LinuxLists.cc - [PATCH v6 0/8] Make ghes

2022-09-12 15:00:32

Subject: [PATCH v6 0/8] Make ghes_edac a proper module

2022-09-12 15:04:50

Subject: [PATCH v6 3/8] EDAC:ghes: Move ghes_edac.force_load to ghes module parameter

ghes_edac_init() is too late to set this module flag ghes_edac.force_load.
Also, other edac drivers should not be able to control this flag.

Move this flag to the module parameter in ghes instead.

Suggested-by: Toshi Kani <[email protected]>
Signed-off-by: Jia He <[email protected]>
Reviewed-by: Toshi Kani <[email protected]>
---
drivers/acpi/apei/ghes.c | 8 ++++++++
drivers/edac/ghes_edac.c | 10 +++-------
include/acpi/apei.h | 2 ++
3 files changed, 13 insertions(+), 7 deletions(-)

diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c
index 8cb65f757d06..b0a6445c6da2 100644
--- a/drivers/acpi/apei/ghes.c
+++ b/drivers/acpi/apei/ghes.c
@@ -109,6 +109,14 @@ static inline bool is_hest_type_generic_v2(struct ghes *ghes)
bool ghes_disable;
module_param_named(disable, ghes_disable, bool, 0);

+/*
+ * "ghes.edac_force_enable" forcibly enables ghes_edac and skips the platform
+ * check.
+ */
+bool ghes_edac_force_enable;
+EXPORT_SYMBOL(ghes_edac_force_enable);
+module_param_named(edac_force_enable, ghes_edac_force_enable, bool, 0);
+
/*
* All error sources notified with HED (Hardware Error Device) share a
* single notifier callback, so they need to be linked and checked one
diff --git a/drivers/edac/ghes_edac.c b/drivers/edac/ghes_edac.c
index 7b8d56a769f6..11a1b5e7e484 100644
--- a/drivers/edac/ghes_edac.c
+++ b/drivers/edac/ghes_edac.c
@@ -54,10 +54,6 @@ static DEFINE_MUTEX(ghes_reg_mutex);
*/
static DEFINE_SPINLOCK(ghes_lock);

-/* "ghes_edac.force_load=1" skips the platform check */
-static bool __read_mostly force_load;
-module_param(force_load, bool, 0);
-
static bool system_scanned;

/* Memory Device - Type 17 of SMBIOS spec */
@@ -408,10 +404,10 @@ int ghes_edac_register(struct ghes *ghes, struct device *dev)
if (IS_ENABLED(CONFIG_X86)) {
/* Check if safe to enable on this system */
idx = acpi_match_platform_list(plat_list);
- if (!force_load && idx < 0)
+ if (!ghes_edac_force_enable && idx < 0)
return -ENODEV;
} else {
- force_load = true;
+ ghes_edac_force_enable = true;
idx = 0;
}

@@ -535,7 +531,7 @@ void ghes_edac_unregister(struct ghes *ghes)
struct mem_ctl_info *mci;
unsigned long flags;

- if (!force_load)
+ if (!ghes_edac_force_enable)
return;

mutex_lock(&ghes_reg_mutex);
diff --git a/include/acpi/apei.h b/include/acpi/apei.h
index dc60f7db5524..ab310393766e 100644
--- a/include/acpi/apei.h
+++ b/include/acpi/apei.h
@@ -27,9 +27,11 @@ extern int hest_disable;
extern int erst_disable;
#ifdef CONFIG_ACPI_APEI_GHES
extern bool ghes_disable;
+extern bool ghes_edac_force_enable;
void __init acpi_ghes_init(void);
#else
#define ghes_disable 1
+#define ghes_edac_force_enable 0
static inline void acpi_ghes_init(void) { }
#endif

--
2.25.1

2022-09-12 15:10:24

by Justin He

[permalink] [raw]

Subject: [PATCH v6 5/8] EDAC/ghes: Make ghes_edac a proper module to remove the dependency on ghes

Commit dc4e8c07e9e2 ("ACPI: APEI: explicit init of HEST and GHES in
apci_init()") introduced a bug that ghes_edac_register() would be invoked
before edac_init(). Because at that time, the bus "edac" hadn't been even
registered, this created sysfs /devices/mc0 instead of
/sys/devices/system/edac/mc/mc0 on an Ampere eMag server.

To remove the dependency of ghes_edac on ghes, make it a proper module. Use
a list to save the probing devices in ghes_probe(), and defer the
ghes_edac_register() to module_init() of the new ghes_edac module by
iterating over the devices list.

The ghes_edac_force_enable check is not needed in ghes_edac_unregister()
since it has been checked in ghes_edac_init().

Co-developed-by: Borislav Petkov <[email protected]>
Signed-off-by: Borislav Petkov <[email protected]>
Signed-off-by: Jia He <[email protected]>
Fixes: dc4e8c07e9e2 ("ACPI: APEI: explicit init of HEST and GHES in apci_init()")
Cc: Shuai Xue <[email protected]>
Acked-by: Toshi Kani <[email protected]>
---
drivers/acpi/apei/ghes.c | 22 +++++++++++++--
drivers/edac/Kconfig | 4 +--
drivers/edac/ghes_edac.c | 59 +++++++++++++++++++++++-----------------
include/acpi/ghes.h | 24 ++++------------
4 files changed, 62 insertions(+), 47 deletions(-)

diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c
index 0c2dbfffe22a..4af269b22bd4 100644
--- a/drivers/acpi/apei/ghes.c
+++ b/drivers/acpi/apei/ghes.c
@@ -130,6 +130,9 @@ static bool ghes_present;
static LIST_HEAD(ghes_hed);
static DEFINE_MUTEX(ghes_list_mutex);

+static LIST_HEAD(ghes_devs);
+static DEFINE_MUTEX(ghes_devs_mutex);
+
/*
* Because the memory area used to transfer hardware error information
* from BIOS to Linux can be determined only in NMI, IRQ or timer
@@ -1388,7 +1391,11 @@ static int ghes_probe(struct platform_device *ghes_dev)

platform_set_drvdata(ghes_dev, ghes);

- ghes_edac_register(ghes, &ghes_dev->dev);
+ ghes->dev = &ghes_dev->dev;
+
+ mutex_lock(&ghes_devs_mutex);
+ list_add_tail(&ghes->elist, &ghes_devs);
+ mutex_unlock(&ghes_devs_mutex);

ghes_present = true;

@@ -1454,7 +1461,9 @@ static int ghes_remove(struct platform_device *ghes_dev)

ghes_fini(ghes);

- ghes_edac_unregister(ghes);
+ mutex_lock(&ghes_devs_mutex);
+ list_del(&ghes->elist);
+ mutex_unlock(&ghes_devs_mutex);

kfree(ghes);

@@ -1537,6 +1546,15 @@ bool ghes_edac_preferred(void)
}
EXPORT_SYMBOL_GPL(ghes_edac_preferred);

+struct list_head *ghes_get_devices(void)
+{
+ if (!ghes_edac_preferred())
+ return NULL;
+
+ return &ghes_devs;
+}
+EXPORT_SYMBOL_GPL(ghes_get_devices);
+
void ghes_register_report_chain(struct notifier_block *nb)
{
atomic_notifier_chain_register(&ghes_report_chain, nb);
diff --git a/drivers/edac/Kconfig b/drivers/edac/Kconfig
index 17562cf1fe97..df45db81858b 100644
--- a/drivers/edac/Kconfig
+++ b/drivers/edac/Kconfig
@@ -53,8 +53,8 @@ config EDAC_DECODE_MCE
has been initialized.

config EDAC_GHES
- bool "Output ACPI APEI/GHES BIOS detected errors via EDAC"
- depends on ACPI_APEI_GHES && (EDAC=y)
+ tristate "Output ACPI APEI/GHES BIOS detected errors via EDAC"
+ depends on ACPI_APEI_GHES
select UEFI_CPER
help
Not all machines support hardware-driven error report. Some of those
diff --git a/drivers/edac/ghes_edac.c b/drivers/edac/ghes_edac.c
index 11a1b5e7e484..129665896a80 100644
--- a/drivers/edac/ghes_edac.c
+++ b/drivers/edac/ghes_edac.c
@@ -56,6 +56,8 @@ static DEFINE_SPINLOCK(ghes_lock);

static bool system_scanned;

+static struct list_head *ghes_devs;
+
/* Memory Device - Type 17 of SMBIOS spec */
struct memdev_dmi_entry {
u8 type;
@@ -383,34 +385,15 @@ static struct notifier_block ghes_edac_mem_err_nb = {
.priority = 0,
};

-/*
- * Known systems that are safe to enable this module.
- */
-static struct acpi_platform_list plat_list[] = {
- {"HPE ", "Server ", 0, ACPI_SIG_FADT, all_versions},
- { } /* End */
-};
-
-int ghes_edac_register(struct ghes *ghes, struct device *dev)
+static int ghes_edac_register(struct device *dev)
{
bool fake = false;
struct mem_ctl_info *mci;
struct ghes_pvt *pvt;
struct edac_mc_layer layers[1];
unsigned long flags;
- int idx = -1;
int rc = 0;

- if (IS_ENABLED(CONFIG_X86)) {
- /* Check if safe to enable on this system */
- idx = acpi_match_platform_list(plat_list);
- if (!ghes_edac_force_enable && idx < 0)
- return -ENODEV;
- } else {
- ghes_edac_force_enable = true;
- idx = 0;
- }
-
/* finish another registration/unregistration instance first */
mutex_lock(&ghes_reg_mutex);

@@ -454,7 +437,7 @@ int ghes_edac_register(struct ghes *ghes, struct device *dev)
pr_info("This system has a very crappy BIOS: It doesn't even list the DIMMS.\n");
pr_info("Its SMBIOS info is wrong. It is doubtful that the error report would\n");
pr_info("work on such system. Use this driver with caution\n");
- } else if (idx < 0) {
+ } else if (ghes_edac_force_enable) {
pr_info("This EDAC driver relies on BIOS to enumerate memory and get error reports.\n");
pr_info("Unfortunately, not all BIOSes reflect the memory layout correctly.\n");
pr_info("So, the end result of using this driver varies from vendor to vendor.\n");
@@ -526,14 +509,11 @@ int ghes_edac_register(struct ghes *ghes, struct device *dev)
return rc;
}

-void ghes_edac_unregister(struct ghes *ghes)
+static void ghes_edac_unregister(struct ghes *ghes)
{
struct mem_ctl_info *mci;
unsigned long flags;

- if (!ghes_edac_force_enable)
- return;
-
mutex_lock(&ghes_reg_mutex);

system_scanned = false;
@@ -562,3 +542,32 @@ void ghes_edac_unregister(struct ghes *ghes)
unlock:
mutex_unlock(&ghes_reg_mutex);
}
+
+static int __init ghes_edac_init(void)
+{
+ struct ghes *g, *g_tmp;
+
+ ghes_devs = ghes_get_devices();
+ if (!ghes_devs)
+ return -ENODEV;
+
+ list_for_each_entry_safe(g, g_tmp, ghes_devs, elist) {
+ ghes_edac_register(g->dev);
+ }
+
+ return 0;
+}
+module_init(ghes_edac_init);
+
+static void __exit ghes_edac_exit(void)
+{
+ struct ghes *g, *g_tmp;
+
+ list_for_each_entry_safe(g, g_tmp, ghes_devs, elist) {
+ ghes_edac_unregister(g);
+ }
+}
+module_exit(ghes_edac_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Output ACPI APEI/GHES BIOS detected errors module via EDAC");
diff --git a/include/acpi/ghes.h b/include/acpi/ghes.h
index 7ce91c0ff3eb..08f12dde99ae 100644
--- a/include/acpi/ghes.h
+++ b/include/acpi/ghes.h
@@ -27,6 +27,8 @@ struct ghes {
struct timer_list timer;
unsigned int irq;
};
+ struct device *dev;
+ struct list_head elist;
};

struct ghes_estatus_node {
@@ -69,30 +71,16 @@ int ghes_register_vendor_record_notifier(struct notifier_block *nb);
* @nb: pointer to the notifier_block structure of the vendor record handler.
*/
void ghes_unregister_vendor_record_notifier(struct notifier_block *nb);
-#endif
-
-int ghes_estatus_pool_init(int num_ghes);
-
-/* From drivers/edac/ghes_edac.c */
-
-#ifdef CONFIG_EDAC_GHES
-int ghes_edac_register(struct ghes *ghes, struct device *dev);
-
-void ghes_edac_unregister(struct ghes *ghes);

+struct list_head *ghes_get_devices(void);
bool ghes_edac_preferred(void);
#else
-static inline int ghes_edac_register(struct ghes *ghes, struct device *dev)
-{
- return -ENODEV;
-}
-
-static inline void ghes_edac_unregister(struct ghes *ghes)
-{
-}
+static inline struct list_head *ghes_get_devices(void) { return NULL; }
static bool ghes_edac_preferred(void) { return false; }
#endif

+int ghes_estatus_pool_init(int num_ghes);
+
static inline int acpi_hest_get_version(struct acpi_hest_generic_data *gdata)
{
return gdata->revision >> 8;
--
2.25.1

2022-09-12 15:33:42

by Justin He

[permalink] [raw]

Subject: [PATCH v6 6/8] EDAC: Add the ghes_edac_preferred check for chipset-specific edac drivers

Add ghes_edac_preferred() check to chipset-specific edac drivers to ensure
that ghes_edac is used on the platform where ghes_edac is preferred over
chipset-specific edac driver.

Unlike the existing edac_get_owner() check, the ghes_edac_preferred()
check works independent to the module_init ordering.

Suggested-by: Toshi Kani <[email protected]>
Signed-off-by: Jia He <[email protected]>
Reviewed-by: Toshi Kani <[email protected]>
---
drivers/edac/amd64_edac.c | 3 +++
drivers/edac/armada_xp_edac.c | 3 +++
drivers/edac/edac_module.h | 1 +
drivers/edac/i10nm_base.c | 3 +++
drivers/edac/igen6_edac.c | 3 +++
drivers/edac/layerscape_edac.c | 3 +++
drivers/edac/pnd2_edac.c | 3 +++
drivers/edac/sb_edac.c | 3 +++
drivers/edac/skx_base.c | 3 +++
drivers/edac/thunderx_edac.c | 3 +++
drivers/edac/xgene_edac.c | 3 +++
11 files changed, 31 insertions(+)

diff --git a/drivers/edac/amd64_edac.c b/drivers/edac/amd64_edac.c
index 2f854feeeb23..5314a934c2bf 100644
--- a/drivers/edac/amd64_edac.c
+++ b/drivers/edac/amd64_edac.c
@@ -4329,6 +4329,9 @@ static int __init amd64_edac_init(void)
int err = -ENODEV;
int i;

+ if (ghes_edac_preferred())
+ return -EBUSY;
+
owner = edac_get_owner();
if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
return -EBUSY;
diff --git a/drivers/edac/armada_xp_edac.c b/drivers/edac/armada_xp_edac.c
index 038abbb83f4b..486532b92ce0 100644
--- a/drivers/edac/armada_xp_edac.c
+++ b/drivers/edac/armada_xp_edac.c
@@ -599,6 +599,9 @@ static int __init armada_xp_edac_init(void)
{
int res;

+ if (ghes_edac_preferred())
+ return -EBUSY;
+
/* only polling is supported */
edac_op_state = EDAC_OPSTATE_POLL;

diff --git a/drivers/edac/edac_module.h b/drivers/edac/edac_module.h
index 96f6de0c8ff6..3826f82de487 100644
--- a/drivers/edac/edac_module.h
+++ b/drivers/edac/edac_module.h
@@ -11,6 +11,7 @@
#ifndef __EDAC_MODULE_H__
#define __EDAC_MODULE_H__

+#include <acpi/ghes.h>
#include "edac_mc.h"
#include "edac_pci.h"
#include "edac_device.h"
diff --git a/drivers/edac/i10nm_base.c b/drivers/edac/i10nm_base.c
index 6cf50ee0b77c..691df9b51622 100644
--- a/drivers/edac/i10nm_base.c
+++ b/drivers/edac/i10nm_base.c
@@ -548,6 +548,9 @@ static int __init i10nm_init(void)

edac_dbg(2, "\n");

+ if (ghes_edac_preferred())
+ return -EBUSY;
+
owner = edac_get_owner();
if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
return -EBUSY;
diff --git a/drivers/edac/igen6_edac.c b/drivers/edac/igen6_edac.c
index a07bbfd075d0..4ac6d0c533ec 100644
--- a/drivers/edac/igen6_edac.c
+++ b/drivers/edac/igen6_edac.c
@@ -1271,6 +1271,9 @@ static int __init igen6_init(void)

edac_dbg(2, "\n");

+ if (ghes_edac_preferred())
+ return -EBUSY;
+
owner = edac_get_owner();
if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
return -ENODEV;
diff --git a/drivers/edac/layerscape_edac.c b/drivers/edac/layerscape_edac.c
index 94cac7686a56..60ff4c6674cd 100644
--- a/drivers/edac/layerscape_edac.c
+++ b/drivers/edac/layerscape_edac.c
@@ -38,6 +38,9 @@ static int __init fsl_ddr_mc_init(void)
{
int res;

+ if (ghes_edac_preferred())
+ return -EBUSY;
+
/* make sure error reporting method is sane */
switch (edac_op_state) {
case EDAC_OPSTATE_POLL:
diff --git a/drivers/edac/pnd2_edac.c b/drivers/edac/pnd2_edac.c
index a20b299f1202..4ddc43e6c420 100644
--- a/drivers/edac/pnd2_edac.c
+++ b/drivers/edac/pnd2_edac.c
@@ -1528,6 +1528,9 @@ static int __init pnd2_init(void)

edac_dbg(2, "\n");

+ if (ghes_edac_preferred())
+ return -EBUSY;
+
owner = edac_get_owner();
if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
return -EBUSY;
diff --git a/drivers/edac/sb_edac.c b/drivers/edac/sb_edac.c
index 9678ab97c7ac..3ff604a5e95a 100644
--- a/drivers/edac/sb_edac.c
+++ b/drivers/edac/sb_edac.c
@@ -3506,6 +3506,9 @@ static int __init sbridge_init(void)

edac_dbg(2, "\n");

+ if (ghes_edac_preferred())
+ return -EBUSY;
+
owner = edac_get_owner();
if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
return -EBUSY;
diff --git a/drivers/edac/skx_base.c b/drivers/edac/skx_base.c
index 1abc020d49ab..286370728552 100644
--- a/drivers/edac/skx_base.c
+++ b/drivers/edac/skx_base.c
@@ -653,6 +653,9 @@ static int __init skx_init(void)

edac_dbg(2, "\n");

+ if (ghes_edac_preferred())
+ return -EBUSY;
+
owner = edac_get_owner();
if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
return -EBUSY;
diff --git a/drivers/edac/thunderx_edac.c b/drivers/edac/thunderx_edac.c
index f13674081cb6..2c4baa6817a9 100644
--- a/drivers/edac/thunderx_edac.c
+++ b/drivers/edac/thunderx_edac.c
@@ -2114,6 +2114,9 @@ static int __init thunderx_edac_init(void)
{
int rc = 0;

+ if (ghes_edac_preferred())
+ return -EBUSY;
+
rc = pci_register_driver(&thunderx_lmc_driver);
if (rc)
return rc;
diff --git a/drivers/edac/xgene_edac.c b/drivers/edac/xgene_edac.c
index 54081403db4f..9aa68220b625 100644
--- a/drivers/edac/xgene_edac.c
+++ b/drivers/edac/xgene_edac.c
@@ -2004,6 +2004,9 @@ static int __init xgene_edac_init(void)
{
int rc;

+ if (ghes_edac_preferred())
+ return -EBUSY;
+
/* Make sure error reporting method is sane */
switch (edac_op_state) {
case EDAC_OPSTATE_POLL:
--
2.25.1

2022-09-12 15:35:13

by Justin He

[permalink] [raw]

Subject: [PATCH v6 8/8] EDAC/igen6: Return consistent errno when another edac driver is enabled

Only a single edac driver can be enabled for EDAC MC. The igen6_init()
should be returned with EBUSY instead of ENODEV, which is consistent with
other edac drivers.

Signed-off-by: Jia He <[email protected]>
---
drivers/edac/igen6_edac.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/edac/igen6_edac.c b/drivers/edac/igen6_edac.c
index 4ac6d0c533ec..4646cb72b9f8 100644
--- a/drivers/edac/igen6_edac.c
+++ b/drivers/edac/igen6_edac.c
@@ -1276,7 +1276,7 @@ static int __init igen6_init(void)

owner = edac_get_owner();
if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
- return -ENODEV;
+ return -EBUSY;

edac_op_state = EDAC_OPSTATE_NMI;

--
2.25.1

2022-09-12 15:37:55

by Justin He

[permalink] [raw]

Subject: [PATCH v6 1/8] efi/cper: export several helpers for ghes_edac to use

Before ghes_edac can be turned back into a proper module again, export
several helpers which are going to be used by it.

Signed-off-by: Jia He <[email protected]>
Acked-by: Ard Biesheuvel <[email protected]>
---
drivers/firmware/efi/cper.c | 3 +++
1 file changed, 3 insertions(+)

diff --git a/drivers/firmware/efi/cper.c b/drivers/firmware/efi/cper.c
index e4e5ea7ce910..053eae13f409 100644
--- a/drivers/firmware/efi/cper.c
+++ b/drivers/firmware/efi/cper.c
@@ -290,6 +290,7 @@ int cper_mem_err_location(struct cper_mem_err_compact *mem, char *msg)

return n;
}
+EXPORT_SYMBOL_GPL(cper_mem_err_location);

int cper_dimm_err_location(struct cper_mem_err_compact *mem, char *msg)
{
@@ -310,6 +311,7 @@ int cper_dimm_err_location(struct cper_mem_err_compact *mem, char *msg)

return n;
}
+EXPORT_SYMBOL_GPL(cper_dimm_err_location);

void cper_mem_err_pack(const struct cper_sec_mem_err *mem,
struct cper_mem_err_compact *cmem)
@@ -331,6 +333,7 @@ void cper_mem_err_pack(const struct cper_sec_mem_err *mem,
cmem->mem_array_handle = mem->mem_array_handle;
cmem->mem_dev_handle = mem->mem_dev_handle;
}
+EXPORT_SYMBOL_GPL(cper_mem_err_pack);

const char *cper_mem_err_unpack(struct trace_seq *p,
struct cper_mem_err_compact *cmem)
--
2.25.1

2022-09-20 18:09:30

by Borislav Petkov

[permalink] [raw]

Subject: Re: [PATCH v6 5/8] EDAC/ghes: Make ghes_edac a proper module to remove the dependency on ghes

On Mon, Sep 12, 2022 at 02:40:02PM +0000, Jia He wrote:
> Commit dc4e8c07e9e2 ("ACPI: APEI: explicit init of HEST and GHES in
> apci_init()") introduced a bug that ghes_edac_register() would be invoked
> before edac_init(). Because at that time, the bus "edac" hadn't been even
> registered, this created sysfs /devices/mc0 instead of
> /sys/devices/system/edac/mc/mc0 on an Ampere eMag server.
>
> To remove the dependency of ghes_edac on ghes, make it a proper module. Use
> a list to save the probing devices in ghes_probe(), and defer the
> ghes_edac_register() to module_init() of the new ghes_edac module by
> iterating over the devices list.
>
> The ghes_edac_force_enable check is not needed in ghes_edac_unregister()
> since it has been checked in ghes_edac_init().
>
> Co-developed-by: Borislav Petkov <[email protected]>
> Signed-off-by: Borislav Petkov <[email protected]>
> Signed-off-by: Jia He <[email protected]>
> Fixes: dc4e8c07e9e2 ("ACPI: APEI: explicit init of HEST and GHES in apci_init()")
> Cc: Shuai Xue <[email protected]>
> Acked-by: Toshi Kani <[email protected]>
> ---
> drivers/acpi/apei/ghes.c | 22 +++++++++++++--
> drivers/edac/Kconfig | 4 +--
> drivers/edac/ghes_edac.c | 59 +++++++++++++++++++++++-----------------
> include/acpi/ghes.h | 24 ++++------------
> 4 files changed, 62 insertions(+), 47 deletions(-)

So those last three patches look unnecessarily complex. You don't
need to export ghes_edac_force_enable and you don't need that
ghes_edac_preferred() thing.

IOW, I'd like to see the below split into two patches:

1. Add ghes_get_devices() to ghes.c along with moving the
ghes_edac_force_enable param to ghes.c

2. Add init() and exit() module functions and fixup Kconfig

There's no need for ghes_present - ghes.c has either registered devices
or not.

And ghes_edac is preferred if ghes_get_devices() gives you a list of
devices or not.

IOW, it all boils down to whether the list returned is NULL or not.

HTH.

---
diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c
index 8cb65f757d06..27c72b175e4b 100644
--- a/drivers/acpi/apei/ghes.c
+++ b/drivers/acpi/apei/ghes.c
@@ -109,6 +109,13 @@ static inline bool is_hest_type_generic_v2(struct ghes *ghes)
bool ghes_disable;
module_param_named(disable, ghes_disable, bool, 0);

+/*
+ * "ghes.edac_force_enable" forcibly enables ghes_edac and skips the platform
+ * check.
+ */
+static bool ghes_edac_force_enable;
+module_param_named(edac_force_enable, ghes_edac_force_enable, bool, 0);
+
/*
* All error sources notified with HED (Hardware Error Device) share a
* single notifier callback, so they need to be linked and checked one
@@ -120,6 +127,9 @@ module_param_named(disable, ghes_disable, bool, 0);
static LIST_HEAD(ghes_hed);
static DEFINE_MUTEX(ghes_list_mutex);

+static LIST_HEAD(ghes_devs);
+static DEFINE_MUTEX(ghes_devs_mutex);
+
/*
* Because the memory area used to transfer hardware error information
* from BIOS to Linux can be determined only in NMI, IRQ or timer
@@ -1378,7 +1388,11 @@ static int ghes_probe(struct platform_device *ghes_dev)

platform_set_drvdata(ghes_dev, ghes);

- ghes_edac_register(ghes, &ghes_dev->dev);
+ ghes->dev = &ghes_dev->dev;
+
+ mutex_lock(&ghes_devs_mutex);
+ list_add_tail(&ghes->elist, &ghes_devs);
+ mutex_unlock(&ghes_devs_mutex);

/* Handle any pending errors right away */
spin_lock_irqsave(&ghes_notify_lock_irq, flags);
@@ -1442,7 +1456,9 @@ static int ghes_remove(struct platform_device *ghes_dev)

ghes_fini(ghes);

- ghes_edac_unregister(ghes);
+ mutex_lock(&ghes_devs_mutex);
+ list_del(&ghes->elist);
+ mutex_unlock(&ghes_devs_mutex);

kfree(ghes);

@@ -1500,6 +1516,32 @@ void __init acpi_ghes_init(void)
pr_info(GHES_PFX "Failed to enable APEI firmware first mode.\n");
}

+/*
+ * Known x86 systems that prefer GHES error reporting:
+ */
+static struct acpi_platform_list plat_list[] = {
+ {"HPE ", "Server ", 0, ACPI_SIG_FADT, all_versions},
+ { } /* End */
+};
+
+struct list_head *ghes_get_devices(void)
+{
+ int idx = -1;
+
+ if (IS_ENABLED(CONFIG_X86)) {
+ idx = acpi_match_platform_list(plat_list);
+ if (idx < 0) {
+ if (!ghes_edac_force_enable)
+ return NULL;
+
+ pr_warn_once("Force-loading ghes_edac on an unsupported platform. You're on your own!\n");
+ }
+ }
+
+ return &ghes_devs;
+}
+EXPORT_SYMBOL_GPL(ghes_get_devices);
+
void ghes_register_report_chain(struct notifier_block *nb)
{
atomic_notifier_chain_register(&ghes_report_chain, nb);
diff --git a/drivers/edac/Kconfig b/drivers/edac/Kconfig
index 17562cf1fe97..df45db81858b 100644
--- a/drivers/edac/Kconfig
+++ b/drivers/edac/Kconfig
@@ -53,8 +53,8 @@ config EDAC_DECODE_MCE
has been initialized.

config EDAC_GHES
- bool "Output ACPI APEI/GHES BIOS detected errors via EDAC"
- depends on ACPI_APEI_GHES && (EDAC=y)
+ tristate "Output ACPI APEI/GHES BIOS detected errors via EDAC"
+ depends on ACPI_APEI_GHES
select UEFI_CPER
help
Not all machines support hardware-driven error report. Some of those
diff --git a/drivers/edac/ghes_edac.c b/drivers/edac/ghes_edac.c
index 7b8d56a769f6..8460108e8623 100644
--- a/drivers/edac/ghes_edac.c
+++ b/drivers/edac/ghes_edac.c
@@ -54,12 +54,10 @@ static DEFINE_MUTEX(ghes_reg_mutex);
*/
static DEFINE_SPINLOCK(ghes_lock);

-/* "ghes_edac.force_load=1" skips the platform check */
-static bool __read_mostly force_load;
-module_param(force_load, bool, 0);
-
static bool system_scanned;

+static struct list_head *ghes_devs;
+
/* Memory Device - Type 17 of SMBIOS spec */
struct memdev_dmi_entry {
u8 type;
@@ -387,34 +385,15 @@ static struct notifier_block ghes_edac_mem_err_nb = {
.priority = 0,
};

-/*
- * Known systems that are safe to enable this module.
- */
-static struct acpi_platform_list plat_list[] = {
- {"HPE ", "Server ", 0, ACPI_SIG_FADT, all_versions},
- { } /* End */
-};
-
-int ghes_edac_register(struct ghes *ghes, struct device *dev)
+static int ghes_edac_register(struct device *dev)
{
bool fake = false;
struct mem_ctl_info *mci;
struct ghes_pvt *pvt;
struct edac_mc_layer layers[1];
unsigned long flags;
- int idx = -1;
int rc = 0;

- if (IS_ENABLED(CONFIG_X86)) {
- /* Check if safe to enable on this system */
- idx = acpi_match_platform_list(plat_list);
- if (!force_load && idx < 0)
- return -ENODEV;
- } else {
- force_load = true;
- idx = 0;
- }
-
/* finish another registration/unregistration instance first */
mutex_lock(&ghes_reg_mutex);

@@ -458,15 +437,10 @@ int ghes_edac_register(struct ghes *ghes, struct device *dev)
pr_info("This system has a very crappy BIOS: It doesn't even list the DIMMS.\n");
pr_info("Its SMBIOS info is wrong. It is doubtful that the error report would\n");
pr_info("work on such system. Use this driver with caution\n");
- } else if (idx < 0) {
- pr_info("This EDAC driver relies on BIOS to enumerate memory and get error reports.\n");
- pr_info("Unfortunately, not all BIOSes reflect the memory layout correctly.\n");
- pr_info("So, the end result of using this driver varies from vendor to vendor.\n");
- pr_info("If you find incorrect reports, please contact your hardware vendor\n");
- pr_info("to correct its BIOS.\n");
- pr_info("This system has %d DIMM sockets.\n", ghes_hw.num_dimms);
}

+ pr_info("This system has %d DIMM sockets.\n", ghes_hw.num_dimms);
+
if (!fake) {
struct dimm_info *src, *dst;
int i = 0;
@@ -530,14 +504,11 @@ int ghes_edac_register(struct ghes *ghes, struct device *dev)
return rc;
}

-void ghes_edac_unregister(struct ghes *ghes)
+static void ghes_edac_unregister(struct ghes *ghes)
{
struct mem_ctl_info *mci;
unsigned long flags;

- if (!force_load)
- return;
-
mutex_lock(&ghes_reg_mutex);

system_scanned = false;
@@ -566,3 +537,32 @@ void ghes_edac_unregister(struct ghes *ghes)
unlock:
mutex_unlock(&ghes_reg_mutex);
}
+
+static int __init ghes_edac_init(void)
+{
+ struct ghes *g, *g_tmp;
+
+ ghes_devs = ghes_get_devices();
+ if (!ghes_devs)
+ return -ENODEV;
+
+ list_for_each_entry_safe(g, g_tmp, ghes_devs, elist) {
+ ghes_edac_register(g->dev);
+ }
+
+ return 0;
+}
+module_init(ghes_edac_init);
+
+static void __exit ghes_edac_exit(void)
+{
+ struct ghes *g, *g_tmp;
+
+ list_for_each_entry_safe(g, g_tmp, ghes_devs, elist) {
+ ghes_edac_unregister(g);
+ }
+}
+module_exit(ghes_edac_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Output ACPI APEI/GHES BIOS detected errors module via EDAC");
diff --git a/include/acpi/ghes.h b/include/acpi/ghes.h
index 5cbd38b6e4e1..2e785d3554d8 100644
--- a/include/acpi/ghes.h
+++ b/include/acpi/ghes.h
@@ -27,6 +27,8 @@ struct ghes {
struct timer_list timer;
unsigned int irq;
};
+ struct device *dev;
+ struct list_head elist;
};

struct ghes_estatus_node {
@@ -69,28 +71,14 @@ int ghes_register_vendor_record_notifier(struct notifier_block *nb);
* @nb: pointer to the notifier_block structure of the vendor record handler.
*/
void ghes_unregister_vendor_record_notifier(struct notifier_block *nb);
-#endif
-
-int ghes_estatus_pool_init(int num_ghes);
-
-/* From drivers/edac/ghes_edac.c */
-
-#ifdef CONFIG_EDAC_GHES
-int ghes_edac_register(struct ghes *ghes, struct device *dev);
-
-void ghes_edac_unregister(struct ghes *ghes);

+struct list_head *ghes_get_devices(void);
#else
-static inline int ghes_edac_register(struct ghes *ghes, struct device *dev)
-{
- return -ENODEV;
-}
-
-static inline void ghes_edac_unregister(struct ghes *ghes)
-{
-}
+static inline struct list_head *ghes_get_devices(void) { return NULL; }
#endif

+int ghes_estatus_pool_init(int num_ghes);
+
static inline int acpi_hest_get_version(struct acpi_hest_generic_data *gdata)
{
return gdata->revision >> 8;

--
Regards/Gruss,
Boris.

https://people.kernel.org/tglx/notes-about-netiquette

2022-09-22 08:30:18

by Justin He

[permalink] [raw]

Subject: RE: [PATCH v6 5/8] EDAC/ghes: Make ghes_edac a proper module to remove the dependency on ghes

Hi Borislav

> -----Original Message-----
> From: Borislav Petkov <[email protected]>
> Sent: Wednesday, September 21, 2022 1:22 AM
> To: Justin He <[email protected]>
> Cc: Len Brown <[email protected]>; James Morse <[email protected]>;
> Tony Luck <[email protected]>; Mauro Carvalho Chehab
> <[email protected]>; Robert Richter <[email protected]>; Robert Moore
> <[email protected]>; Qiuxu Zhuo <[email protected]>; Yazen
> Ghannam <[email protected]>; Jan Luebbe <[email protected]>;
> Khuong Dinh <[email protected]>; Kani Toshi
> <[email protected]>; Ard Biesheuvel <[email protected]>;
> [email protected]; [email protected];
> [email protected]; [email protected]; Rafael J . Wysocki
> <[email protected]>; Shuai Xue <[email protected]>; Jarkko
> Sakkinen <[email protected]>; [email protected]; nd <[email protected]>
> Subject: Re: [PATCH v6 5/8] EDAC/ghes: Make ghes_edac a proper module to
> remove the dependency on ghes
>
> On Mon, Sep 12, 2022 at 02:40:02PM +0000, Jia He wrote:
> > Commit dc4e8c07e9e2 ("ACPI: APEI: explicit init of HEST and GHES in
> > apci_init()") introduced a bug that ghes_edac_register() would be
> > invoked before edac_init(). Because at that time, the bus "edac"
> > hadn't been even registered, this created sysfs /devices/mc0 instead
> > of
> > /sys/devices/system/edac/mc/mc0 on an Ampere eMag server.
> >
> > To remove the dependency of ghes_edac on ghes, make it a proper
> > module. Use a list to save the probing devices in ghes_probe(), and
> > defer the
> > ghes_edac_register() to module_init() of the new ghes_edac module by
> > iterating over the devices list.
> >
> > The ghes_edac_force_enable check is not needed in
> > ghes_edac_unregister() since it has been checked in ghes_edac_init().
> >
> > Co-developed-by: Borislav Petkov <[email protected]>
> > Signed-off-by: Borislav Petkov <[email protected]>
> > Signed-off-by: Jia He <[email protected]>
> > Fixes: dc4e8c07e9e2 ("ACPI: APEI: explicit init of HEST and GHES in
> > apci_init()")
> > Cc: Shuai Xue <[email protected]>
> > Acked-by: Toshi Kani <[email protected]>
> > ---
> > drivers/acpi/apei/ghes.c | 22 +++++++++++++--
> > drivers/edac/Kconfig | 4 +--
> > drivers/edac/ghes_edac.c | 59
> +++++++++++++++++++++++-----------------
> > include/acpi/ghes.h | 24 ++++------------
> > 4 files changed, 62 insertions(+), 47 deletions(-)
>
> So those last three patches look unnecessarily complex. You don't need to
> export ghes_edac_force_enable and you don't need that
> ghes_edac_preferred() thing.
>
> IOW, I'd like to see the below split into two patches:
>
> 1. Add ghes_get_devices() to ghes.c along with moving the
> ghes_edac_force_enable param to ghes.c
>
> 2. Add init() and exit() module functions and fixup Kconfig
>
> There's no need for ghes_present - ghes.c has either registered devices or
> not.
>

If there is no ghes_present flag.
What if ghes.disable is passed to kernel boot parameter and then ghes_edac is
loaded by modprobe?
Thus, ghes_edac can be loaded even if ghes is disabled. (ghes_dev list is null)

The ghes_present is to avoid this case.

What do u think of it?

--
Cheers,
Justin (Jia He)

2022-09-23 16:41:02

by Borislav Petkov

[permalink] [raw]

Subject: Re: [PATCH v6 5/8] EDAC/ghes: Make ghes_edac a proper module to remove the dependency on ghes

On Thu, Sep 22, 2022 at 08:26:16AM +0000, Justin He wrote:
> If there is no ghes_present flag.
> What if ghes.disable is passed to kernel boot parameter and then ghes_edac is
> loaded by modprobe?
> Thus, ghes_edac can be loaded even if ghes is disabled. (ghes_dev list is null)

Yes, and what happens if ghes_dev is NULL?

The other drivers would do in their init function:

struct list_head *ghes_devs;

ghes_devs = ghes_get_devices();
if (ghes_devs)
return -ENODEV;

/* Continue init */

and then load in that case because user has disabled GHES and thus no
ghes_edac either. So the platform-specific one loads.

Right?

--
Regards/Gruss,
Boris.

https://people.kernel.org/tglx/notes-about-netiquette

2022-09-25 15:21:39

by Justin He

[permalink] [raw]

Subject: RE: [PATCH v6 5/8] EDAC/ghes: Make ghes_edac a proper module to remove the dependency on ghes

Hi Borislav

> -----Original Message-----
> From: Borislav Petkov <[email protected]>
> Sent: Friday, September 23, 2022 11:43 PM
> To: Justin He <[email protected]>
> Cc: Len Brown <[email protected]>; James Morse <[email protected]>;
> Tony Luck <[email protected]>; Mauro Carvalho Chehab
> <[email protected]>; Robert Richter <[email protected]>; Robert Moore
> <[email protected]>; Qiuxu Zhuo <[email protected]>; Yazen
> Ghannam <[email protected]>; Jan Luebbe <[email protected]>;
> Khuong Dinh <[email protected]>; Kani Toshi
> <[email protected]>; Ard Biesheuvel <[email protected]>;
> [email protected]; [email protected];
> [email protected]; [email protected]; Rafael J . Wysocki
> <[email protected]>; Shuai Xue <[email protected]>; Jarkko
> Sakkinen <[email protected]>; [email protected]; nd <[email protected]>
> Subject: Re: [PATCH v6 5/8] EDAC/ghes: Make ghes_edac a proper module to
> remove the dependency on ghes
>
> On Thu, Sep 22, 2022 at 08:26:16AM +0000, Justin He wrote:
> > If there is no ghes_present flag.
> > What if ghes.disable is passed to kernel boot parameter and then
> > ghes_edac is loaded by modprobe?
> > Thus, ghes_edac can be loaded even if ghes is disabled. (ghes_dev list
> > is null)
>
> Yes, and what happens if ghes_dev is NULL?
>
Okay, I am fine if it is allowed to register ghes_dev with
> The other drivers would do in their init function:
>
> struct list_head *ghes_devs;
>
> ghes_devs = ghes_get_devices();
> if (ghes_devs)
> return -ENODEV;
>
> /* Continue init */
>
> and then load in that case because user has disabled GHES and thus no
> ghes_edac either. So the platform-specific one loads.

Okay, I am fine with removing ghes_present and ghes_edac_preferred if
It is ok to register empty ghes_dev list without prompting an error

--
Cheers,
Justin (Jia He)

2022-09-26 12:52:26

by Borislav Petkov

[permalink] [raw]

Subject: Re: [PATCH v6 5/8] EDAC/ghes: Make ghes_edac a proper module to remove the dependency on ghes

On Sun, Sep 25, 2022 at 03:07:12PM +0000, Justin He wrote:
> Okay, I am fine with removing ghes_present and ghes_edac_preferred if
> It is ok to register empty ghes_dev list without prompting an error

What "prompting an error"?

In that case, you simply return a NULL list which says that GHES hasn't
been initialized...

--
Regards/Gruss,
Boris.

https://people.kernel.org/tglx/notes-about-netiquette