR_FW_CFG_FILE, NULL, NULL,
         NULL, &(ags->ghes_addr_le), sizeof(ags->ghes_addr_le), false);
 
+    fw_cfg_add_file_callback(s, ACPI_HEST_ADDR_FW_CFG_FILE, NULL, NULL,
+        NULL, &(ags->hest_addr_le), sizeof(ags->hest_addr_le), false);
+
     ags->present = true;
 }
 
include/hw/acpi/ghes.h +1 -0
index 674f6958e905..28b956acb19a 100644
@@ -63,6 +63,7 @@ enum {
 };
 
 typedef struct AcpiGhesState {
+    uint64_t hest_addr_le;
     uint64_t ghes_addr_le;
     bool present; /* True if GHES is present at all on this board */
 } AcpiGhesState;
#3 lore raw

Re: [PATCH v9 01/12] acpi/ghes: add a firmware file with HEST address

Em Wed, 11 Sep 2024 15:51:08 +0200
Igor Mammedov <[email protected]> escreveu:

Quoted 6 lines
Quoted 3 lines · On Sun, 25 Aug 2024 05:45:56 +0200 Mauro Carvalho Chehab…

Store HEST table address at GPA, placing its content at hest_addr_le variable.

Quoted 8 lines
Quoted 1 line

This looks good to me.

in addition to this, it needs a patch on top to make sure that we migrate hest_addr_le.
See a08a64627b6b 'ACPI: Record the Generic Error Status Block address' and fixes on top of that for an example.

Hmm... If I understood such change well, vmstate_ghes_state() will use this structure as basis to do migration:

	/* ghes.h */
	typedef struct AcpiGhesState {
	    uint64_t hest_addr_le;
	    uint64_t ghes_addr_le;
	    bool present; /* True if GHES is present at all on this board */
	} AcpiGhesState;

	/* generic_event_device.c */
	static const VMStateDescription vmstate_ghes_state = {
	    .name = "acpi-ged/ghes",
	    .version_id = 1,
	    .minimum_version_id = 1,
	    .needed = ghes_needed,
	    .fields      = (VMStateField[]) {
	        VMSTATE_STRUCT(ghes_state, AcpiGedState, 1,
	                       vmstate_ghes_state, AcpiGhesState),
	        VMSTATE_END_OF_LIST()
	    }
	};

	/* hw/arm/virt-acpi-build.c */
	void virt_acpi_setup(VirtMachineState *vms)
	{
	    ...
	    if (vms->ras) {
	        assert(vms->acpi_dev);
	        acpi_ged_state = ACPI_GED(vms->acpi_dev);
	        acpi_ghes_add_fw_cfg(&acpi_ged_state->ghes_state,
	                             vms->fw_cfg, tables.hardware_errors);
	    }

Which relies on acpi_ghes_add_fw_cfg() function to setup callbacks for the migration:

	/* ghes.c */
	void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
	                          GArray *hardware_error)
	{
	    /* Create a read-only fw_cfg file for GHES */
	    fw_cfg_add_file(s, ACPI_HW_ERROR_FW_CFG_FILE, hardware_error->data,
	                    hardware_error->len);

	    /* Create a read-write fw_cfg file for Address */
	    fw_cfg_add_file_callback(s, ACPI_HW_ERROR_ADDR_FW_CFG_FILE, NULL, NULL,
	        NULL, &(ags->ghes_addr_le), sizeof(ags->ghes_addr_le), false);

	    fw_cfg_add_file_callback(s, ACPI_HEST_ADDR_FW_CFG_FILE, NULL, NULL,
	        NULL, &(ags->hest_addr_le), sizeof(ags->hest_addr_le), false);

	    ags->present = true;
	}

It sounds to me that both ghes_addr_le and hest_addr_le will be migrated altogether.

Did I miss something?

Thanks,
Mauro

#4 lore raw

Re: [PATCH v9 01/12] acpi/ghes: add a firmware file with HEST address

Quoted 32 lines · On Fri, 13 Sep 2024 07:44:35 +0200 Mauro Carvalho Chehab…

Em Wed, 11 Sep 2024 15:51:08 +0200
Igor Mammedov <[email protected]> escreveu:

Quoted 6 lines
Quoted 3 lines · On Sun, 25 Aug 2024 05:45:56 +0200 Mauro Carvalho Chehab…

Store HEST table address at GPA, placing its content at hest_addr_le variable.

Quoted 8 lines
Quoted 1 line

This looks good to me.

in addition to this, it needs a patch on top to make sure that we migrate hest_addr_le.
See a08a64627b6b 'ACPI: Record the Generic Error Status Block address' and fixes on top of that for an example.

Hmm... If I understood such change well, vmstate_ghes_state() will use this structure as basis to do migration:

	/* ghes.h */
	typedef struct AcpiGhesState {
	    uint64_t hest_addr_le;
	    uint64_t ghes_addr_le;
	    bool present; /* True if GHES is present at all on this board */
	} AcpiGhesState;

	/* generic_event_device.c */
	static const VMStateDescription vmstate_ghes_state = {
	    .name = "acpi-ged/ghes",
	    .version_id = 1,
Quoted 8 lines
	    .minimum_version_id = 1,
	    .needed = ghes_needed,
	    .fields      = (VMStateField[]) {
	        VMSTATE_STRUCT(ghes_state, AcpiGedState, 1,
	                       vmstate_ghes_state, AcpiGhesState),
	        VMSTATE_END_OF_LIST()
	    }
	};

current code looks like that:

static const VMStateDescription vmstate_ghes = {                                 
    .name = "acpi-ghes",                                                         
    .version_id = 1,                                                             
    .minimum_version_id = 1,                                                     
    .fields = (const VMStateField[]) {                                           
        VMSTATE_UINT64(ghes_addr_le, AcpiGhesState),   <<===                         
        VMSTATE_END_OF_LIST()                                                    
    },                                                                           
};                                                                               

static bool ghes_needed(void *opaque)

{                                                                                
    AcpiGedState *s = opaque;                                                    
    return s->ghes_state.ghes_addr_le;                                           
}                                                                                
static const VMStateDescription vmstate_ghes_state = {                           
    .name = "acpi-ged/ghes",                                                     
    .version_id = 1,                                                             
    .minimum_version_id = 1,                                                     
    .needed = ghes_needed,                                                       
    .fields = (const VMStateField[]) {                                           
        VMSTATE_STRUCT(ghes_state, AcpiGedState, 1,                              
                       vmstate_ghes, AcpiGhesState),                             
        VMSTATE_END_OF_LIST()                                                    
    }                                                                            
};  

where
VMSTATE_UINT64(ghes_addr_le, AcpiGhesState),
explicitly defines field(s) within structure to be sent over wire.

we need to add a conditional field for ghes_addr_le
which will be sent only with new machine types, but not with old ones to avoid migration breakage.

I don't know much about migration, but maybe we can get away with similar condition as in ghes_needed(), or enabling QMP error injection based on machine type version.

Or maybe adding a CLI option to enable QMP error injection in which case the explicit option would serve as a trigger enable QMP command and to migrate hest_addr_le.
It might be even better this way, since machine wouldn't need to carry extra error source that will be used only for testing and practically never in production VMs (aka reduced attack surface).

You can easily test it locally:
new-qemu: with your patches
old-qemu: qemu-9.1

and then try to do forth & back migration for following cases:
1. (ping-pong case with OLD firmware/ACPI tables)

     start old-qemu with 9.1 machine type ->
       migrate to file ->
       start new-qemu with 9.1 machine type -> restore from file ->
       migrate to file ->
       start old-qemu with 9.1 machine type ->restore from file ->

2. (ping-pong case with NEW firmware/ACPI tables) do the same as #1 but starting with new-qemu binary

(from upstream pov #2 is optional, but not implementing it is pain for downstream so it's better to have it if it's not too much work)

Quoted 32 lines
	/* hw/arm/virt-acpi-build.c */
	void virt_acpi_setup(VirtMachineState *vms)
	{
	    ...
	    if (vms->ras) {
	        assert(vms->acpi_dev);
	        acpi_ged_state = ACPI_GED(vms->acpi_dev);
	        acpi_ghes_add_fw_cfg(&acpi_ged_state->ghes_state,
	                             vms->fw_cfg, tables.hardware_errors);
	    }

Which relies on acpi_ghes_add_fw_cfg() function to setup callbacks for the migration:

	/* ghes.c */
	void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
	                          GArray *hardware_error)
	{
	    /* Create a read-only fw_cfg file for GHES */
	    fw_cfg_add_file(s, ACPI_HW_ERROR_FW_CFG_FILE, hardware_error->data,
	                    hardware_error->len);

	    /* Create a read-write fw_cfg file for Address */
	    fw_cfg_add_file_callback(s, ACPI_HW_ERROR_ADDR_FW_CFG_FILE, NULL, NULL,
	        NULL, &(ags->ghes_addr_le), sizeof(ags->ghes_addr_le), false);

	    fw_cfg_add_file_callback(s, ACPI_HEST_ADDR_FW_CFG_FILE, NULL, NULL,
	        NULL, &(ags->hest_addr_le), sizeof(ags->hest_addr_le), false);

	    ags->present = true;
	}
Quoted 2 lines

It sounds to me that both ghes_addr_le and hest_addr_le will be migrated altogether.

fwcfg callbacks are irrelevant to migration, they tell firmware what to do with specified addresses (in our case, write into ags->hest_addr_le address of HEST), so that HW (qemu) would know where firmware placed the table. But that's about all it does.

Quoted 6 lines

Did I miss something?

Thanks,
Mauro

#5 lore raw

Re: [PATCH v9 01/12] acpi/ghes: add a firmware file with HEST address

Hi Igor,

Em Fri, 13 Sep 2024 15:25:18 +0200
Igor Mammedov <[email protected]> escreveu:

Quoted 81 lines
Quoted 19 lines
Quoted 4 lines

in addition to this, it needs a patch on top to make sure that we migrate hest_addr_le.
See a08a64627b6b 'ACPI: Record the Generic Error Status Block address' and fixes on top of that for an example.

Hmm... If I understood such change well, vmstate_ghes_state() will use this structure as basis to do migration:

	/* ghes.h */
	typedef struct AcpiGhesState {
	    uint64_t hest_addr_le;
	    uint64_t ghes_addr_le;
	    bool present; /* True if GHES is present at all on this board */
	} AcpiGhesState;

	/* generic_event_device.c */
	static const VMStateDescription vmstate_ghes_state = {
	    .name = "acpi-ged/ghes",
	    .version_id = 1,
Quoted 8 lines
	    .minimum_version_id = 1,
	    .needed = ghes_needed,
	    .fields      = (VMStateField[]) {
	        VMSTATE_STRUCT(ghes_state, AcpiGedState, 1,
	                       vmstate_ghes_state, AcpiGhesState),
	        VMSTATE_END_OF_LIST()
	    }
	};  

current code looks like that:

static const VMStateDescription vmstate_ghes = {                                 
    .name = "acpi-ghes",                                                         
    .version_id = 1,                                                             
    .minimum_version_id = 1,                                                     
    .fields = (const VMStateField[]) {                                           
        VMSTATE_UINT64(ghes_addr_le, AcpiGhesState),   <<===                         
        VMSTATE_END_OF_LIST()                                                    
    },                                                                           
};                                                                               

static bool ghes_needed(void *opaque)

{                                                                                
    AcpiGedState *s = opaque;                                                    
    return s->ghes_state.ghes_addr_le;                                           
}                                                                                
static const VMStateDescription vmstate_ghes_state = {                           
    .name = "acpi-ged/ghes",                                                     
    .version_id = 1,                                                             
    .minimum_version_id = 1,                                                     
    .needed = ghes_needed,                                                       
    .fields = (const VMStateField[]) {                                           
        VMSTATE_STRUCT(ghes_state, AcpiGedState, 1,                              
                       vmstate_ghes, AcpiGhesState),                             
        VMSTATE_END_OF_LIST()                                                    
    }                                                                            
};  

where
VMSTATE_UINT64(ghes_addr_le, AcpiGhesState),
explicitly defines field(s) within structure to be sent over wire.

we need to add a conditional field for ghes_addr_le
which will be sent only with new machine types, but not with old ones to avoid migration breakage.

I don't know much about migration, but maybe we can get away with similar condition as in ghes_needed(), or enabling QMP error injection based on machine type version.

Or maybe adding a CLI option to enable QMP error injection in which case the explicit option would serve as a trigger enable QMP command and to migrate hest_addr_le.
It might be even better this way, since machine wouldn't need to carry extra error source that will be used only for testing and practically never in production VMs (aka reduced attack surface).

You can easily test it locally:
new-qemu: with your patches
old-qemu: qemu-9.1

Quoted 6 lines

and then try to do forth & back migration for following cases:
1. (ping-pong case with OLD firmware/ACPI tables)

     start old-qemu with 9.1 machine type ->
       migrate to file ->
       start new-qemu with 9.1 machine type -> restore from file ->
       migrate to file ->

As I never used migration, I'm a little stuck with the command line parameters.

I guess I got the one to do the migration at the monitor:

(qemu) migrate file://tmp/migrate

But no idea how to start a machine using a saved state.

Quoted 8 lines
       start old-qemu with 9.1 machine type ->restore from file ->

2. (ping-pong case with NEW firmware/ACPI tables) do the same as #1 but starting with new-qemu binary

(from upstream pov #2 is optional, but not implementing it is pain for downstream so it's better to have it if it's not too much work)

If I understood the migration documentation, every when new fields are added, we should increment .version_id. If new version is not backward-compatible, .minimum_version_id is also incremented.

So, for a migration-compatible code with a 9.1 VM, the code needs to handle the case where hest_addr_le is not defined, e. g. use offsets relative to ghes_addr_le, just like the current version, e.g.:

    uint64_t cper_addr, read_ack_start_addr;

    AcpiGedState *acpi_ged_state =
        ACPI_GED(object_resolve_path_type("", TYPE_ACPI_GED, NULL));
    AcpiGhesState *ags = &acpi_ged_state->ghes_state;

    if (!ags->hest_addr_le) {
        // Backward-compatible migration code
        uint64_t base = le64_to_cpu(ags->ghes_addr_le);

        *read_ack_start_addr = base +
            ACPI_GHES_ERROR_SOURCE_COUNT * sizeof(uint64_t) +
            error_source_to_index[notify] * sizeof(uint64_t);

        *cper_addr = base +
            ACPI_GHES_ERROR_SOURCE_COUNT * sizeof(uint64_t) +
            ACPI_GHES_ERROR_SOURCE_COUNT * sizeof(uint64_t) +
            error_source_to_index[notify] * ACPI_GHES_MAX_RAW_DATA_LENGTH;
   } else {
        // Use the new logic from ags->hest_addr_le

}

There are two problems with that:

1. On your reviews, if I understood right, the code above is not migration safe. So, while implementing it would be technically correct, migration still won't work;

2. With the new code, ACPI_GHES_ERROR_SOURCE_COUNT is not defined anymore, as the size of the error source structure can be different on different architectures, being 2 for new VMs and 1 for old ones.

Basically the new code gets it right because it can see a pointer to the HEST table, so it can get the number from there:

	hest_addr = le64_to_cpu(ags->hest_addr_le);
	cpu_physical_memory_read(hest_addr, &num_sources, sizeof(num_sources));

But, without hest_addr_le, getting num_sources is not possible.

An alternative would be to add a hacky code that works only for arm machines (as new versions may support more archs).

Something like:

	#define V1_ARM_ACPI_GHES_ERROR_SOURCE_COUNT 1
	#define V2_ARM_ACPI_GHES_ERROR_SOURCE_COUNT 2

And have a hardcoded logic that would work before/after this changeset but may break on newer versions, if the number of source IDs change, if we add other HEST types, etc.

Now, assuming that such hack would work, it sounds too hacky to my taste.

So, IMO it is a lot safer to not support migrations from v1 (only ghes_addr_le), using a patch like the enclosed one to ensure that.

Btw, checking existing migration structs, it sounds that for almost all structures, .version_id is identical to .minimum_version_id, meaning that migration between different versions aren't supported on most cases.

Thanks,
Mauro

---

[PATCH] acpi/generic_event_device: Update GHES migration to cover hest addr

The GHES migration logic at GED should now support HEST table location too.

Increase migration version and change needed to check for both ghes_addr_le and hest_addr_le.

hw/acpi/generic_event_device.c +6 -5
index b4c83a089a02..efae0ff62c7b 100644
@@ -351,10 +351,11 @@ static const VMStateDescription vmstate_ged_state = {
 
 static const VMStateDescription vmstate_ghes = {
     .name = "acpi-ghes",
-    .version_id = 1,
-    .minimum_version_id = 1,
+    .version_id = 2,
+    .minimum_version_id = 2,
     .fields = (const VMStateField[]) {
         VMSTATE_UINT64(ghes_addr_le, AcpiGhesState),
+        VMSTATE_UINT64(hest_addr_le, AcpiGhesState),
         VMSTATE_END_OF_LIST()
     },
 };
@@ -362,13 +363,13 @@ static const VMStateDescription vmstate_ghes = {
 static bool ghes_needed(void *opaque)
 {
     AcpiGedState *s = opaque;
-    return s->ghes_state.ghes_addr_le;
+    return s->ghes_state.ghes_addr_le && s->ghes_state.hest_addr_le;
 }
 
 static const VMStateDescription vmstate_ghes_state = {
     .name = "acpi-ged/ghes",
-    .version_id = 1,
-    .minimum_version_id = 1,
+    .version_id = 2,
+    .minimum_version_id = 2,
     .needed = ghes_needed,
     .fields = (const VMStateField[]) {
         VMSTATE_STRUCT(ghes_state, AcpiGedState, 1,
#6 lore raw

Re: [PATCH v9 01/12] acpi/ghes: add a firmware file with HEST address

Quoted 93 lines · On Sat, 14 Sep 2024 07:33:14 +0200 Mauro Carvalho Chehab…

Hi Igor,

Em Fri, 13 Sep 2024 15:25:18 +0200
Igor Mammedov <[email protected]> escreveu:

Quoted 81 lines
Quoted 19 lines
Quoted 4 lines

in addition to this, it needs a patch on top to make sure that we migrate hest_addr_le.
See a08a64627b6b 'ACPI: Record the Generic Error Status Block address' and fixes on top of that for an example.

Hmm... If I understood such change well, vmstate_ghes_state() will use this structure as basis to do migration:

	/* ghes.h */
	typedef struct AcpiGhesState {
	    uint64_t hest_addr_le;
	    uint64_t ghes_addr_le;
	    bool present; /* True if GHES is present at all on this board */
	} AcpiGhesState;

	/* generic_event_device.c */
	static const VMStateDescription vmstate_ghes_state = {
	    .name = "acpi-ged/ghes",
	    .version_id = 1,
Quoted 8 lines
	    .minimum_version_id = 1,
	    .needed = ghes_needed,
	    .fields      = (VMStateField[]) {
	        VMSTATE_STRUCT(ghes_state, AcpiGedState, 1,
	                       vmstate_ghes_state, AcpiGhesState),
	        VMSTATE_END_OF_LIST()
	    }
	};    

current code looks like that:

static const VMStateDescription vmstate_ghes = {                                 
    .name = "acpi-ghes",                                                         
    .version_id = 1,                                                             
    .minimum_version_id = 1,                                                     
    .fields = (const VMStateField[]) {                                           
        VMSTATE_UINT64(ghes_addr_le, AcpiGhesState),   <<===                         
        VMSTATE_END_OF_LIST()                                                    
    },                                                                           
};                                                                               

static bool ghes_needed(void *opaque)

{                                                                                
    AcpiGedState *s = opaque;                                                    
    return s->ghes_state.ghes_addr_le;                                           
}                                                                                
static const VMStateDescription vmstate_ghes_state = {                           
    .name = "acpi-ged/ghes",                                                     
    .version_id = 1,                                                             
    .minimum_version_id = 1,                                                     
    .needed = ghes_needed,                                                       
    .fields = (const VMStateField[]) {                                           
        VMSTATE_STRUCT(ghes_state, AcpiGedState, 1,                              
                       vmstate_ghes, AcpiGhesState),                             
        VMSTATE_END_OF_LIST()                                                    
    }                                                                            
};  

where
VMSTATE_UINT64(ghes_addr_le, AcpiGhesState),
explicitly defines field(s) within structure to be sent over wire.

we need to add a conditional field for ghes_addr_le
which will be sent only with new machine types, but not with old ones to avoid migration breakage.

I don't know much about migration, but maybe we can get away with similar condition as in ghes_needed(), or enabling QMP error injection based on machine type version.

Or maybe adding a CLI option to enable QMP error injection in which case the explicit option would serve as a trigger enable QMP command and to migrate hest_addr_le.
It might be even better this way, since machine wouldn't need to carry extra error source that will be used only for testing and practically never in production VMs (aka reduced attack surface).

You can easily test it locally:
new-qemu: with your patches
old-qemu: qemu-9.1

Quoted 6 lines

and then try to do forth & back migration for following cases:
1. (ping-pong case with OLD firmware/ACPI tables)

     start old-qemu with 9.1 machine type ->
       migrate to file ->
       start new-qemu with 9.1 machine type -> restore from file ->
       migrate to file ->  
Quoted 8 lines

As I never used migration, I'm a little stuck with the command line parameters.

I guess I got the one to do the migration at the monitor:

(qemu) migrate file://tmp/migrate

But no idea how to start a machine using a saved state.

see https://www.linux-kvm.org/page/Migration
'savevm/loadvm to an external state file (using pseudo-migration)' section

Quoted 73 lines
Quoted 9 lines
       start old-qemu with 9.1 machine type ->restore from file ->

2. (ping-pong case with NEW firmware/ACPI tables) do the same as #1 but starting with new-qemu binary

(from upstream pov #2 is optional, but not implementing it is pain for downstream so it's better to have it if it's not too much work)

If I understood the migration documentation, every when new fields are added, we should increment .version_id. If new version is not backward-compatible, .minimum_version_id is also incremented.

So, for a migration-compatible code with a 9.1 VM, the code needs to handle the case where hest_addr_le is not defined, e. g. use offsets relative to ghes_addr_le, just like the current version, e.g.:

    uint64_t cper_addr, read_ack_start_addr;

    AcpiGedState *acpi_ged_state =
        ACPI_GED(object_resolve_path_type("", TYPE_ACPI_GED, NULL));
    AcpiGhesState *ags = &acpi_ged_state->ghes_state;

    if (!ags->hest_addr_le) {
        // Backward-compatible migration code
        uint64_t base = le64_to_cpu(ags->ghes_addr_le);

        *read_ack_start_addr = base +
            ACPI_GHES_ERROR_SOURCE_COUNT * sizeof(uint64_t) +
            error_source_to_index[notify] * sizeof(uint64_t);

        *cper_addr = base +
            ACPI_GHES_ERROR_SOURCE_COUNT * sizeof(uint64_t) +
            ACPI_GHES_ERROR_SOURCE_COUNT * sizeof(uint64_t) +
            error_source_to_index[notify] * ACPI_GHES_MAX_RAW_DATA_LENGTH;
   } else {
        // Use the new logic from ags->hest_addr_le

}

There are two problems with that:

1. On your reviews, if I understood right, the code above is not migration safe. So, while implementing it would be technically correct, migration still won't work;

2. With the new code, ACPI_GHES_ERROR_SOURCE_COUNT is not defined anymore, as the size of the error source structure can be different on different architectures, being 2 for new VMs and 1 for old ones.

Basically the new code gets it right because it can see a pointer to the HEST table, so it can get the number from there:

	hest_addr = le64_to_cpu(ags->hest_addr_le);
	cpu_physical_memory_read(hest_addr, &num_sources, sizeof(num_sources));

But, without hest_addr_le, getting num_sources is not possible.

An alternative would be to add a hacky code that works only for arm machines (as new versions may support more archs).

Something like:

	#define V1_ARM_ACPI_GHES_ERROR_SOURCE_COUNT 1
	#define V2_ARM_ACPI_GHES_ERROR_SOURCE_COUNT 2

And have a hardcoded logic that would work before/after this changeset but may break on newer versions, if the number of source IDs change, if we add other HEST types, etc.

Now, assuming that such hack would work, it sounds too hacky to my taste.

Quoted 6 lines

So, IMO it is a lot safer to not support migrations from v1 (only ghes_addr_le), using a patch like the enclosed one to ensure that.

Btw, checking existing migration structs, it sounds that for almost all structures, .version_id is identical to .minimum_version_id, meaning that migration between different versions aren't supported on most cases.

let me try to find more examples on how to implement migration bits hopefully more comprehensible.

Quoted 53 lines

Thanks,
Mauro

---

[PATCH] acpi/generic_event_device: Update GHES migration to cover hest addr

The GHES migration logic at GED should now support HEST table location too.

Increase migration version and change needed to check for both ghes_addr_le and hest_addr_le.

hw/acpi/generic_event_device.c +6 -5
index b4c83a089a02..efae0ff62c7b 100644
@@ -351,10 +351,11 @@ static const VMStateDescription vmstate_ged_state = {
 
 static const VMStateDescription vmstate_ghes = {
     .name = "acpi-ghes",
-    .version_id = 1,
-    .minimum_version_id = 1,
+    .version_id = 2,
+    .minimum_version_id = 2,
     .fields = (const VMStateField[]) {
         VMSTATE_UINT64(ghes_addr_le, AcpiGhesState),
+        VMSTATE_UINT64(hest_addr_le, AcpiGhesState),
         VMSTATE_END_OF_LIST()
     },
 };
@@ -362,13 +363,13 @@ static const VMStateDescription vmstate_ghes = {
 static bool ghes_needed(void *opaque)
 {
     AcpiGedState *s = opaque;
-    return s->ghes_state.ghes_addr_le;
+    return s->ghes_state.ghes_addr_le && s->ghes_state.hest_addr_le;
 }
 
 static const VMStateDescription vmstate_ghes_state = {
     .name = "acpi-ged/ghes",
-    .version_id = 1,
-    .minimum_version_id = 1,
+    .version_id = 2,
+    .minimum_version_id = 2,
     .needed = ghes_needed,
     .fields = (const VMStateField[]) {
         VMSTATE_STRUCT(ghes_state, AcpiGedState, 1,
#7 lore raw

Re: [PATCH v9 01/12] acpi/ghes: add a firmware file with HEST address

Em Mon, 16 Sep 2024 13:05:06 +0200
Igor Mammedov <[email protected]> escreveu:

Quoted 5 lines
Quoted 1 line

But no idea how to start a machine using a saved state.

see https://www.linux-kvm.org/page/Migration
'savevm/loadvm to an external state file (using pseudo-migration)' section

It didn't work. Is migration currently working between 9.1 and 9.2?

I did a compilation of qemu version v9.1.0-rc0 and saved the state.

Then, on vanilla 9.2 (changeset 01dc65a3bc26), I tried to restore the state with both "virt" and "virt-9.1". None worked:

$ qemu-system-aarch64 -m 4g,maxmem=8G,slots=8 -M type=virt-9.1,nvdimm=on,gic-version=3,ras=on -cpu max -smp 4 -numa node,nodeid=0,cpus=0-3,memdev=mem0 --nographic -monitor telnet:127.0.0.1:1234,server,nowait -incoming "exec: gzip -c -d statefile.gz" -no-reboot -bios /new_devel/edac/emulator/QEMU_EFI-silent.fd -kernel /new_devel/edac/work/arm64_build/arch/arm64/boot/Image.gz -device pcie-root-port,id=root_port1 -device virtio-blk-pci,drive=hd -device virtio-net-pci,netdev=mynet,id=bob -drive if=none,file=/new_devel/edac/emulator/debian.qcow2,format=qcow2,id=hd -object memory-backend-ram,size=4G,id=mem0 -netdev type=user,id=mynet,hostfwd=tcp::5555-:22 -qmp tcp:localhost:4445,server=on,wait=off -append 'earlycon nomodeset root=/dev/vda1 fsck.mode=skip tp_printk maxcpus=4'
qemu-system-aarch64: error while loading state for instance 0x0 of device 'cpu'
qemu-system-aarch64: load of migration failed: Operation not permitted
$ qemu-system-aarch64 -m 4g,maxmem=8G,slots=8 -M type=virt,nvdimm=on,gic-version=3,ras=on -cpu max -smp 4 -numa node,nodeid=0,cpus=0-3,memdev=mem0 --nographic -monitor telnet:127.0.0.1:1234,server,nowait -incoming "exec: gzip -c -d statefile.gz" -no-reboot -bios /new_devel/edac/emulator/QEMU_EFI-silent.fd -kernel /new_devel/edac/work/arm64_build/arch/arm64/boot/Image.gz -device pcie-root-port,id=root_port1 -device virtio-blk-pci,drive=hd -device virtio-net-pci,netdev=mynet,id=bob -drive if=none,file=/new_devel/edac/emulator/debian.qcow2,format=qcow2,id=hd -object memory-backend-ram,size=4G,id=mem0 -netdev type=user,id=mynet,hostfwd=tcp::5555-:22 -qmp tcp:localhost:4445,server=on,wait=off -append 'earlycon nomodeset root=/dev/vda1 fsck.mode=skip tp_printk maxcpus=4'
qemu-system-aarch64: Machine type received is 'virt-9.1' and local is 'virt-9.2'
qemu-system-aarch64: load of migration failed: Invalid argument

Did I made something wrong?

Regards,
Mauro

#8 lore raw

Re: [PATCH v9 01/12] acpi/ghes: add a firmware file with HEST address

Quoted 5 lines · On Tue, 1 Oct 2024 10:54:26 +0200 Mauro Carvalho Chehab…

Em Mon, 16 Sep 2024 13:05:06 +0200
Igor Mammedov <[email protected]> escreveu:

Quoted 2 lines
Quoted 1 line

But no idea how to start a machine using a saved state.

Quoted 7 lines
Quoted 3 lines

see https://www.linux-kvm.org/page/Migration
'savevm/loadvm to an external state file (using pseudo-migration)' section

It didn't work. Is migration currently working between 9.1 and 9.2?

I did a compilation of qemu version v9.1.0-rc0 and saved the state.

but it's better to use actually released 9.1 source code. that should be effectively virt-9.1 machine type (unless something changed between rc0 and release version)

show us CLI on source side (i.e. qemu instance where you saved state, it should be the same as below modulo '-incoming' option)

Quoted 8 lines

Then, on vanilla 9.2 (changeset 01dc65a3bc26), I tried to restore the state with both "virt" and "virt-9.1". None worked:

$ qemu-system-aarch64 -m 4g,maxmem=8G,slots=8 -M type=virt-9.1,nvdimm=on,gic-version=3,ras=on -cpu max -smp 4 -numa node,nodeid=0,cpus=0-3,memdev=mem0 --nographic -monitor telnet:127.0.0.1:1234,server,nowait -incoming "exec: gzip -c -d statefile.gz" -no-reboot -bios /new_devel/edac/emulator/QEMU_EFI-silent.fd -kernel /new_devel/edac/work/arm64_build/arch/arm64/boot/Image.gz -device pcie-root-port,id=root_port1 -device virtio-blk-pci,drive=hd -device virtio-net-pci,netdev=mynet,id=bob -drive if=none,file=/new_devel/edac/emulator/debian.qcow2,format=qcow2,id=hd -object memory-backend-ram,size=4G,id=mem0 -netdev type=user,id=mynet,hostfwd=tcp::5555-:22 -qmp tcp:localhost:4445,server=on,wait=off -append 'earlycon nomodeset root=/dev/vda1 fsck.mode=skip tp_printk maxcpus=4'
qemu-system-aarch64: error while loading state for instance 0x0 of device 'cpu'
qemu-system-aarch64: load of migration failed: Operation not permitted

and that should've worked

Quoted 3 lines
$ qemu-system-aarch64 -m 4g,maxmem=8G,slots=8 -M type=virt,nvdimm=on,gic-version=3,ras=on -cpu max -smp 4 -numa node,nodeid=0,cpus=0-3,memdev=mem0 --nographic -monitor telnet:127.0.0.1:1234,server,nowait -incoming "exec: gzip -c -d statefile.gz" -no-reboot -bios /new_devel/edac/emulator/QEMU_EFI-silent.fd -kernel /new_devel/edac/work/arm64_build/arch/arm64/boot/Image.gz -device pcie-root-port,id=root_port1 -device virtio-blk-pci,drive=hd -device virtio-net-pci,netdev=mynet,id=bob -drive if=none,file=/new_devel/edac/emulator/debian.qcow2,format=qcow2,id=hd -object memory-backend-ram,size=4G,id=mem0 -netdev type=user,id=mynet,hostfwd=tcp::5555-:22 -qmp tcp:localhost:4445,server=on,wait=off -append 'earlycon nomodeset root=/dev/vda1 fsck.mode=skip tp_printk maxcpus=4'
qemu-system-aarch64: Machine type received is 'virt-9.1' and local is 'virt-9.2'
qemu-system-aarch64: load of migration failed: Invalid argument

that isn't meant to work.

Quoted 2 lines

Did I made something wrong?

let's see src CLI first, and then ask someone who knows more about migration and how to troubleshoot it

Quoted 3 lines

Regards,
Mauro

#9 lore raw

PATCH v9 2/12 acpi/ghes: rework the logic to handle HEST source ID

In reply to Mauro Carvalho Chehab

The current logic is based on a lot of duct tape, with offsets calculated based on one define with the number of source IDs and an enum.

Rewrite the logic in a way that it would be more resilient of code changes, by moving the source ID count to an enum and make the offset calculus more explicit.

Such change was inspired on a patch from Jonathan Cameron splitting the logic to get the CPER address on a separate function, as this will be needed to support generic error injection.

---

Changes from v8:
- Non-rename/cleanup changes merged altogether;
- source ID is now more generic, defined per guest target. That should make easier to add support for 86.

---
 hw/acpi/ghes.c           | 275 ++++++++++++++++++++++++---------------
 hw/arm/virt-acpi-build.c |  10 +-
 include/hw/acpi/ghes.h   |  18 +--
 include/hw/arm/virt.h    |   7 +
 target/arm/kvm.c         |   3 +-
 5 files changed, 198 insertions(+), 115 deletions(-)
hw/acpi/ghes.c +173 -102
index 529c14e3289f..965fb1b36587 100644
@@ -35,9 +35,6 @@
 /* The max size in bytes for one error block */
 #define ACPI_GHES_MAX_RAW_DATA_LENGTH   (1 * KiB)
 
-/* Now only support ARMv8 SEA notification type error source */
-#define ACPI_GHES_ERROR_SOURCE_COUNT        1
-
 /* Generic Hardware Error Source version 2 */
 #define ACPI_GHES_SOURCE_GENERIC_ERROR_V2   10
 
@@ -64,6 +61,19 @@
  */
 #define ACPI_GHES_GESB_SIZE                 20
 
+/*
+ * Offsets with regards to the start of the HEST table stored at
+ * ags->hest_addr_le, according with the memory layout map at
+ * docs/specs/acpi_hest_ghes.rst.
+ */
+
+/* ACPI 6.2: 18.3.2.8 Generic Hardware Error Source version 2 */
+#define HEST_GHES_V2_TABLE_SIZE  92
+#define GHES_ACK_OFFSET          (64 + GAS_ADDR_OFFSET)
+
+/* ACPI 6.2: 18.3.2.7: Generic Hardware Error Source */
+#define GHES_ERR_ST_ADDR_OFFSET  (20 + GAS_ADDR_OFFSET)
+
 /*
  * Values for error_severity field
  */
@@ -185,51 +195,30 @@ static void acpi_ghes_build_append_mem_cper(GArray *table,
     build_append_int_noprefix(table, 0, 7);
 }
 
-static int acpi_ghes_record_mem_error(uint64_t error_block_address,
-                                      uint64_t error_physical_addr)
+static void
+ghes_gen_err_data_uncorrectable_recoverable(GArray *block,
+                                            const uint8_t *section_type,
+                                            int data_length)
 {
-    GArray *block;
-
-    /* Memory Error Section Type */
-    const uint8_t uefi_cper_mem_sec[] =
-          UUID_LE(0xA5BC1114, 0x6F64, 0x4EDE, 0xB8, 0x63, 0x3E, 0x83, \
-                  0xED, 0x7C, 0x83, 0xB1);
-
     /* invalid fru id: ACPI 4.0: 17.3.2.6.1 Generic Error Data,
      * Table 17-13 Generic Error Data Entry
      */
     QemuUUID fru_id = {};
-    uint32_t data_length;
 
-    block = g_array_new(false, true /* clear */, 1);
-
-    /* This is the length if adding a new generic error data entry*/
-    data_length = ACPI_GHES_DATA_LENGTH + ACPI_GHES_MEM_CPER_LENGTH;
     /*
-     * It should not run out of the preallocated memory if adding a new generic
-     * error data entry
+     * Calculate the size with this block. No need to check for
+     * too big CPER, as CPER size is checked at ghes_record_cper_errors()
      */
-    assert((data_length + ACPI_GHES_GESB_SIZE) <=
-            ACPI_GHES_MAX_RAW_DATA_LENGTH);
+    data_length += ACPI_GHES_GESB_SIZE;
 
     /* Build the new generic error status block header */
     acpi_ghes_generic_error_status(block, ACPI_GEBS_UNCORRECTABLE,
         0, 0, data_length, ACPI_CPER_SEV_RECOVERABLE);
 
     /* Build this new generic error data entry header */
-    acpi_ghes_generic_error_data(block, uefi_cper_mem_sec,
+    acpi_ghes_generic_error_data(block, section_type,
         ACPI_CPER_SEV_RECOVERABLE, 0, 0,
         ACPI_GHES_MEM_CPER_LENGTH, fru_id, 0);
-
-    /* Build the memory section CPER for above new generic error data entry */
-    acpi_ghes_build_append_mem_cper(block, error_physical_addr);
-
-    /* Write the generic error data entry into guest memory */
-    cpu_physical_memory_write(error_block_address, block->data, block->len);
-
-    g_array_free(block, true);
-
-    return 0;
 }
 
 /*
@@ -237,17 +226,18 @@ static int acpi_ghes_record_mem_error(uint64_t error_block_address,
  * Initialize "etc/hardware_errors" and "etc/hardware_errors_addr" fw_cfg blobs.
  * See docs/specs/acpi_hest_ghes.rst for blobs format.
  */
-void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker)
+static void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker,
+                                   int num_sources)
 {
     int i, error_status_block_offset;
 
     /* Build error_block_address */
-    for (i = 0; i < ACPI_GHES_ERROR_SOURCE_COUNT; i++) {
+    for (i = 0; i < num_sources; i++) {
         build_append_int_noprefix(hardware_errors, 0, sizeof(uint64_t));
     }
 
     /* Build read_ack_register */
-    for (i = 0; i < ACPI_GHES_ERROR_SOURCE_COUNT; i++) {
+    for (i = 0; i < num_sources; i++) {
         /*
          * Initialize the value of read_ack_register to 1, so GHES can be
          * writable after (re)boot.
@@ -262,13 +252,13 @@ void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker)
 
     /* Reserve space for Error Status Data Block */
     acpi_data_push(hardware_errors,
-        ACPI_GHES_MAX_RAW_DATA_LENGTH * ACPI_GHES_ERROR_SOURCE_COUNT);
+        ACPI_GHES_MAX_RAW_DATA_LENGTH * num_sources);
 
     /* Tell guest firmware to place hardware_errors blob into RAM */
     bios_linker_loader_alloc(linker, ACPI_GHES_ERRORS_FW_CFG_FILE,
                              hardware_errors, sizeof(uint64_t), false);
 
-    for (i = 0; i < ACPI_GHES_ERROR_SOURCE_COUNT; i++) {
+    for (i = 0; i < num_sources; i++) {
         /*
          * Tell firmware to patch error_block_address entries to point to
          * corresponding "Generic Error Status Block"
@@ -283,14 +273,20 @@ void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker)
      * tell firmware to write hardware_errors GPA into
      * hardware_errors_addr fw_cfg, once the former has been initialized.
      */
-    bios_linker_loader_write_pointer(linker, ACPI_GHES_DATA_ADDR_FW_CFG_FILE,
-        0, sizeof(uint64_t), ACPI_GHES_ERRORS_FW_CFG_FILE, 0);
+    bios_linker_loader_write_pointer(linker, ACPI_GHES_DATA_ADDR_FW_CFG_FILE, 0,
+                                     sizeof(uint64_t),
+                                     ACPI_GHES_ERRORS_FW_CFG_FILE, 0);
 }
 
 /* Build Generic Hardware Error Source version 2 (GHESv2) */
-static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker)
+static void build_ghes_v2(GArray *table_data,
+                          BIOSLinker *linker,
+                          enum AcpiGhesNotifyType notify,
+                          uint16_t source_id,
+                          int num_sources)
 {
     uint64_t address_offset;
+
     /*
      * Type:
      * Generic Hardware Error Source version 2(GHESv2 - Type 10)
@@ -317,21 +313,13 @@ static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker)
     build_append_gas(table_data, AML_AS_SYSTEM_MEMORY, 0x40, 0,
                      4 /* QWord access */, 0);
     bios_linker_loader_add_pointer(linker, ACPI_BUILD_TABLE_FILE,
-        address_offset + GAS_ADDR_OFFSET, sizeof(uint64_t),
-        ACPI_GHES_ERRORS_FW_CFG_FILE, source_id * sizeof(uint64_t));
+                                   address_offset + GAS_ADDR_OFFSET,
+                                   sizeof(uint64_t),
+                                   ACPI_GHES_ERRORS_FW_CFG_FILE,
+                                   source_id * sizeof(uint64_t));
 
-    switch (source_id) {
-    case ACPI_HEST_SRC_ID_SEA:
-        /*
-         * Notification Structure
-         * Now only enable ARMv8 SEA notification type
-         */
-        build_ghes_hw_error_notification(table_data, ACPI_GHES_NOTIFY_SEA);
-        break;
-    default:
-        error_report("Not support this error source");
-        abort();
-    }
+    /* Notification Structure */
+    build_ghes_hw_error_notification(table_data, notify);
 
     /* Error Status Block Length */
     build_append_int_noprefix(table_data, ACPI_GHES_MAX_RAW_DATA_LENGTH, 4);
@@ -345,9 +333,11 @@ static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker)
     build_append_gas(table_data, AML_AS_SYSTEM_MEMORY, 0x40, 0,
                      4 /* QWord access */, 0);
     bios_linker_loader_add_pointer(linker, ACPI_BUILD_TABLE_FILE,
-        address_offset + GAS_ADDR_OFFSET,
-        sizeof(uint64_t), ACPI_GHES_ERRORS_FW_CFG_FILE,
-        (ACPI_GHES_ERROR_SOURCE_COUNT + source_id) * sizeof(uint64_t));
+                                   address_offset + GAS_ADDR_OFFSET,
+                                   sizeof(uint64_t),
+                                   ACPI_GHES_ERRORS_FW_CFG_FILE,
+                                   (num_sources + source_id) *
+                                   sizeof(uint64_t));
 
     /*
      * Read Ack Preserve field
@@ -360,19 +350,28 @@ static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker)
 }
 
 /* Build Hardware Error Source Table */
-void acpi_build_hest(GArray *table_data, BIOSLinker *linker,
+void acpi_build_hest(GArray *table_data, GArray *hardware_errors,
+                     BIOSLinker *linker,
+                     const uint16_t * const notify,
+                     int num_sources,
                      const char *oem_id, const char *oem_table_id)
 {
     AcpiTable table = { .sig = "HEST", .rev = 1,
                         .oem_id = oem_id, .oem_table_id = oem_table_id };
+    int i;
+
+    build_ghes_error_table(hardware_errors, linker, num_sources);
 
     acpi_table_begin(&table, table_data);
 
+    /* Beginning at the HEST Error Source struct count and data */
     int hest_offset = table_data->len;
 
     /* Error Source Count */
-    build_append_int_noprefix(table_data, ACPI_GHES_ERROR_SOURCE_COUNT, 4);
-    build_ghes_v2(table_data, ACPI_HEST_SRC_ID_SEA, linker);
+    build_append_int_noprefix(table_data, num_sources, 4);
+    for (i = 0; i < num_sources; i++) {
+        build_ghes_v2(table_data, linker, notify[i], i, num_sources);
+    }
 
     acpi_table_end(linker, &table);
 
@@ -403,60 +402,132 @@ void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
     ags->present = true;
 }
 
-int acpi_ghes_record_errors(uint8_t source_id, uint64_t physical_address)
+void ghes_record_cper_errors(const void *cper, size_t len,
+                             uint16_t source_id, Error **errp)
 {
-    uint64_t error_block_addr, read_ack_register_addr, read_ack_register = 0;
-    uint64_t start_addr;
-    bool ret = -1;
+    uint64_t hest_read_ack_start_addr, read_ack_start_addr;
+    uint64_t hest_addr, cper_addr, err_source_struct;
+    uint64_t hest_err_block_addr, error_block_addr;
+    uint32_t num_sources, i;
     AcpiGedState *acpi_ged_state;
     AcpiGhesState *ags;
+    uint64_t read_ack;
 
-    assert(source_id < ACPI_HEST_SRC_ID_RESERVED);
+    if (len > ACPI_GHES_MAX_RAW_DATA_LENGTH) {
+        error_setg(errp, "GHES CPER record is too big: %ld", len);
+    }
 
     acpi_ged_state = ACPI_GED(object_resolve_path_type("", TYPE_ACPI_GED,
                                                        NULL));
     g_assert(acpi_ged_state);
     ags = &acpi_ged_state->ghes_state;
 
-    start_addr = le64_to_cpu(ags->ghes_addr_le);
-
-    if (physical_address) {
-
-        if (source_id < ACPI_HEST_SRC_ID_RESERVED) {
-            start_addr += source_id * sizeof(uint64_t);
-        }
-
-        cpu_physical_memory_read(start_addr, &error_block_addr,
-                                 sizeof(error_block_addr));
-
-        error_block_addr = le64_to_cpu(error_block_addr);
-
-        read_ack_register_addr = start_addr +
-            ACPI_GHES_ERROR_SOURCE_COUNT * sizeof(uint64_t);
-
-        cpu_physical_memory_read(read_ack_register_addr,
-                                 &read_ack_register, sizeof(read_ack_register));
-
-        /* zero means OSPM does not acknowledge the error */
-        if (!read_ack_register) {
-            error_report("OSPM does not acknowledge previous error,"
-                " so can not record CPER for current error anymore");
-        } else if (error_block_addr) {
-            read_ack_register = cpu_to_le64(0);
-            /*
-             * Clear the Read Ack Register, OSPM will write it to 1 when
-             * it acknowledges this error.
-             */
-            cpu_physical_memory_write(read_ack_register_addr,
-                &read_ack_register, sizeof(uint64_t));
-
-            ret = acpi_ghes_record_mem_error(error_block_addr,
-                                             physical_address);
-        } else
-            error_report("can not find Generic Error Status Block");
+    hest_addr = le64_to_cpu(ags->hest_addr_le);
+
+    cpu_physical_memory_read(hest_addr, &num_sources, sizeof(num_sources));
+
+    if (source_id >= num_sources) {
+        error_setg(errp,
+                   "GHES: Source %d not found. Only %d sources are defined",
+                   source_id, num_sources);
+        return;
+    }
+    err_source_struct = hest_addr + sizeof(num_sources);
+
+    for (i = 0; i < num_sources; i++) {
+        uint64_t addr = err_source_struct;
+        uint16_t type, src_id;
+
+        cpu_physical_memory_read(addr, &type, sizeof(type));
+
+        /* For now, we only know the size of GHESv2 table */
+        assert(type == ACPI_GHES_SOURCE_GENERIC_ERROR_V2);
+
+        /* It is GHES. Compare CPER source address */
+        addr += sizeof(type);
+        cpu_physical_memory_read(addr, &src_id, sizeof(src_id));
+
+        if (src_id == source_id)
+            break;
+
+        err_source_struct += HEST_GHES_V2_TABLE_SIZE;
+    }
+    if (i == num_sources) {
+        error_setg(errp, "HEST: Source %d not found.", source_id);
+        return;
+    }
+
+    /* Check if BIOS addr pointers were properly generated */
+
+    hest_err_block_addr = err_source_struct + GHES_ERR_ST_ADDR_OFFSET;
+    hest_read_ack_start_addr = err_source_struct + GHES_ACK_OFFSET;
+
+    cpu_physical_memory_read(hest_err_block_addr, &error_block_addr,
+                             sizeof(error_block_addr));
+
+    cpu_physical_memory_read(error_block_addr, &cper_addr,
+                             sizeof(error_block_addr));
+
+    cpu_physical_memory_read(hest_read_ack_start_addr, &read_ack_start_addr,
+			     sizeof(read_ack_start_addr));
+
+    /* Update ACK offset to notify about a new error */
+
+    cpu_physical_memory_read(read_ack_start_addr,
+                             &read_ack, sizeof(read_ack));
+
+    /* zero means OSPM does not acknowledge the error */
+    if (!read_ack) {
+        error_setg(errp,
+                   "Last CPER record was not acknowledged yet");
+        read_ack = 1;
+        cpu_physical_memory_write(read_ack_start_addr,
+                                  &read_ack, sizeof(read_ack));
+        return;
+    }
+
+    read_ack = cpu_to_le64(0);
+    cpu_physical_memory_write(read_ack_start_addr,
+                              &read_ack, sizeof(read_ack));
+
+    /* Write the generic error data entry into guest memory */
+    cpu_physical_memory_write(cper_addr, cper, len);
+}
+
+int acpi_ghes_record_errors(int source_id, uint64_t physical_address)
+{
+    /* Memory Error Section Type */
+    const uint8_t guid[] =
+          UUID_LE(0xA5BC1114, 0x6F64, 0x4EDE, 0xB8, 0x63, 0x3E, 0x83, \
+                  0xED, 0x7C, 0x83, 0xB1);
+    Error *errp = NULL;
+    GArray *block;
+
+    if (!physical_address) {
+        error_report("can not find Generic Error Status Block for source id %d",
+                     source_id);
+        return -1;
+    }
+
+    block = g_array_new(false, true /* clear */, 1);
+
+    ghes_gen_err_data_uncorrectable_recoverable(block, guid,
+                                                ACPI_GHES_MAX_RAW_DATA_LENGTH);
+
+    /* Build the memory section CPER for above new generic error data entry */
+    acpi_ghes_build_append_mem_cper(block, physical_address);
+
+    /* Report the error */
+    ghes_record_cper_errors(block->data, block->len, source_id, &errp);
+
+    g_array_free(block, true);
+
+    if (errp) {
+        error_report_err(errp);
+        return -1;
     }
 
-    return ret;
+    return 0;
 }
 
 bool acpi_ghes_present(void)
hw/arm/virt-acpi-build.c +7 -3
index f76fb117adff..39100c2822c2 100644
@@ -890,6 +890,10 @@ static void acpi_align_size(GArray *blob, unsigned align)
     g_array_set_size(blob, ROUND_UP(acpi_data_len(blob), align));
 }
 
+static const uint16_t hest_ghes_notify[] = {
+    [ARM_ACPI_HEST_SRC_ID_SEA] = ACPI_GHES_NOTIFY_SEA,
+};
+
 static
 void virt_acpi_build(VirtMachineState *vms, AcpiBuildTables *tables)
 {
@@ -943,10 +947,10 @@ void virt_acpi_build(VirtMachineState *vms, AcpiBuildTables *tables)
     build_dbg2(tables_blob, tables->linker, vms);
 
     if (vms->ras) {
-        build_ghes_error_table(tables->hardware_errors, tables->linker);
         acpi_add_table(table_offsets, tables_blob);
-        acpi_build_hest(tables_blob, tables->linker, vms->oem_id,
-                        vms->oem_table_id);
+        acpi_build_hest(tables_blob, tables->hardware_errors, tables->linker,
+                        hest_ghes_notify, sizeof(hest_ghes_notify),
+                        vms->oem_id, vms->oem_table_id);
     }
 
     if (ms->numa_state->num_nodes > 0) {
include/hw/acpi/ghes.h +9 -9
index 28b956acb19a..4b5af86ec077 100644
@@ -23,6 +23,7 @@
 #define ACPI_GHES_H
 
 #include "hw/acpi/bios-linker-loader.h"
+#include "qapi/error.h"
 
 /*
  * Values for Hardware Error Notification Type field
@@ -56,24 +57,23 @@ enum AcpiGhesNotifyType {
     ACPI_GHES_NOTIFY_RESERVED = 12
 };
 
-enum {
-    ACPI_HEST_SRC_ID_SEA = 0,
-    /* future ids go here */
-    ACPI_HEST_SRC_ID_RESERVED,
-};
-
 typedef struct AcpiGhesState {
     uint64_t hest_addr_le;
     uint64_t ghes_addr_le;
     bool present; /* True if GHES is present at all on this board */
 } AcpiGhesState;
 
-void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker);
-void acpi_build_hest(GArray *table_data, BIOSLinker *linker,
+void acpi_build_hest(GArray *table_data, GArray *hardware_errors,
+                     BIOSLinker *linker,
+                     const uint16_t * const notify,
+                     int num_sources,
                      const char *oem_id, const char *oem_table_id);
 void acpi_ghes_add_fw_cfg(AcpiGhesState *vms, FWCfgState *s,
                           GArray *hardware_errors);
-int acpi_ghes_record_errors(uint8_t notify, uint64_t error_physical_addr);
+int acpi_ghes_record_errors(int source_id,
+                            uint64_t error_physical_addr);
+void ghes_record_cper_errors(const void *cper, size_t len,
+                             uint16_t source_id, Error **errp);
 
 /**
  * acpi_ghes_present: Report whether ACPI GHES table is present
include/hw/arm/virt.h +7 -0
index a4d937ed45ac..d62d8d9db5ae 100644
@@ -189,6 +189,13 @@ OBJECT_DECLARE_TYPE(VirtMachineState, VirtMachineClass, VIRT_MACHINE)
 void virt_acpi_setup(VirtMachineState *vms);
 bool virt_is_acpi_enabled(VirtMachineState *vms);
 
+/*
+ * ID numbers used to fill HEST source ID field
+ */
+enum {
+    ARM_ACPI_HEST_SRC_ID_SEA,
+};
+
 /* Return number of redistributors that fit in the specified region */
 static uint32_t virt_redist_capacity(VirtMachineState *vms, int region)
 {
target/arm/kvm.c +2 -1
index 849e2e21b304..8c4c8263b85a 100644
@@ -2373,7 +2373,8 @@ void kvm_arch_on_sigbus_vcpu(CPUState *c, int code, void *addr)
              */
             if (code == BUS_MCEERR_AR) {
                 kvm_cpu_synchronize_state(c);
-                if (!acpi_ghes_record_errors(ACPI_HEST_SRC_ID_SEA, paddr)) {
+                if (!acpi_ghes_record_errors(ARM_ACPI_HEST_SRC_ID_SEA,
+                                             paddr)) {
                     kvm_inject_arm_sea(c);
                 } else {
                     error_report("failed to record the error");
Signature · 1 line
2.46.0
#10 lore raw

Re: [PATCH v9 02/12] acpi/ghes: rework the logic to handle HEST source ID

Quoted 4 lines · On Sun, 25 Aug 2024 05:45:57 +0200 Mauro Carvalho Chehab…

The current logic is based on a lot of duct tape, with offsets calculated based on one define with the number of source IDs and an enum.

Quoted 8 lines

Rewrite the logic in a way that it would be more resilient of code changes, by moving the source ID count to an enum and make the offset calculus more explicit.

Such change was inspired on a patch from Jonathan Cameron splitting the logic to get the CPER address on a separate function, as this will be needed to support generic error injection.

patch is too large and does too many things at once, see inline suggestions on how to split it in more manageable chunks.
(I'll mark preferred patch order with numbers)

Quoted 24 lines

---

Changes from v8:
- Non-rename/cleanup changes merged altogether;
- source ID is now more generic, defined per guest target. That should make easier to add support for 86.

---
 hw/acpi/ghes.c           | 275 ++++++++++++++++++++++++---------------
 hw/arm/virt-acpi-build.c |  10 +-
 include/hw/acpi/ghes.h   |  18 +--
 include/hw/arm/virt.h    |   7 +
 target/arm/kvm.c         |   3 +-
 5 files changed, 198 insertions(+), 115 deletions(-)
hw/acpi/ghes.c +0 -0
index 529c14e3289f..965fb1b36587 100644

hw/acpi/ghes.c line 39

Quoted 5 lines
hw/acpi/ghes.c +0 -2
@@ -35,9 +35,6 @@
 /* The max size in bytes for one error block */
 #define ACPI_GHES_MAX_RAW_DATA_LENGTH   (1 * KiB)
 
-/* Now only support ARMv8 SEA notification type error source */
-#define ACPI_GHES_ERROR_SOURCE_COUNT        1

[patch 4] getting rid of this and introducing num_sources
(aka variable size HEST)

hw/acpi/ghes.c line 68

Quoted 12 lines
hw/acpi/ghes.c +5 -0
 /* Generic Hardware Error Source version 2 */
 #define ACPI_GHES_SOURCE_GENERIC_ERROR_V2   10
 
@@ -64,6 +61,19 @@
  */
 #define ACPI_GHES_GESB_SIZE                 20
 
+/*
+ * Offsets with regards to the start of the HEST table stored at
+ * ags->hest_addr_le, according with the memory layout map at
+ * docs/specs/acpi_hest_ghes.rst.
+ */

perhaps mention in comment/commit message, that hest lookup is implemented only GHESv2 error sources.

That will work as far we do forward migration only (i.e. old qemu -> new qemu), which is what upstream supports.

However it won't work for backward migration (new qemu -> old qemu) since old one doesn't know about new non-GHESv2 sources. And that means we would need to introduce compat knobs for every new non-GHESv2 source is added. Which is easy to overlook and it adds up to maintenance.
(You've already described zoo of types ACPI spec has in v8 review, but I don't thing it's too complex to implement lookup of all known types. compared to headache we would have with compat settings if anyone remembers)

I won't insist on adding all known sources lookup in this series, if you agree to do it as a patch on top of this series within this dev cycle (~2 months time-frame).

hw/acpi/ghes.c line 70

Quoted 1 line
hw/acpi/ghes.c +1 -0
+/* ACPI 6.2: 18.3.2.8 Generic Hardware Error Source version 2 */

+ ,Table 18-383

hw/acpi/ghes.c line 74

Quoted 4 lines
hw/acpi/ghes.c +4 -0
+#define HEST_GHES_V2_TABLE_SIZE  92
+#define GHES_ACK_OFFSET          (64 + GAS_ADDR_OFFSET)
+
+/* ACPI 6.2: 18.3.2.7: Generic Hardware Error Source */

Table 18-380 'Error Status Address' field

Quoted 6 lines
hw/acpi/ghes.c +2 -0
+#define GHES_ERR_ST_ADDR_OFFSET  (20 + GAS_ADDR_OFFSET)
+
 /*
  * Values for error_severity field
  */

hw/acpi/ghes.c static void acpi_ghes_build_append_mem_cper(GArray *table, · line 201

Quoted 9 lines
hw/acpi/ghes.c +4 -2
@@ -185,51 +195,30 @@ static void acpi_ghes_build_append_mem_cper(GArray *table,
     build_append_int_noprefix(table, 0, 7);
 }
 
-static int acpi_ghes_record_mem_error(uint64_t error_block_address,
-                                      uint64_t error_physical_addr)
+static void
+ghes_gen_err_data_uncorrectable_recoverable(GArray *block,
+                                            const uint8_t *section_type,
+                                            int data_length)

[patch 2] splitting acpi_ghes_record_mem_error() on reusable and mem specific
code

Quoted 90 lines
hw/acpi/ghes.c +10 -32
 {
-    GArray *block;
-
-    /* Memory Error Section Type */
-    const uint8_t uefi_cper_mem_sec[] =
-          UUID_LE(0xA5BC1114, 0x6F64, 0x4EDE, 0xB8, 0x63, 0x3E, 0x83, \
-                  0xED, 0x7C, 0x83, 0xB1);
-
     /* invalid fru id: ACPI 4.0: 17.3.2.6.1 Generic Error Data,
      * Table 17-13 Generic Error Data Entry
      */
     QemuUUID fru_id = {};
-    uint32_t data_length;
 
-    block = g_array_new(false, true /* clear */, 1);
-
-    /* This is the length if adding a new generic error data entry*/
-    data_length = ACPI_GHES_DATA_LENGTH + ACPI_GHES_MEM_CPER_LENGTH;
     /*
-     * It should not run out of the preallocated memory if adding a new generic
-     * error data entry
+     * Calculate the size with this block. No need to check for
+     * too big CPER, as CPER size is checked at ghes_record_cper_errors()
      */
-    assert((data_length + ACPI_GHES_GESB_SIZE) <=
-            ACPI_GHES_MAX_RAW_DATA_LENGTH);
+    data_length += ACPI_GHES_GESB_SIZE;
 
     /* Build the new generic error status block header */
     acpi_ghes_generic_error_status(block, ACPI_GEBS_UNCORRECTABLE,
         0, 0, data_length, ACPI_CPER_SEV_RECOVERABLE);
 
     /* Build this new generic error data entry header */
-    acpi_ghes_generic_error_data(block, uefi_cper_mem_sec,
+    acpi_ghes_generic_error_data(block, section_type,
         ACPI_CPER_SEV_RECOVERABLE, 0, 0,
         ACPI_GHES_MEM_CPER_LENGTH, fru_id, 0);
-
-    /* Build the memory section CPER for above new generic error data entry */
-    acpi_ghes_build_append_mem_cper(block, error_physical_addr);
-
-    /* Write the generic error data entry into guest memory */
-    cpu_physical_memory_write(error_block_address, block->data, block->len);
-
-    g_array_free(block, true);
-
-    return 0;
 }
 
 /*
@@ -237,17 +226,18 @@ static int acpi_ghes_record_mem_error(uint64_t error_block_address,
  * Initialize "etc/hardware_errors" and "etc/hardware_errors_addr" fw_cfg blobs.
  * See docs/specs/acpi_hest_ghes.rst for blobs format.
  */
-void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker)
+static void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker,
+                                   int num_sources)
 {
     int i, error_status_block_offset;
 
     /* Build error_block_address */
-    for (i = 0; i < ACPI_GHES_ERROR_SOURCE_COUNT; i++) {
+    for (i = 0; i < num_sources; i++) {
         build_append_int_noprefix(hardware_errors, 0, sizeof(uint64_t));
     }
 
     /* Build read_ack_register */
-    for (i = 0; i < ACPI_GHES_ERROR_SOURCE_COUNT; i++) {
+    for (i = 0; i < num_sources; i++) {
         /*
          * Initialize the value of read_ack_register to 1, so GHES can be
          * writable after (re)boot.
@@ -262,13 +252,13 @@ void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker)
 
     /* Reserve space for Error Status Data Block */
     acpi_data_push(hardware_errors,
-        ACPI_GHES_MAX_RAW_DATA_LENGTH * ACPI_GHES_ERROR_SOURCE_COUNT);
+        ACPI_GHES_MAX_RAW_DATA_LENGTH * num_sources);
 
     /* Tell guest firmware to place hardware_errors blob into RAM */
     bios_linker_loader_alloc(linker, ACPI_GHES_ERRORS_FW_CFG_FILE,
                              hardware_errors, sizeof(uint64_t), false);
 
-    for (i = 0; i < ACPI_GHES_ERROR_SOURCE_COUNT; i++) {
+    for (i = 0; i < num_sources; i++) {
         /*
          * Tell firmware to patch error_block_address entries to point to
          * corresponding "Generic Error Status Block"

hw/acpi/ghes.c void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker) · line 278

Quoted 8 lines
hw/acpi/ghes.c +3 -2
@@ -283,14 +273,20 @@ void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker)
      * tell firmware to write hardware_errors GPA into
      * hardware_errors_addr fw_cfg, once the former has been initialized.
      */
-    bios_linker_loader_write_pointer(linker, ACPI_GHES_DATA_ADDR_FW_CFG_FILE,
-        0, sizeof(uint64_t), ACPI_GHES_ERRORS_FW_CFG_FILE, 0);
+    bios_linker_loader_write_pointer(linker, ACPI_GHES_DATA_ADDR_FW_CFG_FILE, 0,
+                                     sizeof(uint64_t),
+                                     ACPI_GHES_ERRORS_FW_CFG_FILE, 0);

[patch 1] all indent changes in its own patch, or just drop them altogether

Quoted 16 lines
hw/acpi/ghes.c +6 -1
 }
 
 /* Build Generic Hardware Error Source version 2 (GHESv2) */
-static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker)
+static void build_ghes_v2(GArray *table_data,
+                          BIOSLinker *linker,
+                          enum AcpiGhesNotifyType notify,
+                          uint16_t source_id,
+                          int num_sources)
 {
     uint64_t address_offset;
+
     /*
      * Type:
      * Generic Hardware Error Source version 2(GHESv2 - Type 10)

hw/acpi/ghes.c static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker) · line 319

Quoted 9 lines
hw/acpi/ghes.c +4 -2
@@ -317,21 +313,13 @@ static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker)
     build_append_gas(table_data, AML_AS_SYSTEM_MEMORY, 0x40, 0,
                      4 /* QWord access */, 0);
     bios_linker_loader_add_pointer(linker, ACPI_BUILD_TABLE_FILE,
-        address_offset + GAS_ADDR_OFFSET, sizeof(uint64_t),
-        ACPI_GHES_ERRORS_FW_CFG_FILE, source_id * sizeof(uint64_t));
+                                   address_offset + GAS_ADDR_OFFSET,
+                                   sizeof(uint64_t),
+                                   ACPI_GHES_ERRORS_FW_CFG_FILE,
+                                   source_id * sizeof(uint64_t));

ditto

Quoted 19 lines
hw/acpi/ghes.c +2 -12
-    switch (source_id) {
-    case ACPI_HEST_SRC_ID_SEA:
-        /*
-         * Notification Structure
-         * Now only enable ARMv8 SEA notification type
-         */
-        build_ghes_hw_error_notification(table_data, ACPI_GHES_NOTIFY_SEA);
-        break;
-    default:
-        error_report("Not support this error source");
-        abort();
-    }
+    /* Notification Structure */
+    build_ghes_hw_error_notification(table_data, notify);
 
     /* Error Status Block Length */
     build_append_int_noprefix(table_data, ACPI_GHES_MAX_RAW_DATA_LENGTH, 4);
@@ -345,9 +333,11 @@ static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker)
     build_append_gas(table_data, AML_AS_SYSTEM_MEMORY, 0x40, 0,

hw/acpi/ghes.c static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker) · line 340

Quoted 10 lines
hw/acpi/ghes.c +5 -3
                      4 /* QWord access */, 0);
     bios_linker_loader_add_pointer(linker, ACPI_BUILD_TABLE_FILE,
-        address_offset + GAS_ADDR_OFFSET,
-        sizeof(uint64_t), ACPI_GHES_ERRORS_FW_CFG_FILE,
-        (ACPI_GHES_ERROR_SOURCE_COUNT + source_id) * sizeof(uint64_t));
+                                   address_offset + GAS_ADDR_OFFSET,
+                                   sizeof(uint64_t),
+                                   ACPI_GHES_ERRORS_FW_CFG_FILE,
+                                   (num_sources + source_id) *
+                                   sizeof(uint64_t));

ditto

Quoted 35 lines
hw/acpi/ghes.c +12 -3
     /*
      * Read Ack Preserve field
@@ -360,19 +350,28 @@ static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker)
 }
 
 /* Build Hardware Error Source Table */
-void acpi_build_hest(GArray *table_data, BIOSLinker *linker,
+void acpi_build_hest(GArray *table_data, GArray *hardware_errors,
+                     BIOSLinker *linker,
+                     const uint16_t * const notify,
+                     int num_sources,
                      const char *oem_id, const char *oem_table_id)
 {
     AcpiTable table = { .sig = "HEST", .rev = 1,
                         .oem_id = oem_id, .oem_table_id = oem_table_id };
+    int i;
+
+    build_ghes_error_table(hardware_errors, linker, num_sources);
 
     acpi_table_begin(&table, table_data);
 
+    /* Beginning at the HEST Error Source struct count and data */
     int hest_offset = table_data->len;
 
     /* Error Source Count */
-    build_append_int_noprefix(table_data, ACPI_GHES_ERROR_SOURCE_COUNT, 4);
-    build_ghes_v2(table_data, ACPI_HEST_SRC_ID_SEA, linker);
+    build_append_int_noprefix(table_data, num_sources, 4);
+    for (i = 0; i < num_sources; i++) {
+        build_ghes_v2(table_data, linker, notify[i], i, num_sources);
+    }
 
     acpi_table_end(linker, &table);
 

hw/acpi/ghes.c void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s, · line 406

Quoted 6 lines
hw/acpi/ghes.c +2 -1
@@ -403,60 +402,132 @@ void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
     ags->present = true;
 }
 
-int acpi_ghes_record_errors(uint8_t source_id, uint64_t physical_address)
+void ghes_record_cper_errors(const void *cper, size_t len,
+                             uint16_t source_id, Error **errp)

[patch 3] switching to hest source id lookup method

Quoted 173 lines
hw/acpi/ghes.c +112 -41
 {
-    uint64_t error_block_addr, read_ack_register_addr, read_ack_register = 0;
-    uint64_t start_addr;
-    bool ret = -1;
+    uint64_t hest_read_ack_start_addr, read_ack_start_addr;
+    uint64_t hest_addr, cper_addr, err_source_struct;
+    uint64_t hest_err_block_addr, error_block_addr;
+    uint32_t num_sources, i;
     AcpiGedState *acpi_ged_state;
     AcpiGhesState *ags;
+    uint64_t read_ack;
 
-    assert(source_id < ACPI_HEST_SRC_ID_RESERVED);
+    if (len > ACPI_GHES_MAX_RAW_DATA_LENGTH) {
+        error_setg(errp, "GHES CPER record is too big: %ld", len);
+    }
 
     acpi_ged_state = ACPI_GED(object_resolve_path_type("", TYPE_ACPI_GED,
                                                        NULL));
     g_assert(acpi_ged_state);
     ags = &acpi_ged_state->ghes_state;
 
-    start_addr = le64_to_cpu(ags->ghes_addr_le);
-
-    if (physical_address) {
-
-        if (source_id < ACPI_HEST_SRC_ID_RESERVED) {
-            start_addr += source_id * sizeof(uint64_t);
-        }
-
-        cpu_physical_memory_read(start_addr, &error_block_addr,
-                                 sizeof(error_block_addr));
-
-        error_block_addr = le64_to_cpu(error_block_addr);
-
-        read_ack_register_addr = start_addr +
-            ACPI_GHES_ERROR_SOURCE_COUNT * sizeof(uint64_t);
-
-        cpu_physical_memory_read(read_ack_register_addr,
-                                 &read_ack_register, sizeof(read_ack_register));
-
-        /* zero means OSPM does not acknowledge the error */
-        if (!read_ack_register) {
-            error_report("OSPM does not acknowledge previous error,"
-                " so can not record CPER for current error anymore");
-        } else if (error_block_addr) {
-            read_ack_register = cpu_to_le64(0);
-            /*
-             * Clear the Read Ack Register, OSPM will write it to 1 when
-             * it acknowledges this error.
-             */
-            cpu_physical_memory_write(read_ack_register_addr,
-                &read_ack_register, sizeof(uint64_t));
-
-            ret = acpi_ghes_record_mem_error(error_block_addr,
-                                             physical_address);
-        } else
-            error_report("can not find Generic Error Status Block");
+    hest_addr = le64_to_cpu(ags->hest_addr_le);
+
+    cpu_physical_memory_read(hest_addr, &num_sources, sizeof(num_sources));
+
+    if (source_id >= num_sources) {
+        error_setg(errp,
+                   "GHES: Source %d not found. Only %d sources are defined",
+                   source_id, num_sources);
+        return;
+    }
+    err_source_struct = hest_addr + sizeof(num_sources);
+
+    for (i = 0; i < num_sources; i++) {
+        uint64_t addr = err_source_struct;
+        uint16_t type, src_id;
+
+        cpu_physical_memory_read(addr, &type, sizeof(type));
+
+        /* For now, we only know the size of GHESv2 table */
+        assert(type == ACPI_GHES_SOURCE_GENERIC_ERROR_V2);
+
+        /* It is GHES. Compare CPER source address */
+        addr += sizeof(type);
+        cpu_physical_memory_read(addr, &src_id, sizeof(src_id));
+
+        if (src_id == source_id)
+            break;
+
+        err_source_struct += HEST_GHES_V2_TABLE_SIZE;
+    }
+    if (i == num_sources) {
+        error_setg(errp, "HEST: Source %d not found.", source_id);
+        return;
+    }
+
+    /* Check if BIOS addr pointers were properly generated */
+
+    hest_err_block_addr = err_source_struct + GHES_ERR_ST_ADDR_OFFSET;
+    hest_read_ack_start_addr = err_source_struct + GHES_ACK_OFFSET;
+
+    cpu_physical_memory_read(hest_err_block_addr, &error_block_addr,
+                             sizeof(error_block_addr));
+
+    cpu_physical_memory_read(error_block_addr, &cper_addr,
+                             sizeof(error_block_addr));
+
+    cpu_physical_memory_read(hest_read_ack_start_addr, &read_ack_start_addr,
+			     sizeof(read_ack_start_addr));
+
+    /* Update ACK offset to notify about a new error */
+
+    cpu_physical_memory_read(read_ack_start_addr,
+                             &read_ack, sizeof(read_ack));
+
+    /* zero means OSPM does not acknowledge the error */
+    if (!read_ack) {
+        error_setg(errp,
+                   "Last CPER record was not acknowledged yet");
+        read_ack = 1;
+        cpu_physical_memory_write(read_ack_start_addr,
+                                  &read_ack, sizeof(read_ack));
+        return;
+    }
+
+    read_ack = cpu_to_le64(0);
+    cpu_physical_memory_write(read_ack_start_addr,
+                              &read_ack, sizeof(read_ack));
+
+    /* Write the generic error data entry into guest memory */
+    cpu_physical_memory_write(cper_addr, cper, len);
+}
+
+int acpi_ghes_record_errors(int source_id, uint64_t physical_address)
+{
+    /* Memory Error Section Type */
+    const uint8_t guid[] =
+          UUID_LE(0xA5BC1114, 0x6F64, 0x4EDE, 0xB8, 0x63, 0x3E, 0x83, \
+                  0xED, 0x7C, 0x83, 0xB1);
+    Error *errp = NULL;
+    GArray *block;
+
+    if (!physical_address) {
+        error_report("can not find Generic Error Status Block for source id %d",
+                     source_id);
+        return -1;
+    }
+
+    block = g_array_new(false, true /* clear */, 1);
+
+    ghes_gen_err_data_uncorrectable_recoverable(block, guid,
+                                                ACPI_GHES_MAX_RAW_DATA_LENGTH);
+
+    /* Build the memory section CPER for above new generic error data entry */
+    acpi_ghes_build_append_mem_cper(block, physical_address);
+
+    /* Report the error */
+    ghes_record_cper_errors(block->data, block->len, source_id, &errp);
+
+    g_array_free(block, true);
+
+    if (errp) {
+        error_report_err(errp);
+        return -1;
     }
 
-    return ret;
+    return 0;
 }
 
 bool acpi_ghes_present(void)
hw/arm/virt-acpi-build.c +0 -0
index f76fb117adff..39100c2822c2 100644

hw/arm/virt-acpi-build.c static void acpi_align_size(GArray *blob, unsigned align) · line 895

Quoted 6 lines
hw/arm/virt-acpi-build.c +3 -0
@@ -890,6 +890,10 @@ static void acpi_align_size(GArray *blob, unsigned align)
     g_array_set_size(blob, ROUND_UP(acpi_data_len(blob), align));
 }
 
+static const uint16_t hest_ghes_notify[] = {
+    [ARM_ACPI_HEST_SRC_ID_SEA] = ACPI_GHES_NOTIFY_SEA,
+};

I agree that machine/platform shall opt in for a specific source id, but I'm not sure about whether we need platform specific source ids, it seems to complicate things needlessly.

For example if one would define different src_id for error injection for ARM and X86, then we would somehow need to take that in account when QMP command X called so it would use correct ID

Maybe this needs it's own patch with a commit message that would explain need for this approach (but so far I'm not seeing the point).

PS:
I'd prefer common/shared SRC_ID registry, from which boards would pick applicable ones.

Quoted 49 lines
hw/arm/virt-acpi-build.c +4 -3
+
 static
 void virt_acpi_build(VirtMachineState *vms, AcpiBuildTables *tables)
 {
@@ -943,10 +947,10 @@ void virt_acpi_build(VirtMachineState *vms, AcpiBuildTables *tables)
     build_dbg2(tables_blob, tables->linker, vms);
 
     if (vms->ras) {
-        build_ghes_error_table(tables->hardware_errors, tables->linker);
         acpi_add_table(table_offsets, tables_blob);
-        acpi_build_hest(tables_blob, tables->linker, vms->oem_id,
-                        vms->oem_table_id);
+        acpi_build_hest(tables_blob, tables->hardware_errors, tables->linker,
+                        hest_ghes_notify, sizeof(hest_ghes_notify),
+                        vms->oem_id, vms->oem_table_id);
     }
 
     if (ms->numa_state->num_nodes > 0) {
include/hw/acpi/ghes.h +2 -8
index 28b956acb19a..4b5af86ec077 100644
@@ -23,6 +23,7 @@
 #define ACPI_GHES_H
 
 #include "hw/acpi/bios-linker-loader.h"
+#include "qapi/error.h"
 
 /*
  * Values for Hardware Error Notification Type field
@@ -56,24 +57,23 @@ enum AcpiGhesNotifyType {
     ACPI_GHES_NOTIFY_RESERVED = 12
 };
 
-enum {
-    ACPI_HEST_SRC_ID_SEA = 0,
-    /* future ids go here */
-    ACPI_HEST_SRC_ID_RESERVED,
-};
-
 typedef struct AcpiGhesState {
     uint64_t hest_addr_le;
     uint64_t ghes_addr_le;
     bool present; /* True if GHES is present at all on this board */
 } AcpiGhesState;
 
-void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker);
-void acpi_build_hest(GArray *table_data, BIOSLinker *linker,
+void acpi_build_hest(GArray *table_data, GArray *hardware_errors,

include/hw/acpi/ghes.h enum AcpiGhesNotifyType { · line 75

Quoted 10 lines
include/hw/acpi/ghes.h +6 -1
+                     BIOSLinker *linker,
+                     const uint16_t * const notify,
+                     int num_sources,
                      const char *oem_id, const char *oem_table_id);
 void acpi_ghes_add_fw_cfg(AcpiGhesState *vms, FWCfgState *s,
                           GArray *hardware_errors);
-int acpi_ghes_record_errors(uint8_t notify, uint64_t error_physical_addr);
+int acpi_ghes_record_errors(int source_id,
+                            uint64_t error_physical_addr);
+void ghes_record_cper_errors(const void *cper, size_t len,

use GArray for cper so you won't have to pass down len

Quoted 36 lines
include/hw/acpi/ghes.h +1 -0
+                             uint16_t source_id, Error **errp);
 
 /**
  * acpi_ghes_present: Report whether ACPI GHES table is present
include/hw/arm/virt.h +7 -0
index a4d937ed45ac..d62d8d9db5ae 100644
@@ -189,6 +189,13 @@ OBJECT_DECLARE_TYPE(VirtMachineState, VirtMachineClass, VIRT_MACHINE)
 void virt_acpi_setup(VirtMachineState *vms);
 bool virt_is_acpi_enabled(VirtMachineState *vms);
 
+/*
+ * ID numbers used to fill HEST source ID field
+ */
+enum {
+    ARM_ACPI_HEST_SRC_ID_SEA,
+};
+
 /* Return number of redistributors that fit in the specified region */
 static uint32_t virt_redist_capacity(VirtMachineState *vms, int region)
 {
target/arm/kvm.c +2 -1
index 849e2e21b304..8c4c8263b85a 100644
@@ -2373,7 +2373,8 @@ void kvm_arch_on_sigbus_vcpu(CPUState *c, int code, void *addr)
              */
             if (code == BUS_MCEERR_AR) {
                 kvm_cpu_synchronize_state(c);
-                if (!acpi_ghes_record_errors(ACPI_HEST_SRC_ID_SEA, paddr)) {
+                if (!acpi_ghes_record_errors(ARM_ACPI_HEST_SRC_ID_SEA,
+                                             paddr)) {
                     kvm_inject_arm_sea(c);
                 } else {
                     error_report("failed to record the error");
#11 lore raw

Re: [PATCH v9 02/12] acpi/ghes: rework the logic to handle HEST source ID

Em Wed, 11 Sep 2024 17:01:57 +0200
Igor Mammedov <[email protected]> escreveu:

Quoted 16 lines
Quoted 4 lines · On Sun, 25 Aug 2024 05:45:57 +0200 Mauro Carvalho Chehab…

The current logic is based on a lot of duct tape, with offsets calculated based on one define with the number of source IDs and an enum.

Quoted 8 lines

Rewrite the logic in a way that it would be more resilient of code changes, by moving the source ID count to an enum and make the offset calculus more explicit.

Such change was inspired on a patch from Jonathan Cameron splitting the logic to get the CPER address on a separate function, as this will be needed to support generic error injection.

Quoted 4 lines

patch is too large and does too many things at once, see inline suggestions on how to split it in more manageable chunks.
(I'll mark preferred patch order with numbers)

I ended adding more patches to make changes more logic.

Quoted 28 lines
Quoted 28 lines

---

Changes from v8:
- Non-rename/cleanup changes merged altogether;
- source ID is now more generic, defined per guest target. That should make easier to add support for 86.

---
 hw/acpi/ghes.c           | 275 ++++++++++++++++++++++++---------------
 hw/arm/virt-acpi-build.c |  10 +-
 include/hw/acpi/ghes.h   |  18 +--
 include/hw/arm/virt.h    |   7 +
 target/arm/kvm.c         |   3 +-
 5 files changed, 198 insertions(+), 115 deletions(-)
hw/acpi/ghes.c +0 -0
index 529c14e3289f..965fb1b36587 100644
@@ -35,9 +35,6 @@
 /* The max size in bytes for one error block */
 #define ACPI_GHES_MAX_RAW_DATA_LENGTH   (1 * KiB)
 
Quoted 5 lines

hw/acpi/ghes.c line 39

Quoted 2 lines
hw/acpi/ghes.c +0 -2
-/* Now only support ARMv8 SEA notification type error source */
-#define ACPI_GHES_ERROR_SOURCE_COUNT        1  

[patch 4] getting rid of this and introducing num_sources
(aka variable size HEST)

ok.

Quoted 8 lines
Quoted 8 lines
hw/acpi/ghes.c +0 -0
 /* Generic Hardware Error Source version 2 */
 #define ACPI_GHES_SOURCE_GENERIC_ERROR_V2   10
 
@@ -64,6 +61,19 @@
  */
 #define ACPI_GHES_GESB_SIZE                 20
 
Quoted 7 lines

hw/acpi/ghes.c line 68

Quoted 5 lines
hw/acpi/ghes.c +5 -0
+/*
+ * Offsets with regards to the start of the HEST table stored at
+ * ags->hest_addr_le, according with the memory layout map at
+ * docs/specs/acpi_hest_ghes.rst.
+ */  

perhaps mention in comment/commit message, that hest lookup is implemented only GHESv2 error sources.

Ok, will add a comment, but IMO, it fits better at the routine which handles HEST error sources, so I added this there:

    /*
     * Currently, HEST Error source navigates only for GHESv2 tables
     */
    for (i = 0; i < ACPI_GHES_ERROR_SOURCE_COUNT; i++) {
        uint64_t addr = err_source_struct;
        uint16_t type, src_id;

...

Quoted 14 lines

That will work as far we do forward migration only (i.e. old qemu -> new qemu), which is what upstream supports.

However it won't work for backward migration (new qemu -> old qemu) since old one doesn't know about new non-GHESv2 sources. And that means we would need to introduce compat knobs for every new non-GHESv2 source is added. Which is easy to overlook and it adds up to maintenance.
(You've already described zoo of types ACPI spec has in v8 review, but I don't thing it's too complex to implement lookup of all known types. compared to headache we would have with compat settings if anyone remembers)

Quoted 3 lines

I won't insist on adding all known sources lookup in this series, if you agree to do it as a patch on top of this series within this dev cycle (~2 months time-frame).

Seems fine to me to place it at the dev cycle.

Quoted 10 lines
Quoted 1 line

+/* ACPI 6.2: 18.3.2.8 Generic Hardware Error Source version 2 */

+ ,Table 18-383

hw/acpi/ghes.c line 74

Quoted 4 lines
hw/acpi/ghes.c +4 -0
+#define HEST_GHES_V2_TABLE_SIZE  92
+#define GHES_ACK_OFFSET          (64 + GAS_ADDR_OFFSET)
+
+/* ACPI 6.2: 18.3.2.7: Generic Hardware Error Source */  

Table 18-380 'Error Status Address' field

Actually on ACPI 6.2, those tables are 18-382 and 18-379.

I'll change the above to reflect that:

/* ACPI 6.2: 18.3.2.8 Generic Hardware Error Source version 2
* Table 18-382 Generic Hardware Error Source version 2 (GHESv2) Structure */
#define HEST_GHES_V2_TABLE_SIZE 92
#define GHES_ACK_OFFSET (64 + GAS_ADDR_OFFSET)

/* ACPI 6.2: 18.3.2.7: Generic Hardware Error Source
* Table 18-379: 'Error Status Address' field

Quoted 10 lines
Quoted 10 lines
hw/acpi/ghes.c +2 -0
+#define GHES_ERR_ST_ADDR_OFFSET  (20 + GAS_ADDR_OFFSET)
+
 /*
  * Values for error_severity field
  */
@@ -185,51 +195,30 @@ static void acpi_ghes_build_append_mem_cper(GArray *table,
     build_append_int_noprefix(table, 0, 7);
 }
 
Quoted 8 lines

hw/acpi/ghes.c static void acpi_ghes_build_append_mem_cper(GArray *table, · line 201

Quoted 6 lines
hw/acpi/ghes.c +4 -2
-static int acpi_ghes_record_mem_error(uint64_t error_block_address,
-                                      uint64_t error_physical_addr)
+static void
+ghes_gen_err_data_uncorrectable_recoverable(GArray *block,
+                                            const uint8_t *section_type,
+                                            int data_length)  

[patch 2] splitting acpi_ghes_record_mem_error() on reusable and mem specific
code

Ok, will move to a separate patch after this one.

Quoted 90 lines
Quoted 89 lines
hw/acpi/ghes.c +10 -32
 {
-    GArray *block;
-
-    /* Memory Error Section Type */
-    const uint8_t uefi_cper_mem_sec[] =
-          UUID_LE(0xA5BC1114, 0x6F64, 0x4EDE, 0xB8, 0x63, 0x3E, 0x83, \
-                  0xED, 0x7C, 0x83, 0xB1);
-
     /* invalid fru id: ACPI 4.0: 17.3.2.6.1 Generic Error Data,
      * Table 17-13 Generic Error Data Entry
      */
     QemuUUID fru_id = {};
-    uint32_t data_length;
 
-    block = g_array_new(false, true /* clear */, 1);
-
-    /* This is the length if adding a new generic error data entry*/
-    data_length = ACPI_GHES_DATA_LENGTH + ACPI_GHES_MEM_CPER_LENGTH;
     /*
-     * It should not run out of the preallocated memory if adding a new generic
-     * error data entry
+     * Calculate the size with this block. No need to check for
+     * too big CPER, as CPER size is checked at ghes_record_cper_errors()
      */
-    assert((data_length + ACPI_GHES_GESB_SIZE) <=
-            ACPI_GHES_MAX_RAW_DATA_LENGTH);
+    data_length += ACPI_GHES_GESB_SIZE;
 
     /* Build the new generic error status block header */
     acpi_ghes_generic_error_status(block, ACPI_GEBS_UNCORRECTABLE,
         0, 0, data_length, ACPI_CPER_SEV_RECOVERABLE);
 
     /* Build this new generic error data entry header */
-    acpi_ghes_generic_error_data(block, uefi_cper_mem_sec,
+    acpi_ghes_generic_error_data(block, section_type,
         ACPI_CPER_SEV_RECOVERABLE, 0, 0,
         ACPI_GHES_MEM_CPER_LENGTH, fru_id, 0);
-
-    /* Build the memory section CPER for above new generic error data entry */
-    acpi_ghes_build_append_mem_cper(block, error_physical_addr);
-
-    /* Write the generic error data entry into guest memory */
-    cpu_physical_memory_write(error_block_address, block->data, block->len);
-
-    g_array_free(block, true);
-
-    return 0;
 }
 
 /*
@@ -237,17 +226,18 @@ static int acpi_ghes_record_mem_error(uint64_t error_block_address,
  * Initialize "etc/hardware_errors" and "etc/hardware_errors_addr" fw_cfg blobs.
  * See docs/specs/acpi_hest_ghes.rst for blobs format.
  */
-void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker)
+static void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker,
+                                   int num_sources)
 {
     int i, error_status_block_offset;
 
     /* Build error_block_address */
-    for (i = 0; i < ACPI_GHES_ERROR_SOURCE_COUNT; i++) {
+    for (i = 0; i < num_sources; i++) {
         build_append_int_noprefix(hardware_errors, 0, sizeof(uint64_t));
     }
 
     /* Build read_ack_register */
-    for (i = 0; i < ACPI_GHES_ERROR_SOURCE_COUNT; i++) {
+    for (i = 0; i < num_sources; i++) {
         /*
          * Initialize the value of read_ack_register to 1, so GHES can be
          * writable after (re)boot.
@@ -262,13 +252,13 @@ void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker)
 
     /* Reserve space for Error Status Data Block */
     acpi_data_push(hardware_errors,
-        ACPI_GHES_MAX_RAW_DATA_LENGTH * ACPI_GHES_ERROR_SOURCE_COUNT);
+        ACPI_GHES_MAX_RAW_DATA_LENGTH * num_sources);
 
     /* Tell guest firmware to place hardware_errors blob into RAM */
     bios_linker_loader_alloc(linker, ACPI_GHES_ERRORS_FW_CFG_FILE,
                              hardware_errors, sizeof(uint64_t), false);
 
-    for (i = 0; i < ACPI_GHES_ERROR_SOURCE_COUNT; i++) {
+    for (i = 0; i < num_sources; i++) {
         /*
          * Tell firmware to patch error_block_address entries to point to
          * corresponding "Generic Error Status Block"  
Quoted 11 lines

hw/acpi/ghes.c void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker) · line 278

Quoted 9 lines
hw/acpi/ghes.c +3 -2
@@ -283,14 +273,20 @@ void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker)
      * tell firmware to write hardware_errors GPA into
      * hardware_errors_addr fw_cfg, once the former has been initialized.
      */
-    bios_linker_loader_write_pointer(linker, ACPI_GHES_DATA_ADDR_FW_CFG_FILE,
-        0, sizeof(uint64_t), ACPI_GHES_ERRORS_FW_CFG_FILE, 0);
+    bios_linker_loader_write_pointer(linker, ACPI_GHES_DATA_ADDR_FW_CFG_FILE, 0,
+                                     sizeof(uint64_t),
+                                     ACPI_GHES_ERRORS_FW_CFG_FILE, 0);  

[patch 1] all indent changes in its own patch, or just drop them altogether

I'll drop the pure reformat changes from this patch. They'll be placed at the patches that rename ACPI_GHES_*_FW_CFG_*.

Same for other occurrences.

Quoted 1 line
Quoted 1 line

...

Quoted 38 lines
Quoted 38 lines

/*
* Read Ack Preserve field

hw/acpi/ghes.c +12 -3
@@ -360,19 +350,28 @@ static void build_ghes_v2(GArray *table_data, int source_id, BIOSLinker *linker)
 }
 
 /* Build Hardware Error Source Table */
-void acpi_build_hest(GArray *table_data, BIOSLinker *linker,
+void acpi_build_hest(GArray *table_data, GArray *hardware_errors,
+                     BIOSLinker *linker,
+                     const uint16_t * const notify,
+                     int num_sources,
                      const char *oem_id, const char *oem_table_id)
 {
     AcpiTable table = { .sig = "HEST", .rev = 1,
                         .oem_id = oem_id, .oem_table_id = oem_table_id };
+    int i;
+
+    build_ghes_error_table(hardware_errors, linker, num_sources);
 
     acpi_table_begin(&table, table_data);
 
+    /* Beginning at the HEST Error Source struct count and data */
     int hest_offset = table_data->len;
 
     /* Error Source Count */
-    build_append_int_noprefix(table_data, ACPI_GHES_ERROR_SOURCE_COUNT, 4);
-    build_ghes_v2(table_data, ACPI_HEST_SRC_ID_SEA, linker);
+    build_append_int_noprefix(table_data, num_sources, 4);
+    for (i = 0; i < num_sources; i++) {
+        build_ghes_v2(table_data, linker, notify[i], i, num_sources);
+    }
 
     acpi_table_end(linker, &table);
 
@@ -403,60 +402,132 @@ void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
     ags->present = true;
 }
 
Quoted 5 lines

hw/acpi/ghes.c void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s, · line 406

Quoted 3 lines
hw/acpi/ghes.c +2 -1
-int acpi_ghes_record_errors(uint8_t source_id, uint64_t physical_address)
+void ghes_record_cper_errors(const void *cper, size_t len,
+                             uint16_t source_id, Error **errp)  

[patch 3] switching to hest source id lookup method

Ok.

Quoted 188 lines
Quoted 173 lines
hw/acpi/ghes.c +112 -41
 {
-    uint64_t error_block_addr, read_ack_register_addr, read_ack_register = 0;
-    uint64_t start_addr;
-    bool ret = -1;
+    uint64_t hest_read_ack_start_addr, read_ack_start_addr;
+    uint64_t hest_addr, cper_addr, err_source_struct;
+    uint64_t hest_err_block_addr, error_block_addr;
+    uint32_t num_sources, i;
     AcpiGedState *acpi_ged_state;
     AcpiGhesState *ags;
+    uint64_t read_ack;
 
-    assert(source_id < ACPI_HEST_SRC_ID_RESERVED);
+    if (len > ACPI_GHES_MAX_RAW_DATA_LENGTH) {
+        error_setg(errp, "GHES CPER record is too big: %ld", len);
+    }
 
     acpi_ged_state = ACPI_GED(object_resolve_path_type("", TYPE_ACPI_GED,
                                                        NULL));
     g_assert(acpi_ged_state);
     ags = &acpi_ged_state->ghes_state;
 
-    start_addr = le64_to_cpu(ags->ghes_addr_le);
-
-    if (physical_address) {
-
-        if (source_id < ACPI_HEST_SRC_ID_RESERVED) {
-            start_addr += source_id * sizeof(uint64_t);
-        }
-
-        cpu_physical_memory_read(start_addr, &error_block_addr,
-                                 sizeof(error_block_addr));
-
-        error_block_addr = le64_to_cpu(error_block_addr);
-
-        read_ack_register_addr = start_addr +
-            ACPI_GHES_ERROR_SOURCE_COUNT * sizeof(uint64_t);
-
-        cpu_physical_memory_read(read_ack_register_addr,
-                                 &read_ack_register, sizeof(read_ack_register));
-
-        /* zero means OSPM does not acknowledge the error */
-        if (!read_ack_register) {
-            error_report("OSPM does not acknowledge previous error,"
-                " so can not record CPER for current error anymore");
-        } else if (error_block_addr) {
-            read_ack_register = cpu_to_le64(0);
-            /*
-             * Clear the Read Ack Register, OSPM will write it to 1 when
-             * it acknowledges this error.
-             */
-            cpu_physical_memory_write(read_ack_register_addr,
-                &read_ack_register, sizeof(uint64_t));
-
-            ret = acpi_ghes_record_mem_error(error_block_addr,
-                                             physical_address);
-        } else
-            error_report("can not find Generic Error Status Block");
+    hest_addr = le64_to_cpu(ags->hest_addr_le);
+
+    cpu_physical_memory_read(hest_addr, &num_sources, sizeof(num_sources));
+
+    if (source_id >= num_sources) {
+        error_setg(errp,
+                   "GHES: Source %d not found. Only %d sources are defined",
+                   source_id, num_sources);
+        return;
+    }
+    err_source_struct = hest_addr + sizeof(num_sources);
+
+    for (i = 0; i < num_sources; i++) {
+        uint64_t addr = err_source_struct;
+        uint16_t type, src_id;
+
+        cpu_physical_memory_read(addr, &type, sizeof(type));
+
+        /* For now, we only know the size of GHESv2 table */
+        assert(type == ACPI_GHES_SOURCE_GENERIC_ERROR_V2);
+
+        /* It is GHES. Compare CPER source address */
+        addr += sizeof(type);
+        cpu_physical_memory_read(addr, &src_id, sizeof(src_id));
+
+        if (src_id == source_id)
+            break;
+
+        err_source_struct += HEST_GHES_V2_TABLE_SIZE;
+    }
+    if (i == num_sources) {
+        error_setg(errp, "HEST: Source %d not found.", source_id);
+        return;
+    }
+
+    /* Check if BIOS addr pointers were properly generated */
+
+    hest_err_block_addr = err_source_struct + GHES_ERR_ST_ADDR_OFFSET;
+    hest_read_ack_start_addr = err_source_struct + GHES_ACK_OFFSET;
+
+    cpu_physical_memory_read(hest_err_block_addr, &error_block_addr,
+                             sizeof(error_block_addr));
+
+    cpu_physical_memory_read(error_block_addr, &cper_addr,
+                             sizeof(error_block_addr));
+
+    cpu_physical_memory_read(hest_read_ack_start_addr, &read_ack_start_addr,
+			     sizeof(read_ack_start_addr));
+
+    /* Update ACK offset to notify about a new error */
+
+    cpu_physical_memory_read(read_ack_start_addr,
+                             &read_ack, sizeof(read_ack));
+
+    /* zero means OSPM does not acknowledge the error */
+    if (!read_ack) {
+        error_setg(errp,
+                   "Last CPER record was not acknowledged yet");
+        read_ack = 1;
+        cpu_physical_memory_write(read_ack_start_addr,
+                                  &read_ack, sizeof(read_ack));
+        return;
+    }
+
+    read_ack = cpu_to_le64(0);
+    cpu_physical_memory_write(read_ack_start_addr,
+                              &read_ack, sizeof(read_ack));
+
+    /* Write the generic error data entry into guest memory */
+    cpu_physical_memory_write(cper_addr, cper, len);
+}
+
+int acpi_ghes_record_errors(int source_id, uint64_t physical_address)
+{
+    /* Memory Error Section Type */
+    const uint8_t guid[] =
+          UUID_LE(0xA5BC1114, 0x6F64, 0x4EDE, 0xB8, 0x63, 0x3E, 0x83, \
+                  0xED, 0x7C, 0x83, 0xB1);
+    Error *errp = NULL;
+    GArray *block;
+
+    if (!physical_address) {
+        error_report("can not find Generic Error Status Block for source id %d",
+                     source_id);
+        return -1;
+    }
+
+    block = g_array_new(false, true /* clear */, 1);
+
+    ghes_gen_err_data_uncorrectable_recoverable(block, guid,
+                                                ACPI_GHES_MAX_RAW_DATA_LENGTH);
+
+    /* Build the memory section CPER for above new generic error data entry */
+    acpi_ghes_build_append_mem_cper(block, physical_address);
+
+    /* Report the error */
+    ghes_record_cper_errors(block->data, block->len, source_id, &errp);
+
+    g_array_free(block, true);
+
+    if (errp) {
+        error_report_err(errp);
+        return -1;
     }
 
-    return ret;
+    return 0;
 }
 
 bool acpi_ghes_present(void)
hw/arm/virt-acpi-build.c +0 -0
index f76fb117adff..39100c2822c2 100644

hw/arm/virt-acpi-build.c static void acpi_align_size(GArray *blob, unsigned align) · line 895

Quoted 6 lines
hw/arm/virt-acpi-build.c +3 -0
@@ -890,6 +890,10 @@ static void acpi_align_size(GArray *blob, unsigned align)
     g_array_set_size(blob, ROUND_UP(acpi_data_len(blob), align));
 }
 
+static const uint16_t hest_ghes_notify[] = {
+    [ARM_ACPI_HEST_SRC_ID_SEA] = ACPI_GHES_NOTIFY_SEA,
+};  

I agree that machine/platform shall opt in for a specific source id, but I'm not sure about whether we need platform specific source ids, it seems to complicate things needlessly.

For example if one would define different src_id for error injection for ARM and X86, then we would somehow need to take that in account when QMP command X called so it would use correct ID

Quoted 6 lines

Maybe this needs it's own patch with a commit message that would explain need for this approach (but so far I'm not seeing the point).

PS:
I'd prefer common/shared SRC_ID registry, from which boards would pick applicable ones.

I'll use a different approach, adding this to ghes.h:

	/*
	 * ID numbers used to fill HEST source ID field
	 */
	enum AcpiGhesSourceID {
	    ACPI_HEST_SRC_ID_SYNC,
	    ACPI_HEST_SRC_ID_QMP,       /* Use it only for QMP injected errors */
	};

	typedef struct AcpiNotificationSourceId {
	    enum AcpiGhesSourceID source_id;
	    enum AcpiGhesNotifyType notify;
	} AcpiNotificationSourceId;

And, at the binding logic (at arm/virt-acpi-build):

	static const AcpiNotificationSourceId hest_ghes_notify[] = {
	    {ACPI_HEST_SRC_ID_SYNC, ACPI_GHES_NOTIFY_SEA},
	    {ACPI_HEST_SRC_ID_QMP, ACPI_GHES_NOTIFY_GPIO},
	};
	...
	acpi_build_hest(tables_blob, tables->hardware_errors, tables->linker,
                        hest_ghes_notify, sizeof(hest_ghes_notify),
                        vms->oem_id, vms->oem_table_id);

For x86, with just QMP implemented, this will be:

	static const AcpiNotificationSourceId hest_ghes_notify[] = {
	    {ACPI_HEST_SRC_ID_QMP, ACPI_GHES_NOTIFY_SCI},
	};
	...
	acpi_build_hest(tables_blob, tables->hardware_errors, tables->linker,
                        hest_ghes_notify, sizeof(hest_ghes_notify),
                        vms->oem_id, vms->oem_table_id);

As the current logic doesn't assume anymore that source_id is an index, but instead searches for it along the error structures, such logic works fine and allows each arch to define what IDs they'll use and what notification is associated to each one of them.

Quoted 46 lines
Quoted 46 lines
hw/arm/virt-acpi-build.c +4 -3
+
 static
 void virt_acpi_build(VirtMachineState *vms, AcpiBuildTables *tables)
 {
@@ -943,10 +947,10 @@ void virt_acpi_build(VirtMachineState *vms, AcpiBuildTables *tables)
     build_dbg2(tables_blob, tables->linker, vms);
 
     if (vms->ras) {
-        build_ghes_error_table(tables->hardware_errors, tables->linker);
         acpi_add_table(table_offsets, tables_blob);
-        acpi_build_hest(tables_blob, tables->linker, vms->oem_id,
-                        vms->oem_table_id);
+        acpi_build_hest(tables_blob, tables->hardware_errors, tables->linker,
+                        hest_ghes_notify, sizeof(hest_ghes_notify),
+                        vms->oem_id, vms->oem_table_id);
     }
 
     if (ms->numa_state->num_nodes > 0) {
include/hw/acpi/ghes.h +1 -6
index 28b956acb19a..4b5af86ec077 100644
@@ -23,6 +23,7 @@
 #define ACPI_GHES_H
 
 #include "hw/acpi/bios-linker-loader.h"
+#include "qapi/error.h"
 
 /*
  * Values for Hardware Error Notification Type field
@@ -56,24 +57,23 @@ enum AcpiGhesNotifyType {
     ACPI_GHES_NOTIFY_RESERVED = 12
 };
 
-enum {
-    ACPI_HEST_SRC_ID_SEA = 0,
-    /* future ids go here */
-    ACPI_HEST_SRC_ID_RESERVED,
-};
-
 typedef struct AcpiGhesState {
     uint64_t hest_addr_le;
     uint64_t ghes_addr_le;
     bool present; /* True if GHES is present at all on this board */
 } AcpiGhesState;
 
Quoted 15 lines

include/hw/acpi/ghes.h enum AcpiGhesNotifyType { · line 75

Quoted 13 lines
include/hw/acpi/ghes.h +7 -3
-void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker);
-void acpi_build_hest(GArray *table_data, BIOSLinker *linker,
+void acpi_build_hest(GArray *table_data, GArray *hardware_errors,
+                     BIOSLinker *linker,
+                     const uint16_t * const notify,
+                     int num_sources,
                      const char *oem_id, const char *oem_table_id);
 void acpi_ghes_add_fw_cfg(AcpiGhesState *vms, FWCfgState *s,
                           GArray *hardware_errors);
-int acpi_ghes_record_errors(uint8_t notify, uint64_t error_physical_addr);
+int acpi_ghes_record_errors(int source_id,
+                            uint64_t error_physical_addr);
+void ghes_record_cper_errors(const void *cper, size_t len,  

use GArray for cper so you won't have to pass down len

Here, it would work fine, but when adding hw/acpi/ghes_cper.c, the logic there is:

    cper = qbase64_decode(qmp_cper, -1, &len, errp);
    if (!cper) {
        error_setg(errp, "missing GHES CPER payload");
        return;
    }

    ghes_record_cper_errors(cper, len, ACPI_HEST_SRC_ID_QMP, errp);

If I use a GArray, it would mean an extra memory allocation for no good reason.

So, IMO, better to keep passing buffer and length.

Regards,
Mauro

#12 lore raw

PATCH v9 3/12 acpi/ghes: rename etc/hardware_error file macros

In reply to Mauro Carvalho Chehab

Now that we have also have a file to store HEST data location, which is part of GHES, better name the file where CPER records are stored.

No functional changes.

---
 hw/acpi/ghes.c | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)
hw/acpi/ghes.c +11 -11
index 965fb1b36587..3190eb954de4 100644
@@ -28,8 +28,8 @@
 #include "hw/nvram/fw_cfg.h"
 #include "qemu/uuid.h"
 
-#define ACPI_GHES_ERRORS_FW_CFG_FILE        "etc/hardware_errors"
-#define ACPI_GHES_DATA_ADDR_FW_CFG_FILE     "etc/hardware_errors_addr"
+#define ACPI_HW_ERROR_FW_CFG_FILE           "etc/hardware_errors"
+#define ACPI_HW_ERROR_ADDR_FW_CFG_FILE      "etc/hardware_errors_addr"
 #define ACPI_HEST_ADDR_FW_CFG_FILE          "etc/acpi_table_hest_addr"
 
 /* The max size in bytes for one error block */
@@ -255,7 +255,7 @@ static void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker,
         ACPI_GHES_MAX_RAW_DATA_LENGTH * num_sources);
 
     /* Tell guest firmware to place hardware_errors blob into RAM */
-    bios_linker_loader_alloc(linker, ACPI_GHES_ERRORS_FW_CFG_FILE,
+    bios_linker_loader_alloc(linker, ACPI_HW_ERROR_FW_CFG_FILE,
                              hardware_errors, sizeof(uint64_t), false);
 
     for (i = 0; i < num_sources; i++) {
@@ -264,8 +264,8 @@ static void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker,
          * corresponding "Generic Error Status Block"
          */
         bios_linker_loader_add_pointer(linker,
-            ACPI_GHES_ERRORS_FW_CFG_FILE, sizeof(uint64_t) * i,
-            sizeof(uint64_t), ACPI_GHES_ERRORS_FW_CFG_FILE,
+            ACPI_HW_ERROR_FW_CFG_FILE, sizeof(uint64_t) * i,
+            sizeof(uint64_t), ACPI_HW_ERROR_FW_CFG_FILE,
             error_status_block_offset + i * ACPI_GHES_MAX_RAW_DATA_LENGTH);
     }
 
@@ -273,9 +273,9 @@ static void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker,
      * tell firmware to write hardware_errors GPA into
      * hardware_errors_addr fw_cfg, once the former has been initialized.
      */
-    bios_linker_loader_write_pointer(linker, ACPI_GHES_DATA_ADDR_FW_CFG_FILE, 0,
+    bios_linker_loader_write_pointer(linker, ACPI_HW_ERROR_ADDR_FW_CFG_FILE, 0,
                                      sizeof(uint64_t),
-                                     ACPI_GHES_ERRORS_FW_CFG_FILE, 0);
+                                     ACPI_HW_ERROR_FW_CFG_FILE, 0);
 }
 
 /* Build Generic Hardware Error Source version 2 (GHESv2) */
@@ -315,7 +315,7 @@ static void build_ghes_v2(GArray *table_data,
     bios_linker_loader_add_pointer(linker, ACPI_BUILD_TABLE_FILE,
                                    address_offset + GAS_ADDR_OFFSET,
                                    sizeof(uint64_t),
-                                   ACPI_GHES_ERRORS_FW_CFG_FILE,
+                                   ACPI_HW_ERROR_FW_CFG_FILE,
                                    source_id * sizeof(uint64_t));
 
     /* Notification Structure */
@@ -335,7 +335,7 @@ static void build_ghes_v2(GArray *table_data,
     bios_linker_loader_add_pointer(linker, ACPI_BUILD_TABLE_FILE,
                                    address_offset + GAS_ADDR_OFFSET,
                                    sizeof(uint64_t),
-                                   ACPI_GHES_ERRORS_FW_CFG_FILE,
+                                   ACPI_HW_ERROR_FW_CFG_FILE,
                                    (num_sources + source_id) *
                                    sizeof(uint64_t));
 
@@ -389,11 +389,11 @@ void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
                           GArray *hardware_error)
 {
     /* Create a read-only fw_cfg file for GHES */
-    fw_cfg_add_file(s, ACPI_GHES_ERRORS_FW_CFG_FILE, hardware_error->data,
+    fw_cfg_add_file(s, ACPI_HW_ERROR_FW_CFG_FILE, hardware_error->data,
                     hardware_error->len);
 
     /* Create a read-write fw_cfg file for Address */
-    fw_cfg_add_file_callback(s, ACPI_GHES_DATA_ADDR_FW_CFG_FILE, NULL, NULL,
+    fw_cfg_add_file_callback(s, ACPI_HW_ERROR_ADDR_FW_CFG_FILE, NULL, NULL,
         NULL, &(ags->ghes_addr_le), sizeof(ags->ghes_addr_le), false);
 
     fw_cfg_add_file_callback(s, ACPI_HEST_ADDR_FW_CFG_FILE, NULL, NULL,
Signature · 1 line
2.46.0
Reviewed-by on v9 3/12
#13 lore raw

Re: [PATCH v9 03/12] acpi/ghes: rename etc/hardware_error file macros

Quoted 7 lines · On Sun, 25 Aug 2024 05:45:58 +0200 Mauro Carvalho Chehab…

Now that we have also have a file to store HEST data location, which is part of GHES, better name the file where CPER records are stored.

No functional changes.

Quoted 83 lines
---
 hw/acpi/ghes.c | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)
hw/acpi/ghes.c +11 -11
index 965fb1b36587..3190eb954de4 100644
@@ -28,8 +28,8 @@
 #include "hw/nvram/fw_cfg.h"
 #include "qemu/uuid.h"
 
-#define ACPI_GHES_ERRORS_FW_CFG_FILE        "etc/hardware_errors"
-#define ACPI_GHES_DATA_ADDR_FW_CFG_FILE     "etc/hardware_errors_addr"
+#define ACPI_HW_ERROR_FW_CFG_FILE           "etc/hardware_errors"
+#define ACPI_HW_ERROR_ADDR_FW_CFG_FILE      "etc/hardware_errors_addr"
 #define ACPI_HEST_ADDR_FW_CFG_FILE          "etc/acpi_table_hest_addr"
 
 /* The max size in bytes for one error block */
@@ -255,7 +255,7 @@ static void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker,
         ACPI_GHES_MAX_RAW_DATA_LENGTH * num_sources);
 
     /* Tell guest firmware to place hardware_errors blob into RAM */
-    bios_linker_loader_alloc(linker, ACPI_GHES_ERRORS_FW_CFG_FILE,
+    bios_linker_loader_alloc(linker, ACPI_HW_ERROR_FW_CFG_FILE,
                              hardware_errors, sizeof(uint64_t), false);
 
     for (i = 0; i < num_sources; i++) {
@@ -264,8 +264,8 @@ static void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker,
          * corresponding "Generic Error Status Block"
          */
         bios_linker_loader_add_pointer(linker,
-            ACPI_GHES_ERRORS_FW_CFG_FILE, sizeof(uint64_t) * i,
-            sizeof(uint64_t), ACPI_GHES_ERRORS_FW_CFG_FILE,
+            ACPI_HW_ERROR_FW_CFG_FILE, sizeof(uint64_t) * i,
+            sizeof(uint64_t), ACPI_HW_ERROR_FW_CFG_FILE,
             error_status_block_offset + i * ACPI_GHES_MAX_RAW_DATA_LENGTH);
     }
 
@@ -273,9 +273,9 @@ static void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker,
      * tell firmware to write hardware_errors GPA into
      * hardware_errors_addr fw_cfg, once the former has been initialized.
      */
-    bios_linker_loader_write_pointer(linker, ACPI_GHES_DATA_ADDR_FW_CFG_FILE, 0,
+    bios_linker_loader_write_pointer(linker, ACPI_HW_ERROR_ADDR_FW_CFG_FILE, 0,
                                      sizeof(uint64_t),
-                                     ACPI_GHES_ERRORS_FW_CFG_FILE, 0);
+                                     ACPI_HW_ERROR_FW_CFG_FILE, 0);
 }
 
 /* Build Generic Hardware Error Source version 2 (GHESv2) */
@@ -315,7 +315,7 @@ static void build_ghes_v2(GArray *table_data,
     bios_linker_loader_add_pointer(linker, ACPI_BUILD_TABLE_FILE,
                                    address_offset + GAS_ADDR_OFFSET,
                                    sizeof(uint64_t),
-                                   ACPI_GHES_ERRORS_FW_CFG_FILE,
+                                   ACPI_HW_ERROR_FW_CFG_FILE,
                                    source_id * sizeof(uint64_t));
 
     /* Notification Structure */
@@ -335,7 +335,7 @@ static void build_ghes_v2(GArray *table_data,
     bios_linker_loader_add_pointer(linker, ACPI_BUILD_TABLE_FILE,
                                    address_offset + GAS_ADDR_OFFSET,
                                    sizeof(uint64_t),
-                                   ACPI_GHES_ERRORS_FW_CFG_FILE,
+                                   ACPI_HW_ERROR_FW_CFG_FILE,
                                    (num_sources + source_id) *
                                    sizeof(uint64_t));
 
@@ -389,11 +389,11 @@ void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
                           GArray *hardware_error)
 {
     /* Create a read-only fw_cfg file for GHES */
-    fw_cfg_add_file(s, ACPI_GHES_ERRORS_FW_CFG_FILE, hardware_error->data,
+    fw_cfg_add_file(s, ACPI_HW_ERROR_FW_CFG_FILE, hardware_error->data,
                     hardware_error->len);
 
     /* Create a read-write fw_cfg file for Address */
-    fw_cfg_add_file_callback(s, ACPI_GHES_DATA_ADDR_FW_CFG_FILE, NULL, NULL,
+    fw_cfg_add_file_callback(s, ACPI_HW_ERROR_ADDR_FW_CFG_FILE, NULL, NULL,
         NULL, &(ags->ghes_addr_le), sizeof(ags->ghes_addr_le), false);
 
     fw_cfg_add_file_callback(s, ACPI_HEST_ADDR_FW_CFG_FILE, NULL, NULL,
#14 lore raw

PATCH v9 4/12 acpi/ghes: better name GHES memory error function

In reply to Mauro Carvalho Chehab

The current function used to generate GHES data is specific for memory errors. Give a better name for it, as we now have a generic function as well.

---
 hw/acpi/ghes-stub.c    | 2 +-
 hw/acpi/ghes.c         | 2 +-
 include/hw/acpi/ghes.h | 4 ++--
 target/arm/kvm.c       | 2 +-
 4 files changed, 5 insertions(+), 5 deletions(-)
hw/acpi/ghes-stub.c +1 -1
index c315de1802d6..dd41b3fd91df 100644
@@ -11,7 +11,7 @@
 #include "qemu/osdep.h"
 #include "hw/acpi/ghes.h"
 
-int acpi_ghes_record_errors(uint8_t source_id, uint64_t physical_address)
+int acpi_ghes_memory_errors(uint8_t source_id, uint64_t physical_address)
 {
     return -1;
 }
hw/acpi/ghes.c +1 -1
index 3190eb954de4..10ed9c0614ff 100644
@@ -494,7 +494,7 @@ void ghes_record_cper_errors(const void *cper, size_t len,
     cpu_physical_memory_write(cper_addr, cper, len);
 }
 
-int acpi_ghes_record_errors(int source_id, uint64_t physical_address)
+int acpi_ghes_memory_errors(int source_id, uint64_t physical_address)
 {
     /* Memory Error Section Type */
     const uint8_t guid[] =
include/hw/acpi/ghes.h +2 -2
index 4b5af86ec077..be53b7c53c91 100644
@@ -70,7 +70,7 @@ void acpi_build_hest(GArray *table_data, GArray *hardware_errors,
                      const char *oem_id, const char *oem_table_id);
 void acpi_ghes_add_fw_cfg(AcpiGhesState *vms, FWCfgState *s,
                           GArray *hardware_errors);
-int acpi_ghes_record_errors(int source_id,
+int acpi_ghes_memory_errors(int source_id,
                             uint64_t error_physical_addr);
 void ghes_record_cper_errors(const void *cper, size_t len,
                              uint16_t source_id, Error **errp);
@@ -79,7 +79,7 @@ void ghes_record_cper_errors(const void *cper, size_t len,
  * acpi_ghes_present: Report whether ACPI GHES table is present
  *
  * Returns: true if the system has an ACPI GHES table and it is
- * safe to call acpi_ghes_record_errors() to record a memory error.
+ * safe to call acpi_ghes_memory_errors() to record a memory error.
  */
 bool acpi_ghes_present(void);
 #endif
target/arm/kvm.c +1 -1
index 8c4c8263b85a..8e63e9a59a5e 100644
@@ -2373,7 +2373,7 @@ void kvm_arch_on_sigbus_vcpu(CPUState *c, int code, void *addr)
              */
             if (code == BUS_MCEERR_AR) {
                 kvm_cpu_synchronize_state(c);
-                if (!acpi_ghes_record_errors(ARM_ACPI_HEST_SRC_ID_SEA,
+                if (!acpi_ghes_memory_errors(ARM_ACPI_HEST_SRC_ID_SEA,
                                              paddr)) {
                     kvm_inject_arm_sea(c);
                 } else {
Signature · 1 line
2.46.0
Reviewed-by on v9 4/12
#15 lore raw

Re: [PATCH v9 04/12] acpi/ghes: better name GHES memory error function

Quoted 5 lines · On Sun, 25 Aug 2024 05:45:59 +0200 Mauro Carvalho Chehab…

The current function used to generate GHES data is specific for memory errors. Give a better name for it, as we now have a generic function as well.

Quoted 68 lines
---
 hw/acpi/ghes-stub.c    | 2 +-
 hw/acpi/ghes.c         | 2 +-
 include/hw/acpi/ghes.h | 4 ++--
 target/arm/kvm.c       | 2 +-
 4 files changed, 5 insertions(+), 5 deletions(-)
hw/acpi/ghes-stub.c +1 -1
index c315de1802d6..dd41b3fd91df 100644
@@ -11,7 +11,7 @@
 #include "qemu/osdep.h"
 #include "hw/acpi/ghes.h"
 
-int acpi_ghes_record_errors(uint8_t source_id, uint64_t physical_address)
+int acpi_ghes_memory_errors(uint8_t source_id, uint64_t physical_address)
 {
     return -1;
 }
hw/acpi/ghes.c +1 -1
index 3190eb954de4..10ed9c0614ff 100644
@@ -494,7 +494,7 @@ void ghes_record_cper_errors(const void *cper, size_t len,
     cpu_physical_memory_write(cper_addr, cper, len);
 }
 
-int acpi_ghes_record_errors(int source_id, uint64_t physical_address)
+int acpi_ghes_memory_errors(int source_id, uint64_t physical_address)
 {
     /* Memory Error Section Type */
     const uint8_t guid[] =
include/hw/acpi/ghes.h +2 -2
index 4b5af86ec077..be53b7c53c91 100644
@@ -70,7 +70,7 @@ void acpi_build_hest(GArray *table_data, GArray *hardware_errors,
                      const char *oem_id, const char *oem_table_id);
 void acpi_ghes_add_fw_cfg(AcpiGhesState *vms, FWCfgState *s,
                           GArray *hardware_errors);
-int acpi_ghes_record_errors(int source_id,
+int acpi_ghes_memory_errors(int source_id,
                             uint64_t error_physical_addr);
 void ghes_record_cper_errors(const void *cper, size_t len,
                              uint16_t source_id, Error **errp);
@@ -79,7 +79,7 @@ void ghes_record_cper_errors(const void *cper, size_t len,
  * acpi_ghes_present: Report whether ACPI GHES table is present
  *
  * Returns: true if the system has an ACPI GHES table and it is
- * safe to call acpi_ghes_record_errors() to record a memory error.
+ * safe to call acpi_ghes_memory_errors() to record a memory error.
  */
 bool acpi_ghes_present(void);
 #endif
target/arm/kvm.c +1 -1
index 8c4c8263b85a..8e63e9a59a5e 100644
@@ -2373,7 +2373,7 @@ void kvm_arch_on_sigbus_vcpu(CPUState *c, int code, void *addr)
              */
             if (code == BUS_MCEERR_AR) {
                 kvm_cpu_synchronize_state(c);
-                if (!acpi_ghes_record_errors(ARM_ACPI_HEST_SRC_ID_SEA,
+                if (!acpi_ghes_memory_errors(ARM_ACPI_HEST_SRC_ID_SEA,
                                              paddr)) {
                     kvm_inject_arm_sea(c);
                 } else {
#16 lore raw

PATCH v9 5/12 acpi/ghes: add a notifier to notify when error data is ready

In reply to Mauro Carvalho Chehab

Some error injection notify methods are async, like GPIO notify. Add a notifier to be used when the error record is ready to be sent to the guest OS.

---
 hw/acpi/ghes.c         | 5 +++++
 include/hw/acpi/ghes.h | 3 +++
 2 files changed, 8 insertions(+)
hw/acpi/ghes.c +5 -0
index 10ed9c0614ff..2b7103a678a1 100644
@@ -402,6 +402,9 @@ void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s,
     ags->present = true;
 }
 
+NotifierList acpi_generic_error_notifiers =
+    NOTIFIER_LIST_INITIALIZER(error_device_notifiers);
+
 void ghes_record_cper_errors(const void *cper, size_t len,
                              uint16_t source_id, Error **errp)
 {
@@ -492,6 +495,8 @@ void ghes_record_cper_errors(const void *cper, size_t len,
 
     /* Write the generic error data entry into guest memory */
     cpu_physical_memory_write(cper_addr, cper, len);
+
+    notifier_list_notify(&acpi_generic_error_notifiers, NULL);
 }
 
 int acpi_ghes_memory_errors(int source_id, uint64_t physical_address)
include/hw/acpi/ghes.h +3 -0
index be53b7c53c91..b1ec9795270f 100644
@@ -24,6 +24,9 @@
 
 #include "hw/acpi/bios-linker-loader.h"
 #include "qapi/error.h"
+#include "qemu/notify.h"
+
+extern NotifierList acpi_generic_error_notifiers;
 
 /*
  * Values for Hardware Error Notification Type field
Signature · 1 line
2.46.0
#17 lore raw

PATCH v9 6/12 acpi/generic_event_device: add an APEI error device

In reply to Mauro Carvalho Chehab

Adds a generic error device to handle generic hardware error events as specified at ACPI 6.5 specification at 18.3.2.7.2:
https://uefi.org/specs/ACPI/6.5/18_Platform_Error_Interfaces.html#event-notification-for-generic-error-sources
using HID PNP0C33.

The PNP0C33 device is used to report hardware errors to the guest via ACPI APEI Generic Hardware Error Source (GHES).

---
 hw/acpi/aml-build.c                    | 10 ++++++++++
 hw/acpi/generic_event_device.c         |  8 ++++++++
 include/hw/acpi/acpi_dev_interface.h   |  1 +
 include/hw/acpi/aml-build.h            |  2 ++
 include/hw/acpi/generic_event_device.h |  1 +
 5 files changed, 22 insertions(+)
hw/acpi/aml-build.c +10 -0
index 6d4517cfbe3d..222a933a8760 100644
@@ -2520,3 +2520,13 @@ Aml *aml_i2c_serial_bus_device(uint16_t address, const char *resource_source)
 
     return var;
 }
+
+/* ACPI 5.0b: 18.3.2.6.2 Event Notification For Generic Error Sources */
+Aml *aml_error_device(void)
+{
+    Aml *dev = aml_device(ACPI_APEI_ERROR_DEVICE);
+    aml_append(dev, aml_name_decl("_HID", aml_string("PNP0C33")));
+    aml_append(dev, aml_name_decl("_UID", aml_int(0)));
+
+    return dev;
+}
hw/acpi/generic_event_device.c +8 -0
index 15b4c3ebbf24..b4c83a089a02 100644
@@ -26,6 +26,7 @@ static const uint32_t ged_supported_events[] = {
     ACPI_GED_PWR_DOWN_EVT,
     ACPI_GED_NVDIMM_HOTPLUG_EVT,
     ACPI_GED_CPU_HOTPLUG_EVT,
+    ACPI_GED_ERROR_EVT,
 };
 
 /*
@@ -116,6 +117,11 @@ void build_ged_aml(Aml *table, const char *name, HotplugHandler *hotplug_dev,
                            aml_notify(aml_name(ACPI_POWER_BUTTON_DEVICE),
                                       aml_int(0x80)));
                 break;
+            case ACPI_GED_ERROR_EVT:
+                aml_append(if_ctx,
+                           aml_notify(aml_name(ACPI_APEI_ERROR_DEVICE),
+                                      aml_int(0x80)));
+                break;
             case ACPI_GED_NVDIMM_HOTPLUG_EVT:
                 aml_append(if_ctx,
                            aml_notify(aml_name("\\_SB.NVDR"),
@@ -295,6 +301,8 @@ static void acpi_ged_send_event(AcpiDeviceIf *adev, AcpiEventStatusBits ev)
         sel = ACPI_GED_MEM_HOTPLUG_EVT;
     } else if (ev & ACPI_POWER_DOWN_STATUS) {
         sel = ACPI_GED_PWR_DOWN_EVT;
+    } else if (ev & ACPI_GENERIC_ERROR) {
+        sel = ACPI_GED_ERROR_EVT;
     } else if (ev & ACPI_NVDIMM_HOTPLUG_STATUS) {
         sel = ACPI_GED_NVDIMM_HOTPLUG_EVT;
     } else if (ev & ACPI_CPU_HOTPLUG_STATUS) {
include/hw/acpi/acpi_dev_interface.h +1 -0
index 68d9d15f50aa..8294f8f0ccca 100644
@@ -13,6 +13,7 @@ typedef enum {
     ACPI_NVDIMM_HOTPLUG_STATUS = 16,
     ACPI_VMGENID_CHANGE_STATUS = 32,
     ACPI_POWER_DOWN_STATUS = 64,
+    ACPI_GENERIC_ERROR = 128,
 } AcpiEventStatusBits;
 
 #define TYPE_ACPI_DEVICE_IF "acpi-device-interface"
include/hw/acpi/aml-build.h +2 -0
index a3784155cb33..44d1a6af0c69 100644
@@ -252,6 +252,7 @@ struct CrsRangeSet {
 /* Consumer/Producer */
 #define AML_SERIAL_BUS_FLAG_CONSUME_ONLY        (1 << 1)
 
+#define ACPI_APEI_ERROR_DEVICE   "GEDD"
 /**
  * init_aml_allocator:
  *
@@ -382,6 +383,7 @@ Aml *aml_dma(AmlDmaType typ, AmlDmaBusMaster bm, AmlTransferSize sz,
              uint8_t channel);
 Aml *aml_sleep(uint64_t msec);
 Aml *aml_i2c_serial_bus_device(uint16_t address, const char *resource_source);
+Aml *aml_error_device(void);
 
 /* Block AML object primitives */
 Aml *aml_scope(const char *name_format, ...) G_GNUC_PRINTF(1, 2);
include/hw/acpi/generic_event_device.h +1 -0
index 40af3550b56d..9ace8fe70328 100644
@@ -98,6 +98,7 @@ OBJECT_DECLARE_SIMPLE_TYPE(AcpiGedState, ACPI_GED)
 #define ACPI_GED_PWR_DOWN_EVT      0x2
 #define ACPI_GED_NVDIMM_HOTPLUG_EVT 0x4
 #define ACPI_GED_CPU_HOTPLUG_EVT    0x8
+#define ACPI_GED_ERROR_EVT          0x10
 
 typedef struct GEDState {
     MemoryRegion evt;
Signature · 1 line
2.46.0
#18 lore raw

PATCH v9 7/12 arm/virt: Wire up a GED error device for ACPI / GHES

In reply to Mauro Carvalho Chehab

Adds support to ARM virtualization to allow handling generic error ACPI Event via GED & error source device.

It is aligned with Linux Kernel patch:
https://lore.kernel.org/lkml/[email protected]/

---

Changes from v8:

- Added a call to the function that produces GHES generic records, as this is now added earlier in this series.

---
 hw/arm/virt-acpi-build.c |  1 +
 hw/arm/virt.c            | 12 +++++++++++-
 include/hw/arm/virt.h    |  1 +
 3 files changed, 13 insertions(+), 1 deletion(-)
hw/arm/virt-acpi-build.c +1 -0
index 39100c2822c2..9c36da3c831f 100644
@@ -858,6 +858,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker, VirtMachineState *vms)
     }
 
     acpi_dsdt_add_power_button(scope);
+    aml_append(scope, aml_error_device());
 #ifdef CONFIG_TPM
     acpi_dsdt_add_tpm(scope, vms);
 #endif
hw/arm/virt.c +11 -1
index 687fe0bb8bc9..22448e5c5b73 100644
@@ -677,7 +677,7 @@ static inline DeviceState *create_acpi_ged(VirtMachineState *vms)
     DeviceState *dev;
     MachineState *ms = MACHINE(vms);
     int irq = vms->irqmap[VIRT_ACPI_GED];
-    uint32_t event = ACPI_GED_PWR_DOWN_EVT;
+    uint32_t event = ACPI_GED_PWR_DOWN_EVT | ACPI_GED_ERROR_EVT;
 
     if (ms->ram_slots) {
         event |= ACPI_GED_MEM_HOTPLUG_EVT;
@@ -1009,6 +1009,13 @@ static void virt_powerdown_req(Notifier *n, void *opaque)
     }
 }
 
+static void virt_generic_error_req(Notifier *n, void *opaque)
+{
+    VirtMachineState *s = container_of(n, VirtMachineState, generic_error_notifier);
+
+    acpi_send_event(s->acpi_dev, ACPI_GENERIC_ERROR);
+}
+
 static void create_gpio_keys(char *fdt, DeviceState *pl061_dev,
                              uint32_t phandle)
 {
@@ -2385,6 +2392,9 @@ static void machvirt_init(MachineState *machine)
 
     if (has_ged && aarch64 && firmware_loaded && virt_is_acpi_enabled(vms)) {
         vms->acpi_dev = create_acpi_ged(vms);
+        vms->generic_error_notifier.notify = virt_generic_error_req;
+        notifier_list_add(&acpi_generic_error_notifiers,
+                          &vms->generic_error_notifier);
     } else {
         create_gpio_devices(vms, VIRT_GPIO, sysmem);
     }
include/hw/arm/virt.h +1 -0
index d62d8d9db5ae..1c682d43fdac 100644
@@ -175,6 +175,7 @@ struct VirtMachineState {
     DeviceState *gic;
     DeviceState *acpi_dev;
     Notifier powerdown_notifier;
+    Notifier generic_error_notifier;
     PCIBus *bus;
     char *oem_id;
     char *oem_table_id;
Signature · 1 line
2.46.0
#19 lore raw

PATCH v9 8/12 qapi/acpi-hest: add an interface to do generic CPER error injection

In reply to Mauro Carvalho Chehab

Creates a QMP command to be used for generic ACPI APEI hardware error injection (HEST) via GHESv2, and add support for it for ARM guests.

Error injection uses ACPI_HEST_SRC_ID_QMP source ID to be platform independent. This is mapped at arch virt bindings, depending on the types supported by QEMU and by the BIOS. So, on ARM, this is supported via ACPI_GHES_NOTIFY_GPIO notification type.

This patch is co-authored:
- original ghes logic to inject a simple ARM record by Shiju Jose;
- generic logic to handle block addresses by Jonathan Cameron;
- generic GHESv2 error inject by Mauro Carvalho Chehab;

---
 MAINTAINERS              |  7 +++++++
 hw/acpi/Kconfig          |  5 +++++
 hw/acpi/ghes_cper.c      | 32 ++++++++++++++++++++++++++++++++
 hw/acpi/ghes_cper_stub.c | 19 +++++++++++++++++++
 hw/acpi/meson.build      |  2 ++
 hw/arm/Kconfig           |  5 +++++
 hw/arm/virt-acpi-build.c |  1 +
 include/hw/acpi/ghes.h   |  4 ++++
 include/hw/arm/virt.h    |  2 ++
 qapi/acpi-hest.json      | 36 ++++++++++++++++++++++++++++++++++++
 qapi/meson.build         |  1 +
 qapi/qapi-schema.json    |  1 +
 12 files changed, 115 insertions(+)
 create mode 100644 hw/acpi/ghes_cper.c
 create mode 100644 hw/acpi/ghes_cper_stub.c
 create mode 100644 qapi/acpi-hest.json
MAINTAINERS +7 -0
index 3584d6a6c6da..1d8091818899 100644
@@ -2077,6 +2077,13 @@ F: hw/acpi/ghes.c
 F: include/hw/acpi/ghes.h
 F: docs/specs/acpi_hest_ghes.rst
 
+ACPI/HEST/GHES/ARM processor CPER
+R: Mauro Carvalho Chehab <[email protected]>
+S: Maintained
+F: hw/arm/ghes_cper.c
+F: hw/acpi/ghes_cper_stub.c
+F: qapi/acpi-hest.json
+
 ppc4xx
 L: [email protected]
 S: Orphan
hw/acpi/Kconfig +5 -0
index e07d3204eb36..73ffbb82c150 100644
@@ -51,6 +51,11 @@ config ACPI_APEI
     bool
     depends on ACPI
 
+config GHES_CPER
+    bool
+    depends on ACPI_APEI
+    default y
+
 config ACPI_PCI
     bool
     depends on ACPI && PCI
hw/acpi/ghes_cper.c new file +32 -0
new file mode 100644
index 000000000000..a10a5e7ab29b
@@ -0,0 +1,32 @@
+/*
+ * CPER payload parser for error injection
+ *
+ * Copyright(C) 2024 Huawei LTD.
+ *
+ * This code is licensed under the GPL version 2 or later. See the
+ * COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+
+#include "qemu/base64.h"
+#include "qemu/error-report.h"
+#include "qemu/uuid.h"
+#include "qapi/qapi-commands-acpi-hest.h"
+#include "hw/acpi/ghes.h"
+
+void qmp_ghes_cper(const char *qmp_cper, Error **errp)
+{
+
+    uint8_t *cper;
+    size_t  len;
+
+    cper = qbase64_decode(qmp_cper, -1, &len, errp);
+    if (!cper) {
+        error_setg(errp, "missing GHES CPER payload");
+        return;
+    }
+
+    ghes_record_cper_errors(cper, len, ACPI_HEST_SRC_ID_QMP, errp);
+}
hw/acpi/ghes_cper_stub.c new file +19 -0
new file mode 100644
index 000000000000..36138c462ac9
@@ -0,0 +1,19 @@
+/*
+ * Stub interface for CPER payload parser for error injection
+ *
+ * Copyright(C) 2024 Huawei LTD.
+ *
+ * This code is licensed under the GPL version 2 or later. See the
+ * COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qapi/error.h"
+#include "qapi/qapi-commands-acpi-hest.h"
+#include "hw/acpi/ghes.h"
+
+void qmp_ghes_cper(const char *cper, Error **errp)
+{
+    error_setg(errp, "GHES QMP error inject is not compiled in");
+}
hw/acpi/meson.build +2 -0
index fa5c07db9068..6cbf430eb66d 100644
@@ -34,4 +34,6 @@ endif
 system_ss.add(when: 'CONFIG_ACPI', if_false: files('acpi-stub.c', 'aml-build-stub.c', 'ghes-stub.c', 'acpi_interface.c'))
 system_ss.add(when: 'CONFIG_ACPI_PCI_BRIDGE', if_false: files('pci-bridge-stub.c'))
 system_ss.add_all(when: 'CONFIG_ACPI', if_true: acpi_ss)
+system_ss.add(when: 'CONFIG_GHES_CPER', if_true: files('ghes_cper.c'))
+system_ss.add(when: 'CONFIG_GHES_CPER', if_false: files('ghes_cper_stub.c'))
 system_ss.add(files('acpi-qmp-cmds.c'))
hw/arm/Kconfig +5 -0
index 1ad60da7aa2d..bed6ba27d715 100644
@@ -712,3 +712,8 @@ config ARMSSE
     select UNIMP
     select SSE_COUNTER
     select SSE_TIMER
+
+config GHES_CPER
+    bool
+    depends on ARM
+    default y if AARCH64
hw/arm/virt-acpi-build.c +1 -0
index 9c36da3c831f..a6b03b16d922 100644
@@ -893,6 +893,7 @@ static void acpi_align_size(GArray *blob, unsigned align)
 
 static const uint16_t hest_ghes_notify[] = {
     [ARM_ACPI_HEST_SRC_ID_SEA] = ACPI_GHES_NOTIFY_SEA,
+    [ARM_ACPI_HEST_SRC_ID_GPIO] = ACPI_GHES_NOTIFY_GPIO,
 };
 
 static
include/hw/acpi/ghes.h +4 -0
index b1ec9795270f..ea6b33e133d6 100644
@@ -66,6 +66,10 @@ typedef struct AcpiGhesState {
     bool present; /* True if GHES is present at all on this board */
 } AcpiGhesState;
 
+
+/* Source ID associated with qapi/acpi-hest.json QMP error injection */
+#define ACPI_HEST_SRC_ID_QMP   0
+
 void acpi_build_hest(GArray *table_data, GArray *hardware_errors,
                      BIOSLinker *linker,
                      const uint16_t * const notify,
include/hw/arm/virt.h +2 -0
index 1c682d43fdac..8b042053dfa1 100644
@@ -33,6 +33,7 @@
 #include "exec/hwaddr.h"
 #include "qemu/notify.h"
 #include "hw/boards.h"
+#include "hw/acpi/ghes.h"
 #include "hw/arm/boot.h"
 #include "hw/arm/bsa.h"
 #include "hw/block/flash.h"
@@ -194,6 +195,7 @@ bool virt_is_acpi_enabled(VirtMachineState *vms);
  * ID numbers used to fill HEST source ID field
  */
 enum {
+    ARM_ACPI_HEST_SRC_ID_GPIO = ACPI_HEST_SRC_ID_QMP,
     ARM_ACPI_HEST_SRC_ID_SEA,
 };
 
qapi/acpi-hest.json new file +36 -0
new file mode 100644
index 000000000000..0255695c95ad
@@ -0,0 +1,36 @@
+# -*- Mode: Python -*-
+# vim: filetype=python
+
+##
+# = GHESv2 CPER Error Injection
+#
+# Defined since ACPI Specification 6.1,
+# section 18.3.2.8 Generic Hardware Error Source version 2. See:
+#
+# https://uefi.org/sites/default/files/resources/ACPI_6_1.pdf
+##
+
+
+##
+# @ghes-cper:
+#
+# Inject a CPER error data to be filled according to ACPI 6.1
+# spec via GHESv2.
+#
+# @cper: contains a base64 encoded string with raw data for a single CPER
+#     record with Generic Error Status Block, Generic Error Data Entry and
+#     generic error data payload, as described at
+#     https://uefi.org/specs/UEFI/2.10/Apx_N_Common_Platform_Error_Record.html#format
+#
+# Features:
+#
+# @unstable: This command is experimental.
+#
+# Since: 9.2
+##
+{ 'command': 'ghes-cper',
+  'data': {
+    'cper': 'str'
+  },
+  'features': [ 'unstable' ]
+}
qapi/meson.build +1 -0
index e7bc54e5d047..35cea6147262 100644
@@ -59,6 +59,7 @@ qapi_all_modules = [
 if have_system
   qapi_all_modules += [
     'acpi',
+    'acpi-hest',
     'audio',
     'cryptodev',
     'qdev',
qapi/qapi-schema.json +1 -0
index b1581988e4eb..baf19ab73afe 100644
@@ -75,6 +75,7 @@
 { 'include': 'misc-target.json' }
 { 'include': 'audio.json' }
 { 'include': 'acpi.json' }
+{ 'include': 'acpi-hest.json' }
 { 'include': 'pci.json' }
 { 'include': 'stats.json' }
 { 'include': 'virtio.json' }
Signature · 1 line
2.46.0
#20 lore raw

Re: [PATCH v9 08/12] qapi/acpi-hest: add an interface to do generic CPER error injection

Quoted 211 lines · Mauro Carvalho Chehab

Creates a QMP command to be used for generic ACPI APEI hardware error injection (HEST) via GHESv2, and add support for it for ARM guests.

Error injection uses ACPI_HEST_SRC_ID_QMP source ID to be platform independent. This is mapped at arch virt bindings, depending on the types supported by QEMU and by the BIOS. So, on ARM, this is supported via ACPI_GHES_NOTIFY_GPIO notification type.

This patch is co-authored:
- original ghes logic to inject a simple ARM record by Shiju Jose;
- generic logic to handle block addresses by Jonathan Cameron;
- generic GHESv2 error inject by Mauro Carvalho Chehab;

---
 MAINTAINERS              |  7 +++++++
 hw/acpi/Kconfig          |  5 +++++
 hw/acpi/ghes_cper.c      | 32 ++++++++++++++++++++++++++++++++
 hw/acpi/ghes_cper_stub.c | 19 +++++++++++++++++++
 hw/acpi/meson.build      |  2 ++
 hw/arm/Kconfig           |  5 +++++
 hw/arm/virt-acpi-build.c |  1 +
 include/hw/acpi/ghes.h   |  4 ++++
 include/hw/arm/virt.h    |  2 ++
 qapi/acpi-hest.json      | 36 ++++++++++++++++++++++++++++++++++++
 qapi/meson.build         |  1 +
 qapi/qapi-schema.json    |  1 +
 12 files changed, 115 insertions(+)
 create mode 100644 hw/acpi/ghes_cper.c
 create mode 100644 hw/acpi/ghes_cper_stub.c
 create mode 100644 qapi/acpi-hest.json
MAINTAINERS +7 -0
index 3584d6a6c6da..1d8091818899 100644
@@ -2077,6 +2077,13 @@ F: hw/acpi/ghes.c
 F: include/hw/acpi/ghes.h
 F: docs/specs/acpi_hest_ghes.rst
 
+ACPI/HEST/GHES/ARM processor CPER
+R: Mauro Carvalho Chehab <[email protected]>
+S: Maintained
+F: hw/arm/ghes_cper.c
+F: hw/acpi/ghes_cper_stub.c
+F: qapi/acpi-hest.json
+
 ppc4xx
 L: [email protected]
 S: Orphan
hw/acpi/Kconfig +5 -0
index e07d3204eb36..73ffbb82c150 100644
@@ -51,6 +51,11 @@ config ACPI_APEI
     bool
     depends on ACPI
 
+config GHES_CPER
+    bool
+    depends on ACPI_APEI
+    default y
+
 config ACPI_PCI
     bool
     depends on ACPI && PCI
hw/acpi/ghes_cper.c new file +32 -0
new file mode 100644
index 000000000000..a10a5e7ab29b
@@ -0,0 +1,32 @@
+/*
+ * CPER payload parser for error injection
+ *
+ * Copyright(C) 2024 Huawei LTD.
+ *
+ * This code is licensed under the GPL version 2 or later. See the
+ * COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+
+#include "qemu/base64.h"
+#include "qemu/error-report.h"
+#include "qemu/uuid.h"
+#include "qapi/qapi-commands-acpi-hest.h"
+#include "hw/acpi/ghes.h"
+
+void qmp_ghes_cper(const char *qmp_cper, Error **errp)
+{
+
+    uint8_t *cper;
+    size_t  len;
+
+    cper = qbase64_decode(qmp_cper, -1, &len, errp);
+    if (!cper) {
+        error_setg(errp, "missing GHES CPER payload");
+        return;
+    }
+
+    ghes_record_cper_errors(cper, len, ACPI_HEST_SRC_ID_QMP, errp);
+}
hw/acpi/ghes_cper_stub.c new file +19 -0
new file mode 100644
index 000000000000..36138c462ac9
@@ -0,0 +1,19 @@
+/*
+ * Stub interface for CPER payload parser for error injection
+ *
+ * Copyright(C) 2024 Huawei LTD.
+ *
+ * This code is licensed under the GPL version 2 or later. See the
+ * COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qapi/error.h"
+#include "qapi/qapi-commands-acpi-hest.h"
+#include "hw/acpi/ghes.h"
+
+void qmp_ghes_cper(const char *cper, Error **errp)
+{
+    error_setg(errp, "GHES QMP error inject is not compiled in");
+}
hw/acpi/meson.build +2 -0
index fa5c07db9068..6cbf430eb66d 100644
@@ -34,4 +34,6 @@ endif
 system_ss.add(when: 'CONFIG_ACPI', if_false: files('acpi-stub.c', 'aml-build-stub.c', 'ghes-stub.c', 'acpi_interface.c'))
 system_ss.add(when: 'CONFIG_ACPI_PCI_BRIDGE', if_false: files('pci-bridge-stub.c'))
 system_ss.add_all(when: 'CONFIG_ACPI', if_true: acpi_ss)
+system_ss.add(when: 'CONFIG_GHES_CPER', if_true: files('ghes_cper.c'))
+system_ss.add(when: 'CONFIG_GHES_CPER', if_false: files('ghes_cper_stub.c'))
 system_ss.add(files('acpi-qmp-cmds.c'))
hw/arm/Kconfig +5 -0
index 1ad60da7aa2d..bed6ba27d715 100644
@@ -712,3 +712,8 @@ config ARMSSE
     select UNIMP
     select SSE_COUNTER
     select SSE_TIMER
+
+config GHES_CPER
+    bool
+    depends on ARM
+    default y if AARCH64
hw/arm/virt-acpi-build.c +1 -0
index 9c36da3c831f..a6b03b16d922 100644
@@ -893,6 +893,7 @@ static void acpi_align_size(GArray *blob, unsigned align)
 
 static const uint16_t hest_ghes_notify[] = {
     [ARM_ACPI_HEST_SRC_ID_SEA] = ACPI_GHES_NOTIFY_SEA,
+    [ARM_ACPI_HEST_SRC_ID_GPIO] = ACPI_GHES_NOTIFY_GPIO,
 };
 
 static
include/hw/acpi/ghes.h +4 -0
index b1ec9795270f..ea6b33e133d6 100644
@@ -66,6 +66,10 @@ typedef struct AcpiGhesState {
     bool present; /* True if GHES is present at all on this board */
 } AcpiGhesState;
 
+
+/* Source ID associated with qapi/acpi-hest.json QMP error injection */
+#define ACPI_HEST_SRC_ID_QMP   0
+
 void acpi_build_hest(GArray *table_data, GArray *hardware_errors,
                      BIOSLinker *linker,
                      const uint16_t * const notify,
include/hw/arm/virt.h +2 -0
index 1c682d43fdac..8b042053dfa1 100644
@@ -33,6 +33,7 @@
 #include "exec/hwaddr.h"
 #include "qemu/notify.h"
 #include "hw/boards.h"
+#include "hw/acpi/ghes.h"
 #include "hw/arm/boot.h"
 #include "hw/arm/bsa.h"
 #include "hw/block/flash.h"
@@ -194,6 +195,7 @@ bool virt_is_acpi_enabled(VirtMachineState *vms);
  * ID numbers used to fill HEST source ID field
  */
 enum {
+    ARM_ACPI_HEST_SRC_ID_GPIO = ACPI_HEST_SRC_ID_QMP,
     ARM_ACPI_HEST_SRC_ID_SEA,
 };
 
qapi/acpi-hest.json new file +0 -0
new file mode 100644
index 000000000000..0255695c95ad

qapi/acpi-hest.json line 5

Quoted 5 lines
qapi/acpi-hest.json new file +5 -0
@@ -0,0 +1,36 @@
+# -*- Mode: Python -*-
+# vim: filetype=python
+
+##
+# = GHESv2 CPER Error Injection

This makes it a section next to section ACPI. Should it be a subjection of ACPI? If yes, add a second =.

qapi/acpi-hest.json line 15

Quoted 10 lines
qapi/acpi-hest.json new file +10 -0
+#
+# Defined since ACPI Specification 6.1,
+# section 18.3.2.8 Generic Hardware Error Source version 2. See:
+#
+# https://uefi.org/sites/default/files/resources/ACPI_6_1.pdf
+##
+
+
+##
+# @ghes-cper:

Can we pick a command name that provides more of a clue to the uninitiated on what the command does? inject-ghes-cper? inject-ghes-cper-error?

qapi/acpi-hest.json line 18

Quoted 3 lines
qapi/acpi-hest.json new file +3 -0
+#
+# Inject a CPER error data to be filled according to ACPI 6.1
+# spec via GHESv2.

It's either "a datum", or "data". Scratch "data"?

What do you mean by "to be filled"? Who's doing the filling?

What about something like "Inject an error with additional ACPI 6.1 GHES v2 error information"?

qapi/acpi-hest.json line 23

Quoted 5 lines
qapi/acpi-hest.json new file +5 -0
+#
+# @cper: contains a base64 encoded string with raw data for a single CPER
+#     record with Generic Error Status Block, Generic Error Data Entry and
+#     generic error data payload, as described at
+#     https://uefi.org/specs/UEFI/2.10/Apx_N_Common_Platform_Error_Record.html#format

docs/devel/qapi-code-gen.rst:

For legibility, wrap text paragraphs so every line is at most 70 characters long.

Thus:

# @cper: contains a base64 encoded string with raw data for a single

   #     CPER record with Generic Error Status Block, Generic Error Data
   #     Entry and generic error data payload, as described at
   #     https://uefi.org/specs/UEFI/2.10/Apx_N_Common_Platform_Error_Record.html#format
Quoted 37 lines
qapi/acpi-hest.json new file +13 -0
+#
+# Features:
+#
+# @unstable: This command is experimental.
+#
+# Since: 9.2
+##
+{ 'command': 'ghes-cper',
+  'data': {
+    'cper': 'str'
+  },
+  'features': [ 'unstable' ]
+}
qapi/meson.build +1 -0
index e7bc54e5d047..35cea6147262 100644
@@ -59,6 +59,7 @@ qapi_all_modules = [
 if have_system
   qapi_all_modules += [
     'acpi',
+    'acpi-hest',
     'audio',
     'cryptodev',
     'qdev',
qapi/qapi-schema.json +1 -0
index b1581988e4eb..baf19ab73afe 100644
@@ -75,6 +75,7 @@
 { 'include': 'misc-target.json' }
 { 'include': 'audio.json' }
 { 'include': 'acpi.json' }
+{ 'include': 'acpi-hest.json' }
 { 'include': 'pci.json' }
 { 'include': 'stats.json' }
 { 'include': 'virtio.json' }
#21 lore raw

PATCH v9 9/12 docs: acpi_hest_ghes: fix documentation for CPER size

In reply to Mauro Carvalho Chehab

While the spec defines a CPER size of 4KiB for each record, currently it is set to 1KiB. Fix the documentation and add a pointer to the macro name there, as this may help to keep it updated.

---
 docs/specs/acpi_hest_ghes.rst | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
docs/specs/acpi_hest_ghes.rst +4 -2
index 68f1fbe0a4af..c3e9f8d9a702 100644
@@ -67,8 +67,10 @@ Design Details
 (3) The address registers table contains N Error Block Address entries
     and N Read Ack Register entries. The size for each entry is 8-byte.
     The Error Status Data Block table contains N Error Status Data Block
-    entries. The size for each entry is 4096(0x1000) bytes. The total size
-    for the "etc/hardware_errors" fw_cfg blob is (N * 8 * 2 + N * 4096) bytes.
+    entries. The size for each entry is defined at the source code as
+    ACPI_GHES_MAX_RAW_DATA_LENGTH (currently 1024 bytes). The total size
+    for the "etc/hardware_errors" fw_cfg blob is
+    (N * 8 * 2 + N * ACPI_GHES_MAX_RAW_DATA_LENGTH) bytes.
     N is the number of the kinds of hardware error sources.
 
 (4) QEMU generates the ACPI linker/loader script for the firmware. The
Signature · 1 line
2.46.0
#22 lore raw

PATCH v9 10/12 scripts/ghes_inject: add a script to generate GHES error inject

In reply to Mauro Carvalho Chehab

Using the QMP GHESv2 API requires preparing a raw data array containing a CPER record.

Add a helper script with subcommands to prepare such data.

Currently, only ARM Processor error CPER record is supported.

---
 MAINTAINERS                    |   3 +
 scripts/arm_processor_error.py | 377 ++++++++++++++++++
 scripts/ghes_inject.py         |  51 +++
 scripts/qmp_helper.py          | 702 +++++++++++++++++++++++++++++++++
 4 files changed, 1133 insertions(+)
 create mode 100644 scripts/arm_processor_error.py
 create mode 100755 scripts/ghes_inject.py
 create mode 100644 scripts/qmp_helper.py
MAINTAINERS +3 -0
index 1d8091818899..249ed2858198 100644
@@ -2083,6 +2083,9 @@ S: Maintained
 F: hw/arm/ghes_cper.c
 F: hw/acpi/ghes_cper_stub.c
 F: qapi/acpi-hest.json
+F: scripts/ghes_inject.py
+F: scripts/arm_processor_error.py
+F: scripts/qmp_helper.py
 
 ppc4xx
 L: [email protected]
scripts/arm_processor_error.py new file +377 -0

377 lines not on this page · raw message

scripts/ghes_inject.py new file +51 -0

51 lines not on this page · raw message

scripts/qmp_helper.py new file +702 -0

702 lines not on this page · raw message

Signature · 1 line
2.46.0
#23 lore raw

PATCH v9 11/12 target/arm: add an experimental mpidr arm cpu property object

In reply to Mauro Carvalho Chehab

Accurately injecting an ARM Processor error ACPI/APEI GHES error record requires the value of the ARM Multiprocessor Affinity Register (mpidr).

While ARM implements it, this is currently not visible.

Add a field at CPU storing it, and place it at arm_cpu_properties as experimental, thus allowing it to be queried via QMP using qom-get function.

---
 target/arm/cpu.c    |  1 +
 target/arm/cpu.h    |  1 +
 target/arm/helper.c | 10 ++++++++--
 3 files changed, 10 insertions(+), 2 deletions(-)
target/arm/cpu.c +1 -0
index 19191c239181..30fcf0a10f46 100644
@@ -2619,6 +2619,7 @@ static ObjectClass *arm_cpu_class_by_name(const char *cpu_model)
 
 static Property arm_cpu_properties[] = {
     DEFINE_PROP_UINT64("midr", ARMCPU, midr, 0),
+    DEFINE_PROP_UINT64("x-mpidr", ARMCPU, mpidr, 0),
     DEFINE_PROP_UINT64("mp-affinity", ARMCPU,
                         mp_affinity, ARM64_AFFINITY_INVALID),
     DEFINE_PROP_INT32("node-id", ARMCPU, node_id, CPU_UNSET_NUMA_NODE_ID),
target/arm/cpu.h +1 -0
index 9a3fd595621f..3ad4de793409 100644
@@ -1033,6 +1033,7 @@ struct ArchCPU {
         uint64_t reset_pmcr_el0;
     } isar;
     uint64_t midr;
+    uint64_t mpidr;
     uint32_t revidr;
     uint32_t reset_fpsid;
     uint64_t ctr;
target/arm/helper.c +8 -2
index 0a582c1cd3b3..d6e7aa069489 100644
@@ -4690,7 +4690,7 @@ static uint64_t mpidr_read_val(CPUARMState *env)
     return mpidr;
 }
 
-static uint64_t mpidr_read(CPUARMState *env, const ARMCPRegInfo *ri)
+static uint64_t mpidr_read(CPUARMState *env)
 {
     unsigned int cur_el = arm_current_el(env);
 
@@ -4700,6 +4700,11 @@ static uint64_t mpidr_read(CPUARMState *env, const ARMCPRegInfo *ri)
     return mpidr_read_val(env);
 }
 
+static uint64_t mpidr_read_ri(CPUARMState *env, const ARMCPRegInfo *ri)
+{
+    return mpidr_read(env);
+}
+
 static const ARMCPRegInfo lpae_cp_reginfo[] = {
     /* NOP AMAIR0/1 */
     { .name = "AMAIR0", .state = ARM_CP_STATE_BOTH,
@@ -9721,7 +9726,7 @@ void register_cp_regs_for_features(ARMCPU *cpu)
             { .name = "MPIDR_EL1", .state = ARM_CP_STATE_BOTH,
               .opc0 = 3, .crn = 0, .crm = 0, .opc1 = 0, .opc2 = 5,
               .fgt = FGT_MPIDR_EL1,
-              .access = PL1_R, .readfn = mpidr_read, .type = ARM_CP_NO_RAW },
+              .access = PL1_R, .readfn = mpidr_read_ri, .type = ARM_CP_NO_RAW },
         };
 #ifdef CONFIG_USER_ONLY
         static const ARMCPRegUserSpaceInfo mpidr_user_cp_reginfo[] = {
@@ -9731,6 +9736,7 @@ void register_cp_regs_for_features(ARMCPU *cpu)
         modify_arm_cp_regs(mpidr_cp_reginfo, mpidr_user_cp_reginfo);
 #endif
         define_arm_cp_regs(cpu, mpidr_cp_reginfo);
+        cpu->mpidr = mpidr_read(env);
     }
 
     if (arm_feature(env, ARM_FEATURE_AUXCR)) {
Signature · 1 line
2.46.0
#24 lore raw

Re: [PATCH v9 11/12] target/arm: add an experimental mpidr arm cpu property object

Quoted 11 lines · Mauro Carvalho Chehab

Accurately injecting an ARM Processor error ACPI/APEI GHES error record requires the value of the ARM Multiprocessor Affinity Register (mpidr).

While ARM implements it, this is currently not visible.

Add a field at CPU storing it, and place it at arm_cpu_properties as experimental, thus allowing it to be queried via QMP using qom-get function.

Quoted 6 lines
 static Property arm_cpu_properties[] = {
     DEFINE_PROP_UINT64("midr", ARMCPU, midr, 0),
+    DEFINE_PROP_UINT64("x-mpidr", ARMCPU, mpidr, 0),
     DEFINE_PROP_UINT64("mp-affinity", ARMCPU,
                         mp_affinity, ARM64_AFFINITY_INVALID),
     DEFINE_PROP_INT32("node-id", ARMCPU, node_id, CPU_UNSET_NUMA_NODE_ID),

Why do we need this? Why is it experimental? The later patch seems to use it via QMP, which I'm not super enthusiastic about -- the preexisting mpidr and mp-affinity properties are there for code that is creating CPU objects to configure the CPU object, not as a query interface for QOM.

thanks
-- PMM

#25 lore raw

Re: [PATCH v9 11/12] target/arm: add an experimental mpidr arm cpu property object

Em Sun, 25 Aug 2024 12:34:14 +0100
Peter Maydell <[email protected]> escreveu:

Quoted 13 lines
Quoted 10 lines · Mauro Carvalho Chehab

Accurately injecting an ARM Processor error ACPI/APEI GHES error record requires the value of the ARM Multiprocessor Affinity Register (mpidr).

While ARM implements it, this is currently not visible.

Add a field at CPU storing it, and place it at arm_cpu_properties as experimental, thus allowing it to be queried via QMP using qom-get function.

Quoted 8 lines
Quoted 6 lines
 static Property arm_cpu_properties[] = {
     DEFINE_PROP_UINT64("midr", ARMCPU, midr, 0),
+    DEFINE_PROP_UINT64("x-mpidr", ARMCPU, mpidr, 0),
     DEFINE_PROP_UINT64("mp-affinity", ARMCPU,
                         mp_affinity, ARM64_AFFINITY_INVALID),
     DEFINE_PROP_INT32("node-id", ARMCPU, node_id, CPU_UNSET_NUMA_NODE_ID),  

Why do we need this?

The ACPI HEST tables, in particular when using GHESv2 provide several kinds of errors. Among them, we have ARM Processor Error, as defined at UEFI 2.10 spec (and earlier versions), the Common Platform Error Record (CPER) is defined as:

https://uefi.org/specs/UEFI/2.10/Apx_N_Common_Platform_Error_Record.html?highlight=ghes#arm-processor-error-section

There are two fields that are part of the CPER record. One of them is mandatory (MIDR); the other one is optional, but needed to decode another field.

So, basically those errors need them.

Quoted 1 line

Why is it experimental?

This was a suggestion from Igor. As for now the QAPI for external error injection is experimental, It makes sense to me to keep it experimental as well.

Quoted 5 lines

The later patch
seems to use it via QMP, which I'm not super enthusiastic about -- the preexisting mpidr and mp-affinity properties are there for code that is creating CPU objects to configure the CPU object, not as a query interface for QOM.

I saw that. Basically the decoding by OS guest depends on MPIDR, as explained at the description of Error affinity level field:

"For errors that can be attributed to a specific affinity level, this field defines the affinity level at which the error was produced, detected, and/or consumed. This is a value between 0 and 3. All other values (4-255) are reserved

For example, a vendor may choose to define affinity levels as follows:
Level 0: errors that can be precisely attributed to a specific CPU (e.g. due to a synchronous external abort)
Level 1: Cache parity and/or ECC errors detected at cache of affinity level 1 (e.g. only attributed to higher level cache due to prefetching and/or error propagation)

NOTE: Detailed meanings and groupings of affinity level are chip and/or platform specific. The affinity level described here must be consistent with the platform definitions used MPIDR. For cache/TLB errors, the cache/TLB level is provided by the cache/TLB error structure, which may differ from affinity level."

Regards,
Mauro

#26 lore raw

Re: [PATCH v9 11/12] target/arm: add an experimental mpidr arm cpu property object

Quoted 31 lines · Mauro Carvalho Chehab

Em Sun, 25 Aug 2024 12:34:14 +0100
Peter Maydell <[email protected]> escreveu:

Quoted 13 lines
Quoted 10 lines · Mauro Carvalho Chehab

Accurately injecting an ARM Processor error ACPI/APEI GHES error record requires the value of the ARM Multiprocessor Affinity Register (mpidr).

While ARM implements it, this is currently not visible.

Add a field at CPU storing it, and place it at arm_cpu_properties as experimental, thus allowing it to be queried via QMP using qom-get function.

Quoted 8 lines
Quoted 6 lines
 static Property arm_cpu_properties[] = {
     DEFINE_PROP_UINT64("midr", ARMCPU, midr, 0),
+    DEFINE_PROP_UINT64("x-mpidr", ARMCPU, mpidr, 0),
     DEFINE_PROP_UINT64("mp-affinity", ARMCPU,
                         mp_affinity, ARM64_AFFINITY_INVALID),
     DEFINE_PROP_INT32("node-id", ARMCPU, node_id, CPU_UNSET_NUMA_NODE_ID),

Why do we need this?

The ACPI HEST tables, in particular when using GHESv2 provide several kinds of errors. Among them, we have ARM Processor Error, as defined at UEFI 2.10 spec (and earlier versions), the Common Platform Error Record (CPER) is defined as:

Quoted 7 lines

https://uefi.org/specs/UEFI/2.10/Apx_N_Common_Platform_Error_Record.html?highlight=ghes#arm-processor-error-section

There are two fields that are part of the CPER record. One of them is mandatory (MIDR); the other one is optional, but needed to decode another field.

So, basically those errors need them.

OK, but why do scripts outside of QEMU need the information, as opposed to telling QEMU "hey, generate an error" and QEMU knowing the format to use? Do we have any other QMP APIs where something external provides raw ACPI data like this?

-- PMM

#27 lore raw

Re: [PATCH v9 11/12] target/arm: add an experimental mpidr arm cpu property object

Em Fri, 30 Aug 2024 17:27:27 +0100
Peter Maydell <[email protected]> escreveu:

Quoted 39 lines
Quoted 37 lines · Mauro Carvalho Chehab

Em Sun, 25 Aug 2024 12:34:14 +0100
Peter Maydell <[email protected]> escreveu:

Quoted 13 lines
Quoted 10 lines · Mauro Carvalho Chehab

Accurately injecting an ARM Processor error ACPI/APEI GHES error record requires the value of the ARM Multiprocessor Affinity Register (mpidr).

While ARM implements it, this is currently not visible.

Add a field at CPU storing it, and place it at arm_cpu_properties as experimental, thus allowing it to be queried via QMP using qom-get function.

Quoted 8 lines
Quoted 6 lines
 static Property arm_cpu_properties[] = {
     DEFINE_PROP_UINT64("midr", ARMCPU, midr, 0),
+    DEFINE_PROP_UINT64("x-mpidr", ARMCPU, mpidr, 0),
     DEFINE_PROP_UINT64("mp-affinity", ARMCPU,
                         mp_affinity, ARM64_AFFINITY_INVALID),
     DEFINE_PROP_INT32("node-id", ARMCPU, node_id, CPU_UNSET_NUMA_NODE_ID),  

Why do we need this?

The ACPI HEST tables, in particular when using GHESv2 provide several kinds of errors. Among them, we have ARM Processor Error, as defined at UEFI 2.10 spec (and earlier versions), the Common Platform Error Record (CPER) is defined as:

https://uefi.org/specs/UEFI/2.10/Apx_N_Common_Platform_Error_Record.html?highlight=ghes#arm-processor-error-section

There are two fields that are part of the CPER record. One of them is mandatory (MIDR); the other one is optional, but needed to decode another field.

Quoted 7 lines
Quoted 1 line

So, basically those errors need them.

OK, but why do scripts outside of QEMU need the information, as opposed to telling QEMU "hey, generate an error" and QEMU knowing the format to use? Do we have any other QMP APIs where something external provides raw ACPI data like this?

This was discussed during the review of this patch series.

See, the ACPI Platform Error Interfaces (APEI) code currently in QEMU implements limited support for ACPI HEST - Hardware Error Source Table [1].

[1] https://uefi.org/specs/ACPI/6.5/18_Platform_Error_Interfaces.html#acpi-error-source

HEST consists of, currently, 9 error types (plus 3 obsoleted ones). Among them, there is support for generic errors via GHES and GHESv2 types. While not officially obsoleted, GHES is superseded by GHESv2.

GHESv2 (and GHES) has a section type field to identify which error type it is [2]. Currently, there are +10 defined UUIDs for the section type.

[2] https://uefi.org/specs/UEFI/2.10/Apx_N_Common_Platform_Error_Record.html#section-descriptor

The current code on ghes.c implements GHESv2 support for a single type (memory error), received from the host OS via SIGBUS.

Testing such code and injecting such error is not easy, as the host OS needs to send a SIGBUS to the guest, this reflecting an error at the main OS. Such code also has several limitations.

-

At the first three versions of this patch set, the code was just doing like what you said: it was adding an error injection for a HEST GHESv2 ARM Processor Error. So the error record (CPER) were produced in QEMU using some optional parameters passed via QMP to change fields when needed. With such approach, QEMU could use directly the value from MIDR and MPIDR.

The main disadvantage is that, to make full support of HEST, a lot of code will be needed to add support for every GHESv2 type and for every GHESv2 section type. So, the feedback we had were to re-implement it into a generic way.

The generic CPER error inject approach (since v4 of this series), has soma advantages:

- it is easy to do fuzz testing, as the entire CPER is built via a python script;
- no need to modify QEMU to support other GHESv2 types of record and to support other types of processors;
- GHESv2 fields can also be dynamically generated;
- It shouldn't be hard to change the code to support other types of HEST table (currently, only GHESv2 is supported).

The disadvantage is that queries are needed to pick configuration and register values from the current emulation to do error injection. For ARM Processor Error, it means that MPIDR and MIDR, are needed. Other processors and other error types will also require to query other data from QEMU, either using already-existing QMP code or by adding new ones.

Yet, the amount of code for such queries seem to be smaller than the amount of code to be added for every single GHESv2/HEST type.

-

Worth saying that QEMU may still require internal HEST/GHES errors to be able to reflect at the guests hardware problems detected at the host OS.

So, for instance, if a host OS memory is poisoned due to hardware errors, QEMU and guests need to know, in order to kill processes affected by a bad memory.

Regards,
Mauro

#28 lore raw

PATCH v9 12/12 scripts/arm_processor_error.py: retrieve mpidr if not filled

In reply to Mauro Carvalho Chehab

Add support to retrieve mpidr value via qom-get.

---
 scripts/arm_processor_error.py | 27 +++++++++++++++++++--------
 1 file changed, 19 insertions(+), 8 deletions(-)
scripts/arm_processor_error.py +19 -8
index 62e0c5662232..0a16d4f0d8b1 100644
@@ -5,12 +5,10 @@
 #
 # Copyright (C) 2024 Mauro Carvalho Chehab <[email protected]>
 
-# TODO: current implementation has dummy defaults.
-#
-# For a better implementation, a QMP addition/call is needed to
-# retrieve some data for ARM Processor Error injection:
-#
-#   - ARM registers: power_state, mpidr.
+# Note: currently it lacks a method to fill the ARM Processor Error CPER
+# psci field from emulation. On a real hardware, this is filled only
+# when a CPU is not running. Implementing support for it to simulate a
+# real hardware is not trivial.
 
 import argparse
 import re
@@ -174,11 +172,24 @@ def send_cper(self, args):
         else:
             cper["running-state"] = 0
 
+        if args.mpidr:
+            cper["mpidr-el1"] = arg["mpidr"]
+        elif cpus:
+            cmd_arg = {
+                'path': cpus[0],
+                'property': "x-mpidr"
+            }
+            ret = qmp_cmd.send_cmd("qom-get", cmd_arg, may_open=True)
+            if isinstance(ret, int):
+                cper["mpidr-el1"] = ret
+            else:
+                cper["mpidr-el1"] = 0
+
         if arm_valid_init:
             if args.affinity:
                 cper["valid"] |= self.arm_valid_bits["affinity"]
 
-            if args.mpidr:
+            if "mpidr-el1" in cper:
                 cper["valid"] |= self.arm_valid_bits["mpidr"]
 
             if "running-state" in cper:
@@ -362,7 +373,7 @@ def send_cper(self, args):
                 if isinstance(ret, int):
                     arg["midr-el1"] = ret
 
-        util.data_add(data, arg.get("mpidr-el1", 0), 8)
+        util.data_add(data, cper["mpidr-el1"], 8)
         util.data_add(data, arg.get("midr-el1", 0), 8)
         util.data_add(data, cper["running-state"], 4)
         util.data_add(data, arg.get("psci-state", 0), 4)
Signature · 1 line
2.46.0